commit 6cb6ab1d6d99a27d6243a1c8ea42f3adf8516cb5 Author: ModelHub XC Date: Sat Apr 11 13:10:00 2026 +0800 初始化项目,由ModelHub XC社区提供模型 Model: Neelectric/Llama-3.1-8B-Instruct_SFT_sciencefisher_v00.13 Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..52373fe --- /dev/null +++ b/.gitattributes @@ -0,0 +1,36 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..3163899 --- /dev/null +++ b/README.md @@ -0,0 +1,60 @@ +--- +base_model: meta-llama/Llama-3.1-8B-Instruct +datasets: Neelectric/MoT_science_Llama3_4096toks +library_name: transformers +model_name: Llama-3.1-8B-Instruct_SFT_sciencefisher_v00.13 +tags: +- generated_from_trainer +- trl +- open-r1 +- sft +licence: license +--- + +# Model Card for Llama-3.1-8B-Instruct_SFT_sciencefisher_v00.13 + +This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on the [Neelectric/MoT_science_Llama3_4096toks](https://huggingface.co/datasets/Neelectric/MoT_science_Llama3_4096toks) dataset. +It has been trained using [TRL](https://github.com/huggingface/trl). + +## Quick start + +```python +from transformers import pipeline + +question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?" +generator = pipeline("text-generation", model="Neelectric/Llama-3.1-8B-Instruct_SFT_sciencefisher_v00.13", device="cuda") +output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0] +print(output["generated_text"]) +``` + +## Training procedure + +[Visualize in Weights & Biases](https://wandb.ai/neelectric/open-r1_science/runs/1ca6zerq) + + + +This model was trained with SFT. + +### Framework versions + +- TRL: 1.0.0.dev0 +- Transformers: 4.57.6 +- Pytorch: 2.9.0 +- Datasets: 4.8.4 +- Tokenizers: 0.22.2 + +## Citations + + + +Cite TRL as: + +```bibtex +@software{vonwerra2020trl, + title = {{TRL: Transformers Reinforcement Learning}}, + author = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin}, + license = {Apache-2.0}, + url = {https://github.com/huggingface/trl}, + year = {2020} +} +``` \ No newline at end of file diff --git a/all_results.json b/all_results.json new file mode 100644 index 0000000..a49f1b3 --- /dev/null +++ b/all_results.json @@ -0,0 +1,9 @@ +{ + "ewc_loss": 1.055002212524414e-05, + "total_flos": 3.1829322264618205e+19, + "train_loss": 1.0248253059881185, + "train_runtime": 37191.1487, + "train_samples": 145693, + "train_samples_per_second": 11.752, + "train_steps_per_second": 0.735 +} \ No newline at end of file diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000..0ab931a --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,121 @@ +{{- bos_token }} +{%- if custom_tools is defined %} + {%- set tools = custom_tools %} +{%- endif %} +{%- if not tools_in_user_message is defined %} + {%- set tools_in_user_message = true %} +{%- endif %} +{%- if not date_string is defined %} + {%- set date_string = "26 Jul 2024" %} +{%- endif %} +{%- if not tools is defined %} + {%- set tools = none %} +{%- endif %} + +{#- This block extracts the system message, so we can slot it into the right place. #} +{%- if messages[0]['role'] == 'system' %} + {%- set system_message = messages[0]['content']|trim %} + {%- set messages = messages[1:] %} +{%- else %} + {%- set system_message = "You are a helpful AI Assistant that provides well-reasoned and detailed responses. You first think about the reasoning process as an internal monologue and then provide the user with the answer. Respond in the following format: +... + + +... +" %} +{%- endif %} + +{#- System message + builtin tools #} +{{- "<|start_header_id|>system<|end_header_id|>\n\n" }} +{%- if builtin_tools is defined or tools is not none %} + {{- "Environment: ipython\n" }} +{%- endif %} +{%- if builtin_tools is defined %} + {{- "Tools: " + builtin_tools | reject('equalto', 'code_interpreter') | join(", ") + "\n\n"}} +{%- endif %} +{{- "Cutting Knowledge Date: December 2023\n" }} +{{- "Today Date: " + date_string + "\n\n" }} +{%- if tools is not none and not tools_in_user_message %} + {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }} + {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }} + {{- "Do not use variables.\n\n" }} + {%- for t in tools %} + {{- t | tojson(indent=4) }} + {{- "\n\n" }} + {%- endfor %} +{%- endif %} +{{- system_message }} +{{- "<|eot_id|>" }} + +{#- Custom tools are passed in a user message with some extra guidance #} +{%- if tools_in_user_message and not tools is none %} + {#- Extract the first user message so we can plug it in here #} + {%- if messages | length != 0 %} + {%- set first_user_message = messages[0]['content']|trim %} + {%- set messages = messages[1:] %} + {%- else %} + {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }} + {%- endif %} + {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}} + {{- "Given the following functions, please respond with a JSON for a function call " }} + {{- "with its proper arguments that best answers the given prompt.\n\n" }} + {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }} + {{- "Do not use variables.\n\n" }} + {%- for t in tools %} + {{- t | tojson(indent=4) }} + {{- "\n\n" }} + {%- endfor %} + {{- first_user_message + "<|eot_id|>"}} +{%- endif %} + +{%- for message in messages %} + {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %} + {%- if message['role'] == 'assistant' %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }} + {% generation %} + {{- message['content'] | trim + '<|eot_id|>' }} + {% endgeneration %} + {%- else %} + {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' }} + {%- endif %} + {%- elif 'tool_calls' in message %} + {%- if not message.tool_calls|length == 1 %} + {{- raise_exception("This model only supports single tool-calls at once!") }} + {%- endif %} + {%- set tool_call = message.tool_calls[0].function %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}} + {% generation %} + {%- if builtin_tools is defined and tool_call.name in builtin_tools %} + {{- "<|python_tag|>" + tool_call.name + ".call(" }} + {%- for arg_name, arg_val in tool_call.arguments | items %} + {{- arg_name + '="' + arg_val + '"' }} + {%- if not loop.last %} + {{- ", " }} + {%- endif %} + {%- endfor %} + {{- ")" }} + {%- else %} + {{- '{"name": "' + tool_call.name + '", ' }} + {{- '"parameters": ' }} + {{- tool_call.arguments | tojson }} + {{- "}" }} + {%- endif %} + {%- if builtin_tools is defined %} + {{- "<|eom_id|>" }} + {%- else %} + {{- "<|eot_id|>" }} + {%- endif %} + {% endgeneration %} + {%- elif message.role == "tool" or message.role == "ipython" %} + {{- "<|start_header_id|>ipython<|end_header_id|>\n\n" }} + {%- if message.content is mapping or message.content is iterable %} + {{- message.content | tojson }} + {%- else %} + {{- message.content }} + {%- endif %} + {{- "<|eot_id|>" }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }} +{%- endif %} \ No newline at end of file diff --git a/config.json b/config.json new file mode 100644 index 0000000..06df27b --- /dev/null +++ b/config.json @@ -0,0 +1,36 @@ +{ + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "dtype": "bfloat16", + "eos_token_id": 128009, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 14336, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 32, + "num_hidden_layers": 32, + "num_key_value_heads": 8, + "pad_token_id": 128009, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 8.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": false, + "transformers_version": "4.57.6", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000..50f6077 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,9 @@ +{ + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128009, + "pad_token_id": 128009, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.57.6" +} diff --git a/model-00001-of-00004.safetensors b/model-00001-of-00004.safetensors new file mode 100644 index 0000000..fa2354c --- /dev/null +++ b/model-00001-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0db2ea3abf3d10f1d847260625793e486263a367e56a5d6592695028a648de6 +size 4976698672 diff --git a/model-00002-of-00004.safetensors b/model-00002-of-00004.safetensors new file mode 100644 index 0000000..09f66b1 --- /dev/null +++ b/model-00002-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cc95cc7a7764781e168a27a0d9421fe69ea87b35de60cad0421ecc3da9033fce +size 4999802720 diff --git a/model-00003-of-00004.safetensors b/model-00003-of-00004.safetensors new file mode 100644 index 0000000..1d80a63 --- /dev/null +++ b/model-00003-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a96176d278aea99e40f9069fddf6755fde37e581f9efa1453942be3374f7428f +size 4915916176 diff --git a/model-00004-of-00004.safetensors b/model-00004-of-00004.safetensors new file mode 100644 index 0000000..3f2dadd --- /dev/null +++ b/model-00004-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e799d042469808c04e0a437426d7a73f9b9e51b9982d2b28fc7e7883f255bcb5 +size 1168138808 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000..5c64f1e --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,299 @@ +{ + "metadata": { + "total_parameters": 8030261248, + "total_size": 16060522496 + }, + "weight_map": { + "lm_head.weight": "model-00004-of-00004.safetensors", + "model.embed_tokens.weight": "model-00001-of-00004.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.28.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.29.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.29.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.30.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.30.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.30.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.30.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.30.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.31.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.31.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.31.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.31.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.31.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.norm.weight": "model-00004-of-00004.safetensors" + } +} diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000..9d4773c --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,11 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": "<|eot_id|>", + "pad_token": "<|eot_id|>" +} diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000..1c1d8d5 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..3beeacc --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,2063 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|eot_id|>", + "extra_special_tokens": {}, + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "<|eot_id|>", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/train_results.json b/train_results.json new file mode 100644 index 0000000..a49f1b3 --- /dev/null +++ b/train_results.json @@ -0,0 +1,9 @@ +{ + "ewc_loss": 1.055002212524414e-05, + "total_flos": 3.1829322264618205e+19, + "train_loss": 1.0248253059881185, + "train_runtime": 37191.1487, + "train_samples": 145693, + "train_samples_per_second": 11.752, + "train_steps_per_second": 0.735 +} \ No newline at end of file diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000..34c546a --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,273224 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 27318, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00010981770261366133, + "ewc_loss": 0.0, + "grad_norm": 6.513922691345215, + "learning_rate": 0.0, + "loss": 1.5235, + "mean_token_accuracy": 0.6157128810882568, + "num_tokens": 30018.0, + "step": 1 + }, + { + "epoch": 0.00021963540522732265, + "ewc_loss": 0.0, + "grad_norm": 6.409045219421387, + "learning_rate": 3.6603221083455343e-10, + "loss": 1.4062, + "mean_token_accuracy": 0.628243088722229, + "num_tokens": 58744.0, + "step": 2 + }, + { + "epoch": 0.000329453107840984, + "ewc_loss": 5.488773498207866e-19, + "grad_norm": 7.285305023193359, + "learning_rate": 7.320644216691069e-10, + "loss": 1.4743, + "mean_token_accuracy": 0.6152434945106506, + "num_tokens": 84293.0, + "step": 3 + }, + { + "epoch": 0.0004392708104546453, + "ewc_loss": 2.5912431922403556e-17, + "grad_norm": 7.035626411437988, + "learning_rate": 1.0980966325036603e-09, + "loss": 1.5503, + "mean_token_accuracy": 0.5980188846588135, + "num_tokens": 113639.0, + "step": 4 + }, + { + "epoch": 0.0005490885130683066, + "ewc_loss": 5.93275428784068e-16, + "grad_norm": 7.09148645401001, + "learning_rate": 1.4641288433382137e-09, + "loss": 1.515, + "mean_token_accuracy": 0.60816490650177, + "num_tokens": 139615.0, + "step": 5 + }, + { + "epoch": 0.000658906215681968, + "ewc_loss": 7.008282842946301e-16, + "grad_norm": 6.862255096435547, + "learning_rate": 1.8301610541727671e-09, + "loss": 1.6124, + "mean_token_accuracy": 0.5816854238510132, + "num_tokens": 168855.0, + "step": 6 + }, + { + "epoch": 0.0007687239182956292, + "ewc_loss": 2.192690473634684e-15, + "grad_norm": 7.647518634796143, + "learning_rate": 2.1961932650073206e-09, + "loss": 1.523, + "mean_token_accuracy": 0.6033315062522888, + "num_tokens": 192650.0, + "step": 7 + }, + { + "epoch": 0.0008785416209092906, + "ewc_loss": 1.0824674490095276e-14, + "grad_norm": 6.311500072479248, + "learning_rate": 2.562225475841874e-09, + "loss": 1.5186, + "mean_token_accuracy": 0.604823112487793, + "num_tokens": 225812.0, + "step": 8 + }, + { + "epoch": 0.000988359323522952, + "ewc_loss": 1.9206858326015208e-14, + "grad_norm": 8.877950668334961, + "learning_rate": 2.9282576866764274e-09, + "loss": 1.5182, + "mean_token_accuracy": 0.6049196124076843, + "num_tokens": 245223.0, + "step": 9 + }, + { + "epoch": 0.0010981770261366132, + "ewc_loss": 2.3869795029440866e-14, + "grad_norm": 6.8741984367370605, + "learning_rate": 3.294289897510981e-09, + "loss": 1.566, + "mean_token_accuracy": 0.5959545373916626, + "num_tokens": 273619.0, + "step": 10 + }, + { + "epoch": 0.0012079947287502745, + "ewc_loss": 7.327471962526033e-14, + "grad_norm": 7.1762309074401855, + "learning_rate": 3.6603221083455343e-09, + "loss": 1.3786, + "mean_token_accuracy": 0.6409702301025391, + "num_tokens": 297418.0, + "step": 11 + }, + { + "epoch": 0.001317812431363936, + "ewc_loss": 9.725553695716371e-14, + "grad_norm": 6.06724739074707, + "learning_rate": 4.026354319180088e-09, + "loss": 1.3934, + "mean_token_accuracy": 0.6353172063827515, + "num_tokens": 328966.0, + "step": 12 + }, + { + "epoch": 0.0014276301339775973, + "ewc_loss": 1.341149413747189e-13, + "grad_norm": 7.5492401123046875, + "learning_rate": 4.392386530014641e-09, + "loss": 1.389, + "mean_token_accuracy": 0.6386103630065918, + "num_tokens": 353003.0, + "step": 13 + }, + { + "epoch": 0.0015374478365912585, + "ewc_loss": 1.6342482922482304e-13, + "grad_norm": 7.253172874450684, + "learning_rate": 4.758418740849194e-09, + "loss": 1.4652, + "mean_token_accuracy": 0.6212809681892395, + "num_tokens": 377957.0, + "step": 14 + }, + { + "epoch": 0.0016472655392049198, + "ewc_loss": 5.933031843596837e-13, + "grad_norm": 8.110651016235352, + "learning_rate": 5.124450951683748e-09, + "loss": 1.4742, + "mean_token_accuracy": 0.6143144965171814, + "num_tokens": 398957.0, + "step": 15 + }, + { + "epoch": 0.0017570832418185812, + "ewc_loss": 7.318590178329032e-13, + "grad_norm": 7.203577995300293, + "learning_rate": 5.490483162518302e-09, + "loss": 1.5625, + "mean_token_accuracy": 0.5943835377693176, + "num_tokens": 424867.0, + "step": 16 + }, + { + "epoch": 0.0018669009444322424, + "ewc_loss": 8.242295734817162e-13, + "grad_norm": 6.981134414672852, + "learning_rate": 5.856515373352855e-09, + "loss": 1.447, + "mean_token_accuracy": 0.6198843121528625, + "num_tokens": 450753.0, + "step": 17 + }, + { + "epoch": 0.001976718647045904, + "ewc_loss": 1.0089706847793423e-12, + "grad_norm": 6.850698471069336, + "learning_rate": 6.222547584187408e-09, + "loss": 1.453, + "mean_token_accuracy": 0.6167309284210205, + "num_tokens": 478721.0, + "step": 18 + }, + { + "epoch": 0.002086536349659565, + "ewc_loss": 1.1510792319313623e-12, + "grad_norm": 8.587575912475586, + "learning_rate": 6.588579795021962e-09, + "loss": 1.5912, + "mean_token_accuracy": 0.5925484299659729, + "num_tokens": 499326.0, + "step": 19 + }, + { + "epoch": 0.0021963540522732263, + "ewc_loss": 3.581135388230905e-12, + "grad_norm": 7.6291303634643555, + "learning_rate": 6.954612005856515e-09, + "loss": 1.4425, + "mean_token_accuracy": 0.6252572536468506, + "num_tokens": 522720.0, + "step": 20 + }, + { + "epoch": 0.002306171754886888, + "ewc_loss": 4.945377440890297e-12, + "grad_norm": 6.308166980743408, + "learning_rate": 7.3206442166910685e-09, + "loss": 1.4212, + "mean_token_accuracy": 0.6313834190368652, + "num_tokens": 551422.0, + "step": 21 + }, + { + "epoch": 0.002415989457500549, + "ewc_loss": 5.9969806898152456e-12, + "grad_norm": 8.054349899291992, + "learning_rate": 7.686676427525622e-09, + "loss": 1.5619, + "mean_token_accuracy": 0.6015436053276062, + "num_tokens": 572912.0, + "step": 22 + }, + { + "epoch": 0.0025258071601142102, + "ewc_loss": 6.480149750132114e-12, + "grad_norm": 5.989625453948975, + "learning_rate": 8.052708638360176e-09, + "loss": 1.4107, + "mean_token_accuracy": 0.6290940046310425, + "num_tokens": 604172.0, + "step": 23 + }, + { + "epoch": 0.002635624862727872, + "ewc_loss": 7.275957614183426e-12, + "grad_norm": 7.531951427459717, + "learning_rate": 8.418740849194728e-09, + "loss": 1.4466, + "mean_token_accuracy": 0.6204436421394348, + "num_tokens": 627461.0, + "step": 24 + }, + { + "epoch": 0.002745442565341533, + "ewc_loss": 8.01492205937393e-12, + "grad_norm": 6.363536834716797, + "learning_rate": 8.784773060029282e-09, + "loss": 1.4484, + "mean_token_accuracy": 0.6278089284896851, + "num_tokens": 657516.0, + "step": 25 + }, + { + "epoch": 0.0028552602679551946, + "ewc_loss": 8.86757334228605e-12, + "grad_norm": 6.289531230926514, + "learning_rate": 9.150805270863836e-09, + "loss": 1.5441, + "mean_token_accuracy": 0.5948554873466492, + "num_tokens": 691937.0, + "step": 26 + }, + { + "epoch": 0.0029650779705688557, + "ewc_loss": 1.318767317570746e-11, + "grad_norm": 8.475261688232422, + "learning_rate": 9.516837481698389e-09, + "loss": 1.4641, + "mean_token_accuracy": 0.6256919503211975, + "num_tokens": 711900.0, + "step": 27 + }, + { + "epoch": 0.003074895673182517, + "ewc_loss": 3.1604940886609256e-11, + "grad_norm": 7.216952323913574, + "learning_rate": 9.882869692532943e-09, + "loss": 1.4915, + "mean_token_accuracy": 0.6120486259460449, + "num_tokens": 737085.0, + "step": 28 + }, + { + "epoch": 0.0031847133757961785, + "ewc_loss": 3.956301952712238e-11, + "grad_norm": 8.050946235656738, + "learning_rate": 1.0248901903367497e-08, + "loss": 1.5747, + "mean_token_accuracy": 0.6034277081489563, + "num_tokens": 759817.0, + "step": 29 + }, + { + "epoch": 0.0032945310784098397, + "ewc_loss": 4.433786671143025e-11, + "grad_norm": 7.564850330352783, + "learning_rate": 1.0614934114202049e-08, + "loss": 1.418, + "mean_token_accuracy": 0.6270666122436523, + "num_tokens": 783353.0, + "step": 30 + }, + { + "epoch": 0.003404348781023501, + "ewc_loss": 4.8203219193965197e-11, + "grad_norm": 8.319191932678223, + "learning_rate": 1.0980966325036603e-08, + "loss": 1.4705, + "mean_token_accuracy": 0.6181522607803345, + "num_tokens": 804197.0, + "step": 31 + }, + { + "epoch": 0.0035141664836371624, + "ewc_loss": 5.070432962384075e-11, + "grad_norm": 6.864448070526123, + "learning_rate": 1.1346998535871157e-08, + "loss": 1.5125, + "mean_token_accuracy": 0.6061661243438721, + "num_tokens": 830796.0, + "step": 32 + }, + { + "epoch": 0.0036239841862508236, + "ewc_loss": 5.3432813729159534e-11, + "grad_norm": 6.5702409744262695, + "learning_rate": 1.171303074670571e-08, + "loss": 1.4162, + "mean_token_accuracy": 0.6406641006469727, + "num_tokens": 858092.0, + "step": 33 + }, + { + "epoch": 0.0037338018888644848, + "ewc_loss": 5.820766091346741e-11, + "grad_norm": 7.31693696975708, + "learning_rate": 1.2079062957540264e-08, + "loss": 1.4551, + "mean_token_accuracy": 0.6137213706970215, + "num_tokens": 882474.0, + "step": 34 + }, + { + "epoch": 0.0038436195914781464, + "ewc_loss": 6.275513442233205e-11, + "grad_norm": 8.507896423339844, + "learning_rate": 1.2445095168374816e-08, + "loss": 1.4566, + "mean_token_accuracy": 0.6248745918273926, + "num_tokens": 902361.0, + "step": 35 + }, + { + "epoch": 0.003953437294091808, + "ewc_loss": 6.730260793119669e-11, + "grad_norm": 7.381107807159424, + "learning_rate": 1.281112737920937e-08, + "loss": 1.5079, + "mean_token_accuracy": 0.6060892343521118, + "num_tokens": 927171.0, + "step": 36 + }, + { + "epoch": 0.004063254996705469, + "ewc_loss": 7.23048287909478e-11, + "grad_norm": 7.0726237297058105, + "learning_rate": 1.3177159590043924e-08, + "loss": 1.472, + "mean_token_accuracy": 0.6123760938644409, + "num_tokens": 952048.0, + "step": 37 + }, + { + "epoch": 0.00417307269931913, + "ewc_loss": 1.7007550923153758e-10, + "grad_norm": 6.869122505187988, + "learning_rate": 1.3543191800878477e-08, + "loss": 1.5743, + "mean_token_accuracy": 0.5971644520759583, + "num_tokens": 981040.0, + "step": 38 + }, + { + "epoch": 0.004282890401932792, + "ewc_loss": 2.4920154828578234e-10, + "grad_norm": 6.885048866271973, + "learning_rate": 1.390922401171303e-08, + "loss": 1.4834, + "mean_token_accuracy": 0.6067895889282227, + "num_tokens": 1008756.0, + "step": 39 + }, + { + "epoch": 0.004392708104546453, + "ewc_loss": 2.9103830456733704e-10, + "grad_norm": 7.084139347076416, + "learning_rate": 1.4275256222547585e-08, + "loss": 1.4985, + "mean_token_accuracy": 0.6015279293060303, + "num_tokens": 1035140.0, + "step": 40 + }, + { + "epoch": 0.004502525807160114, + "ewc_loss": 3.183231456205249e-10, + "grad_norm": 7.735499382019043, + "learning_rate": 1.4641288433382137e-08, + "loss": 1.5253, + "mean_token_accuracy": 0.6023635268211365, + "num_tokens": 1058792.0, + "step": 41 + }, + { + "epoch": 0.004612343509773776, + "ewc_loss": 3.3287506084889174e-10, + "grad_norm": 5.63086462020874, + "learning_rate": 1.500732064421669e-08, + "loss": 1.4629, + "mean_token_accuracy": 0.6143208742141724, + "num_tokens": 1096383.0, + "step": 42 + }, + { + "epoch": 0.0047221612123874365, + "ewc_loss": 3.510649548843503e-10, + "grad_norm": 6.751671314239502, + "learning_rate": 1.5373352855051243e-08, + "loss": 1.5744, + "mean_token_accuracy": 0.5922353267669678, + "num_tokens": 1125313.0, + "step": 43 + }, + { + "epoch": 0.004831978915001098, + "ewc_loss": 3.637978807091713e-10, + "grad_norm": 6.237889766693115, + "learning_rate": 1.57393850658858e-08, + "loss": 1.4454, + "mean_token_accuracy": 0.6222771406173706, + "num_tokens": 1155650.0, + "step": 44 + }, + { + "epoch": 0.00494179661761476, + "ewc_loss": 3.765308065339923e-10, + "grad_norm": 7.540345191955566, + "learning_rate": 1.610541727672035e-08, + "loss": 1.4681, + "mean_token_accuracy": 0.6189680099487305, + "num_tokens": 1178843.0, + "step": 45 + }, + { + "epoch": 0.0050516143202284204, + "ewc_loss": 3.9108272176235914e-10, + "grad_norm": 6.901407241821289, + "learning_rate": 1.6471449487554904e-08, + "loss": 1.5244, + "mean_token_accuracy": 0.6020889282226562, + "num_tokens": 1206271.0, + "step": 46 + }, + { + "epoch": 0.005161432022842082, + "ewc_loss": 4.2382453102618456e-10, + "grad_norm": 5.8072428703308105, + "learning_rate": 1.6837481698389456e-08, + "loss": 1.4707, + "mean_token_accuracy": 0.6159953474998474, + "num_tokens": 1241401.0, + "step": 47 + }, + { + "epoch": 0.005271249725455744, + "ewc_loss": 4.4565240386873484e-10, + "grad_norm": 7.14586067199707, + "learning_rate": 1.7203513909224012e-08, + "loss": 1.4963, + "mean_token_accuracy": 0.6109057664871216, + "num_tokens": 1267455.0, + "step": 48 + }, + { + "epoch": 0.005381067428069404, + "ewc_loss": 4.6202330850064754e-10, + "grad_norm": 6.748571872711182, + "learning_rate": 1.7569546120058564e-08, + "loss": 1.3768, + "mean_token_accuracy": 0.6367806792259216, + "num_tokens": 1293955.0, + "step": 49 + }, + { + "epoch": 0.005490885130683066, + "ewc_loss": 4.656612873077393e-10, + "grad_norm": 7.710287094116211, + "learning_rate": 1.7935578330893117e-08, + "loss": 1.4392, + "mean_token_accuracy": 0.629240870475769, + "num_tokens": 1316419.0, + "step": 50 + }, + { + "epoch": 0.0056007028332967276, + "ewc_loss": 4.838511813431978e-10, + "grad_norm": 7.730334758758545, + "learning_rate": 1.8301610541727673e-08, + "loss": 1.5094, + "mean_token_accuracy": 0.6069326400756836, + "num_tokens": 1338607.0, + "step": 51 + }, + { + "epoch": 0.005710520535910389, + "ewc_loss": 5.056790541857481e-10, + "grad_norm": 6.848974227905273, + "learning_rate": 1.8667642752562225e-08, + "loss": 1.4854, + "mean_token_accuracy": 0.609703779220581, + "num_tokens": 1366081.0, + "step": 52 + }, + { + "epoch": 0.00582033823852405, + "ewc_loss": 1.0113581083714962e-09, + "grad_norm": 7.352273464202881, + "learning_rate": 1.9033674963396777e-08, + "loss": 1.4823, + "mean_token_accuracy": 0.615078330039978, + "num_tokens": 1390090.0, + "step": 53 + }, + { + "epoch": 0.0059301559411377115, + "ewc_loss": 1.5934347175061703e-09, + "grad_norm": 7.363993167877197, + "learning_rate": 1.9399707174231333e-08, + "loss": 1.4922, + "mean_token_accuracy": 0.6113717555999756, + "num_tokens": 1413964.0, + "step": 54 + }, + { + "epoch": 0.006039973643751373, + "ewc_loss": 1.82626536116004e-09, + "grad_norm": 6.466087818145752, + "learning_rate": 1.9765739385065885e-08, + "loss": 1.5149, + "mean_token_accuracy": 0.6084822416305542, + "num_tokens": 1442849.0, + "step": 55 + }, + { + "epoch": 0.006149791346365034, + "ewc_loss": 2.0081643015146255e-09, + "grad_norm": 6.1061811447143555, + "learning_rate": 2.0131771595900438e-08, + "loss": 1.572, + "mean_token_accuracy": 0.589905858039856, + "num_tokens": 1478493.0, + "step": 56 + }, + { + "epoch": 0.006259609048978695, + "ewc_loss": 2.168235369026661e-09, + "grad_norm": 8.05908489227295, + "learning_rate": 2.0497803806734994e-08, + "loss": 1.5257, + "mean_token_accuracy": 0.6168287992477417, + "num_tokens": 1498811.0, + "step": 57 + }, + { + "epoch": 0.006369426751592357, + "ewc_loss": 2.2264430299401283e-09, + "grad_norm": 6.663083076477051, + "learning_rate": 2.0863836017569546e-08, + "loss": 1.4465, + "mean_token_accuracy": 0.6246469616889954, + "num_tokens": 1525932.0, + "step": 58 + }, + { + "epoch": 0.006479244454206018, + "ewc_loss": 2.2992026060819626e-09, + "grad_norm": 7.264111042022705, + "learning_rate": 2.1229868228404098e-08, + "loss": 1.491, + "mean_token_accuracy": 0.6089786887168884, + "num_tokens": 1550438.0, + "step": 59 + }, + { + "epoch": 0.006589062156819679, + "ewc_loss": 2.35741026699543e-09, + "grad_norm": 8.481172561645508, + "learning_rate": 2.1595900439238654e-08, + "loss": 1.5014, + "mean_token_accuracy": 0.6132295727729797, + "num_tokens": 1569961.0, + "step": 60 + }, + { + "epoch": 0.006698879859433341, + "ewc_loss": 2.4301698431372643e-09, + "grad_norm": 5.948199272155762, + "learning_rate": 2.1961932650073206e-08, + "loss": 1.4962, + "mean_token_accuracy": 0.6055710315704346, + "num_tokens": 1604631.0, + "step": 61 + }, + { + "epoch": 0.006808697562047002, + "ewc_loss": 2.4883775040507317e-09, + "grad_norm": 7.483115196228027, + "learning_rate": 2.232796486090776e-08, + "loss": 1.4846, + "mean_token_accuracy": 0.6134116053581238, + "num_tokens": 1628098.0, + "step": 62 + }, + { + "epoch": 0.006918515264660663, + "ewc_loss": 2.561137080192566e-09, + "grad_norm": 6.5155253410339355, + "learning_rate": 2.2693997071742314e-08, + "loss": 1.48, + "mean_token_accuracy": 0.6204023361206055, + "num_tokens": 1656281.0, + "step": 63 + }, + { + "epoch": 0.007028332967274325, + "ewc_loss": 2.561137080192566e-09, + "grad_norm": 7.267293453216553, + "learning_rate": 2.3060029282576867e-08, + "loss": 1.5127, + "mean_token_accuracy": 0.6164207458496094, + "num_tokens": 1680754.0, + "step": 64 + }, + { + "epoch": 0.007138150669887986, + "ewc_loss": 2.648448571562767e-09, + "grad_norm": 7.465707778930664, + "learning_rate": 2.342606149341142e-08, + "loss": 1.5118, + "mean_token_accuracy": 0.6088693141937256, + "num_tokens": 1704446.0, + "step": 65 + }, + { + "epoch": 0.007247968372501647, + "ewc_loss": 2.735760062932968e-09, + "grad_norm": 6.723734378814697, + "learning_rate": 2.3792093704245972e-08, + "loss": 1.2967, + "mean_token_accuracy": 0.6491153836250305, + "num_tokens": 1728711.0, + "step": 66 + }, + { + "epoch": 0.007357786075115309, + "ewc_loss": 2.8085196390748024e-09, + "grad_norm": 7.746275901794434, + "learning_rate": 2.4158125915080527e-08, + "loss": 1.421, + "mean_token_accuracy": 0.6308141946792603, + "num_tokens": 1750331.0, + "step": 67 + }, + { + "epoch": 0.0074676037777289695, + "ewc_loss": 2.8958311304450035e-09, + "grad_norm": 6.179199695587158, + "learning_rate": 2.452415812591508e-08, + "loss": 1.4363, + "mean_token_accuracy": 0.6182031631469727, + "num_tokens": 1781737.0, + "step": 68 + }, + { + "epoch": 0.007577421480342631, + "ewc_loss": 2.9831426218152046e-09, + "grad_norm": 6.383514881134033, + "learning_rate": 2.4890190336749632e-08, + "loss": 1.4811, + "mean_token_accuracy": 0.611550509929657, + "num_tokens": 1811021.0, + "step": 69 + }, + { + "epoch": 0.007687239182956293, + "ewc_loss": 3.055902197957039e-09, + "grad_norm": 7.354954242706299, + "learning_rate": 2.5256222547584188e-08, + "loss": 1.5341, + "mean_token_accuracy": 0.5974128246307373, + "num_tokens": 1835564.0, + "step": 70 + }, + { + "epoch": 0.007797056885569954, + "ewc_loss": 3.0995579436421394e-09, + "grad_norm": 10.396937370300293, + "learning_rate": 2.562225475841874e-08, + "loss": 1.5251, + "mean_token_accuracy": 0.6137760281562805, + "num_tokens": 1850951.0, + "step": 71 + }, + { + "epoch": 0.007906874588183616, + "ewc_loss": 3.1868694350123405e-09, + "grad_norm": 6.91112756729126, + "learning_rate": 2.5988286969253293e-08, + "loss": 1.537, + "mean_token_accuracy": 0.6032552719116211, + "num_tokens": 1877690.0, + "step": 72 + }, + { + "epoch": 0.008016692290797276, + "ewc_loss": 3.3760443329811096e-09, + "grad_norm": 6.820820331573486, + "learning_rate": 2.6354319180087848e-08, + "loss": 1.5772, + "mean_token_accuracy": 0.5920628309249878, + "num_tokens": 1904455.0, + "step": 73 + }, + { + "epoch": 0.008126509993410937, + "ewc_loss": 5.617039278149605e-09, + "grad_norm": 7.558678150177002, + "learning_rate": 2.67203513909224e-08, + "loss": 1.6077, + "mean_token_accuracy": 0.5876302123069763, + "num_tokens": 1928049.0, + "step": 74 + }, + { + "epoch": 0.008236327696024599, + "ewc_loss": 8.032657206058502e-09, + "grad_norm": 7.876151084899902, + "learning_rate": 2.7086383601756953e-08, + "loss": 1.4989, + "mean_token_accuracy": 0.6141358017921448, + "num_tokens": 1948422.0, + "step": 75 + }, + { + "epoch": 0.00834614539863826, + "ewc_loss": 9.778887033462524e-09, + "grad_norm": 6.777517795562744, + "learning_rate": 2.745241581259151e-08, + "loss": 1.4675, + "mean_token_accuracy": 0.6176923513412476, + "num_tokens": 1974329.0, + "step": 76 + }, + { + "epoch": 0.008455963101251922, + "ewc_loss": 1.0652001947164536e-08, + "grad_norm": 6.557216167449951, + "learning_rate": 2.781844802342606e-08, + "loss": 1.4228, + "mean_token_accuracy": 0.6314253807067871, + "num_tokens": 2000531.0, + "step": 77 + }, + { + "epoch": 0.008565780803865584, + "ewc_loss": 1.1117663234472275e-08, + "grad_norm": 6.85694694519043, + "learning_rate": 2.8184480234260614e-08, + "loss": 1.4781, + "mean_token_accuracy": 0.6129987239837646, + "num_tokens": 2023823.0, + "step": 78 + }, + { + "epoch": 0.008675598506479244, + "ewc_loss": 1.1525116860866547e-08, + "grad_norm": 7.049142360687256, + "learning_rate": 2.855051244509517e-08, + "loss": 1.5138, + "mean_token_accuracy": 0.6031516790390015, + "num_tokens": 2048734.0, + "step": 79 + }, + { + "epoch": 0.008785416209092905, + "ewc_loss": 1.1932570487260818e-08, + "grad_norm": 6.388280868530273, + "learning_rate": 2.8916544655929722e-08, + "loss": 1.4142, + "mean_token_accuracy": 0.6266201734542847, + "num_tokens": 2075741.0, + "step": 80 + }, + { + "epoch": 0.008895233911706567, + "ewc_loss": 1.210719347000122e-08, + "grad_norm": 7.067680835723877, + "learning_rate": 2.9282576866764274e-08, + "loss": 1.5166, + "mean_token_accuracy": 0.6030541062355042, + "num_tokens": 2100247.0, + "step": 81 + }, + { + "epoch": 0.009005051614320228, + "ewc_loss": 1.2223608791828156e-08, + "grad_norm": 7.391300678253174, + "learning_rate": 2.964860907759883e-08, + "loss": 1.5996, + "mean_token_accuracy": 0.6031736731529236, + "num_tokens": 2122939.0, + "step": 82 + }, + { + "epoch": 0.00911486931693389, + "ewc_loss": 1.2514647096395493e-08, + "grad_norm": 8.456125259399414, + "learning_rate": 3.001464128843338e-08, + "loss": 1.499, + "mean_token_accuracy": 0.6026207804679871, + "num_tokens": 2141424.0, + "step": 83 + }, + { + "epoch": 0.009224687019547552, + "ewc_loss": 1.2689270079135895e-08, + "grad_norm": 5.9617815017700195, + "learning_rate": 3.038067349926794e-08, + "loss": 1.3979, + "mean_token_accuracy": 0.627334713935852, + "num_tokens": 2171269.0, + "step": 84 + }, + { + "epoch": 0.009334504722161213, + "ewc_loss": 1.2747477740049362e-08, + "grad_norm": 6.476998805999756, + "learning_rate": 3.074670571010249e-08, + "loss": 1.3792, + "mean_token_accuracy": 0.6459253430366516, + "num_tokens": 2195892.0, + "step": 85 + }, + { + "epoch": 0.009444322424774873, + "ewc_loss": 1.2747477740049362e-08, + "grad_norm": 5.842731952667236, + "learning_rate": 3.111273792093704e-08, + "loss": 1.4783, + "mean_token_accuracy": 0.6203322410583496, + "num_tokens": 2227992.0, + "step": 86 + }, + { + "epoch": 0.009554140127388535, + "ewc_loss": 1.2747477740049362e-08, + "grad_norm": 6.27002477645874, + "learning_rate": 3.14787701317716e-08, + "loss": 1.5348, + "mean_token_accuracy": 0.6110748052597046, + "num_tokens": 2256995.0, + "step": 87 + }, + { + "epoch": 0.009663957830002196, + "ewc_loss": 1.2863893061876297e-08, + "grad_norm": 5.891787528991699, + "learning_rate": 3.184480234260615e-08, + "loss": 1.4483, + "mean_token_accuracy": 0.6223856806755066, + "num_tokens": 2286492.0, + "step": 88 + }, + { + "epoch": 0.009773775532615858, + "ewc_loss": 1.280568540096283e-08, + "grad_norm": 7.36253023147583, + "learning_rate": 3.22108345534407e-08, + "loss": 1.5041, + "mean_token_accuracy": 0.6073181629180908, + "num_tokens": 2309004.0, + "step": 89 + }, + { + "epoch": 0.00988359323522952, + "ewc_loss": 1.280568540096283e-08, + "grad_norm": 6.005519866943359, + "learning_rate": 3.257686676427526e-08, + "loss": 1.4033, + "mean_token_accuracy": 0.6207277774810791, + "num_tokens": 2336480.0, + "step": 90 + }, + { + "epoch": 0.009993410937843181, + "ewc_loss": 1.2980308383703232e-08, + "grad_norm": 6.306613922119141, + "learning_rate": 3.294289897510981e-08, + "loss": 1.4475, + "mean_token_accuracy": 0.615303635597229, + "num_tokens": 2363526.0, + "step": 91 + }, + { + "epoch": 0.010103228640456841, + "ewc_loss": 1.3096723705530167e-08, + "grad_norm": 5.923891067504883, + "learning_rate": 3.3308931185944364e-08, + "loss": 1.3985, + "mean_token_accuracy": 0.6279661655426025, + "num_tokens": 2393847.0, + "step": 92 + }, + { + "epoch": 0.010213046343070502, + "ewc_loss": 1.3329554349184036e-08, + "grad_norm": 6.266355037689209, + "learning_rate": 3.367496339677891e-08, + "loss": 1.4983, + "mean_token_accuracy": 0.6066314578056335, + "num_tokens": 2421639.0, + "step": 93 + }, + { + "epoch": 0.010322864045684164, + "ewc_loss": 1.367880031466484e-08, + "grad_norm": 6.652030944824219, + "learning_rate": 3.404099560761347e-08, + "loss": 1.4514, + "mean_token_accuracy": 0.6239076852798462, + "num_tokens": 2445896.0, + "step": 94 + }, + { + "epoch": 0.010432681748297826, + "ewc_loss": 1.3795215636491776e-08, + "grad_norm": 6.22355318069458, + "learning_rate": 3.4407027818448024e-08, + "loss": 1.5577, + "mean_token_accuracy": 0.6006054282188416, + "num_tokens": 2475814.0, + "step": 95 + }, + { + "epoch": 0.010542499450911487, + "ewc_loss": 1.4028046280145645e-08, + "grad_norm": 6.803187847137451, + "learning_rate": 3.477306002928257e-08, + "loss": 1.3838, + "mean_token_accuracy": 0.6400576829910278, + "num_tokens": 2499512.0, + "step": 96 + }, + { + "epoch": 0.010652317153525149, + "ewc_loss": 1.4668330550193787e-08, + "grad_norm": 6.226831436157227, + "learning_rate": 3.513909224011713e-08, + "loss": 1.5603, + "mean_token_accuracy": 0.5937842130661011, + "num_tokens": 2528412.0, + "step": 97 + }, + { + "epoch": 0.010762134856138809, + "ewc_loss": 1.4901161193847656e-08, + "grad_norm": 6.771093845367432, + "learning_rate": 3.5505124450951685e-08, + "loss": 1.5048, + "mean_token_accuracy": 0.61011803150177, + "num_tokens": 2553855.0, + "step": 98 + }, + { + "epoch": 0.01087195255875247, + "ewc_loss": 1.501757651567459e-08, + "grad_norm": 7.714377403259277, + "learning_rate": 3.5871156661786234e-08, + "loss": 1.5375, + "mean_token_accuracy": 0.6077075004577637, + "num_tokens": 2574996.0, + "step": 99 + }, + { + "epoch": 0.010981770261366132, + "ewc_loss": 1.548323780298233e-08, + "grad_norm": 6.412229061126709, + "learning_rate": 3.623718887262079e-08, + "loss": 1.4908, + "mean_token_accuracy": 0.6122196912765503, + "num_tokens": 2601684.0, + "step": 100 + }, + { + "epoch": 0.011091587963979794, + "ewc_loss": 1.594889909029007e-08, + "grad_norm": 6.630273818969727, + "learning_rate": 3.6603221083455345e-08, + "loss": 1.538, + "mean_token_accuracy": 0.6010323166847229, + "num_tokens": 2628612.0, + "step": 101 + }, + { + "epoch": 0.011201405666593455, + "ewc_loss": 1.8510036170482635e-08, + "grad_norm": 6.502976417541504, + "learning_rate": 3.6969253294289894e-08, + "loss": 1.437, + "mean_token_accuracy": 0.6383704543113708, + "num_tokens": 2652776.0, + "step": 102 + }, + { + "epoch": 0.011311223369207117, + "ewc_loss": 2.2817403078079224e-08, + "grad_norm": 6.54626989364624, + "learning_rate": 3.733528550512445e-08, + "loss": 1.4049, + "mean_token_accuracy": 0.6316857933998108, + "num_tokens": 2677807.0, + "step": 103 + }, + { + "epoch": 0.011421041071820778, + "ewc_loss": 2.8405338525772095e-08, + "grad_norm": 7.894402503967285, + "learning_rate": 3.7701317715959006e-08, + "loss": 1.5519, + "mean_token_accuracy": 0.6000187397003174, + "num_tokens": 2697989.0, + "step": 104 + }, + { + "epoch": 0.011530858774434438, + "ewc_loss": 3.282912075519562e-08, + "grad_norm": 6.640040397644043, + "learning_rate": 3.8067349926793555e-08, + "loss": 1.539, + "mean_token_accuracy": 0.6022933125495911, + "num_tokens": 2722341.0, + "step": 105 + }, + { + "epoch": 0.0116406764770481, + "ewc_loss": 3.608874976634979e-08, + "grad_norm": 5.948328018188477, + "learning_rate": 3.843338213762811e-08, + "loss": 1.5174, + "mean_token_accuracy": 0.6029834151268005, + "num_tokens": 2753675.0, + "step": 106 + }, + { + "epoch": 0.011750494179661761, + "ewc_loss": 3.725290298461914e-08, + "grad_norm": 6.409243106842041, + "learning_rate": 3.8799414348462666e-08, + "loss": 1.4644, + "mean_token_accuracy": 0.6136347055435181, + "num_tokens": 2778974.0, + "step": 107 + }, + { + "epoch": 0.011860311882275423, + "ewc_loss": 3.795139491558075e-08, + "grad_norm": 5.960975170135498, + "learning_rate": 3.9165446559297215e-08, + "loss": 1.5168, + "mean_token_accuracy": 0.6036599278450012, + "num_tokens": 2808967.0, + "step": 108 + }, + { + "epoch": 0.011970129584889085, + "ewc_loss": 3.771856427192688e-08, + "grad_norm": 6.255977153778076, + "learning_rate": 3.953147877013177e-08, + "loss": 1.5109, + "mean_token_accuracy": 0.6004844307899475, + "num_tokens": 2835393.0, + "step": 109 + }, + { + "epoch": 0.012079947287502746, + "ewc_loss": 3.864988684654236e-08, + "grad_norm": 5.722349166870117, + "learning_rate": 3.9897510980966327e-08, + "loss": 1.4718, + "mean_token_accuracy": 0.6122874021530151, + "num_tokens": 2863678.0, + "step": 110 + }, + { + "epoch": 0.012189764990116406, + "ewc_loss": 3.818422555923462e-08, + "grad_norm": 6.055007457733154, + "learning_rate": 4.0263543191800876e-08, + "loss": 1.444, + "mean_token_accuracy": 0.6229895949363708, + "num_tokens": 2890376.0, + "step": 111 + }, + { + "epoch": 0.012299582692730068, + "ewc_loss": 3.795139491558075e-08, + "grad_norm": 5.957438945770264, + "learning_rate": 4.062957540263543e-08, + "loss": 1.441, + "mean_token_accuracy": 0.6229817867279053, + "num_tokens": 2917128.0, + "step": 112 + }, + { + "epoch": 0.01240940039534373, + "ewc_loss": 3.725290298461914e-08, + "grad_norm": 6.434499740600586, + "learning_rate": 4.099560761346999e-08, + "loss": 1.4754, + "mean_token_accuracy": 0.6101410388946533, + "num_tokens": 2940812.0, + "step": 113 + }, + { + "epoch": 0.01251921809795739, + "ewc_loss": 3.771856427192688e-08, + "grad_norm": 5.820069313049316, + "learning_rate": 4.1361639824304536e-08, + "loss": 1.502, + "mean_token_accuracy": 0.6022881269454956, + "num_tokens": 2969881.0, + "step": 114 + }, + { + "epoch": 0.012629035800571052, + "ewc_loss": 3.818422555923462e-08, + "grad_norm": 6.2141618728637695, + "learning_rate": 4.172767203513909e-08, + "loss": 1.5014, + "mean_token_accuracy": 0.6083742380142212, + "num_tokens": 2996049.0, + "step": 115 + }, + { + "epoch": 0.012738853503184714, + "ewc_loss": 3.864988684654236e-08, + "grad_norm": 5.346731185913086, + "learning_rate": 4.209370424597365e-08, + "loss": 1.3711, + "mean_token_accuracy": 0.6322212219238281, + "num_tokens": 3025829.0, + "step": 116 + }, + { + "epoch": 0.012848671205798374, + "ewc_loss": 3.841705620288849e-08, + "grad_norm": 6.303293704986572, + "learning_rate": 4.2459736456808197e-08, + "loss": 1.4889, + "mean_token_accuracy": 0.6107179522514343, + "num_tokens": 3051012.0, + "step": 117 + }, + { + "epoch": 0.012958488908412035, + "ewc_loss": 3.958120942115784e-08, + "grad_norm": 6.083418846130371, + "learning_rate": 4.282576866764275e-08, + "loss": 1.5296, + "mean_token_accuracy": 0.6053721904754639, + "num_tokens": 3077750.0, + "step": 118 + }, + { + "epoch": 0.013068306611025697, + "ewc_loss": 3.934837877750397e-08, + "grad_norm": 6.154629707336426, + "learning_rate": 4.319180087847731e-08, + "loss": 1.4789, + "mean_token_accuracy": 0.6091103553771973, + "num_tokens": 3103271.0, + "step": 119 + }, + { + "epoch": 0.013178124313639359, + "ewc_loss": 4.0046870708465576e-08, + "grad_norm": 6.875816345214844, + "learning_rate": 4.355783308931186e-08, + "loss": 1.5063, + "mean_token_accuracy": 0.6012482643127441, + "num_tokens": 3124511.0, + "step": 120 + }, + { + "epoch": 0.01328794201625302, + "ewc_loss": 4.0745362639427185e-08, + "grad_norm": 6.015988826751709, + "learning_rate": 4.392386530014641e-08, + "loss": 1.4026, + "mean_token_accuracy": 0.6361804604530334, + "num_tokens": 3150451.0, + "step": 121 + }, + { + "epoch": 0.013397759718866682, + "ewc_loss": 4.1211023926734924e-08, + "grad_norm": 5.768141746520996, + "learning_rate": 4.428989751098097e-08, + "loss": 1.5195, + "mean_token_accuracy": 0.603537380695343, + "num_tokens": 3180151.0, + "step": 122 + }, + { + "epoch": 0.013507577421480343, + "ewc_loss": 4.1443854570388794e-08, + "grad_norm": 5.973391056060791, + "learning_rate": 4.465592972181552e-08, + "loss": 1.5133, + "mean_token_accuracy": 0.605829119682312, + "num_tokens": 3207702.0, + "step": 123 + }, + { + "epoch": 0.013617395124094003, + "ewc_loss": 4.1676685214042664e-08, + "grad_norm": 7.483236312866211, + "learning_rate": 4.502196193265007e-08, + "loss": 1.3939, + "mean_token_accuracy": 0.6375589370727539, + "num_tokens": 3226168.0, + "step": 124 + }, + { + "epoch": 0.013727212826707665, + "ewc_loss": 4.284083843231201e-08, + "grad_norm": 5.790709972381592, + "learning_rate": 4.538799414348463e-08, + "loss": 1.4773, + "mean_token_accuracy": 0.6082464456558228, + "num_tokens": 3255714.0, + "step": 125 + }, + { + "epoch": 0.013837030529321327, + "ewc_loss": 4.377216100692749e-08, + "grad_norm": 5.905766487121582, + "learning_rate": 4.575402635431918e-08, + "loss": 1.5153, + "mean_token_accuracy": 0.608139157295227, + "num_tokens": 3283574.0, + "step": 126 + }, + { + "epoch": 0.013946848231934988, + "ewc_loss": 4.516914486885071e-08, + "grad_norm": 6.089626789093018, + "learning_rate": 4.6120058565153734e-08, + "loss": 1.4589, + "mean_token_accuracy": 0.6190888285636902, + "num_tokens": 3309128.0, + "step": 127 + }, + { + "epoch": 0.01405666593454865, + "ewc_loss": 4.7031790018081665e-08, + "grad_norm": 6.837204456329346, + "learning_rate": 4.648609077598828e-08, + "loss": 1.5121, + "mean_token_accuracy": 0.6059969067573547, + "num_tokens": 3331577.0, + "step": 128 + }, + { + "epoch": 0.014166483637162311, + "ewc_loss": 4.889443516731262e-08, + "grad_norm": 6.092732906341553, + "learning_rate": 4.685212298682284e-08, + "loss": 1.5673, + "mean_token_accuracy": 0.5897543430328369, + "num_tokens": 3358884.0, + "step": 129 + }, + { + "epoch": 0.014276301339775971, + "ewc_loss": 4.959292709827423e-08, + "grad_norm": 7.663801193237305, + "learning_rate": 4.7218155197657394e-08, + "loss": 1.5928, + "mean_token_accuracy": 0.5891657471656799, + "num_tokens": 3378122.0, + "step": 130 + }, + { + "epoch": 0.014386119042389633, + "ewc_loss": 5.122274160385132e-08, + "grad_norm": 6.388889789581299, + "learning_rate": 4.7584187408491943e-08, + "loss": 1.547, + "mean_token_accuracy": 0.6023693084716797, + "num_tokens": 3404642.0, + "step": 131 + }, + { + "epoch": 0.014495936745003294, + "ewc_loss": 5.2619725465774536e-08, + "grad_norm": 6.257548809051514, + "learning_rate": 4.79502196193265e-08, + "loss": 1.4407, + "mean_token_accuracy": 0.6163528561592102, + "num_tokens": 3429186.0, + "step": 132 + }, + { + "epoch": 0.014605754447616956, + "ewc_loss": 5.4249539971351624e-08, + "grad_norm": 5.563029766082764, + "learning_rate": 4.8316251830161055e-08, + "loss": 1.4771, + "mean_token_accuracy": 0.6184221506118774, + "num_tokens": 3460810.0, + "step": 133 + }, + { + "epoch": 0.014715572150230618, + "ewc_loss": 5.611218512058258e-08, + "grad_norm": 5.5722856521606445, + "learning_rate": 4.8682284040995604e-08, + "loss": 1.4436, + "mean_token_accuracy": 0.6179695129394531, + "num_tokens": 3491285.0, + "step": 134 + }, + { + "epoch": 0.01482538985284428, + "ewc_loss": 5.704350769519806e-08, + "grad_norm": 5.94484806060791, + "learning_rate": 4.904831625183016e-08, + "loss": 1.4381, + "mean_token_accuracy": 0.6206265687942505, + "num_tokens": 3517893.0, + "step": 135 + }, + { + "epoch": 0.014935207555457939, + "ewc_loss": 5.9371814131736755e-08, + "grad_norm": 6.0293097496032715, + "learning_rate": 4.9414348462664715e-08, + "loss": 1.4892, + "mean_token_accuracy": 0.6066820621490479, + "num_tokens": 3545614.0, + "step": 136 + }, + { + "epoch": 0.0150450252580716, + "ewc_loss": 6.05359673500061e-08, + "grad_norm": 6.420803070068359, + "learning_rate": 4.9780380673499264e-08, + "loss": 1.5323, + "mean_token_accuracy": 0.5983564853668213, + "num_tokens": 3571156.0, + "step": 137 + }, + { + "epoch": 0.015154842960685262, + "ewc_loss": 6.193295121192932e-08, + "grad_norm": 6.359198570251465, + "learning_rate": 5.014641288433382e-08, + "loss": 1.4279, + "mean_token_accuracy": 0.6233689785003662, + "num_tokens": 3595596.0, + "step": 138 + }, + { + "epoch": 0.015264660663298924, + "ewc_loss": 6.426125764846802e-08, + "grad_norm": 5.6222310066223145, + "learning_rate": 5.0512445095168376e-08, + "loss": 1.4406, + "mean_token_accuracy": 0.6330134868621826, + "num_tokens": 3625562.0, + "step": 139 + }, + { + "epoch": 0.015374478365912585, + "ewc_loss": 6.705522537231445e-08, + "grad_norm": 6.554638385772705, + "learning_rate": 5.0878477306002925e-08, + "loss": 1.4939, + "mean_token_accuracy": 0.6201046705245972, + "num_tokens": 3651717.0, + "step": 140 + }, + { + "epoch": 0.015484296068526247, + "ewc_loss": 7.12461769580841e-08, + "grad_norm": 6.620650768280029, + "learning_rate": 5.124450951683748e-08, + "loss": 1.3762, + "mean_token_accuracy": 0.6366029977798462, + "num_tokens": 3674773.0, + "step": 141 + }, + { + "epoch": 0.015594113771139909, + "ewc_loss": 7.729977369308472e-08, + "grad_norm": 6.450253009796143, + "learning_rate": 5.1610541727672036e-08, + "loss": 1.56, + "mean_token_accuracy": 0.5893372893333435, + "num_tokens": 3701717.0, + "step": 142 + }, + { + "epoch": 0.01570393147375357, + "ewc_loss": 8.195638656616211e-08, + "grad_norm": 5.646212100982666, + "learning_rate": 5.1976573938506585e-08, + "loss": 1.4068, + "mean_token_accuracy": 0.6266642808914185, + "num_tokens": 3729901.0, + "step": 143 + }, + { + "epoch": 0.015813749176367232, + "ewc_loss": 8.800998330116272e-08, + "grad_norm": 5.647825717926025, + "learning_rate": 5.234260614934114e-08, + "loss": 1.5065, + "mean_token_accuracy": 0.595840334892273, + "num_tokens": 3757405.0, + "step": 144 + }, + { + "epoch": 0.01592356687898089, + "ewc_loss": 9.452924132347107e-08, + "grad_norm": 5.420976638793945, + "learning_rate": 5.2708638360175697e-08, + "loss": 1.3692, + "mean_token_accuracy": 0.6370492577552795, + "num_tokens": 3787289.0, + "step": 145 + }, + { + "epoch": 0.01603338458159455, + "ewc_loss": 9.96515154838562e-08, + "grad_norm": 5.748161792755127, + "learning_rate": 5.3074670571010246e-08, + "loss": 1.4043, + "mean_token_accuracy": 0.6332556009292603, + "num_tokens": 3811919.0, + "step": 146 + }, + { + "epoch": 0.016143202284208215, + "ewc_loss": 1.0337680578231812e-07, + "grad_norm": 5.827054977416992, + "learning_rate": 5.34407027818448e-08, + "loss": 1.3464, + "mean_token_accuracy": 0.6397305727005005, + "num_tokens": 3835967.0, + "step": 147 + }, + { + "epoch": 0.016253019986821875, + "ewc_loss": 1.0617077350616455e-07, + "grad_norm": 5.895974159240723, + "learning_rate": 5.380673499267936e-08, + "loss": 1.3921, + "mean_token_accuracy": 0.6276445984840393, + "num_tokens": 3859268.0, + "step": 148 + }, + { + "epoch": 0.016362837689435538, + "ewc_loss": 1.0896474123001099e-07, + "grad_norm": 4.863615989685059, + "learning_rate": 5.4172767203513906e-08, + "loss": 1.4429, + "mean_token_accuracy": 0.6140576601028442, + "num_tokens": 3892417.0, + "step": 149 + }, + { + "epoch": 0.016472655392049198, + "ewc_loss": 1.1129304766654968e-07, + "grad_norm": 6.407237529754639, + "learning_rate": 5.453879941434846e-08, + "loss": 1.3682, + "mean_token_accuracy": 0.6370266675949097, + "num_tokens": 3913323.0, + "step": 150 + }, + { + "epoch": 0.01658247309466286, + "ewc_loss": 1.1315569281578064e-07, + "grad_norm": 4.69478702545166, + "learning_rate": 5.490483162518302e-08, + "loss": 1.2638, + "mean_token_accuracy": 0.6565554738044739, + "num_tokens": 3942390.0, + "step": 151 + }, + { + "epoch": 0.01669229079727652, + "ewc_loss": 1.1455267667770386e-07, + "grad_norm": 5.488808631896973, + "learning_rate": 5.527086383601757e-08, + "loss": 1.4484, + "mean_token_accuracy": 0.6132785081863403, + "num_tokens": 3968060.0, + "step": 152 + }, + { + "epoch": 0.01680210849989018, + "ewc_loss": 1.1641532182693481e-07, + "grad_norm": 4.87542200088501, + "learning_rate": 5.563689604685212e-08, + "loss": 1.3261, + "mean_token_accuracy": 0.6453108787536621, + "num_tokens": 3996149.0, + "step": 153 + }, + { + "epoch": 0.016911926202503844, + "ewc_loss": 1.1734664440155029e-07, + "grad_norm": 5.14293098449707, + "learning_rate": 5.600292825768668e-08, + "loss": 1.4605, + "mean_token_accuracy": 0.608967661857605, + "num_tokens": 4022708.0, + "step": 154 + }, + { + "epoch": 0.017021743905117504, + "ewc_loss": 1.1781230568885803e-07, + "grad_norm": 5.467360019683838, + "learning_rate": 5.636896046852123e-08, + "loss": 1.4953, + "mean_token_accuracy": 0.6105849146842957, + "num_tokens": 4048984.0, + "step": 155 + }, + { + "epoch": 0.017131561607731167, + "ewc_loss": 1.2014061212539673e-07, + "grad_norm": 4.996084213256836, + "learning_rate": 5.673499267935578e-08, + "loss": 1.3827, + "mean_token_accuracy": 0.6298693418502808, + "num_tokens": 4075654.0, + "step": 156 + }, + { + "epoch": 0.017241379310344827, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 4.9786763191223145, + "learning_rate": 5.710102489019034e-08, + "loss": 1.5389, + "mean_token_accuracy": 0.5869988203048706, + "num_tokens": 4106959.0, + "step": 157 + }, + { + "epoch": 0.017351197012958487, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 4.945926189422607, + "learning_rate": 5.746705710102489e-08, + "loss": 1.4279, + "mean_token_accuracy": 0.6243952512741089, + "num_tokens": 4135143.0, + "step": 158 + }, + { + "epoch": 0.01746101471557215, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 4.849398136138916, + "learning_rate": 5.7833089311859443e-08, + "loss": 1.37, + "mean_token_accuracy": 0.636972188949585, + "num_tokens": 4162641.0, + "step": 159 + }, + { + "epoch": 0.01757083241818581, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 4.887083530426025, + "learning_rate": 5.8199121522694e-08, + "loss": 1.3279, + "mean_token_accuracy": 0.6353504061698914, + "num_tokens": 4189265.0, + "step": 160 + }, + { + "epoch": 0.017680650120799474, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 5.735540866851807, + "learning_rate": 5.856515373352855e-08, + "loss": 1.4954, + "mean_token_accuracy": 0.6059099435806274, + "num_tokens": 4209765.0, + "step": 161 + }, + { + "epoch": 0.017790467823413134, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 5.1875739097595215, + "learning_rate": 5.8931185944363104e-08, + "loss": 1.3901, + "mean_token_accuracy": 0.6252871751785278, + "num_tokens": 4232444.0, + "step": 162 + }, + { + "epoch": 0.017900285526026797, + "ewc_loss": 1.210719347000122e-07, + "grad_norm": 4.4461798667907715, + "learning_rate": 5.929721815519766e-08, + "loss": 1.394, + "mean_token_accuracy": 0.6196003556251526, + "num_tokens": 4263314.0, + "step": 163 + }, + { + "epoch": 0.018010103228640457, + "ewc_loss": 1.2293457984924316e-07, + "grad_norm": 4.615938663482666, + "learning_rate": 5.966325036603222e-08, + "loss": 1.402, + "mean_token_accuracy": 0.6264450550079346, + "num_tokens": 4292268.0, + "step": 164 + }, + { + "epoch": 0.018119920931254117, + "ewc_loss": 1.2293457984924316e-07, + "grad_norm": 4.951958179473877, + "learning_rate": 6.002928257686676e-08, + "loss": 1.4687, + "mean_token_accuracy": 0.6159982681274414, + "num_tokens": 4317038.0, + "step": 165 + }, + { + "epoch": 0.01822973863386778, + "ewc_loss": 1.2386590242385864e-07, + "grad_norm": 4.770590305328369, + "learning_rate": 6.039531478770131e-08, + "loss": 1.4729, + "mean_token_accuracy": 0.6051391959190369, + "num_tokens": 4344439.0, + "step": 166 + }, + { + "epoch": 0.01833955633648144, + "ewc_loss": 1.2386590242385864e-07, + "grad_norm": 4.891670227050781, + "learning_rate": 6.076134699853588e-08, + "loss": 1.3808, + "mean_token_accuracy": 0.6283180713653564, + "num_tokens": 4371828.0, + "step": 167 + }, + { + "epoch": 0.018449374039095103, + "ewc_loss": 1.2386590242385864e-07, + "grad_norm": 5.1413068771362305, + "learning_rate": 6.112737920937042e-08, + "loss": 1.3006, + "mean_token_accuracy": 0.6415995359420776, + "num_tokens": 4393611.0, + "step": 168 + }, + { + "epoch": 0.018559191741708763, + "ewc_loss": 1.2479722499847412e-07, + "grad_norm": 4.745772838592529, + "learning_rate": 6.149341142020497e-08, + "loss": 1.3806, + "mean_token_accuracy": 0.6269069910049438, + "num_tokens": 4419458.0, + "step": 169 + }, + { + "epoch": 0.018669009444322426, + "ewc_loss": 1.2479722499847412e-07, + "grad_norm": 4.750580310821533, + "learning_rate": 6.185944363103954e-08, + "loss": 1.334, + "mean_token_accuracy": 0.6416382789611816, + "num_tokens": 4444110.0, + "step": 170 + }, + { + "epoch": 0.018778827146936086, + "ewc_loss": 1.257285475730896e-07, + "grad_norm": 5.382021903991699, + "learning_rate": 6.222547584187409e-08, + "loss": 1.3704, + "mean_token_accuracy": 0.6271952390670776, + "num_tokens": 4465140.0, + "step": 171 + }, + { + "epoch": 0.018888644849549746, + "ewc_loss": 1.2759119272232056e-07, + "grad_norm": 4.862536907196045, + "learning_rate": 6.259150805270863e-08, + "loss": 1.3918, + "mean_token_accuracy": 0.6224688291549683, + "num_tokens": 4489605.0, + "step": 172 + }, + { + "epoch": 0.01899846255216341, + "ewc_loss": 1.2945383787155151e-07, + "grad_norm": 4.624092102050781, + "learning_rate": 6.29575402635432e-08, + "loss": 1.3394, + "mean_token_accuracy": 0.6347394585609436, + "num_tokens": 4515095.0, + "step": 173 + }, + { + "epoch": 0.01910828025477707, + "ewc_loss": 1.2945383787155151e-07, + "grad_norm": 4.523379325866699, + "learning_rate": 6.332357247437773e-08, + "loss": 1.374, + "mean_token_accuracy": 0.624242901802063, + "num_tokens": 4541160.0, + "step": 174 + }, + { + "epoch": 0.019218097957390733, + "ewc_loss": 1.2945383787155151e-07, + "grad_norm": 5.112313747406006, + "learning_rate": 6.36896046852123e-08, + "loss": 1.3612, + "mean_token_accuracy": 0.6266942024230957, + "num_tokens": 4563441.0, + "step": 175 + }, + { + "epoch": 0.019327915660004392, + "ewc_loss": 1.3317912817001343e-07, + "grad_norm": 4.744810581207275, + "learning_rate": 6.405563689604684e-08, + "loss": 1.3098, + "mean_token_accuracy": 0.639885425567627, + "num_tokens": 4586881.0, + "step": 176 + }, + { + "epoch": 0.019437733362618052, + "ewc_loss": 1.341104507446289e-07, + "grad_norm": 4.465086936950684, + "learning_rate": 6.44216691068814e-08, + "loss": 1.3928, + "mean_token_accuracy": 0.6229970455169678, + "num_tokens": 4616811.0, + "step": 177 + }, + { + "epoch": 0.019547551065231716, + "ewc_loss": 1.3597309589385986e-07, + "grad_norm": 4.6948699951171875, + "learning_rate": 6.478770131771596e-08, + "loss": 1.2347, + "mean_token_accuracy": 0.6595771312713623, + "num_tokens": 4639573.0, + "step": 178 + }, + { + "epoch": 0.019657368767845376, + "ewc_loss": 1.3690441846847534e-07, + "grad_norm": 4.684775352478027, + "learning_rate": 6.515373352855052e-08, + "loss": 1.4218, + "mean_token_accuracy": 0.6183117628097534, + "num_tokens": 4666392.0, + "step": 179 + }, + { + "epoch": 0.01976718647045904, + "ewc_loss": 1.3969838619232178e-07, + "grad_norm": 4.844146728515625, + "learning_rate": 6.551976573938505e-08, + "loss": 1.3207, + "mean_token_accuracy": 0.6365014314651489, + "num_tokens": 4690288.0, + "step": 180 + }, + { + "epoch": 0.0198770041730727, + "ewc_loss": 1.4062970876693726e-07, + "grad_norm": 4.5302205085754395, + "learning_rate": 6.588579795021962e-08, + "loss": 1.36, + "mean_token_accuracy": 0.6240625381469727, + "num_tokens": 4716583.0, + "step": 181 + }, + { + "epoch": 0.019986821875686362, + "ewc_loss": 1.3969838619232178e-07, + "grad_norm": 4.484119892120361, + "learning_rate": 6.625183016105416e-08, + "loss": 1.291, + "mean_token_accuracy": 0.6487938761711121, + "num_tokens": 4741315.0, + "step": 182 + }, + { + "epoch": 0.020096639578300022, + "ewc_loss": 1.424923539161682e-07, + "grad_norm": 4.2660722732543945, + "learning_rate": 6.661786237188873e-08, + "loss": 1.4536, + "mean_token_accuracy": 0.613031268119812, + "num_tokens": 4771862.0, + "step": 183 + }, + { + "epoch": 0.020206457280913682, + "ewc_loss": 1.4621764421463013e-07, + "grad_norm": 4.781135082244873, + "learning_rate": 6.698389458272328e-08, + "loss": 1.4028, + "mean_token_accuracy": 0.6224547624588013, + "num_tokens": 4798609.0, + "step": 184 + }, + { + "epoch": 0.020316274983527345, + "ewc_loss": 1.471489667892456e-07, + "grad_norm": 4.905125141143799, + "learning_rate": 6.734992679355783e-08, + "loss": 1.3444, + "mean_token_accuracy": 0.6295778155326843, + "num_tokens": 4821516.0, + "step": 185 + }, + { + "epoch": 0.020426092686141005, + "ewc_loss": 1.471489667892456e-07, + "grad_norm": 4.619053363800049, + "learning_rate": 6.771595900439237e-08, + "loss": 1.3248, + "mean_token_accuracy": 0.6414509415626526, + "num_tokens": 4845487.0, + "step": 186 + }, + { + "epoch": 0.02053591038875467, + "ewc_loss": 1.4808028936386108e-07, + "grad_norm": 4.877774715423584, + "learning_rate": 6.808199121522694e-08, + "loss": 1.3916, + "mean_token_accuracy": 0.6294417381286621, + "num_tokens": 4868981.0, + "step": 187 + }, + { + "epoch": 0.020645728091368328, + "ewc_loss": 1.51805579662323e-07, + "grad_norm": 4.285820007324219, + "learning_rate": 6.844802342606149e-08, + "loss": 1.3648, + "mean_token_accuracy": 0.624253511428833, + "num_tokens": 4895095.0, + "step": 188 + }, + { + "epoch": 0.02075554579398199, + "ewc_loss": 1.5273690223693848e-07, + "grad_norm": 4.480078220367432, + "learning_rate": 6.881405563689605e-08, + "loss": 1.4291, + "mean_token_accuracy": 0.6134353876113892, + "num_tokens": 4920215.0, + "step": 189 + }, + { + "epoch": 0.02086536349659565, + "ewc_loss": 1.5459954738616943e-07, + "grad_norm": 4.178596019744873, + "learning_rate": 6.91800878477306e-08, + "loss": 1.3935, + "mean_token_accuracy": 0.6193321943283081, + "num_tokens": 4950113.0, + "step": 190 + }, + { + "epoch": 0.02097518119920931, + "ewc_loss": 1.7881393432617188e-07, + "grad_norm": 16.944215774536133, + "learning_rate": 6.954612005856515e-08, + "loss": 1.3925, + "mean_token_accuracy": 0.6195124983787537, + "num_tokens": 4970901.0, + "step": 191 + }, + { + "epoch": 0.021084998901822975, + "ewc_loss": 1.5832483768463135e-07, + "grad_norm": 4.6001434326171875, + "learning_rate": 6.99121522693997e-08, + "loss": 1.3313, + "mean_token_accuracy": 0.6374809145927429, + "num_tokens": 4996117.0, + "step": 192 + }, + { + "epoch": 0.021194816604436634, + "ewc_loss": 1.601874828338623e-07, + "grad_norm": 4.611410617828369, + "learning_rate": 7.027818448023426e-08, + "loss": 1.3665, + "mean_token_accuracy": 0.6259975433349609, + "num_tokens": 5020994.0, + "step": 193 + }, + { + "epoch": 0.021304634307050298, + "ewc_loss": 1.6205012798309326e-07, + "grad_norm": 4.07274866104126, + "learning_rate": 7.064421669106881e-08, + "loss": 1.4239, + "mean_token_accuracy": 0.6179133653640747, + "num_tokens": 5053438.0, + "step": 194 + }, + { + "epoch": 0.021414452009663958, + "ewc_loss": 1.6763806343078613e-07, + "grad_norm": 4.678153038024902, + "learning_rate": 7.101024890190337e-08, + "loss": 1.4073, + "mean_token_accuracy": 0.62064528465271, + "num_tokens": 5077854.0, + "step": 195 + }, + { + "epoch": 0.021524269712277617, + "ewc_loss": 1.685693860054016e-07, + "grad_norm": 4.340275764465332, + "learning_rate": 7.137628111273792e-08, + "loss": 1.4034, + "mean_token_accuracy": 0.6132748126983643, + "num_tokens": 5103447.0, + "step": 196 + }, + { + "epoch": 0.02163408741489128, + "ewc_loss": 1.7136335372924805e-07, + "grad_norm": 4.51183557510376, + "learning_rate": 7.174231332357247e-08, + "loss": 1.4099, + "mean_token_accuracy": 0.6146407723426819, + "num_tokens": 5129474.0, + "step": 197 + }, + { + "epoch": 0.02174390511750494, + "ewc_loss": 1.7229467630386353e-07, + "grad_norm": 5.239439010620117, + "learning_rate": 7.210834553440702e-08, + "loss": 1.3758, + "mean_token_accuracy": 0.6236597895622253, + "num_tokens": 5154331.0, + "step": 198 + }, + { + "epoch": 0.021853722820118604, + "ewc_loss": 1.7601996660232544e-07, + "grad_norm": 3.978452682495117, + "learning_rate": 7.247437774524158e-08, + "loss": 1.3702, + "mean_token_accuracy": 0.629802405834198, + "num_tokens": 5182997.0, + "step": 199 + }, + { + "epoch": 0.021963540522732264, + "ewc_loss": 1.7974525690078735e-07, + "grad_norm": 4.093644618988037, + "learning_rate": 7.284040995607613e-08, + "loss": 1.3071, + "mean_token_accuracy": 0.6624548435211182, + "num_tokens": 5209941.0, + "step": 200 + }, + { + "epoch": 0.022073358225345927, + "ewc_loss": 1.816079020500183e-07, + "grad_norm": 4.552216529846191, + "learning_rate": 7.320644216691069e-08, + "loss": 1.2467, + "mean_token_accuracy": 0.6572819948196411, + "num_tokens": 5235749.0, + "step": 201 + }, + { + "epoch": 0.022183175927959587, + "ewc_loss": 1.8533319234848022e-07, + "grad_norm": 4.413077354431152, + "learning_rate": 7.357247437774524e-08, + "loss": 1.3214, + "mean_token_accuracy": 0.6420587301254272, + "num_tokens": 5259439.0, + "step": 202 + }, + { + "epoch": 0.022292993630573247, + "ewc_loss": 1.8719583749771118e-07, + "grad_norm": 4.586743354797363, + "learning_rate": 7.393850658857979e-08, + "loss": 1.3338, + "mean_token_accuracy": 0.6310532689094543, + "num_tokens": 5283571.0, + "step": 203 + }, + { + "epoch": 0.02240281133318691, + "ewc_loss": 1.8719583749771118e-07, + "grad_norm": 4.774307727813721, + "learning_rate": 7.430453879941434e-08, + "loss": 1.2653, + "mean_token_accuracy": 0.6515313386917114, + "num_tokens": 5306606.0, + "step": 204 + }, + { + "epoch": 0.02251262903580057, + "ewc_loss": 1.909211277961731e-07, + "grad_norm": 4.624813079833984, + "learning_rate": 7.46705710102489e-08, + "loss": 1.3635, + "mean_token_accuracy": 0.6404963731765747, + "num_tokens": 5333475.0, + "step": 205 + }, + { + "epoch": 0.022622446738414233, + "ewc_loss": 1.8905848264694214e-07, + "grad_norm": 4.037083148956299, + "learning_rate": 7.503660322108345e-08, + "loss": 1.4113, + "mean_token_accuracy": 0.6147767901420593, + "num_tokens": 5360729.0, + "step": 206 + }, + { + "epoch": 0.022732264441027893, + "ewc_loss": 1.94646418094635e-07, + "grad_norm": 4.762199878692627, + "learning_rate": 7.540263543191801e-08, + "loss": 1.3002, + "mean_token_accuracy": 0.6477734446525574, + "num_tokens": 5384083.0, + "step": 207 + }, + { + "epoch": 0.022842082143641557, + "ewc_loss": 1.993030309677124e-07, + "grad_norm": 4.362670421600342, + "learning_rate": 7.576866764275256e-08, + "loss": 1.3192, + "mean_token_accuracy": 0.6354765892028809, + "num_tokens": 5407198.0, + "step": 208 + }, + { + "epoch": 0.022951899846255217, + "ewc_loss": 2.0489096641540527e-07, + "grad_norm": 4.658834934234619, + "learning_rate": 7.613469985358711e-08, + "loss": 1.3232, + "mean_token_accuracy": 0.6417742967605591, + "num_tokens": 5433236.0, + "step": 209 + }, + { + "epoch": 0.023061717548868876, + "ewc_loss": 2.1047890186309814e-07, + "grad_norm": 4.570327281951904, + "learning_rate": 7.650073206442166e-08, + "loss": 1.3842, + "mean_token_accuracy": 0.629004716873169, + "num_tokens": 5457306.0, + "step": 210 + }, + { + "epoch": 0.02317153525148254, + "ewc_loss": 2.1141022443771362e-07, + "grad_norm": 4.3195977210998535, + "learning_rate": 7.686676427525622e-08, + "loss": 1.309, + "mean_token_accuracy": 0.6415446996688843, + "num_tokens": 5477673.0, + "step": 211 + }, + { + "epoch": 0.0232813529540962, + "ewc_loss": 2.1513551473617554e-07, + "grad_norm": 3.8871145248413086, + "learning_rate": 7.723279648609077e-08, + "loss": 1.3768, + "mean_token_accuracy": 0.6285342574119568, + "num_tokens": 5506648.0, + "step": 212 + }, + { + "epoch": 0.023391170656709863, + "ewc_loss": 2.2165477275848389e-07, + "grad_norm": 3.857606887817383, + "learning_rate": 7.759882869692533e-08, + "loss": 1.4425, + "mean_token_accuracy": 0.6064967513084412, + "num_tokens": 5537209.0, + "step": 213 + }, + { + "epoch": 0.023500988359323523, + "ewc_loss": 2.2258609533309937e-07, + "grad_norm": 4.39652681350708, + "learning_rate": 7.796486090775987e-08, + "loss": 1.3899, + "mean_token_accuracy": 0.6310348510742188, + "num_tokens": 5559612.0, + "step": 214 + }, + { + "epoch": 0.023610806061937183, + "ewc_loss": 2.4028122425079346e-07, + "grad_norm": 17.159658432006836, + "learning_rate": 7.833089311859443e-08, + "loss": 1.3005, + "mean_token_accuracy": 0.6488912105560303, + "num_tokens": 5581853.0, + "step": 215 + }, + { + "epoch": 0.023720623764550846, + "ewc_loss": 2.2817403078079224e-07, + "grad_norm": 4.377392768859863, + "learning_rate": 7.869692532942898e-08, + "loss": 1.3131, + "mean_token_accuracy": 0.6423176527023315, + "num_tokens": 5605429.0, + "step": 216 + }, + { + "epoch": 0.023830441467164506, + "ewc_loss": 2.300366759300232e-07, + "grad_norm": 4.116897106170654, + "learning_rate": 7.906295754026354e-08, + "loss": 1.2866, + "mean_token_accuracy": 0.6425932049751282, + "num_tokens": 5635630.0, + "step": 217 + }, + { + "epoch": 0.02394025916977817, + "ewc_loss": 2.2817403078079224e-07, + "grad_norm": 4.412391662597656, + "learning_rate": 7.942898975109809e-08, + "loss": 1.3954, + "mean_token_accuracy": 0.6185718774795532, + "num_tokens": 5661702.0, + "step": 218 + }, + { + "epoch": 0.02405007687239183, + "ewc_loss": 2.3189932107925415e-07, + "grad_norm": 4.973567485809326, + "learning_rate": 7.979502196193265e-08, + "loss": 1.2161, + "mean_token_accuracy": 0.6562098860740662, + "num_tokens": 5682274.0, + "step": 219 + }, + { + "epoch": 0.024159894575005492, + "ewc_loss": 2.3469328880310059e-07, + "grad_norm": 3.713430643081665, + "learning_rate": 8.016105417276719e-08, + "loss": 1.3134, + "mean_token_accuracy": 0.6436862945556641, + "num_tokens": 5715819.0, + "step": 220 + }, + { + "epoch": 0.024269712277619152, + "ewc_loss": 2.3655593395233154e-07, + "grad_norm": 4.775468826293945, + "learning_rate": 8.052708638360175e-08, + "loss": 1.3471, + "mean_token_accuracy": 0.6312658786773682, + "num_tokens": 5740534.0, + "step": 221 + }, + { + "epoch": 0.024379529980232812, + "ewc_loss": 2.384185791015625e-07, + "grad_norm": 5.029636859893799, + "learning_rate": 8.08931185944363e-08, + "loss": 1.3926, + "mean_token_accuracy": 0.6221433281898499, + "num_tokens": 5762357.0, + "step": 222 + }, + { + "epoch": 0.024489347682846475, + "ewc_loss": 2.4028122425079346e-07, + "grad_norm": 4.915915012359619, + "learning_rate": 8.125915080527086e-08, + "loss": 1.219, + "mean_token_accuracy": 0.6529673337936401, + "num_tokens": 5787211.0, + "step": 223 + }, + { + "epoch": 0.024599165385460135, + "ewc_loss": 2.4028122425079346e-07, + "grad_norm": 4.5394721031188965, + "learning_rate": 8.162518301610541e-08, + "loss": 1.3468, + "mean_token_accuracy": 0.6398650407791138, + "num_tokens": 5820989.0, + "step": 224 + }, + { + "epoch": 0.0247089830880738, + "ewc_loss": 2.4028122425079346e-07, + "grad_norm": 4.662457466125488, + "learning_rate": 8.199121522693997e-08, + "loss": 1.2996, + "mean_token_accuracy": 0.6353155970573425, + "num_tokens": 5844903.0, + "step": 225 + }, + { + "epoch": 0.02481880079068746, + "ewc_loss": 2.4028122425079346e-07, + "grad_norm": 4.855496406555176, + "learning_rate": 8.235724743777451e-08, + "loss": 1.3002, + "mean_token_accuracy": 0.64515221118927, + "num_tokens": 5867365.0, + "step": 226 + }, + { + "epoch": 0.024928618493301122, + "ewc_loss": 2.4028122425079346e-07, + "grad_norm": 4.355341911315918, + "learning_rate": 8.272327964860907e-08, + "loss": 1.3758, + "mean_token_accuracy": 0.6221778988838196, + "num_tokens": 5895675.0, + "step": 227 + }, + { + "epoch": 0.02503843619591478, + "ewc_loss": 2.421438694000244e-07, + "grad_norm": 4.786472797393799, + "learning_rate": 8.308931185944362e-08, + "loss": 1.3321, + "mean_token_accuracy": 0.6396603584289551, + "num_tokens": 5920138.0, + "step": 228 + }, + { + "epoch": 0.02514825389852844, + "ewc_loss": 2.5331974029541016e-07, + "grad_norm": 9.364675521850586, + "learning_rate": 8.345534407027818e-08, + "loss": 1.3105, + "mean_token_accuracy": 0.6339833736419678, + "num_tokens": 5945763.0, + "step": 229 + }, + { + "epoch": 0.025258071601142105, + "ewc_loss": 2.4586915969848633e-07, + "grad_norm": 5.714087009429932, + "learning_rate": 8.382137628111273e-08, + "loss": 1.2273, + "mean_token_accuracy": 0.6619923114776611, + "num_tokens": 5971811.0, + "step": 230 + }, + { + "epoch": 0.025367889303755765, + "ewc_loss": 2.477318048477173e-07, + "grad_norm": 5.830867290496826, + "learning_rate": 8.41874084919473e-08, + "loss": 1.3626, + "mean_token_accuracy": 0.6323987245559692, + "num_tokens": 5995852.0, + "step": 231 + }, + { + "epoch": 0.025477707006369428, + "ewc_loss": 2.477318048477173e-07, + "grad_norm": 5.191222190856934, + "learning_rate": 8.455344070278183e-08, + "loss": 1.2927, + "mean_token_accuracy": 0.6413793563842773, + "num_tokens": 6018357.0, + "step": 232 + }, + { + "epoch": 0.025587524708983088, + "ewc_loss": 2.477318048477173e-07, + "grad_norm": 4.725028038024902, + "learning_rate": 8.491947291361639e-08, + "loss": 1.3412, + "mean_token_accuracy": 0.6350704431533813, + "num_tokens": 6046503.0, + "step": 233 + }, + { + "epoch": 0.025697342411596748, + "ewc_loss": 2.551823854446411e-07, + "grad_norm": 5.680643081665039, + "learning_rate": 8.528550512445094e-08, + "loss": 1.2729, + "mean_token_accuracy": 0.6456397771835327, + "num_tokens": 6067210.0, + "step": 234 + }, + { + "epoch": 0.02580716011421041, + "ewc_loss": 2.551823854446411e-07, + "grad_norm": 4.915842533111572, + "learning_rate": 8.56515373352855e-08, + "loss": 1.2505, + "mean_token_accuracy": 0.6505640149116516, + "num_tokens": 6091649.0, + "step": 235 + }, + { + "epoch": 0.02591697781682407, + "ewc_loss": 2.60770320892334e-07, + "grad_norm": 5.018431663513184, + "learning_rate": 8.601756954612005e-08, + "loss": 1.3103, + "mean_token_accuracy": 0.6431167125701904, + "num_tokens": 6119052.0, + "step": 236 + }, + { + "epoch": 0.026026795519437734, + "ewc_loss": 2.5704503059387207e-07, + "grad_norm": 5.837022304534912, + "learning_rate": 8.638360175695462e-08, + "loss": 1.2537, + "mean_token_accuracy": 0.657387375831604, + "num_tokens": 6141688.0, + "step": 237 + }, + { + "epoch": 0.026136613222051394, + "ewc_loss": 2.60770320892334e-07, + "grad_norm": 5.44358491897583, + "learning_rate": 8.674963396778915e-08, + "loss": 1.3336, + "mean_token_accuracy": 0.629616379737854, + "num_tokens": 6168591.0, + "step": 238 + }, + { + "epoch": 0.026246430924665057, + "ewc_loss": 2.644956111907959e-07, + "grad_norm": 7.5994720458984375, + "learning_rate": 8.711566617862371e-08, + "loss": 1.2058, + "mean_token_accuracy": 0.6594893932342529, + "num_tokens": 6186942.0, + "step": 239 + }, + { + "epoch": 0.026356248627278717, + "ewc_loss": 2.682209014892578e-07, + "grad_norm": 5.431744575500488, + "learning_rate": 8.748169838945826e-08, + "loss": 1.3014, + "mean_token_accuracy": 0.6414440274238586, + "num_tokens": 6214324.0, + "step": 240 + }, + { + "epoch": 0.026466066329892377, + "ewc_loss": 2.6263296604156494e-07, + "grad_norm": 5.042110919952393, + "learning_rate": 8.784773060029283e-08, + "loss": 1.3193, + "mean_token_accuracy": 0.6367415189743042, + "num_tokens": 6240342.0, + "step": 241 + }, + { + "epoch": 0.02657588403250604, + "ewc_loss": 2.6263296604156494e-07, + "grad_norm": 4.9072980880737305, + "learning_rate": 8.821376281112737e-08, + "loss": 1.3708, + "mean_token_accuracy": 0.6256118416786194, + "num_tokens": 6266070.0, + "step": 242 + }, + { + "epoch": 0.0266857017351197, + "ewc_loss": 2.6635825634002686e-07, + "grad_norm": 5.357433319091797, + "learning_rate": 8.857979502196194e-08, + "loss": 1.3227, + "mean_token_accuracy": 0.636807918548584, + "num_tokens": 6294087.0, + "step": 243 + }, + { + "epoch": 0.026795519437733364, + "ewc_loss": 2.6635825634002686e-07, + "grad_norm": 4.832674980163574, + "learning_rate": 8.894582723279647e-08, + "loss": 1.2876, + "mean_token_accuracy": 0.6398687958717346, + "num_tokens": 6318429.0, + "step": 244 + }, + { + "epoch": 0.026905337140347024, + "ewc_loss": 2.682209014892578e-07, + "grad_norm": 4.545891284942627, + "learning_rate": 8.931185944363104e-08, + "loss": 1.3197, + "mean_token_accuracy": 0.640902578830719, + "num_tokens": 6348101.0, + "step": 245 + }, + { + "epoch": 0.027015154842960687, + "ewc_loss": 2.7194619178771973e-07, + "grad_norm": 4.500149250030518, + "learning_rate": 8.967789165446558e-08, + "loss": 1.2944, + "mean_token_accuracy": 0.6467714309692383, + "num_tokens": 6373529.0, + "step": 246 + }, + { + "epoch": 0.027124972545574347, + "ewc_loss": 2.7008354663848877e-07, + "grad_norm": 4.939931869506836, + "learning_rate": 9.004392386530015e-08, + "loss": 1.4017, + "mean_token_accuracy": 0.617154598236084, + "num_tokens": 6400776.0, + "step": 247 + }, + { + "epoch": 0.027234790248188007, + "ewc_loss": 2.738088369369507e-07, + "grad_norm": 5.297226905822754, + "learning_rate": 9.04099560761347e-08, + "loss": 1.2685, + "mean_token_accuracy": 0.640670657157898, + "num_tokens": 6425702.0, + "step": 248 + }, + { + "epoch": 0.02734460795080167, + "ewc_loss": 2.738088369369507e-07, + "grad_norm": 5.012160301208496, + "learning_rate": 9.077598828696926e-08, + "loss": 1.2738, + "mean_token_accuracy": 0.6483451724052429, + "num_tokens": 6447754.0, + "step": 249 + }, + { + "epoch": 0.02745442565341533, + "ewc_loss": 2.7939677238464355e-07, + "grad_norm": 3.9392127990722656, + "learning_rate": 9.11420204978038e-08, + "loss": 1.2179, + "mean_token_accuracy": 0.6668751835823059, + "num_tokens": 6477535.0, + "step": 250 + }, + { + "epoch": 0.027564243356028993, + "ewc_loss": 2.812594175338745e-07, + "grad_norm": 4.5546650886535645, + "learning_rate": 9.150805270863836e-08, + "loss": 1.2596, + "mean_token_accuracy": 0.6517938375473022, + "num_tokens": 6504584.0, + "step": 251 + }, + { + "epoch": 0.027674061058642653, + "ewc_loss": 2.8312206268310547e-07, + "grad_norm": 5.164485931396484, + "learning_rate": 9.18740849194729e-08, + "loss": 1.1544, + "mean_token_accuracy": 0.6791795492172241, + "num_tokens": 6529726.0, + "step": 252 + }, + { + "epoch": 0.027783878761256313, + "ewc_loss": 2.8870999813079834e-07, + "grad_norm": 7.021096706390381, + "learning_rate": 9.224011713030747e-08, + "loss": 1.3527, + "mean_token_accuracy": 0.6331825256347656, + "num_tokens": 6550265.0, + "step": 253 + }, + { + "epoch": 0.027893696463869976, + "ewc_loss": 2.8312206268310547e-07, + "grad_norm": 5.607412338256836, + "learning_rate": 9.260614934114202e-08, + "loss": 1.4245, + "mean_token_accuracy": 0.6153475046157837, + "num_tokens": 6579733.0, + "step": 254 + }, + { + "epoch": 0.028003514166483636, + "ewc_loss": 2.849847078323364e-07, + "grad_norm": 4.41669225692749, + "learning_rate": 9.297218155197657e-08, + "loss": 1.3128, + "mean_token_accuracy": 0.6407374143600464, + "num_tokens": 6613781.0, + "step": 255 + }, + { + "epoch": 0.0281133318690973, + "ewc_loss": 2.849847078323364e-07, + "grad_norm": 5.472139835357666, + "learning_rate": 9.333821376281111e-08, + "loss": 1.2249, + "mean_token_accuracy": 0.6494632959365845, + "num_tokens": 6635982.0, + "step": 256 + }, + { + "epoch": 0.02822314957171096, + "ewc_loss": 2.868473529815674e-07, + "grad_norm": 6.600491523742676, + "learning_rate": 9.370424597364568e-08, + "loss": 1.2902, + "mean_token_accuracy": 0.6364535689353943, + "num_tokens": 6660180.0, + "step": 257 + }, + { + "epoch": 0.028332967274324623, + "ewc_loss": 2.905726432800293e-07, + "grad_norm": 4.603999137878418, + "learning_rate": 9.407027818448023e-08, + "loss": 1.2895, + "mean_token_accuracy": 0.6388111114501953, + "num_tokens": 6687798.0, + "step": 258 + }, + { + "epoch": 0.028442784976938282, + "ewc_loss": 2.9243528842926025e-07, + "grad_norm": 6.057923316955566, + "learning_rate": 9.443631039531479e-08, + "loss": 1.3145, + "mean_token_accuracy": 0.6413578987121582, + "num_tokens": 6716005.0, + "step": 259 + }, + { + "epoch": 0.028552602679551942, + "ewc_loss": 2.9243528842926025e-07, + "grad_norm": 5.273670673370361, + "learning_rate": 9.480234260614934e-08, + "loss": 1.2945, + "mean_token_accuracy": 0.6423937678337097, + "num_tokens": 6744026.0, + "step": 260 + }, + { + "epoch": 0.028662420382165606, + "ewc_loss": 2.9243528842926025e-07, + "grad_norm": 6.374000549316406, + "learning_rate": 9.516837481698389e-08, + "loss": 1.2635, + "mean_token_accuracy": 0.6442674994468689, + "num_tokens": 6765799.0, + "step": 261 + }, + { + "epoch": 0.028772238084779266, + "ewc_loss": 2.905726432800293e-07, + "grad_norm": 4.989360332489014, + "learning_rate": 9.553440702781844e-08, + "loss": 1.3285, + "mean_token_accuracy": 0.6290304064750671, + "num_tokens": 6792636.0, + "step": 262 + }, + { + "epoch": 0.02888205578739293, + "ewc_loss": 2.905726432800293e-07, + "grad_norm": 7.6237640380859375, + "learning_rate": 9.5900439238653e-08, + "loss": 1.1301, + "mean_token_accuracy": 0.6806138753890991, + "num_tokens": 6814122.0, + "step": 263 + }, + { + "epoch": 0.02899187349000659, + "ewc_loss": 2.9243528842926025e-07, + "grad_norm": 5.209842681884766, + "learning_rate": 9.626647144948755e-08, + "loss": 1.3414, + "mean_token_accuracy": 0.6280863285064697, + "num_tokens": 6844162.0, + "step": 264 + }, + { + "epoch": 0.029101691192620252, + "ewc_loss": 2.942979335784912e-07, + "grad_norm": 6.777639865875244, + "learning_rate": 9.663250366032211e-08, + "loss": 1.1376, + "mean_token_accuracy": 0.6773605346679688, + "num_tokens": 6867856.0, + "step": 265 + }, + { + "epoch": 0.029211508895233912, + "ewc_loss": 2.942979335784912e-07, + "grad_norm": 5.463246822357178, + "learning_rate": 9.699853587115666e-08, + "loss": 1.2382, + "mean_token_accuracy": 0.6630851030349731, + "num_tokens": 6894300.0, + "step": 266 + }, + { + "epoch": 0.029321326597847572, + "ewc_loss": 2.9616057872772217e-07, + "grad_norm": 5.541365146636963, + "learning_rate": 9.736456808199121e-08, + "loss": 1.2527, + "mean_token_accuracy": 0.6507213711738586, + "num_tokens": 6919880.0, + "step": 267 + }, + { + "epoch": 0.029431144300461235, + "ewc_loss": 2.9616057872772217e-07, + "grad_norm": 4.982932090759277, + "learning_rate": 9.773060029282576e-08, + "loss": 1.2961, + "mean_token_accuracy": 0.6449486017227173, + "num_tokens": 6948610.0, + "step": 268 + }, + { + "epoch": 0.029540962003074895, + "ewc_loss": 3.03611159324646e-07, + "grad_norm": 4.882441520690918, + "learning_rate": 9.809663250366032e-08, + "loss": 1.2749, + "mean_token_accuracy": 0.6436885595321655, + "num_tokens": 6974058.0, + "step": 269 + }, + { + "epoch": 0.02965077970568856, + "ewc_loss": 3.03611159324646e-07, + "grad_norm": 6.5734663009643555, + "learning_rate": 9.846266471449487e-08, + "loss": 1.3284, + "mean_token_accuracy": 0.6348215341567993, + "num_tokens": 7001527.0, + "step": 270 + }, + { + "epoch": 0.029760597408302218, + "ewc_loss": 3.0919909477233887e-07, + "grad_norm": 6.581190586090088, + "learning_rate": 9.882869692532943e-08, + "loss": 1.3043, + "mean_token_accuracy": 0.6334043741226196, + "num_tokens": 7025908.0, + "step": 271 + }, + { + "epoch": 0.029870415110915878, + "ewc_loss": 3.110617399215698e-07, + "grad_norm": 6.883568286895752, + "learning_rate": 9.919472913616398e-08, + "loss": 1.2021, + "mean_token_accuracy": 0.6562824845314026, + "num_tokens": 7048417.0, + "step": 272 + }, + { + "epoch": 0.02998023281352954, + "ewc_loss": 3.129243850708008e-07, + "grad_norm": 6.949985504150391, + "learning_rate": 9.956076134699853e-08, + "loss": 1.2375, + "mean_token_accuracy": 0.6524739265441895, + "num_tokens": 7072239.0, + "step": 273 + }, + { + "epoch": 0.0300900505161432, + "ewc_loss": 3.129243850708008e-07, + "grad_norm": 5.349663257598877, + "learning_rate": 9.992679355783308e-08, + "loss": 1.3007, + "mean_token_accuracy": 0.6390265822410583, + "num_tokens": 7097960.0, + "step": 274 + }, + { + "epoch": 0.030199868218756865, + "ewc_loss": 3.129243850708008e-07, + "grad_norm": 6.057145595550537, + "learning_rate": 1.0029282576866764e-07, + "loss": 1.2808, + "mean_token_accuracy": 0.6490392684936523, + "num_tokens": 7129347.0, + "step": 275 + }, + { + "epoch": 0.030309685921370524, + "ewc_loss": 3.166496753692627e-07, + "grad_norm": 6.24724817276001, + "learning_rate": 1.0065885797950219e-07, + "loss": 1.2725, + "mean_token_accuracy": 0.6432185173034668, + "num_tokens": 7155811.0, + "step": 276 + }, + { + "epoch": 0.030419503623984188, + "ewc_loss": 3.166496753692627e-07, + "grad_norm": 8.710041999816895, + "learning_rate": 1.0102489019033675e-07, + "loss": 1.1655, + "mean_token_accuracy": 0.671107292175293, + "num_tokens": 7175864.0, + "step": 277 + }, + { + "epoch": 0.030529321326597848, + "ewc_loss": 3.1851232051849365e-07, + "grad_norm": 7.92246150970459, + "learning_rate": 1.013909224011713e-07, + "loss": 1.1947, + "mean_token_accuracy": 0.6593742370605469, + "num_tokens": 7197905.0, + "step": 278 + }, + { + "epoch": 0.030639139029211507, + "ewc_loss": 3.2223761081695557e-07, + "grad_norm": 8.855232238769531, + "learning_rate": 1.0175695461200585e-07, + "loss": 1.2805, + "mean_token_accuracy": 0.6451390981674194, + "num_tokens": 7223710.0, + "step": 279 + }, + { + "epoch": 0.03074895673182517, + "ewc_loss": 3.2782554626464844e-07, + "grad_norm": 7.208897113800049, + "learning_rate": 1.021229868228404e-07, + "loss": 1.2334, + "mean_token_accuracy": 0.6532043218612671, + "num_tokens": 7245870.0, + "step": 280 + }, + { + "epoch": 0.03085877443443883, + "ewc_loss": 3.259629011154175e-07, + "grad_norm": 6.150954246520996, + "learning_rate": 1.0248901903367496e-07, + "loss": 1.2725, + "mean_token_accuracy": 0.6451698541641235, + "num_tokens": 7271138.0, + "step": 281 + }, + { + "epoch": 0.030968592137052494, + "ewc_loss": 3.2782554626464844e-07, + "grad_norm": 8.976801872253418, + "learning_rate": 1.0285505124450951e-07, + "loss": 1.2897, + "mean_token_accuracy": 0.6480138301849365, + "num_tokens": 7297052.0, + "step": 282 + }, + { + "epoch": 0.031078409839666154, + "ewc_loss": 3.296881914138794e-07, + "grad_norm": 8.36817455291748, + "learning_rate": 1.0322108345534407e-07, + "loss": 1.3572, + "mean_token_accuracy": 0.6280487775802612, + "num_tokens": 7326733.0, + "step": 283 + }, + { + "epoch": 0.031188227542279817, + "ewc_loss": 3.2782554626464844e-07, + "grad_norm": 7.141181945800781, + "learning_rate": 1.0358711566617862e-07, + "loss": 1.3034, + "mean_token_accuracy": 0.6351481676101685, + "num_tokens": 7358626.0, + "step": 284 + }, + { + "epoch": 0.03129804524489348, + "ewc_loss": 3.3527612686157227e-07, + "grad_norm": 7.383785724639893, + "learning_rate": 1.0395314787701317e-07, + "loss": 1.3043, + "mean_token_accuracy": 0.6364835500717163, + "num_tokens": 7385058.0, + "step": 285 + }, + { + "epoch": 0.03140786294750714, + "ewc_loss": 3.296881914138794e-07, + "grad_norm": 7.172998428344727, + "learning_rate": 1.0431918008784772e-07, + "loss": 1.2596, + "mean_token_accuracy": 0.6515467762947083, + "num_tokens": 7414433.0, + "step": 286 + }, + { + "epoch": 0.0315176806501208, + "ewc_loss": 3.334134817123413e-07, + "grad_norm": 6.586865425109863, + "learning_rate": 1.0468521229868228e-07, + "loss": 1.3749, + "mean_token_accuracy": 0.6210986375808716, + "num_tokens": 7446196.0, + "step": 287 + }, + { + "epoch": 0.031627498352734464, + "ewc_loss": 3.3527612686157227e-07, + "grad_norm": 10.33791732788086, + "learning_rate": 1.0505124450951683e-07, + "loss": 1.2353, + "mean_token_accuracy": 0.6475216746330261, + "num_tokens": 7469003.0, + "step": 288 + }, + { + "epoch": 0.03173731605534812, + "ewc_loss": 3.390014171600342e-07, + "grad_norm": 8.473564147949219, + "learning_rate": 1.0541727672035139e-07, + "loss": 1.299, + "mean_token_accuracy": 0.6406275033950806, + "num_tokens": 7494794.0, + "step": 289 + }, + { + "epoch": 0.03184713375796178, + "ewc_loss": 3.427267074584961e-07, + "grad_norm": 7.592014789581299, + "learning_rate": 1.0578330893118593e-07, + "loss": 1.296, + "mean_token_accuracy": 0.6386842727661133, + "num_tokens": 7531151.0, + "step": 290 + }, + { + "epoch": 0.03195695146057544, + "ewc_loss": 3.4458935260772705e-07, + "grad_norm": 6.803127288818359, + "learning_rate": 1.0614934114202049e-07, + "loss": 1.2035, + "mean_token_accuracy": 0.6590300798416138, + "num_tokens": 7556362.0, + "step": 291 + }, + { + "epoch": 0.0320667691631891, + "ewc_loss": 3.4086406230926514e-07, + "grad_norm": 7.434401988983154, + "learning_rate": 1.0651537335285504e-07, + "loss": 1.2919, + "mean_token_accuracy": 0.6367580890655518, + "num_tokens": 7580209.0, + "step": 292 + }, + { + "epoch": 0.03217658686580277, + "ewc_loss": 3.4086406230926514e-07, + "grad_norm": 5.793886661529541, + "learning_rate": 1.068814055636896e-07, + "loss": 1.3489, + "mean_token_accuracy": 0.6268287897109985, + "num_tokens": 7607823.0, + "step": 293 + }, + { + "epoch": 0.03228640456841643, + "ewc_loss": 3.427267074584961e-07, + "grad_norm": 6.006738185882568, + "learning_rate": 1.0724743777452415e-07, + "loss": 1.36, + "mean_token_accuracy": 0.628200352191925, + "num_tokens": 7630083.0, + "step": 294 + }, + { + "epoch": 0.03239622227103009, + "ewc_loss": 3.427267074584961e-07, + "grad_norm": 5.5522613525390625, + "learning_rate": 1.0761346998535871e-07, + "loss": 1.3481, + "mean_token_accuracy": 0.6322265267372131, + "num_tokens": 7657339.0, + "step": 295 + }, + { + "epoch": 0.03250603997364375, + "ewc_loss": 3.46451997756958e-07, + "grad_norm": 6.584111213684082, + "learning_rate": 1.0797950219619325e-07, + "loss": 1.1373, + "mean_token_accuracy": 0.6675947904586792, + "num_tokens": 7678789.0, + "step": 296 + }, + { + "epoch": 0.03261585767625741, + "ewc_loss": 3.4831464290618896e-07, + "grad_norm": 5.057745456695557, + "learning_rate": 1.0834553440702781e-07, + "loss": 1.2178, + "mean_token_accuracy": 0.6626150012016296, + "num_tokens": 7705531.0, + "step": 297 + }, + { + "epoch": 0.032725675378871076, + "ewc_loss": 3.520399332046509e-07, + "grad_norm": 6.813444137573242, + "learning_rate": 1.0871156661786236e-07, + "loss": 1.2422, + "mean_token_accuracy": 0.6514413356781006, + "num_tokens": 7726907.0, + "step": 298 + }, + { + "epoch": 0.032835493081484736, + "ewc_loss": 3.557652235031128e-07, + "grad_norm": 5.336249351501465, + "learning_rate": 1.0907759882869692e-07, + "loss": 1.4086, + "mean_token_accuracy": 0.6286894679069519, + "num_tokens": 7754917.0, + "step": 299 + }, + { + "epoch": 0.032945310784098396, + "ewc_loss": 3.5762786865234375e-07, + "grad_norm": 5.8648576736450195, + "learning_rate": 1.0944363103953147e-07, + "loss": 1.2622, + "mean_token_accuracy": 0.6426950097084045, + "num_tokens": 7780922.0, + "step": 300 + }, + { + "epoch": 0.033055128486712056, + "ewc_loss": 3.594905138015747e-07, + "grad_norm": 5.344520568847656, + "learning_rate": 1.0980966325036604e-07, + "loss": 1.2822, + "mean_token_accuracy": 0.6442726850509644, + "num_tokens": 7807004.0, + "step": 301 + }, + { + "epoch": 0.03316494618932572, + "ewc_loss": 3.6135315895080566e-07, + "grad_norm": 8.726885795593262, + "learning_rate": 1.1017569546120057e-07, + "loss": 1.1629, + "mean_token_accuracy": 0.6645535230636597, + "num_tokens": 7830480.0, + "step": 302 + }, + { + "epoch": 0.03327476389193938, + "ewc_loss": 3.6135315895080566e-07, + "grad_norm": 8.758410453796387, + "learning_rate": 1.1054172767203513e-07, + "loss": 1.2682, + "mean_token_accuracy": 0.6421757936477661, + "num_tokens": 7856537.0, + "step": 303 + }, + { + "epoch": 0.03338458159455304, + "ewc_loss": 3.6135315895080566e-07, + "grad_norm": 7.43403434753418, + "learning_rate": 1.1090775988286968e-07, + "loss": 1.1964, + "mean_token_accuracy": 0.6616547703742981, + "num_tokens": 7875428.0, + "step": 304 + }, + { + "epoch": 0.0334943992971667, + "ewc_loss": 3.688037395477295e-07, + "grad_norm": 7.12901496887207, + "learning_rate": 1.1127379209370424e-07, + "loss": 1.2268, + "mean_token_accuracy": 0.6504001617431641, + "num_tokens": 7899765.0, + "step": 305 + }, + { + "epoch": 0.03360421699978036, + "ewc_loss": 3.91155481338501e-07, + "grad_norm": 18.399810791015625, + "learning_rate": 1.116398243045388e-07, + "loss": 1.2596, + "mean_token_accuracy": 0.646466851234436, + "num_tokens": 7922002.0, + "step": 306 + }, + { + "epoch": 0.03371403470239403, + "ewc_loss": 3.725290298461914e-07, + "grad_norm": 7.139491558074951, + "learning_rate": 1.1200585651537336e-07, + "loss": 1.2672, + "mean_token_accuracy": 0.6518369913101196, + "num_tokens": 7945165.0, + "step": 307 + }, + { + "epoch": 0.03382385240500769, + "ewc_loss": 3.7066638469696045e-07, + "grad_norm": 6.900804042816162, + "learning_rate": 1.1237188872620789e-07, + "loss": 1.1899, + "mean_token_accuracy": 0.6711652278900146, + "num_tokens": 7967060.0, + "step": 308 + }, + { + "epoch": 0.03393367010762135, + "ewc_loss": 3.7066638469696045e-07, + "grad_norm": 5.27346134185791, + "learning_rate": 1.1273792093704245e-07, + "loss": 1.209, + "mean_token_accuracy": 0.6553012728691101, + "num_tokens": 7988205.0, + "step": 309 + }, + { + "epoch": 0.03404348781023501, + "ewc_loss": 3.7066638469696045e-07, + "grad_norm": 5.901983737945557, + "learning_rate": 1.13103953147877e-07, + "loss": 1.1513, + "mean_token_accuracy": 0.6639883518218994, + "num_tokens": 8018516.0, + "step": 310 + }, + { + "epoch": 0.03415330551284867, + "ewc_loss": 3.7997961044311523e-07, + "grad_norm": 6.587949275970459, + "learning_rate": 1.1346998535871157e-07, + "loss": 1.248, + "mean_token_accuracy": 0.6538381576538086, + "num_tokens": 8049713.0, + "step": 311 + }, + { + "epoch": 0.034263123215462335, + "ewc_loss": 3.762543201446533e-07, + "grad_norm": 5.318255424499512, + "learning_rate": 1.1383601756954611e-07, + "loss": 1.2469, + "mean_token_accuracy": 0.652997612953186, + "num_tokens": 8077271.0, + "step": 312 + }, + { + "epoch": 0.034372940918075995, + "ewc_loss": 3.818422555923462e-07, + "grad_norm": 6.588942527770996, + "learning_rate": 1.1420204978038068e-07, + "loss": 1.2557, + "mean_token_accuracy": 0.6547434329986572, + "num_tokens": 8100334.0, + "step": 313 + }, + { + "epoch": 0.034482758620689655, + "ewc_loss": 3.8370490074157715e-07, + "grad_norm": 5.983547687530518, + "learning_rate": 1.1456808199121521e-07, + "loss": 1.2746, + "mean_token_accuracy": 0.66133052110672, + "num_tokens": 8127291.0, + "step": 314 + }, + { + "epoch": 0.034592576323303315, + "ewc_loss": 3.818422555923462e-07, + "grad_norm": 4.591169357299805, + "learning_rate": 1.1493411420204978e-07, + "loss": 1.307, + "mean_token_accuracy": 0.6370600461959839, + "num_tokens": 8158273.0, + "step": 315 + }, + { + "epoch": 0.034702394025916974, + "ewc_loss": 3.8370490074157715e-07, + "grad_norm": 5.485336780548096, + "learning_rate": 1.1530014641288432e-07, + "loss": 1.1741, + "mean_token_accuracy": 0.6625244617462158, + "num_tokens": 8180056.0, + "step": 316 + }, + { + "epoch": 0.03481221172853064, + "ewc_loss": 3.855675458908081e-07, + "grad_norm": 6.828196048736572, + "learning_rate": 1.1566617862371889e-07, + "loss": 1.2494, + "mean_token_accuracy": 0.6537277698516846, + "num_tokens": 8203267.0, + "step": 317 + }, + { + "epoch": 0.0349220294311443, + "ewc_loss": 3.91155481338501e-07, + "grad_norm": 6.160902976989746, + "learning_rate": 1.1603221083455344e-07, + "loss": 1.2399, + "mean_token_accuracy": 0.6556102633476257, + "num_tokens": 8225551.0, + "step": 318 + }, + { + "epoch": 0.03503184713375796, + "ewc_loss": 3.9301812648773193e-07, + "grad_norm": 6.64773416519165, + "learning_rate": 1.16398243045388e-07, + "loss": 1.2915, + "mean_token_accuracy": 0.6331308484077454, + "num_tokens": 8249591.0, + "step": 319 + }, + { + "epoch": 0.03514166483637162, + "ewc_loss": 3.9301812648773193e-07, + "grad_norm": 5.995591640472412, + "learning_rate": 1.1676427525622253e-07, + "loss": 1.2595, + "mean_token_accuracy": 0.642275333404541, + "num_tokens": 8274565.0, + "step": 320 + }, + { + "epoch": 0.03525148253898529, + "ewc_loss": 4.0046870708465576e-07, + "grad_norm": 6.982208728790283, + "learning_rate": 1.171303074670571e-07, + "loss": 1.2788, + "mean_token_accuracy": 0.641364336013794, + "num_tokens": 8301284.0, + "step": 321 + }, + { + "epoch": 0.03536130024159895, + "ewc_loss": 4.0605664253234863e-07, + "grad_norm": 6.590976238250732, + "learning_rate": 1.1749633967789165e-07, + "loss": 1.2517, + "mean_token_accuracy": 0.643413782119751, + "num_tokens": 8322113.0, + "step": 322 + }, + { + "epoch": 0.03547111794421261, + "ewc_loss": 4.0605664253234863e-07, + "grad_norm": 5.89585018157959, + "learning_rate": 1.1786237188872621e-07, + "loss": 1.2852, + "mean_token_accuracy": 0.6375751495361328, + "num_tokens": 8349315.0, + "step": 323 + }, + { + "epoch": 0.03558093564682627, + "ewc_loss": 4.079192876815796e-07, + "grad_norm": 5.742954254150391, + "learning_rate": 1.1822840409956076e-07, + "loss": 1.2446, + "mean_token_accuracy": 0.6482012867927551, + "num_tokens": 8369953.0, + "step": 324 + }, + { + "epoch": 0.03569075334943993, + "ewc_loss": 4.079192876815796e-07, + "grad_norm": 5.085165500640869, + "learning_rate": 1.1859443631039532e-07, + "loss": 1.2768, + "mean_token_accuracy": 0.6444140076637268, + "num_tokens": 8394310.0, + "step": 325 + }, + { + "epoch": 0.035800571052053594, + "ewc_loss": 4.1350722312927246e-07, + "grad_norm": 6.4333577156066895, + "learning_rate": 1.1896046852122986e-07, + "loss": 1.3281, + "mean_token_accuracy": 0.6310571432113647, + "num_tokens": 8419318.0, + "step": 326 + }, + { + "epoch": 0.035910388754667254, + "ewc_loss": 4.172325134277344e-07, + "grad_norm": 5.730651378631592, + "learning_rate": 1.1932650073206443e-07, + "loss": 1.1065, + "mean_token_accuracy": 0.6801391243934631, + "num_tokens": 8443229.0, + "step": 327 + }, + { + "epoch": 0.036020206457280914, + "ewc_loss": 4.1909515857696533e-07, + "grad_norm": 6.042314052581787, + "learning_rate": 1.1969253294289895e-07, + "loss": 1.2584, + "mean_token_accuracy": 0.6497629880905151, + "num_tokens": 8467929.0, + "step": 328 + }, + { + "epoch": 0.03613002415989457, + "ewc_loss": 4.172325134277344e-07, + "grad_norm": 5.96434211730957, + "learning_rate": 1.2005856515373353e-07, + "loss": 1.2221, + "mean_token_accuracy": 0.6545003652572632, + "num_tokens": 8491470.0, + "step": 329 + }, + { + "epoch": 0.03623984186250823, + "ewc_loss": 4.209578037261963e-07, + "grad_norm": 7.78279447555542, + "learning_rate": 1.2042459736456808e-07, + "loss": 1.19, + "mean_token_accuracy": 0.66277676820755, + "num_tokens": 8519715.0, + "step": 330 + }, + { + "epoch": 0.0363496595651219, + "ewc_loss": 4.2282044887542725e-07, + "grad_norm": 7.3487982749938965, + "learning_rate": 1.2079062957540263e-07, + "loss": 1.2496, + "mean_token_accuracy": 0.6456990242004395, + "num_tokens": 8542376.0, + "step": 331 + }, + { + "epoch": 0.03645947726773556, + "ewc_loss": 4.246830940246582e-07, + "grad_norm": 6.533610820770264, + "learning_rate": 1.2115666178623718e-07, + "loss": 1.2605, + "mean_token_accuracy": 0.6516838073730469, + "num_tokens": 8564513.0, + "step": 332 + }, + { + "epoch": 0.03656929497034922, + "ewc_loss": 4.246830940246582e-07, + "grad_norm": 6.404029846191406, + "learning_rate": 1.2152269399707175e-07, + "loss": 1.1811, + "mean_token_accuracy": 0.6676427721977234, + "num_tokens": 8587957.0, + "step": 333 + }, + { + "epoch": 0.03667911267296288, + "ewc_loss": 4.2654573917388916e-07, + "grad_norm": 6.124603271484375, + "learning_rate": 1.2188872620790627e-07, + "loss": 1.2008, + "mean_token_accuracy": 0.6654633283615112, + "num_tokens": 8612155.0, + "step": 334 + }, + { + "epoch": 0.03678893037557654, + "ewc_loss": 4.284083843231201e-07, + "grad_norm": 5.265948295593262, + "learning_rate": 1.2225475841874085e-07, + "loss": 1.2215, + "mean_token_accuracy": 0.6613635420799255, + "num_tokens": 8643677.0, + "step": 335 + }, + { + "epoch": 0.036898748078190206, + "ewc_loss": 4.302710294723511e-07, + "grad_norm": 4.873325824737549, + "learning_rate": 1.226207906295754e-07, + "loss": 1.1775, + "mean_token_accuracy": 0.6633418798446655, + "num_tokens": 8672599.0, + "step": 336 + }, + { + "epoch": 0.037008565780803866, + "ewc_loss": 4.33996319770813e-07, + "grad_norm": 5.719762325286865, + "learning_rate": 1.2298682284040995e-07, + "loss": 1.2706, + "mean_token_accuracy": 0.656209409236908, + "num_tokens": 8696109.0, + "step": 337 + }, + { + "epoch": 0.037118383483417526, + "ewc_loss": 4.3585896492004395e-07, + "grad_norm": 5.343245029449463, + "learning_rate": 1.233528550512445e-07, + "loss": 1.1752, + "mean_token_accuracy": 0.6700279712677002, + "num_tokens": 8725617.0, + "step": 338 + }, + { + "epoch": 0.037228201186031186, + "ewc_loss": 4.3585896492004395e-07, + "grad_norm": 4.98927640914917, + "learning_rate": 1.2371888726207907e-07, + "loss": 1.1854, + "mean_token_accuracy": 0.6651210784912109, + "num_tokens": 8753416.0, + "step": 339 + }, + { + "epoch": 0.03733801888864485, + "ewc_loss": 4.3958425521850586e-07, + "grad_norm": 5.173382759094238, + "learning_rate": 1.240849194729136e-07, + "loss": 1.2499, + "mean_token_accuracy": 0.6434476375579834, + "num_tokens": 8781610.0, + "step": 340 + }, + { + "epoch": 0.03744783659125851, + "ewc_loss": 4.414469003677368e-07, + "grad_norm": 5.431872367858887, + "learning_rate": 1.2445095168374817e-07, + "loss": 1.1832, + "mean_token_accuracy": 0.6560370922088623, + "num_tokens": 8805386.0, + "step": 341 + }, + { + "epoch": 0.03755765429387217, + "ewc_loss": 4.4330954551696777e-07, + "grad_norm": 6.060981273651123, + "learning_rate": 1.2481698389458272e-07, + "loss": 1.3225, + "mean_token_accuracy": 0.6359550952911377, + "num_tokens": 8830009.0, + "step": 342 + }, + { + "epoch": 0.03766747199648583, + "ewc_loss": 4.414469003677368e-07, + "grad_norm": 7.596446990966797, + "learning_rate": 1.2518301610541727e-07, + "loss": 1.1925, + "mean_token_accuracy": 0.6585776805877686, + "num_tokens": 8848499.0, + "step": 343 + }, + { + "epoch": 0.03777728969909949, + "ewc_loss": 4.507601261138916e-07, + "grad_norm": 6.030839920043945, + "learning_rate": 1.2554904831625182e-07, + "loss": 1.1665, + "mean_token_accuracy": 0.6632581949234009, + "num_tokens": 8869974.0, + "step": 344 + }, + { + "epoch": 0.03788710740171316, + "ewc_loss": 4.507601261138916e-07, + "grad_norm": 5.7710466384887695, + "learning_rate": 1.259150805270864e-07, + "loss": 1.2905, + "mean_token_accuracy": 0.6373567581176758, + "num_tokens": 8897952.0, + "step": 345 + }, + { + "epoch": 0.03799692510432682, + "ewc_loss": 4.5262277126312256e-07, + "grad_norm": 4.567122459411621, + "learning_rate": 1.2628111273792094e-07, + "loss": 1.2137, + "mean_token_accuracy": 0.66092449426651, + "num_tokens": 8931394.0, + "step": 346 + }, + { + "epoch": 0.03810674280694048, + "ewc_loss": 4.507601261138916e-07, + "grad_norm": 5.2453436851501465, + "learning_rate": 1.2664714494875547e-07, + "loss": 1.2646, + "mean_token_accuracy": 0.6478144526481628, + "num_tokens": 8962484.0, + "step": 347 + }, + { + "epoch": 0.03821656050955414, + "ewc_loss": 4.544854164123535e-07, + "grad_norm": 5.128142356872559, + "learning_rate": 1.2701317715959004e-07, + "loss": 1.2297, + "mean_token_accuracy": 0.6494109630584717, + "num_tokens": 8988810.0, + "step": 348 + }, + { + "epoch": 0.0383263782121678, + "ewc_loss": 4.5821070671081543e-07, + "grad_norm": 6.612077713012695, + "learning_rate": 1.273792093704246e-07, + "loss": 1.1756, + "mean_token_accuracy": 0.6675543785095215, + "num_tokens": 9011086.0, + "step": 349 + }, + { + "epoch": 0.038436195914781465, + "ewc_loss": 4.5821070671081543e-07, + "grad_norm": 6.011735916137695, + "learning_rate": 1.2774524158125917e-07, + "loss": 1.2515, + "mean_token_accuracy": 0.6505276560783386, + "num_tokens": 9038964.0, + "step": 350 + }, + { + "epoch": 0.038546013617395125, + "ewc_loss": 4.5821070671081543e-07, + "grad_norm": 5.978487491607666, + "learning_rate": 1.281112737920937e-07, + "loss": 1.321, + "mean_token_accuracy": 0.6283475160598755, + "num_tokens": 9065347.0, + "step": 351 + }, + { + "epoch": 0.038655831320008785, + "ewc_loss": 4.600733518600464e-07, + "grad_norm": 6.590922832489014, + "learning_rate": 1.2847730600292824e-07, + "loss": 1.1514, + "mean_token_accuracy": 0.6745767593383789, + "num_tokens": 9089414.0, + "step": 352 + }, + { + "epoch": 0.038765649022622445, + "ewc_loss": 4.6566128730773926e-07, + "grad_norm": 4.599437713623047, + "learning_rate": 1.288433382137628e-07, + "loss": 1.3156, + "mean_token_accuracy": 0.6354637145996094, + "num_tokens": 9120551.0, + "step": 353 + }, + { + "epoch": 0.038875466725236105, + "ewc_loss": 4.6566128730773926e-07, + "grad_norm": 6.105797290802002, + "learning_rate": 1.2920937042459736e-07, + "loss": 1.2174, + "mean_token_accuracy": 0.6582200527191162, + "num_tokens": 9143936.0, + "step": 354 + }, + { + "epoch": 0.03898528442784977, + "ewc_loss": 4.675239324569702e-07, + "grad_norm": 5.547848224639893, + "learning_rate": 1.295754026354319e-07, + "loss": 1.3008, + "mean_token_accuracy": 0.6339598894119263, + "num_tokens": 9171765.0, + "step": 355 + }, + { + "epoch": 0.03909510213046343, + "ewc_loss": 4.7497451305389404e-07, + "grad_norm": 5.5273518562316895, + "learning_rate": 1.2994143484626646e-07, + "loss": 1.2554, + "mean_token_accuracy": 0.6455789804458618, + "num_tokens": 9198275.0, + "step": 356 + }, + { + "epoch": 0.03920491983307709, + "ewc_loss": 4.7497451305389404e-07, + "grad_norm": 8.234600067138672, + "learning_rate": 1.3030746705710104e-07, + "loss": 1.2063, + "mean_token_accuracy": 0.6510463953018188, + "num_tokens": 9221218.0, + "step": 357 + }, + { + "epoch": 0.03931473753569075, + "ewc_loss": 4.805624485015869e-07, + "grad_norm": 6.316937446594238, + "learning_rate": 1.3067349926793558e-07, + "loss": 1.2071, + "mean_token_accuracy": 0.6497524976730347, + "num_tokens": 9249309.0, + "step": 358 + }, + { + "epoch": 0.03942455523830442, + "ewc_loss": 4.805624485015869e-07, + "grad_norm": 7.4320597648620605, + "learning_rate": 1.310395314787701e-07, + "loss": 1.238, + "mean_token_accuracy": 0.6483880281448364, + "num_tokens": 9270701.0, + "step": 359 + }, + { + "epoch": 0.03953437294091808, + "ewc_loss": 4.842877388000488e-07, + "grad_norm": 6.758646488189697, + "learning_rate": 1.3140556368960468e-07, + "loss": 1.2697, + "mean_token_accuracy": 0.6472316980361938, + "num_tokens": 9303156.0, + "step": 360 + }, + { + "epoch": 0.03964419064353174, + "ewc_loss": 4.842877388000488e-07, + "grad_norm": 6.8819451332092285, + "learning_rate": 1.3177159590043923e-07, + "loss": 1.1664, + "mean_token_accuracy": 0.6717756390571594, + "num_tokens": 9325163.0, + "step": 361 + }, + { + "epoch": 0.0397540083461454, + "ewc_loss": 4.880130290985107e-07, + "grad_norm": 8.428961753845215, + "learning_rate": 1.321376281112738e-07, + "loss": 1.1384, + "mean_token_accuracy": 0.6753673553466797, + "num_tokens": 9349474.0, + "step": 362 + }, + { + "epoch": 0.03986382604875906, + "ewc_loss": 4.880130290985107e-07, + "grad_norm": 5.091056823730469, + "learning_rate": 1.3250366032210833e-07, + "loss": 1.2028, + "mean_token_accuracy": 0.6575313806533813, + "num_tokens": 9376227.0, + "step": 363 + }, + { + "epoch": 0.039973643751372724, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.50909948348999, + "learning_rate": 1.3286969253294288e-07, + "loss": 1.1528, + "mean_token_accuracy": 0.6681852340698242, + "num_tokens": 9398802.0, + "step": 364 + }, + { + "epoch": 0.040083461453986384, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.845608711242676, + "learning_rate": 1.3323572474377745e-07, + "loss": 1.236, + "mean_token_accuracy": 0.6516038179397583, + "num_tokens": 9425588.0, + "step": 365 + }, + { + "epoch": 0.040193279156600044, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.464847564697266, + "learning_rate": 1.33601756954612e-07, + "loss": 1.2869, + "mean_token_accuracy": 0.6325814723968506, + "num_tokens": 9457907.0, + "step": 366 + }, + { + "epoch": 0.040303096859213704, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 6.191164493560791, + "learning_rate": 1.3396778916544655e-07, + "loss": 1.1365, + "mean_token_accuracy": 0.6822715997695923, + "num_tokens": 9482988.0, + "step": 367 + }, + { + "epoch": 0.040412914561827364, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 7.37857723236084, + "learning_rate": 1.343338213762811e-07, + "loss": 1.2541, + "mean_token_accuracy": 0.6458632349967957, + "num_tokens": 9503553.0, + "step": 368 + }, + { + "epoch": 0.04052273226444103, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.855843544006348, + "learning_rate": 1.3469985358711565e-07, + "loss": 1.253, + "mean_token_accuracy": 0.6495869159698486, + "num_tokens": 9528274.0, + "step": 369 + }, + { + "epoch": 0.04063254996705469, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.779054164886475, + "learning_rate": 1.3506588579795023e-07, + "loss": 1.2141, + "mean_token_accuracy": 0.6518704891204834, + "num_tokens": 9563887.0, + "step": 370 + }, + { + "epoch": 0.04074236766966835, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 7.1775970458984375, + "learning_rate": 1.3543191800878475e-07, + "loss": 1.2649, + "mean_token_accuracy": 0.6455574035644531, + "num_tokens": 9586556.0, + "step": 371 + }, + { + "epoch": 0.04085218537228201, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.745059967041016, + "learning_rate": 1.3579795021961932e-07, + "loss": 1.2144, + "mean_token_accuracy": 0.6561329960823059, + "num_tokens": 9616303.0, + "step": 372 + }, + { + "epoch": 0.04096200307489567, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.66424560546875, + "learning_rate": 1.3616398243045387e-07, + "loss": 1.2404, + "mean_token_accuracy": 0.6512757539749146, + "num_tokens": 9645693.0, + "step": 373 + }, + { + "epoch": 0.04107182077750934, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 7.591475963592529, + "learning_rate": 1.3653001464128845e-07, + "loss": 1.2335, + "mean_token_accuracy": 0.6557022333145142, + "num_tokens": 9672371.0, + "step": 374 + }, + { + "epoch": 0.041181638480122996, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 6.797176361083984, + "learning_rate": 1.3689604685212297e-07, + "loss": 1.1951, + "mean_token_accuracy": 0.6711779832839966, + "num_tokens": 9705264.0, + "step": 375 + }, + { + "epoch": 0.041291456182736656, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 6.319879531860352, + "learning_rate": 1.3726207906295752e-07, + "loss": 1.1376, + "mean_token_accuracy": 0.6754142642021179, + "num_tokens": 9729726.0, + "step": 376 + }, + { + "epoch": 0.041401273885350316, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.450720310211182, + "learning_rate": 1.376281112737921e-07, + "loss": 1.2613, + "mean_token_accuracy": 0.6519554257392883, + "num_tokens": 9756079.0, + "step": 377 + }, + { + "epoch": 0.04151109158796398, + "ewc_loss": 4.917383193969727e-07, + "grad_norm": 5.76698112487793, + "learning_rate": 1.3799414348462665e-07, + "loss": 1.1189, + "mean_token_accuracy": 0.679732084274292, + "num_tokens": 9778317.0, + "step": 378 + }, + { + "epoch": 0.04162090929057764, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 4.877236366271973, + "learning_rate": 1.383601756954612e-07, + "loss": 1.2498, + "mean_token_accuracy": 0.6478954553604126, + "num_tokens": 9808114.0, + "step": 379 + }, + { + "epoch": 0.0417307269931913, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 5.560779571533203, + "learning_rate": 1.3872620790629574e-07, + "loss": 1.23, + "mean_token_accuracy": 0.6431102752685547, + "num_tokens": 9833909.0, + "step": 380 + }, + { + "epoch": 0.04184054469580496, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 5.883454322814941, + "learning_rate": 1.390922401171303e-07, + "loss": 1.193, + "mean_token_accuracy": 0.6638387441635132, + "num_tokens": 9857091.0, + "step": 381 + }, + { + "epoch": 0.04195036239841862, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 6.085241794586182, + "learning_rate": 1.3945827232796487e-07, + "loss": 1.1588, + "mean_token_accuracy": 0.6657668948173523, + "num_tokens": 9880395.0, + "step": 382 + }, + { + "epoch": 0.04206018010103229, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 6.033981800079346, + "learning_rate": 1.398243045387994e-07, + "loss": 1.2812, + "mean_token_accuracy": 0.6405492424964905, + "num_tokens": 9904483.0, + "step": 383 + }, + { + "epoch": 0.04216999780364595, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 6.05173921585083, + "learning_rate": 1.4019033674963397e-07, + "loss": 1.1625, + "mean_token_accuracy": 0.6652724742889404, + "num_tokens": 9927865.0, + "step": 384 + }, + { + "epoch": 0.04227981550625961, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 6.240216255187988, + "learning_rate": 1.4055636896046852e-07, + "loss": 1.2434, + "mean_token_accuracy": 0.654443621635437, + "num_tokens": 9956712.0, + "step": 385 + }, + { + "epoch": 0.04238963320887327, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 6.257021903991699, + "learning_rate": 1.409224011713031e-07, + "loss": 1.0934, + "mean_token_accuracy": 0.6901869773864746, + "num_tokens": 9979222.0, + "step": 386 + }, + { + "epoch": 0.04249945091148693, + "ewc_loss": 4.954636096954346e-07, + "grad_norm": 6.276026248931885, + "learning_rate": 1.4128843338213761e-07, + "loss": 1.2272, + "mean_token_accuracy": 0.6480135321617126, + "num_tokens": 10007509.0, + "step": 387 + }, + { + "epoch": 0.042609268614100596, + "ewc_loss": 5.029141902923584e-07, + "grad_norm": 5.613750457763672, + "learning_rate": 1.4165446559297216e-07, + "loss": 1.2627, + "mean_token_accuracy": 0.649271547794342, + "num_tokens": 10037020.0, + "step": 388 + }, + { + "epoch": 0.042719086316714255, + "ewc_loss": 5.029141902923584e-07, + "grad_norm": 6.617029190063477, + "learning_rate": 1.4202049780380674e-07, + "loss": 1.2853, + "mean_token_accuracy": 0.6385359168052673, + "num_tokens": 10062076.0, + "step": 389 + }, + { + "epoch": 0.042828904019327915, + "ewc_loss": 5.029141902923584e-07, + "grad_norm": 5.249279975891113, + "learning_rate": 1.423865300146413e-07, + "loss": 1.1991, + "mean_token_accuracy": 0.6593226194381714, + "num_tokens": 10093034.0, + "step": 390 + }, + { + "epoch": 0.042938721721941575, + "ewc_loss": 5.029141902923584e-07, + "grad_norm": 6.008511066436768, + "learning_rate": 1.4275256222547584e-07, + "loss": 1.2307, + "mean_token_accuracy": 0.657219648361206, + "num_tokens": 10117118.0, + "step": 391 + }, + { + "epoch": 0.043048539424555235, + "ewc_loss": 5.029141902923584e-07, + "grad_norm": 4.883378028869629, + "learning_rate": 1.4311859443631039e-07, + "loss": 1.2721, + "mean_token_accuracy": 0.6431287527084351, + "num_tokens": 10151018.0, + "step": 392 + }, + { + "epoch": 0.0431583571271689, + "ewc_loss": 5.029141902923584e-07, + "grad_norm": 7.898329257965088, + "learning_rate": 1.4348462664714493e-07, + "loss": 1.1161, + "mean_token_accuracy": 0.6773233413696289, + "num_tokens": 10168012.0, + "step": 393 + }, + { + "epoch": 0.04326817482978256, + "ewc_loss": 5.140900611877441e-07, + "grad_norm": 6.0772504806518555, + "learning_rate": 1.438506588579795e-07, + "loss": 1.2467, + "mean_token_accuracy": 0.646052360534668, + "num_tokens": 10193267.0, + "step": 394 + }, + { + "epoch": 0.04337799253239622, + "ewc_loss": 5.140900611877441e-07, + "grad_norm": 4.989299774169922, + "learning_rate": 1.4421669106881403e-07, + "loss": 1.1536, + "mean_token_accuracy": 0.6611297130584717, + "num_tokens": 10219769.0, + "step": 395 + }, + { + "epoch": 0.04348781023500988, + "ewc_loss": 5.21540641784668e-07, + "grad_norm": 7.177287578582764, + "learning_rate": 1.445827232796486e-07, + "loss": 1.2608, + "mean_token_accuracy": 0.6425594687461853, + "num_tokens": 10244899.0, + "step": 396 + }, + { + "epoch": 0.04359762793762355, + "ewc_loss": 5.289912223815918e-07, + "grad_norm": 8.441168785095215, + "learning_rate": 1.4494875549048316e-07, + "loss": 1.1998, + "mean_token_accuracy": 0.6576688289642334, + "num_tokens": 10263877.0, + "step": 397 + }, + { + "epoch": 0.04370744564023721, + "ewc_loss": 5.327165126800537e-07, + "grad_norm": 6.38808012008667, + "learning_rate": 1.4531478770131773e-07, + "loss": 1.2588, + "mean_token_accuracy": 0.6440654993057251, + "num_tokens": 10287450.0, + "step": 398 + }, + { + "epoch": 0.04381726334285087, + "ewc_loss": 5.327165126800537e-07, + "grad_norm": 5.50461483001709, + "learning_rate": 1.4568081991215226e-07, + "loss": 1.2132, + "mean_token_accuracy": 0.664949893951416, + "num_tokens": 10314939.0, + "step": 399 + }, + { + "epoch": 0.04392708104546453, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 6.49721622467041, + "learning_rate": 1.460468521229868e-07, + "loss": 1.1037, + "mean_token_accuracy": 0.6850341558456421, + "num_tokens": 10340075.0, + "step": 400 + }, + { + "epoch": 0.04403689874807819, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 6.215656280517578, + "learning_rate": 1.4641288433382138e-07, + "loss": 1.2566, + "mean_token_accuracy": 0.6508988738059998, + "num_tokens": 10364116.0, + "step": 401 + }, + { + "epoch": 0.044146716450691854, + "ewc_loss": 5.327165126800537e-07, + "grad_norm": 5.7598700523376465, + "learning_rate": 1.4677891654465593e-07, + "loss": 1.269, + "mean_token_accuracy": 0.6485041975975037, + "num_tokens": 10391839.0, + "step": 402 + }, + { + "epoch": 0.044256534153305514, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 4.9763031005859375, + "learning_rate": 1.4714494875549048e-07, + "loss": 1.2035, + "mean_token_accuracy": 0.6719236969947815, + "num_tokens": 10419954.0, + "step": 403 + }, + { + "epoch": 0.044366351855919174, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 5.5488176345825195, + "learning_rate": 1.4751098096632503e-07, + "loss": 1.2648, + "mean_token_accuracy": 0.6428497433662415, + "num_tokens": 10444143.0, + "step": 404 + }, + { + "epoch": 0.044476169558532834, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 5.358310222625732, + "learning_rate": 1.4787701317715958e-07, + "loss": 1.1982, + "mean_token_accuracy": 0.6596634387969971, + "num_tokens": 10473697.0, + "step": 405 + }, + { + "epoch": 0.044585987261146494, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 4.945537090301514, + "learning_rate": 1.4824304538799415e-07, + "loss": 1.1917, + "mean_token_accuracy": 0.6631374955177307, + "num_tokens": 10502612.0, + "step": 406 + }, + { + "epoch": 0.04469580496376016, + "ewc_loss": 5.364418029785156e-07, + "grad_norm": 5.192854404449463, + "learning_rate": 1.4860907759882867e-07, + "loss": 1.2353, + "mean_token_accuracy": 0.6495343446731567, + "num_tokens": 10529266.0, + "step": 407 + }, + { + "epoch": 0.04480562266637382, + "ewc_loss": 5.401670932769775e-07, + "grad_norm": 6.497915744781494, + "learning_rate": 1.4897510980966325e-07, + "loss": 1.2155, + "mean_token_accuracy": 0.6543995141983032, + "num_tokens": 10550616.0, + "step": 408 + }, + { + "epoch": 0.04491544036898748, + "ewc_loss": 5.438923835754395e-07, + "grad_norm": 5.160150051116943, + "learning_rate": 1.493411420204978e-07, + "loss": 1.179, + "mean_token_accuracy": 0.6654064059257507, + "num_tokens": 10582682.0, + "step": 409 + }, + { + "epoch": 0.04502525807160114, + "ewc_loss": 5.401670932769775e-07, + "grad_norm": 5.104585647583008, + "learning_rate": 1.4970717423133235e-07, + "loss": 1.1085, + "mean_token_accuracy": 0.6791342496871948, + "num_tokens": 10606229.0, + "step": 410 + }, + { + "epoch": 0.0451350757742148, + "ewc_loss": 5.401670932769775e-07, + "grad_norm": 6.057744026184082, + "learning_rate": 1.500732064421669e-07, + "loss": 1.2569, + "mean_token_accuracy": 0.6437011957168579, + "num_tokens": 10639889.0, + "step": 411 + }, + { + "epoch": 0.04524489347682847, + "ewc_loss": 5.401670932769775e-07, + "grad_norm": 5.486156463623047, + "learning_rate": 1.5043923865300145e-07, + "loss": 1.2625, + "mean_token_accuracy": 0.6390776038169861, + "num_tokens": 10664352.0, + "step": 412 + }, + { + "epoch": 0.04535471117944213, + "ewc_loss": 5.476176738739014e-07, + "grad_norm": 4.419734954833984, + "learning_rate": 1.5080527086383602e-07, + "loss": 1.1713, + "mean_token_accuracy": 0.6695878505706787, + "num_tokens": 10692255.0, + "step": 413 + }, + { + "epoch": 0.04546452888205579, + "ewc_loss": 5.476176738739014e-07, + "grad_norm": 6.063440322875977, + "learning_rate": 1.5117130307467057e-07, + "loss": 1.1959, + "mean_token_accuracy": 0.6647091507911682, + "num_tokens": 10713360.0, + "step": 414 + }, + { + "epoch": 0.045574346584669446, + "ewc_loss": 5.513429641723633e-07, + "grad_norm": 4.662374973297119, + "learning_rate": 1.5153733528550512e-07, + "loss": 1.182, + "mean_token_accuracy": 0.6647757291793823, + "num_tokens": 10743290.0, + "step": 415 + }, + { + "epoch": 0.04568416428728311, + "ewc_loss": 5.513429641723633e-07, + "grad_norm": 6.272580146789551, + "learning_rate": 1.5190336749633967e-07, + "loss": 1.1997, + "mean_token_accuracy": 0.6519140601158142, + "num_tokens": 10771567.0, + "step": 416 + }, + { + "epoch": 0.04579398198989677, + "ewc_loss": 5.513429641723633e-07, + "grad_norm": 4.983972072601318, + "learning_rate": 1.5226939970717422e-07, + "loss": 1.162, + "mean_token_accuracy": 0.6802948713302612, + "num_tokens": 10796435.0, + "step": 417 + }, + { + "epoch": 0.04590379969251043, + "ewc_loss": 5.513429641723633e-07, + "grad_norm": 4.934422016143799, + "learning_rate": 1.526354319180088e-07, + "loss": 1.2251, + "mean_token_accuracy": 0.6621217131614685, + "num_tokens": 10826350.0, + "step": 418 + }, + { + "epoch": 0.04601361739512409, + "ewc_loss": 5.550682544708252e-07, + "grad_norm": 5.475852966308594, + "learning_rate": 1.5300146412884332e-07, + "loss": 1.1625, + "mean_token_accuracy": 0.6686758399009705, + "num_tokens": 10851407.0, + "step": 419 + }, + { + "epoch": 0.04612343509773775, + "ewc_loss": 5.550682544708252e-07, + "grad_norm": 4.302502632141113, + "learning_rate": 1.533674963396779e-07, + "loss": 1.3153, + "mean_token_accuracy": 0.6442840099334717, + "num_tokens": 10885508.0, + "step": 420 + }, + { + "epoch": 0.04623325280035142, + "ewc_loss": 5.550682544708252e-07, + "grad_norm": 5.63292121887207, + "learning_rate": 1.5373352855051244e-07, + "loss": 1.1968, + "mean_token_accuracy": 0.6534098386764526, + "num_tokens": 10911158.0, + "step": 421 + }, + { + "epoch": 0.04634307050296508, + "ewc_loss": 5.550682544708252e-07, + "grad_norm": 5.973572254180908, + "learning_rate": 1.54099560761347e-07, + "loss": 1.1323, + "mean_token_accuracy": 0.6750329732894897, + "num_tokens": 10932147.0, + "step": 422 + }, + { + "epoch": 0.04645288820557874, + "ewc_loss": 5.587935447692871e-07, + "grad_norm": 5.2935872077941895, + "learning_rate": 1.5446559297218154e-07, + "loss": 1.1852, + "mean_token_accuracy": 0.6620231866836548, + "num_tokens": 10952566.0, + "step": 423 + }, + { + "epoch": 0.0465627059081924, + "ewc_loss": 5.587935447692871e-07, + "grad_norm": 7.30365514755249, + "learning_rate": 1.548316251830161e-07, + "loss": 1.1553, + "mean_token_accuracy": 0.6686241626739502, + "num_tokens": 10971181.0, + "step": 424 + }, + { + "epoch": 0.04667252361080606, + "ewc_loss": 5.587935447692871e-07, + "grad_norm": 5.677846431732178, + "learning_rate": 1.5519765739385066e-07, + "loss": 1.1172, + "mean_token_accuracy": 0.6788139343261719, + "num_tokens": 10991562.0, + "step": 425 + }, + { + "epoch": 0.046782341313419726, + "ewc_loss": 5.62518835067749e-07, + "grad_norm": 5.5541791915893555, + "learning_rate": 1.5556368960468521e-07, + "loss": 1.1909, + "mean_token_accuracy": 0.6592897176742554, + "num_tokens": 11016836.0, + "step": 426 + }, + { + "epoch": 0.046892159016033386, + "ewc_loss": 5.662441253662109e-07, + "grad_norm": 4.825276851654053, + "learning_rate": 1.5592972181551974e-07, + "loss": 1.2344, + "mean_token_accuracy": 0.6488593816757202, + "num_tokens": 11040431.0, + "step": 427 + }, + { + "epoch": 0.047001976718647046, + "ewc_loss": 5.662441253662109e-07, + "grad_norm": 4.623035430908203, + "learning_rate": 1.562957540263543e-07, + "loss": 1.1812, + "mean_token_accuracy": 0.6653108596801758, + "num_tokens": 11065612.0, + "step": 428 + }, + { + "epoch": 0.047111794421260705, + "ewc_loss": 5.774199962615967e-07, + "grad_norm": 6.2504706382751465, + "learning_rate": 1.5666178623718886e-07, + "loss": 1.1299, + "mean_token_accuracy": 0.6807101368904114, + "num_tokens": 11090027.0, + "step": 429 + }, + { + "epoch": 0.047221612123874365, + "ewc_loss": 5.774199962615967e-07, + "grad_norm": 4.407691478729248, + "learning_rate": 1.5702781844802344e-07, + "loss": 1.2948, + "mean_token_accuracy": 0.633425235748291, + "num_tokens": 11121097.0, + "step": 430 + }, + { + "epoch": 0.04733142982648803, + "ewc_loss": 5.811452865600586e-07, + "grad_norm": 4.79215669631958, + "learning_rate": 1.5739385065885796e-07, + "loss": 1.2106, + "mean_token_accuracy": 0.6528740525245667, + "num_tokens": 11144282.0, + "step": 431 + }, + { + "epoch": 0.04744124752910169, + "ewc_loss": 5.885958671569824e-07, + "grad_norm": 5.704708576202393, + "learning_rate": 1.5775988286969253e-07, + "loss": 1.2128, + "mean_token_accuracy": 0.6596455574035645, + "num_tokens": 11172508.0, + "step": 432 + }, + { + "epoch": 0.04755106523171535, + "ewc_loss": 5.923211574554443e-07, + "grad_norm": 4.941435813903809, + "learning_rate": 1.5812591508052708e-07, + "loss": 1.1631, + "mean_token_accuracy": 0.6714267730712891, + "num_tokens": 11194936.0, + "step": 433 + }, + { + "epoch": 0.04766088293432901, + "ewc_loss": 5.923211574554443e-07, + "grad_norm": 4.374855041503906, + "learning_rate": 1.5849194729136163e-07, + "loss": 1.0645, + "mean_token_accuracy": 0.6914801001548767, + "num_tokens": 11220873.0, + "step": 434 + }, + { + "epoch": 0.04777070063694268, + "ewc_loss": 5.923211574554443e-07, + "grad_norm": 5.044304847717285, + "learning_rate": 1.5885797950219618e-07, + "loss": 1.2076, + "mean_token_accuracy": 0.6552228927612305, + "num_tokens": 11245691.0, + "step": 435 + }, + { + "epoch": 0.04788051833955634, + "ewc_loss": 5.923211574554443e-07, + "grad_norm": 4.347521781921387, + "learning_rate": 1.5922401171303073e-07, + "loss": 1.2547, + "mean_token_accuracy": 0.6620258092880249, + "num_tokens": 11272386.0, + "step": 436 + }, + { + "epoch": 0.04799033604217, + "ewc_loss": 5.923211574554443e-07, + "grad_norm": 4.272153377532959, + "learning_rate": 1.595900439238653e-07, + "loss": 1.2604, + "mean_token_accuracy": 0.6476004123687744, + "num_tokens": 11301211.0, + "step": 437 + }, + { + "epoch": 0.04810015374478366, + "ewc_loss": 5.997717380523682e-07, + "grad_norm": 4.182567596435547, + "learning_rate": 1.5995607613469986e-07, + "loss": 1.2513, + "mean_token_accuracy": 0.6518019437789917, + "num_tokens": 11331253.0, + "step": 438 + }, + { + "epoch": 0.04820997144739732, + "ewc_loss": 5.997717380523682e-07, + "grad_norm": 3.752122163772583, + "learning_rate": 1.6032210834553438e-07, + "loss": 1.174, + "mean_token_accuracy": 0.6736372709274292, + "num_tokens": 11360900.0, + "step": 439 + }, + { + "epoch": 0.048319789150010985, + "ewc_loss": 6.07222318649292e-07, + "grad_norm": 4.264418125152588, + "learning_rate": 1.6068814055636895e-07, + "loss": 1.288, + "mean_token_accuracy": 0.6355787515640259, + "num_tokens": 11391599.0, + "step": 440 + }, + { + "epoch": 0.048429606852624645, + "ewc_loss": 6.07222318649292e-07, + "grad_norm": 4.1373772621154785, + "learning_rate": 1.610541727672035e-07, + "loss": 1.2129, + "mean_token_accuracy": 0.6586199998855591, + "num_tokens": 11424019.0, + "step": 441 + }, + { + "epoch": 0.048539424555238304, + "ewc_loss": 6.146728992462158e-07, + "grad_norm": 4.8840436935424805, + "learning_rate": 1.6142020497803808e-07, + "loss": 1.1845, + "mean_token_accuracy": 0.6565196514129639, + "num_tokens": 11446707.0, + "step": 442 + }, + { + "epoch": 0.048649242257851964, + "ewc_loss": 6.146728992462158e-07, + "grad_norm": 5.204649925231934, + "learning_rate": 1.617862371888726e-07, + "loss": 1.1416, + "mean_token_accuracy": 0.6697417497634888, + "num_tokens": 11467310.0, + "step": 443 + }, + { + "epoch": 0.048759059960465624, + "ewc_loss": 6.146728992462158e-07, + "grad_norm": 4.048951148986816, + "learning_rate": 1.6215226939970718e-07, + "loss": 1.1944, + "mean_token_accuracy": 0.6600021123886108, + "num_tokens": 11493973.0, + "step": 444 + }, + { + "epoch": 0.04886887766307929, + "ewc_loss": 6.146728992462158e-07, + "grad_norm": 4.920746326446533, + "learning_rate": 1.6251830161054173e-07, + "loss": 1.133, + "mean_token_accuracy": 0.6837158203125, + "num_tokens": 11514611.0, + "step": 445 + }, + { + "epoch": 0.04897869536569295, + "ewc_loss": 6.183981895446777e-07, + "grad_norm": 4.816885471343994, + "learning_rate": 1.6288433382137627e-07, + "loss": 1.2779, + "mean_token_accuracy": 0.645038366317749, + "num_tokens": 11538158.0, + "step": 446 + }, + { + "epoch": 0.04908851306830661, + "ewc_loss": 6.221234798431396e-07, + "grad_norm": 4.72380256652832, + "learning_rate": 1.6325036603221082e-07, + "loss": 1.1934, + "mean_token_accuracy": 0.6626439094543457, + "num_tokens": 11570295.0, + "step": 447 + }, + { + "epoch": 0.04919833077092027, + "ewc_loss": 6.221234798431396e-07, + "grad_norm": 5.060939311981201, + "learning_rate": 1.6361639824304537e-07, + "loss": 1.1631, + "mean_token_accuracy": 0.6704279184341431, + "num_tokens": 11597780.0, + "step": 448 + }, + { + "epoch": 0.04930814847353393, + "ewc_loss": 6.258487701416016e-07, + "grad_norm": 4.4564619064331055, + "learning_rate": 1.6398243045387995e-07, + "loss": 1.2193, + "mean_token_accuracy": 0.6599583625793457, + "num_tokens": 11623199.0, + "step": 449 + }, + { + "epoch": 0.0494179661761476, + "ewc_loss": 6.332993507385254e-07, + "grad_norm": 4.287581920623779, + "learning_rate": 1.643484626647145e-07, + "loss": 1.2551, + "mean_token_accuracy": 0.6404659152030945, + "num_tokens": 11651657.0, + "step": 450 + }, + { + "epoch": 0.04952778387876126, + "ewc_loss": 6.295740604400635e-07, + "grad_norm": 4.430448532104492, + "learning_rate": 1.6471449487554902e-07, + "loss": 1.1933, + "mean_token_accuracy": 0.6581103801727295, + "num_tokens": 11680381.0, + "step": 451 + }, + { + "epoch": 0.04963760158137492, + "ewc_loss": 6.295740604400635e-07, + "grad_norm": 4.2317705154418945, + "learning_rate": 1.650805270863836e-07, + "loss": 1.1516, + "mean_token_accuracy": 0.670866847038269, + "num_tokens": 11705013.0, + "step": 452 + }, + { + "epoch": 0.04974741928398858, + "ewc_loss": 6.332993507385254e-07, + "grad_norm": 4.957801818847656, + "learning_rate": 1.6544655929721814e-07, + "loss": 1.0936, + "mean_token_accuracy": 0.685613751411438, + "num_tokens": 11731140.0, + "step": 453 + }, + { + "epoch": 0.049857236986602244, + "ewc_loss": 6.370246410369873e-07, + "grad_norm": 4.058785915374756, + "learning_rate": 1.6581259150805272e-07, + "loss": 1.3011, + "mean_token_accuracy": 0.634880006313324, + "num_tokens": 11761261.0, + "step": 454 + }, + { + "epoch": 0.0499670546892159, + "ewc_loss": 6.370246410369873e-07, + "grad_norm": 4.70720100402832, + "learning_rate": 1.6617862371888724e-07, + "loss": 1.1729, + "mean_token_accuracy": 0.6736915707588196, + "num_tokens": 11787241.0, + "step": 455 + }, + { + "epoch": 0.05007687239182956, + "ewc_loss": 6.444752216339111e-07, + "grad_norm": 5.756059646606445, + "learning_rate": 1.6654465592972182e-07, + "loss": 1.2078, + "mean_token_accuracy": 0.6581506133079529, + "num_tokens": 11808564.0, + "step": 456 + }, + { + "epoch": 0.05018669009444322, + "ewc_loss": 6.407499313354492e-07, + "grad_norm": 4.369946479797363, + "learning_rate": 1.6691068814055637e-07, + "loss": 1.1771, + "mean_token_accuracy": 0.6678125858306885, + "num_tokens": 11835170.0, + "step": 457 + }, + { + "epoch": 0.05029650779705688, + "ewc_loss": 6.444752216339111e-07, + "grad_norm": 5.114834785461426, + "learning_rate": 1.6727672035139092e-07, + "loss": 1.1795, + "mean_token_accuracy": 0.6670151352882385, + "num_tokens": 11862838.0, + "step": 458 + }, + { + "epoch": 0.05040632549967055, + "ewc_loss": 6.444752216339111e-07, + "grad_norm": 4.920928955078125, + "learning_rate": 1.6764275256222547e-07, + "loss": 1.3103, + "mean_token_accuracy": 0.6405317187309265, + "num_tokens": 11892865.0, + "step": 459 + }, + { + "epoch": 0.05051614320228421, + "ewc_loss": 6.48200511932373e-07, + "grad_norm": 5.034519195556641, + "learning_rate": 1.6800878477306001e-07, + "loss": 1.126, + "mean_token_accuracy": 0.6789013147354126, + "num_tokens": 11913385.0, + "step": 460 + }, + { + "epoch": 0.05062596090489787, + "ewc_loss": 6.444752216339111e-07, + "grad_norm": 4.364758014678955, + "learning_rate": 1.683748169838946e-07, + "loss": 1.2224, + "mean_token_accuracy": 0.6530160903930664, + "num_tokens": 11942433.0, + "step": 461 + }, + { + "epoch": 0.05073577860751153, + "ewc_loss": 6.444752216339111e-07, + "grad_norm": 4.28647518157959, + "learning_rate": 1.6874084919472914e-07, + "loss": 1.1943, + "mean_token_accuracy": 0.6576026082038879, + "num_tokens": 11968602.0, + "step": 462 + }, + { + "epoch": 0.05084559631012519, + "ewc_loss": 6.51925802230835e-07, + "grad_norm": 5.0820770263671875, + "learning_rate": 1.6910688140556366e-07, + "loss": 1.1832, + "mean_token_accuracy": 0.664920449256897, + "num_tokens": 11989652.0, + "step": 463 + }, + { + "epoch": 0.050955414012738856, + "ewc_loss": 6.556510925292969e-07, + "grad_norm": 4.421688079833984, + "learning_rate": 1.6947291361639824e-07, + "loss": 1.2133, + "mean_token_accuracy": 0.6490497589111328, + "num_tokens": 12014428.0, + "step": 464 + }, + { + "epoch": 0.051065231715352516, + "ewc_loss": 6.556510925292969e-07, + "grad_norm": 3.9277889728546143, + "learning_rate": 1.6983894582723279e-07, + "loss": 1.1132, + "mean_token_accuracy": 0.6771771907806396, + "num_tokens": 12041211.0, + "step": 465 + }, + { + "epoch": 0.051175049417966176, + "ewc_loss": 6.556510925292969e-07, + "grad_norm": 3.9403645992279053, + "learning_rate": 1.7020497803806736e-07, + "loss": 1.243, + "mean_token_accuracy": 0.6593117713928223, + "num_tokens": 12071813.0, + "step": 466 + }, + { + "epoch": 0.051284867120579836, + "ewc_loss": 6.556510925292969e-07, + "grad_norm": 5.120719909667969, + "learning_rate": 1.7057101024890188e-07, + "loss": 1.2446, + "mean_token_accuracy": 0.6500012278556824, + "num_tokens": 12099495.0, + "step": 467 + }, + { + "epoch": 0.051394684823193496, + "ewc_loss": 6.668269634246826e-07, + "grad_norm": 5.6009907722473145, + "learning_rate": 1.7093704245973643e-07, + "loss": 1.2427, + "mean_token_accuracy": 0.6435306668281555, + "num_tokens": 12125207.0, + "step": 468 + }, + { + "epoch": 0.05150450252580716, + "ewc_loss": 6.631016731262207e-07, + "grad_norm": 4.948457717895508, + "learning_rate": 1.71303074670571e-07, + "loss": 1.1984, + "mean_token_accuracy": 0.6520974040031433, + "num_tokens": 12146398.0, + "step": 469 + }, + { + "epoch": 0.05161432022842082, + "ewc_loss": 6.668269634246826e-07, + "grad_norm": 4.128112316131592, + "learning_rate": 1.7166910688140556e-07, + "loss": 1.0985, + "mean_token_accuracy": 0.6791354417800903, + "num_tokens": 12170215.0, + "step": 470 + }, + { + "epoch": 0.05172413793103448, + "ewc_loss": 6.705522537231445e-07, + "grad_norm": 4.671850204467773, + "learning_rate": 1.720351390922401e-07, + "loss": 1.2504, + "mean_token_accuracy": 0.6435985565185547, + "num_tokens": 12192610.0, + "step": 471 + }, + { + "epoch": 0.05183395563364814, + "ewc_loss": 6.705522537231445e-07, + "grad_norm": 4.479904651641846, + "learning_rate": 1.7240117130307466e-07, + "loss": 1.1941, + "mean_token_accuracy": 0.6646517515182495, + "num_tokens": 12220138.0, + "step": 472 + }, + { + "epoch": 0.05194377333626181, + "ewc_loss": 6.780028343200684e-07, + "grad_norm": 4.641949653625488, + "learning_rate": 1.7276720351390923e-07, + "loss": 1.2021, + "mean_token_accuracy": 0.6599081754684448, + "num_tokens": 12251326.0, + "step": 473 + }, + { + "epoch": 0.05205359103887547, + "ewc_loss": 6.742775440216064e-07, + "grad_norm": 3.695521593093872, + "learning_rate": 1.7313323572474378e-07, + "loss": 1.2603, + "mean_token_accuracy": 0.6462953686714172, + "num_tokens": 12280837.0, + "step": 474 + }, + { + "epoch": 0.05216340874148913, + "ewc_loss": 6.742775440216064e-07, + "grad_norm": 4.129424571990967, + "learning_rate": 1.734992679355783e-07, + "loss": 1.0818, + "mean_token_accuracy": 0.6892046928405762, + "num_tokens": 12304158.0, + "step": 475 + }, + { + "epoch": 0.05227322644410279, + "ewc_loss": 6.780028343200684e-07, + "grad_norm": 4.483066082000732, + "learning_rate": 1.7386530014641288e-07, + "loss": 1.2571, + "mean_token_accuracy": 0.6430839896202087, + "num_tokens": 12331193.0, + "step": 476 + }, + { + "epoch": 0.05238304414671645, + "ewc_loss": 6.780028343200684e-07, + "grad_norm": 4.35945987701416, + "learning_rate": 1.7423133235724743e-07, + "loss": 1.238, + "mean_token_accuracy": 0.653323769569397, + "num_tokens": 12357251.0, + "step": 477 + }, + { + "epoch": 0.052492861849330115, + "ewc_loss": 6.780028343200684e-07, + "grad_norm": 4.059031009674072, + "learning_rate": 1.74597364568082e-07, + "loss": 1.2527, + "mean_token_accuracy": 0.6460257172584534, + "num_tokens": 12384863.0, + "step": 478 + }, + { + "epoch": 0.052602679551943775, + "ewc_loss": 6.817281246185303e-07, + "grad_norm": 3.7860195636749268, + "learning_rate": 1.7496339677891653e-07, + "loss": 1.112, + "mean_token_accuracy": 0.6797788739204407, + "num_tokens": 12415065.0, + "step": 479 + }, + { + "epoch": 0.052712497254557435, + "ewc_loss": 6.854534149169922e-07, + "grad_norm": 4.641814231872559, + "learning_rate": 1.7532942898975108e-07, + "loss": 1.1259, + "mean_token_accuracy": 0.6749916076660156, + "num_tokens": 12439060.0, + "step": 480 + }, + { + "epoch": 0.052822314957171095, + "ewc_loss": 6.891787052154541e-07, + "grad_norm": 4.162638187408447, + "learning_rate": 1.7569546120058565e-07, + "loss": 1.1343, + "mean_token_accuracy": 0.6687732338905334, + "num_tokens": 12462856.0, + "step": 481 + }, + { + "epoch": 0.052932132659784754, + "ewc_loss": 6.891787052154541e-07, + "grad_norm": 4.631666660308838, + "learning_rate": 1.760614934114202e-07, + "loss": 1.1011, + "mean_token_accuracy": 0.6801316738128662, + "num_tokens": 12487168.0, + "step": 482 + }, + { + "epoch": 0.05304195036239842, + "ewc_loss": 6.966292858123779e-07, + "grad_norm": 4.632079124450684, + "learning_rate": 1.7642752562225475e-07, + "loss": 1.2593, + "mean_token_accuracy": 0.6422936916351318, + "num_tokens": 12516282.0, + "step": 483 + }, + { + "epoch": 0.05315176806501208, + "ewc_loss": 6.92903995513916e-07, + "grad_norm": 4.8424072265625, + "learning_rate": 1.767935578330893e-07, + "loss": 1.1188, + "mean_token_accuracy": 0.6783052682876587, + "num_tokens": 12541665.0, + "step": 484 + }, + { + "epoch": 0.05326158576762574, + "ewc_loss": 7.003545761108398e-07, + "grad_norm": 3.993403434753418, + "learning_rate": 1.7715959004392387e-07, + "loss": 1.2492, + "mean_token_accuracy": 0.6535322666168213, + "num_tokens": 12573589.0, + "step": 485 + }, + { + "epoch": 0.0533714034702394, + "ewc_loss": 7.003545761108398e-07, + "grad_norm": 3.8972549438476562, + "learning_rate": 1.7752562225475842e-07, + "loss": 1.143, + "mean_token_accuracy": 0.6722196936607361, + "num_tokens": 12601106.0, + "step": 486 + }, + { + "epoch": 0.05348122117285306, + "ewc_loss": 7.078051567077637e-07, + "grad_norm": 4.335481643676758, + "learning_rate": 1.7789165446559295e-07, + "loss": 1.166, + "mean_token_accuracy": 0.6701561212539673, + "num_tokens": 12631770.0, + "step": 487 + }, + { + "epoch": 0.05359103887546673, + "ewc_loss": 7.152557373046875e-07, + "grad_norm": 4.955552101135254, + "learning_rate": 1.7825768667642752e-07, + "loss": 1.1505, + "mean_token_accuracy": 0.6623479127883911, + "num_tokens": 12655234.0, + "step": 488 + }, + { + "epoch": 0.05370085657808039, + "ewc_loss": 7.152557373046875e-07, + "grad_norm": 3.841592788696289, + "learning_rate": 1.7862371888726207e-07, + "loss": 1.1997, + "mean_token_accuracy": 0.6588443517684937, + "num_tokens": 12682179.0, + "step": 489 + }, + { + "epoch": 0.05381067428069405, + "ewc_loss": 7.152557373046875e-07, + "grad_norm": 4.210421562194824, + "learning_rate": 1.7898975109809665e-07, + "loss": 1.1609, + "mean_token_accuracy": 0.6672227382659912, + "num_tokens": 12711843.0, + "step": 490 + }, + { + "epoch": 0.05392049198330771, + "ewc_loss": 7.115304470062256e-07, + "grad_norm": 4.470306873321533, + "learning_rate": 1.7935578330893117e-07, + "loss": 1.2022, + "mean_token_accuracy": 0.656453549861908, + "num_tokens": 12741567.0, + "step": 491 + }, + { + "epoch": 0.054030309685921374, + "ewc_loss": 7.152557373046875e-07, + "grad_norm": 4.175143718719482, + "learning_rate": 1.7972181551976572e-07, + "loss": 1.2024, + "mean_token_accuracy": 0.661676824092865, + "num_tokens": 12767598.0, + "step": 492 + }, + { + "epoch": 0.054140127388535034, + "ewc_loss": 7.152557373046875e-07, + "grad_norm": 4.189019680023193, + "learning_rate": 1.800878477306003e-07, + "loss": 1.1562, + "mean_token_accuracy": 0.6738574504852295, + "num_tokens": 12790026.0, + "step": 493 + }, + { + "epoch": 0.054249945091148694, + "ewc_loss": 7.189810276031494e-07, + "grad_norm": 4.042415618896484, + "learning_rate": 1.8045387994143484e-07, + "loss": 1.1618, + "mean_token_accuracy": 0.6715102791786194, + "num_tokens": 12818197.0, + "step": 494 + }, + { + "epoch": 0.05435976279376235, + "ewc_loss": 7.227063179016113e-07, + "grad_norm": 4.735192775726318, + "learning_rate": 1.808199121522694e-07, + "loss": 1.1714, + "mean_token_accuracy": 0.6637181043624878, + "num_tokens": 12840299.0, + "step": 495 + }, + { + "epoch": 0.05446958049637601, + "ewc_loss": 7.264316082000732e-07, + "grad_norm": 5.369107723236084, + "learning_rate": 1.8118594436310394e-07, + "loss": 1.2486, + "mean_token_accuracy": 0.6453890800476074, + "num_tokens": 12865428.0, + "step": 496 + }, + { + "epoch": 0.05457939819898968, + "ewc_loss": 7.227063179016113e-07, + "grad_norm": 3.460883855819702, + "learning_rate": 1.8155197657393852e-07, + "loss": 1.1618, + "mean_token_accuracy": 0.6745012402534485, + "num_tokens": 12892978.0, + "step": 497 + }, + { + "epoch": 0.05468921590160334, + "ewc_loss": 7.264316082000732e-07, + "grad_norm": 3.438774824142456, + "learning_rate": 1.8191800878477307e-07, + "loss": 1.2076, + "mean_token_accuracy": 0.667080283164978, + "num_tokens": 12921515.0, + "step": 498 + }, + { + "epoch": 0.054799033604217, + "ewc_loss": 7.264316082000732e-07, + "grad_norm": 4.439999103546143, + "learning_rate": 1.822840409956076e-07, + "loss": 1.1914, + "mean_token_accuracy": 0.6633288860321045, + "num_tokens": 12947231.0, + "step": 499 + }, + { + "epoch": 0.05490885130683066, + "ewc_loss": 7.301568984985352e-07, + "grad_norm": 4.056319236755371, + "learning_rate": 1.8265007320644216e-07, + "loss": 1.2006, + "mean_token_accuracy": 0.6568018794059753, + "num_tokens": 12975796.0, + "step": 500 + }, + { + "epoch": 0.05501866900944432, + "ewc_loss": 7.338821887969971e-07, + "grad_norm": 3.3077592849731445, + "learning_rate": 1.830161054172767e-07, + "loss": 1.1367, + "mean_token_accuracy": 0.6775511503219604, + "num_tokens": 13005740.0, + "step": 501 + }, + { + "epoch": 0.055128486712057986, + "ewc_loss": 7.338821887969971e-07, + "grad_norm": 4.089972496032715, + "learning_rate": 1.833821376281113e-07, + "loss": 1.1712, + "mean_token_accuracy": 0.6592104434967041, + "num_tokens": 13029608.0, + "step": 502 + }, + { + "epoch": 0.055238304414671646, + "ewc_loss": 7.413327693939209e-07, + "grad_norm": 4.805291652679443, + "learning_rate": 1.837481698389458e-07, + "loss": 1.0812, + "mean_token_accuracy": 0.6889519691467285, + "num_tokens": 13050822.0, + "step": 503 + }, + { + "epoch": 0.055348122117285306, + "ewc_loss": 7.450580596923828e-07, + "grad_norm": 3.7338058948516846, + "learning_rate": 1.8411420204978036e-07, + "loss": 1.0043, + "mean_token_accuracy": 0.7032748460769653, + "num_tokens": 13076150.0, + "step": 504 + }, + { + "epoch": 0.055457939819898966, + "ewc_loss": 7.525086402893066e-07, + "grad_norm": 3.7404394149780273, + "learning_rate": 1.8448023426061494e-07, + "loss": 1.2222, + "mean_token_accuracy": 0.650881290435791, + "num_tokens": 13103581.0, + "step": 505 + }, + { + "epoch": 0.055567757522512626, + "ewc_loss": 7.562339305877686e-07, + "grad_norm": 3.885591506958008, + "learning_rate": 1.8484626647144948e-07, + "loss": 1.1544, + "mean_token_accuracy": 0.6623363494873047, + "num_tokens": 13131904.0, + "step": 506 + }, + { + "epoch": 0.05567757522512629, + "ewc_loss": 7.562339305877686e-07, + "grad_norm": 4.56964635848999, + "learning_rate": 1.8521229868228403e-07, + "loss": 1.061, + "mean_token_accuracy": 0.6890081763267517, + "num_tokens": 13155934.0, + "step": 507 + }, + { + "epoch": 0.05578739292773995, + "ewc_loss": 7.562339305877686e-07, + "grad_norm": 3.8494222164154053, + "learning_rate": 1.8557833089311858e-07, + "loss": 1.1168, + "mean_token_accuracy": 0.6771673560142517, + "num_tokens": 13179650.0, + "step": 508 + }, + { + "epoch": 0.05589721063035361, + "ewc_loss": 7.562339305877686e-07, + "grad_norm": 3.563845157623291, + "learning_rate": 1.8594436310395313e-07, + "loss": 1.208, + "mean_token_accuracy": 0.6535316705703735, + "num_tokens": 13205174.0, + "step": 509 + }, + { + "epoch": 0.05600702833296727, + "ewc_loss": 7.562339305877686e-07, + "grad_norm": 4.083751678466797, + "learning_rate": 1.863103953147877e-07, + "loss": 1.1504, + "mean_token_accuracy": 0.6657792329788208, + "num_tokens": 13231544.0, + "step": 510 + }, + { + "epoch": 0.05611684603558094, + "ewc_loss": 7.636845111846924e-07, + "grad_norm": 4.601687431335449, + "learning_rate": 1.8667642752562223e-07, + "loss": 1.1791, + "mean_token_accuracy": 0.661078929901123, + "num_tokens": 13252845.0, + "step": 511 + }, + { + "epoch": 0.0562266637381946, + "ewc_loss": 7.636845111846924e-07, + "grad_norm": 3.5759427547454834, + "learning_rate": 1.870424597364568e-07, + "loss": 1.2732, + "mean_token_accuracy": 0.637967586517334, + "num_tokens": 13285009.0, + "step": 512 + }, + { + "epoch": 0.05633648144080826, + "ewc_loss": 7.636845111846924e-07, + "grad_norm": 3.5249013900756836, + "learning_rate": 1.8740849194729135e-07, + "loss": 1.0921, + "mean_token_accuracy": 0.687414288520813, + "num_tokens": 13312518.0, + "step": 513 + }, + { + "epoch": 0.05644629914342192, + "ewc_loss": 7.674098014831543e-07, + "grad_norm": 3.963029146194458, + "learning_rate": 1.8777452415812593e-07, + "loss": 1.0926, + "mean_token_accuracy": 0.6876910328865051, + "num_tokens": 13335578.0, + "step": 514 + }, + { + "epoch": 0.05655611684603558, + "ewc_loss": 7.674098014831543e-07, + "grad_norm": 3.139417886734009, + "learning_rate": 1.8814055636896045e-07, + "loss": 1.1986, + "mean_token_accuracy": 0.6660881042480469, + "num_tokens": 13368449.0, + "step": 515 + }, + { + "epoch": 0.056665934548649245, + "ewc_loss": 7.674098014831543e-07, + "grad_norm": 4.279016017913818, + "learning_rate": 1.88506588579795e-07, + "loss": 1.2285, + "mean_token_accuracy": 0.6539916396141052, + "num_tokens": 13397250.0, + "step": 516 + }, + { + "epoch": 0.056775752251262905, + "ewc_loss": 7.674098014831543e-07, + "grad_norm": 4.3342719078063965, + "learning_rate": 1.8887262079062958e-07, + "loss": 1.168, + "mean_token_accuracy": 0.6578351855278015, + "num_tokens": 13422857.0, + "step": 517 + }, + { + "epoch": 0.056885569953876565, + "ewc_loss": 7.674098014831543e-07, + "grad_norm": 3.7356302738189697, + "learning_rate": 1.8923865300146413e-07, + "loss": 1.0941, + "mean_token_accuracy": 0.676680862903595, + "num_tokens": 13447180.0, + "step": 518 + }, + { + "epoch": 0.056995387656490225, + "ewc_loss": 7.711350917816162e-07, + "grad_norm": 4.188251972198486, + "learning_rate": 1.8960468521229868e-07, + "loss": 1.2456, + "mean_token_accuracy": 0.6398528814315796, + "num_tokens": 13469860.0, + "step": 519 + }, + { + "epoch": 0.057105205359103885, + "ewc_loss": 7.7858567237854e-07, + "grad_norm": 3.7144124507904053, + "learning_rate": 1.8997071742313322e-07, + "loss": 1.1935, + "mean_token_accuracy": 0.6689136028289795, + "num_tokens": 13495606.0, + "step": 520 + }, + { + "epoch": 0.05721502306171755, + "ewc_loss": 7.7858567237854e-07, + "grad_norm": 3.6592655181884766, + "learning_rate": 1.9033674963396777e-07, + "loss": 1.1749, + "mean_token_accuracy": 0.6646216511726379, + "num_tokens": 13527790.0, + "step": 521 + }, + { + "epoch": 0.05732484076433121, + "ewc_loss": 7.748603820800781e-07, + "grad_norm": 3.7190873622894287, + "learning_rate": 1.9070278184480235e-07, + "loss": 1.1565, + "mean_token_accuracy": 0.6789453625679016, + "num_tokens": 13551986.0, + "step": 522 + }, + { + "epoch": 0.05743465846694487, + "ewc_loss": 7.7858567237854e-07, + "grad_norm": 3.7632675170898438, + "learning_rate": 1.9106881405563687e-07, + "loss": 1.2032, + "mean_token_accuracy": 0.6558300256729126, + "num_tokens": 13577954.0, + "step": 523 + }, + { + "epoch": 0.05754447616955853, + "ewc_loss": 7.7858567237854e-07, + "grad_norm": 3.9305593967437744, + "learning_rate": 1.9143484626647145e-07, + "loss": 1.2027, + "mean_token_accuracy": 0.6529682874679565, + "num_tokens": 13604870.0, + "step": 524 + }, + { + "epoch": 0.05765429387217219, + "ewc_loss": 7.860362529754639e-07, + "grad_norm": 3.525642156600952, + "learning_rate": 1.91800878477306e-07, + "loss": 1.2191, + "mean_token_accuracy": 0.6584707498550415, + "num_tokens": 13637200.0, + "step": 525 + }, + { + "epoch": 0.05776411157478586, + "ewc_loss": 7.860362529754639e-07, + "grad_norm": 5.086313724517822, + "learning_rate": 1.9216691068814057e-07, + "loss": 1.2115, + "mean_token_accuracy": 0.650030255317688, + "num_tokens": 13656718.0, + "step": 526 + }, + { + "epoch": 0.05787392927739952, + "ewc_loss": 8.009374141693115e-07, + "grad_norm": 4.75492525100708, + "learning_rate": 1.925329428989751e-07, + "loss": 1.1264, + "mean_token_accuracy": 0.6738549470901489, + "num_tokens": 13680498.0, + "step": 527 + }, + { + "epoch": 0.05798374698001318, + "ewc_loss": 7.972121238708496e-07, + "grad_norm": 4.333847999572754, + "learning_rate": 1.9289897510980964e-07, + "loss": 1.0981, + "mean_token_accuracy": 0.6845390200614929, + "num_tokens": 13701339.0, + "step": 528 + }, + { + "epoch": 0.05809356468262684, + "ewc_loss": 7.972121238708496e-07, + "grad_norm": 4.149324893951416, + "learning_rate": 1.9326500732064422e-07, + "loss": 1.1804, + "mean_token_accuracy": 0.664412260055542, + "num_tokens": 13724927.0, + "step": 529 + }, + { + "epoch": 0.058203382385240504, + "ewc_loss": 8.046627044677734e-07, + "grad_norm": 4.309876918792725, + "learning_rate": 1.9363103953147877e-07, + "loss": 1.0859, + "mean_token_accuracy": 0.6866269707679749, + "num_tokens": 13745791.0, + "step": 530 + }, + { + "epoch": 0.058313200087854164, + "ewc_loss": 8.121132850646973e-07, + "grad_norm": 4.491830825805664, + "learning_rate": 1.9399707174231332e-07, + "loss": 1.105, + "mean_token_accuracy": 0.677643895149231, + "num_tokens": 13770891.0, + "step": 531 + }, + { + "epoch": 0.058423017790467824, + "ewc_loss": 8.083879947662354e-07, + "grad_norm": 4.207163333892822, + "learning_rate": 1.9436310395314787e-07, + "loss": 1.1291, + "mean_token_accuracy": 0.6674941778182983, + "num_tokens": 13790625.0, + "step": 532 + }, + { + "epoch": 0.058532835493081484, + "ewc_loss": 8.121132850646973e-07, + "grad_norm": 3.801628589630127, + "learning_rate": 1.9472913616398242e-07, + "loss": 1.0955, + "mean_token_accuracy": 0.6795531511306763, + "num_tokens": 13813346.0, + "step": 533 + }, + { + "epoch": 0.058642653195695144, + "ewc_loss": 8.121132850646973e-07, + "grad_norm": 3.8244943618774414, + "learning_rate": 1.95095168374817e-07, + "loss": 1.2095, + "mean_token_accuracy": 0.6544665098190308, + "num_tokens": 13837250.0, + "step": 534 + }, + { + "epoch": 0.05875247089830881, + "ewc_loss": 8.270144462585449e-07, + "grad_norm": 5.140410900115967, + "learning_rate": 1.9546120058565151e-07, + "loss": 1.1346, + "mean_token_accuracy": 0.6773539781570435, + "num_tokens": 13865730.0, + "step": 535 + }, + { + "epoch": 0.05886228860092247, + "ewc_loss": 8.195638656616211e-07, + "grad_norm": 4.37430477142334, + "learning_rate": 1.958272327964861e-07, + "loss": 1.1216, + "mean_token_accuracy": 0.674970805644989, + "num_tokens": 13893215.0, + "step": 536 + }, + { + "epoch": 0.05897210630353613, + "ewc_loss": 8.270144462585449e-07, + "grad_norm": 3.786372184753418, + "learning_rate": 1.9619326500732064e-07, + "loss": 1.1702, + "mean_token_accuracy": 0.6644796133041382, + "num_tokens": 13919232.0, + "step": 537 + }, + { + "epoch": 0.05908192400614979, + "ewc_loss": 8.307397365570068e-07, + "grad_norm": 3.3816471099853516, + "learning_rate": 1.9655929721815521e-07, + "loss": 1.1779, + "mean_token_accuracy": 0.6633831262588501, + "num_tokens": 13943661.0, + "step": 538 + }, + { + "epoch": 0.05919174170876345, + "ewc_loss": 8.307397365570068e-07, + "grad_norm": 4.176119804382324, + "learning_rate": 1.9692532942898974e-07, + "loss": 1.1811, + "mean_token_accuracy": 0.6578341126441956, + "num_tokens": 13967209.0, + "step": 539 + }, + { + "epoch": 0.05930155941137712, + "ewc_loss": 8.307397365570068e-07, + "grad_norm": 4.122870445251465, + "learning_rate": 1.9729136163982429e-07, + "loss": 1.045, + "mean_token_accuracy": 0.6950101256370544, + "num_tokens": 13987948.0, + "step": 540 + }, + { + "epoch": 0.059411377113990776, + "ewc_loss": 8.344650268554688e-07, + "grad_norm": 3.5263330936431885, + "learning_rate": 1.9765739385065886e-07, + "loss": 1.2103, + "mean_token_accuracy": 0.6586334109306335, + "num_tokens": 14016862.0, + "step": 541 + }, + { + "epoch": 0.059521194816604436, + "ewc_loss": 8.381903171539307e-07, + "grad_norm": 3.609049081802368, + "learning_rate": 1.980234260614934e-07, + "loss": 1.1937, + "mean_token_accuracy": 0.6631184816360474, + "num_tokens": 14041242.0, + "step": 542 + }, + { + "epoch": 0.059631012519218096, + "ewc_loss": 8.344650268554688e-07, + "grad_norm": 3.575523614883423, + "learning_rate": 1.9838945827232796e-07, + "loss": 1.0959, + "mean_token_accuracy": 0.6845457553863525, + "num_tokens": 14064732.0, + "step": 543 + }, + { + "epoch": 0.059740830221831756, + "ewc_loss": 8.419156074523926e-07, + "grad_norm": 3.4827332496643066, + "learning_rate": 1.987554904831625e-07, + "loss": 1.0719, + "mean_token_accuracy": 0.6857932806015015, + "num_tokens": 14090186.0, + "step": 544 + }, + { + "epoch": 0.05985064792444542, + "ewc_loss": 8.419156074523926e-07, + "grad_norm": 3.1926968097686768, + "learning_rate": 1.9912152269399706e-07, + "loss": 1.2224, + "mean_token_accuracy": 0.6542167663574219, + "num_tokens": 14121747.0, + "step": 545 + }, + { + "epoch": 0.05996046562705908, + "ewc_loss": 8.419156074523926e-07, + "grad_norm": 3.1918673515319824, + "learning_rate": 1.9948755490483163e-07, + "loss": 1.2332, + "mean_token_accuracy": 0.6501064300537109, + "num_tokens": 14152053.0, + "step": 546 + }, + { + "epoch": 0.06007028332967274, + "ewc_loss": 8.456408977508545e-07, + "grad_norm": 4.526571750640869, + "learning_rate": 1.9985358711566616e-07, + "loss": 1.1331, + "mean_token_accuracy": 0.679854691028595, + "num_tokens": 14170890.0, + "step": 547 + }, + { + "epoch": 0.0601801010322864, + "ewc_loss": 8.456408977508545e-07, + "grad_norm": 3.0111324787139893, + "learning_rate": 2.0021961932650073e-07, + "loss": 1.1211, + "mean_token_accuracy": 0.6761049628257751, + "num_tokens": 14200001.0, + "step": 548 + }, + { + "epoch": 0.06028991873490007, + "ewc_loss": 8.456408977508545e-07, + "grad_norm": 3.3615975379943848, + "learning_rate": 2.0058565153733528e-07, + "loss": 1.1613, + "mean_token_accuracy": 0.6696596145629883, + "num_tokens": 14225062.0, + "step": 549 + }, + { + "epoch": 0.06039973643751373, + "ewc_loss": 8.493661880493164e-07, + "grad_norm": 3.6655941009521484, + "learning_rate": 2.0095168374816983e-07, + "loss": 1.1328, + "mean_token_accuracy": 0.6711333990097046, + "num_tokens": 14248169.0, + "step": 550 + }, + { + "epoch": 0.06050955414012739, + "ewc_loss": 8.568167686462402e-07, + "grad_norm": 3.3448023796081543, + "learning_rate": 2.0131771595900438e-07, + "loss": 1.1535, + "mean_token_accuracy": 0.6750099658966064, + "num_tokens": 14273309.0, + "step": 551 + }, + { + "epoch": 0.06061937184274105, + "ewc_loss": 8.642673492431641e-07, + "grad_norm": 3.5725252628326416, + "learning_rate": 2.0168374816983893e-07, + "loss": 1.1135, + "mean_token_accuracy": 0.6804121732711792, + "num_tokens": 14300494.0, + "step": 552 + }, + { + "epoch": 0.06072918954535471, + "ewc_loss": 8.642673492431641e-07, + "grad_norm": 3.0936014652252197, + "learning_rate": 2.020497803806735e-07, + "loss": 1.1479, + "mean_token_accuracy": 0.664486289024353, + "num_tokens": 14327576.0, + "step": 553 + }, + { + "epoch": 0.060839007247968376, + "ewc_loss": 8.605420589447021e-07, + "grad_norm": 2.9953174591064453, + "learning_rate": 2.0241581259150805e-07, + "loss": 1.1717, + "mean_token_accuracy": 0.659480094909668, + "num_tokens": 14359413.0, + "step": 554 + }, + { + "epoch": 0.060948824950582035, + "ewc_loss": 8.67992639541626e-07, + "grad_norm": 3.4672257900238037, + "learning_rate": 2.027818448023426e-07, + "loss": 1.2904, + "mean_token_accuracy": 0.6486464738845825, + "num_tokens": 14389772.0, + "step": 555 + }, + { + "epoch": 0.061058642653195695, + "ewc_loss": 8.67992639541626e-07, + "grad_norm": 3.4703903198242188, + "learning_rate": 2.0314787701317715e-07, + "loss": 1.2056, + "mean_token_accuracy": 0.6611325740814209, + "num_tokens": 14416775.0, + "step": 556 + }, + { + "epoch": 0.061168460355809355, + "ewc_loss": 8.67992639541626e-07, + "grad_norm": 3.688498020172119, + "learning_rate": 2.035139092240117e-07, + "loss": 1.2461, + "mean_token_accuracy": 0.6422016620635986, + "num_tokens": 14445547.0, + "step": 557 + }, + { + "epoch": 0.061278278058423015, + "ewc_loss": 8.67992639541626e-07, + "grad_norm": 3.8548717498779297, + "learning_rate": 2.0387994143484627e-07, + "loss": 1.2174, + "mean_token_accuracy": 0.6497800350189209, + "num_tokens": 14469951.0, + "step": 558 + }, + { + "epoch": 0.06138809576103668, + "ewc_loss": 8.717179298400879e-07, + "grad_norm": 3.5884599685668945, + "learning_rate": 2.042459736456808e-07, + "loss": 1.1772, + "mean_token_accuracy": 0.6647446751594543, + "num_tokens": 14492322.0, + "step": 559 + }, + { + "epoch": 0.06149791346365034, + "ewc_loss": 8.791685104370117e-07, + "grad_norm": 3.252530574798584, + "learning_rate": 2.0461200585651537e-07, + "loss": 1.1685, + "mean_token_accuracy": 0.6667274236679077, + "num_tokens": 14517881.0, + "step": 560 + }, + { + "epoch": 0.061607731166264, + "ewc_loss": 8.828938007354736e-07, + "grad_norm": 4.205556869506836, + "learning_rate": 2.0497803806734992e-07, + "loss": 1.1987, + "mean_token_accuracy": 0.664504885673523, + "num_tokens": 14543018.0, + "step": 561 + }, + { + "epoch": 0.06171754886887766, + "ewc_loss": 8.866190910339355e-07, + "grad_norm": 4.039614200592041, + "learning_rate": 2.0534407027818447e-07, + "loss": 1.1792, + "mean_token_accuracy": 0.6593348979949951, + "num_tokens": 14570418.0, + "step": 562 + }, + { + "epoch": 0.06182736657149132, + "ewc_loss": 8.903443813323975e-07, + "grad_norm": 4.091158390045166, + "learning_rate": 2.0571010248901902e-07, + "loss": 1.0352, + "mean_token_accuracy": 0.6946356892585754, + "num_tokens": 14589713.0, + "step": 563 + }, + { + "epoch": 0.06193718427410499, + "ewc_loss": 8.940696716308594e-07, + "grad_norm": 3.4364800453186035, + "learning_rate": 2.0607613469985357e-07, + "loss": 1.1784, + "mean_token_accuracy": 0.6610260605812073, + "num_tokens": 14617418.0, + "step": 564 + }, + { + "epoch": 0.06204700197671865, + "ewc_loss": 8.977949619293213e-07, + "grad_norm": 3.6696090698242188, + "learning_rate": 2.0644216691068814e-07, + "loss": 1.1501, + "mean_token_accuracy": 0.6752034425735474, + "num_tokens": 14645074.0, + "step": 565 + }, + { + "epoch": 0.06215681967933231, + "ewc_loss": 9.08970832824707e-07, + "grad_norm": 3.8980164527893066, + "learning_rate": 2.068081991215227e-07, + "loss": 1.2235, + "mean_token_accuracy": 0.6533102989196777, + "num_tokens": 14671552.0, + "step": 566 + }, + { + "epoch": 0.06226663738194597, + "ewc_loss": 9.126961231231689e-07, + "grad_norm": 3.9027578830718994, + "learning_rate": 2.0717423133235724e-07, + "loss": 1.1449, + "mean_token_accuracy": 0.6717804074287415, + "num_tokens": 14694223.0, + "step": 567 + }, + { + "epoch": 0.062376455084559634, + "ewc_loss": 9.164214134216309e-07, + "grad_norm": 3.8777174949645996, + "learning_rate": 2.075402635431918e-07, + "loss": 1.0919, + "mean_token_accuracy": 0.6893825531005859, + "num_tokens": 14719184.0, + "step": 568 + }, + { + "epoch": 0.062486272787173294, + "ewc_loss": 9.126961231231689e-07, + "grad_norm": 3.228668689727783, + "learning_rate": 2.0790629575402634e-07, + "loss": 1.1505, + "mean_token_accuracy": 0.6713299751281738, + "num_tokens": 14742235.0, + "step": 569 + }, + { + "epoch": 0.06259609048978695, + "ewc_loss": 9.126961231231689e-07, + "grad_norm": 3.5129666328430176, + "learning_rate": 2.0827232796486092e-07, + "loss": 1.2284, + "mean_token_accuracy": 0.6517795324325562, + "num_tokens": 14769348.0, + "step": 570 + }, + { + "epoch": 0.06270590819240061, + "ewc_loss": 9.126961231231689e-07, + "grad_norm": 3.5274462699890137, + "learning_rate": 2.0863836017569544e-07, + "loss": 1.0587, + "mean_token_accuracy": 0.6901646852493286, + "num_tokens": 14794555.0, + "step": 571 + }, + { + "epoch": 0.06281572589501427, + "ewc_loss": 9.126961231231689e-07, + "grad_norm": 3.4053773880004883, + "learning_rate": 2.0900439238653001e-07, + "loss": 1.1494, + "mean_token_accuracy": 0.668853759765625, + "num_tokens": 14817364.0, + "step": 572 + }, + { + "epoch": 0.06292554359762793, + "ewc_loss": 9.08970832824707e-07, + "grad_norm": 3.7347471714019775, + "learning_rate": 2.0937042459736456e-07, + "loss": 1.1819, + "mean_token_accuracy": 0.6667755246162415, + "num_tokens": 14837472.0, + "step": 573 + }, + { + "epoch": 0.0630353613002416, + "ewc_loss": 9.08970832824707e-07, + "grad_norm": 3.0365424156188965, + "learning_rate": 2.097364568081991e-07, + "loss": 1.124, + "mean_token_accuracy": 0.6783350706100464, + "num_tokens": 14865112.0, + "step": 574 + }, + { + "epoch": 0.06314517900285525, + "ewc_loss": 9.126961231231689e-07, + "grad_norm": 3.2981247901916504, + "learning_rate": 2.1010248901903366e-07, + "loss": 1.1601, + "mean_token_accuracy": 0.662319540977478, + "num_tokens": 14894808.0, + "step": 575 + }, + { + "epoch": 0.06325499670546893, + "ewc_loss": 9.201467037200928e-07, + "grad_norm": 4.0308518409729, + "learning_rate": 2.104685212298682e-07, + "loss": 1.0517, + "mean_token_accuracy": 0.6949278116226196, + "num_tokens": 14913195.0, + "step": 576 + }, + { + "epoch": 0.06336481440808259, + "ewc_loss": 9.164214134216309e-07, + "grad_norm": 2.927692413330078, + "learning_rate": 2.1083455344070279e-07, + "loss": 1.1971, + "mean_token_accuracy": 0.6600308418273926, + "num_tokens": 14943929.0, + "step": 577 + }, + { + "epoch": 0.06347463211069625, + "ewc_loss": 9.201467037200928e-07, + "grad_norm": 3.3568813800811768, + "learning_rate": 2.1120058565153734e-07, + "loss": 1.1789, + "mean_token_accuracy": 0.6646564602851868, + "num_tokens": 14969696.0, + "step": 578 + }, + { + "epoch": 0.0635844498133099, + "ewc_loss": 9.313225746154785e-07, + "grad_norm": 3.8800039291381836, + "learning_rate": 2.1156661786237186e-07, + "loss": 1.0203, + "mean_token_accuracy": 0.6963807344436646, + "num_tokens": 14989320.0, + "step": 579 + }, + { + "epoch": 0.06369426751592357, + "ewc_loss": 9.313225746154785e-07, + "grad_norm": 3.214963674545288, + "learning_rate": 2.1193265007320643e-07, + "loss": 1.0858, + "mean_token_accuracy": 0.6836994886398315, + "num_tokens": 15014145.0, + "step": 580 + }, + { + "epoch": 0.06380408521853723, + "ewc_loss": 9.313225746154785e-07, + "grad_norm": 2.942582368850708, + "learning_rate": 2.1229868228404098e-07, + "loss": 1.1651, + "mean_token_accuracy": 0.6619845628738403, + "num_tokens": 15040827.0, + "step": 581 + }, + { + "epoch": 0.06391390292115089, + "ewc_loss": 9.313225746154785e-07, + "grad_norm": 3.132331132888794, + "learning_rate": 2.1266471449487556e-07, + "loss": 1.214, + "mean_token_accuracy": 0.6574066281318665, + "num_tokens": 15069707.0, + "step": 582 + }, + { + "epoch": 0.06402372062376455, + "ewc_loss": 9.313225746154785e-07, + "grad_norm": 3.6988697052001953, + "learning_rate": 2.1303074670571008e-07, + "loss": 1.1513, + "mean_token_accuracy": 0.6671475172042847, + "num_tokens": 15091567.0, + "step": 583 + }, + { + "epoch": 0.0641335383263782, + "ewc_loss": 9.313225746154785e-07, + "grad_norm": 3.6310486793518066, + "learning_rate": 2.1339677891654466e-07, + "loss": 1.2259, + "mean_token_accuracy": 0.6526857614517212, + "num_tokens": 15117176.0, + "step": 584 + }, + { + "epoch": 0.06424335602899188, + "ewc_loss": 9.387731552124023e-07, + "grad_norm": 2.8107481002807617, + "learning_rate": 2.137628111273792e-07, + "loss": 1.1693, + "mean_token_accuracy": 0.6677898168563843, + "num_tokens": 15151656.0, + "step": 585 + }, + { + "epoch": 0.06435317373160554, + "ewc_loss": 9.387731552124023e-07, + "grad_norm": 3.208517551422119, + "learning_rate": 2.1412884333821375e-07, + "loss": 1.0711, + "mean_token_accuracy": 0.6941730976104736, + "num_tokens": 15176031.0, + "step": 586 + }, + { + "epoch": 0.0644629914342192, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 4.068676948547363, + "learning_rate": 2.144948755490483e-07, + "loss": 1.1285, + "mean_token_accuracy": 0.6691128611564636, + "num_tokens": 15197744.0, + "step": 587 + }, + { + "epoch": 0.06457280913683286, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 4.010196685791016, + "learning_rate": 2.1486090775988285e-07, + "loss": 1.1077, + "mean_token_accuracy": 0.6799461841583252, + "num_tokens": 15219377.0, + "step": 588 + }, + { + "epoch": 0.06468262683944652, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 3.1165976524353027, + "learning_rate": 2.1522693997071743e-07, + "loss": 1.1214, + "mean_token_accuracy": 0.6749182939529419, + "num_tokens": 15244196.0, + "step": 589 + }, + { + "epoch": 0.06479244454206018, + "ewc_loss": 9.462237358093262e-07, + "grad_norm": 3.2371773719787598, + "learning_rate": 2.1559297218155198e-07, + "loss": 1.1736, + "mean_token_accuracy": 0.6742652058601379, + "num_tokens": 15273093.0, + "step": 590 + }, + { + "epoch": 0.06490226224467384, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 3.185307502746582, + "learning_rate": 2.159590043923865e-07, + "loss": 1.1296, + "mean_token_accuracy": 0.6762865781784058, + "num_tokens": 15299076.0, + "step": 591 + }, + { + "epoch": 0.0650120799472875, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 3.4077751636505127, + "learning_rate": 2.1632503660322108e-07, + "loss": 1.0487, + "mean_token_accuracy": 0.7001065015792847, + "num_tokens": 15323632.0, + "step": 592 + }, + { + "epoch": 0.06512189764990116, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 2.853393077850342, + "learning_rate": 2.1669106881405563e-07, + "loss": 1.1458, + "mean_token_accuracy": 0.6713336706161499, + "num_tokens": 15356700.0, + "step": 593 + }, + { + "epoch": 0.06523171535251482, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 2.9033210277557373, + "learning_rate": 2.170571010248902e-07, + "loss": 1.0532, + "mean_token_accuracy": 0.6974869966506958, + "num_tokens": 15382486.0, + "step": 594 + }, + { + "epoch": 0.06534153305512849, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 3.934795379638672, + "learning_rate": 2.1742313323572472e-07, + "loss": 1.0797, + "mean_token_accuracy": 0.6904190182685852, + "num_tokens": 15403037.0, + "step": 595 + }, + { + "epoch": 0.06545135075774215, + "ewc_loss": 9.424984455108643e-07, + "grad_norm": 3.6372315883636475, + "learning_rate": 2.177891654465593e-07, + "loss": 1.075, + "mean_token_accuracy": 0.6894652843475342, + "num_tokens": 15427884.0, + "step": 596 + }, + { + "epoch": 0.06556116846035581, + "ewc_loss": 9.499490261077881e-07, + "grad_norm": 2.744227647781372, + "learning_rate": 2.1815519765739385e-07, + "loss": 1.1482, + "mean_token_accuracy": 0.6659747362136841, + "num_tokens": 15461669.0, + "step": 597 + }, + { + "epoch": 0.06567098616296947, + "ewc_loss": 9.5367431640625e-07, + "grad_norm": 3.7888569831848145, + "learning_rate": 2.185212298682284e-07, + "loss": 1.0902, + "mean_token_accuracy": 0.6784982085227966, + "num_tokens": 15484624.0, + "step": 598 + }, + { + "epoch": 0.06578080386558313, + "ewc_loss": 9.5367431640625e-07, + "grad_norm": 3.100031852722168, + "learning_rate": 2.1888726207906295e-07, + "loss": 1.1862, + "mean_token_accuracy": 0.6694333553314209, + "num_tokens": 15508390.0, + "step": 599 + }, + { + "epoch": 0.06589062156819679, + "ewc_loss": 9.5367431640625e-07, + "grad_norm": 3.103928327560425, + "learning_rate": 2.192532942898975e-07, + "loss": 1.1122, + "mean_token_accuracy": 0.6787245273590088, + "num_tokens": 15535744.0, + "step": 600 + }, + { + "epoch": 0.06600043927081045, + "ewc_loss": 9.5367431640625e-07, + "grad_norm": 3.051050901412964, + "learning_rate": 2.1961932650073207e-07, + "loss": 1.1163, + "mean_token_accuracy": 0.6775130033493042, + "num_tokens": 15565036.0, + "step": 601 + }, + { + "epoch": 0.06611025697342411, + "ewc_loss": 9.685754776000977e-07, + "grad_norm": 2.583517551422119, + "learning_rate": 2.1998535871156662e-07, + "loss": 1.21, + "mean_token_accuracy": 0.6567202806472778, + "num_tokens": 15597806.0, + "step": 602 + }, + { + "epoch": 0.06622007467603777, + "ewc_loss": 9.685754776000977e-07, + "grad_norm": 3.3374626636505127, + "learning_rate": 2.2035139092240114e-07, + "loss": 1.1239, + "mean_token_accuracy": 0.6722239255905151, + "num_tokens": 15623941.0, + "step": 603 + }, + { + "epoch": 0.06632989237865144, + "ewc_loss": 9.685754776000977e-07, + "grad_norm": 3.299060583114624, + "learning_rate": 2.2071742313323572e-07, + "loss": 1.1635, + "mean_token_accuracy": 0.6616415977478027, + "num_tokens": 15649843.0, + "step": 604 + }, + { + "epoch": 0.0664397100812651, + "ewc_loss": 9.760260581970215e-07, + "grad_norm": 2.9305593967437744, + "learning_rate": 2.2108345534407027e-07, + "loss": 1.0575, + "mean_token_accuracy": 0.6887045502662659, + "num_tokens": 15677720.0, + "step": 605 + }, + { + "epoch": 0.06654952778387876, + "ewc_loss": 9.834766387939453e-07, + "grad_norm": 3.9448764324188232, + "learning_rate": 2.2144948755490484e-07, + "loss": 1.067, + "mean_token_accuracy": 0.6870230436325073, + "num_tokens": 15704251.0, + "step": 606 + }, + { + "epoch": 0.06665934548649242, + "ewc_loss": 9.760260581970215e-07, + "grad_norm": 3.411473274230957, + "learning_rate": 2.2181551976573937e-07, + "loss": 1.1464, + "mean_token_accuracy": 0.6668318510055542, + "num_tokens": 15729100.0, + "step": 607 + }, + { + "epoch": 0.06676916318910608, + "ewc_loss": 9.760260581970215e-07, + "grad_norm": 3.322129249572754, + "learning_rate": 2.2218155197657394e-07, + "loss": 1.1727, + "mean_token_accuracy": 0.6555952429771423, + "num_tokens": 15753408.0, + "step": 608 + }, + { + "epoch": 0.06687898089171974, + "ewc_loss": 9.909272193908691e-07, + "grad_norm": 3.754793882369995, + "learning_rate": 2.225475841874085e-07, + "loss": 1.087, + "mean_token_accuracy": 0.6894596219062805, + "num_tokens": 15777400.0, + "step": 609 + }, + { + "epoch": 0.0669887985943334, + "ewc_loss": 9.98377799987793e-07, + "grad_norm": 3.6610186100006104, + "learning_rate": 2.2291361639824304e-07, + "loss": 1.201, + "mean_token_accuracy": 0.6640112400054932, + "num_tokens": 15808367.0, + "step": 610 + }, + { + "epoch": 0.06709861629694706, + "ewc_loss": 9.909272193908691e-07, + "grad_norm": 3.5395538806915283, + "learning_rate": 2.232796486090776e-07, + "loss": 1.0303, + "mean_token_accuracy": 0.7064948081970215, + "num_tokens": 15831105.0, + "step": 611 + }, + { + "epoch": 0.06720843399956072, + "ewc_loss": 9.98377799987793e-07, + "grad_norm": 3.3150041103363037, + "learning_rate": 2.2364568081991214e-07, + "loss": 1.139, + "mean_token_accuracy": 0.6719591617584229, + "num_tokens": 15854816.0, + "step": 612 + }, + { + "epoch": 0.06731825170217438, + "ewc_loss": 9.98377799987793e-07, + "grad_norm": 3.657451629638672, + "learning_rate": 2.240117130307467e-07, + "loss": 1.1264, + "mean_token_accuracy": 0.6734627485275269, + "num_tokens": 15873146.0, + "step": 613 + }, + { + "epoch": 0.06742806940478806, + "ewc_loss": 1.0058283805847168e-06, + "grad_norm": 2.904186487197876, + "learning_rate": 2.2437774524158126e-07, + "loss": 1.1752, + "mean_token_accuracy": 0.6684595346450806, + "num_tokens": 15898651.0, + "step": 614 + }, + { + "epoch": 0.06753788710740172, + "ewc_loss": 1.0058283805847168e-06, + "grad_norm": 3.405209541320801, + "learning_rate": 2.2474377745241578e-07, + "loss": 1.1734, + "mean_token_accuracy": 0.656650185585022, + "num_tokens": 15921202.0, + "step": 615 + }, + { + "epoch": 0.06764770481001538, + "ewc_loss": 1.0058283805847168e-06, + "grad_norm": 3.540494918823242, + "learning_rate": 2.2510980966325036e-07, + "loss": 1.0236, + "mean_token_accuracy": 0.6948303580284119, + "num_tokens": 15940488.0, + "step": 616 + }, + { + "epoch": 0.06775752251262904, + "ewc_loss": 1.0132789611816406e-06, + "grad_norm": 3.1251251697540283, + "learning_rate": 2.254758418740849e-07, + "loss": 1.1085, + "mean_token_accuracy": 0.6832973957061768, + "num_tokens": 15967477.0, + "step": 617 + }, + { + "epoch": 0.0678673402152427, + "ewc_loss": 1.0132789611816406e-06, + "grad_norm": 3.089552640914917, + "learning_rate": 2.2584187408491948e-07, + "loss": 1.1239, + "mean_token_accuracy": 0.6742743253707886, + "num_tokens": 15995250.0, + "step": 618 + }, + { + "epoch": 0.06797715791785636, + "ewc_loss": 1.0132789611816406e-06, + "grad_norm": 2.8878979682922363, + "learning_rate": 2.26207906295754e-07, + "loss": 1.2345, + "mean_token_accuracy": 0.6536018252372742, + "num_tokens": 16020883.0, + "step": 619 + }, + { + "epoch": 0.06808697562047002, + "ewc_loss": 1.0132789611816406e-06, + "grad_norm": 2.7887017726898193, + "learning_rate": 2.2657393850658856e-07, + "loss": 1.2489, + "mean_token_accuracy": 0.6498244404792786, + "num_tokens": 16047600.0, + "step": 620 + }, + { + "epoch": 0.06819679332308368, + "ewc_loss": 1.0132789611816406e-06, + "grad_norm": 3.3103251457214355, + "learning_rate": 2.2693997071742313e-07, + "loss": 1.14, + "mean_token_accuracy": 0.670974612236023, + "num_tokens": 16070115.0, + "step": 621 + }, + { + "epoch": 0.06830661102569734, + "ewc_loss": 1.0132789611816406e-06, + "grad_norm": 2.9765055179595947, + "learning_rate": 2.2730600292825768e-07, + "loss": 1.1705, + "mean_token_accuracy": 0.6677532196044922, + "num_tokens": 16096102.0, + "step": 622 + }, + { + "epoch": 0.06841642872831101, + "ewc_loss": 1.0207295417785645e-06, + "grad_norm": 3.039874792098999, + "learning_rate": 2.2767203513909223e-07, + "loss": 1.2055, + "mean_token_accuracy": 0.660908579826355, + "num_tokens": 16122671.0, + "step": 623 + }, + { + "epoch": 0.06852624643092467, + "ewc_loss": 1.0207295417785645e-06, + "grad_norm": 3.0997161865234375, + "learning_rate": 2.2803806734992678e-07, + "loss": 1.0557, + "mean_token_accuracy": 0.6864209175109863, + "num_tokens": 16149145.0, + "step": 624 + }, + { + "epoch": 0.06863606413353833, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.8139874935150146, + "learning_rate": 2.2840409956076135e-07, + "loss": 1.1397, + "mean_token_accuracy": 0.6818225979804993, + "num_tokens": 16178336.0, + "step": 625 + }, + { + "epoch": 0.06874588183615199, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 3.2129502296447754, + "learning_rate": 2.287701317715959e-07, + "loss": 1.1423, + "mean_token_accuracy": 0.6739550828933716, + "num_tokens": 16200808.0, + "step": 626 + }, + { + "epoch": 0.06885569953876565, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.760450601577759, + "learning_rate": 2.2913616398243043e-07, + "loss": 1.2626, + "mean_token_accuracy": 0.6424587965011597, + "num_tokens": 16228780.0, + "step": 627 + }, + { + "epoch": 0.06896551724137931, + "ewc_loss": 1.0207295417785645e-06, + "grad_norm": 3.012667655944824, + "learning_rate": 2.29502196193265e-07, + "loss": 1.197, + "mean_token_accuracy": 0.6514236927032471, + "num_tokens": 16256783.0, + "step": 628 + }, + { + "epoch": 0.06907533494399297, + "ewc_loss": 1.0207295417785645e-06, + "grad_norm": 3.049149513244629, + "learning_rate": 2.2986822840409955e-07, + "loss": 1.1716, + "mean_token_accuracy": 0.6583991050720215, + "num_tokens": 16281018.0, + "step": 629 + }, + { + "epoch": 0.06918515264660663, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 3.0331060886383057, + "learning_rate": 2.3023426061493413e-07, + "loss": 1.0668, + "mean_token_accuracy": 0.6886460781097412, + "num_tokens": 16306575.0, + "step": 630 + }, + { + "epoch": 0.06929497034922029, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.7485828399658203, + "learning_rate": 2.3060029282576865e-07, + "loss": 1.2046, + "mean_token_accuracy": 0.6558257341384888, + "num_tokens": 16337454.0, + "step": 631 + }, + { + "epoch": 0.06940478805183395, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 3.258399486541748, + "learning_rate": 2.309663250366032e-07, + "loss": 1.1416, + "mean_token_accuracy": 0.6655158996582031, + "num_tokens": 16360648.0, + "step": 632 + }, + { + "epoch": 0.06951460575444762, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.7550101280212402, + "learning_rate": 2.3133235724743777e-07, + "loss": 1.1392, + "mean_token_accuracy": 0.6687778234481812, + "num_tokens": 16389325.0, + "step": 633 + }, + { + "epoch": 0.06962442345706128, + "ewc_loss": 1.0207295417785645e-06, + "grad_norm": 3.296785831451416, + "learning_rate": 2.3169838945827232e-07, + "loss": 1.1714, + "mean_token_accuracy": 0.6662108898162842, + "num_tokens": 16412500.0, + "step": 634 + }, + { + "epoch": 0.06973424115967494, + "ewc_loss": 1.0207295417785645e-06, + "grad_norm": 2.4462738037109375, + "learning_rate": 2.3206442166910687e-07, + "loss": 1.2185, + "mean_token_accuracy": 0.646491527557373, + "num_tokens": 16449528.0, + "step": 635 + }, + { + "epoch": 0.0698440588622886, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.7883780002593994, + "learning_rate": 2.3243045387994142e-07, + "loss": 1.0364, + "mean_token_accuracy": 0.7057567238807678, + "num_tokens": 16475464.0, + "step": 636 + }, + { + "epoch": 0.06995387656490226, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.830719470977783, + "learning_rate": 2.32796486090776e-07, + "loss": 1.1286, + "mean_token_accuracy": 0.6735982298851013, + "num_tokens": 16503191.0, + "step": 637 + }, + { + "epoch": 0.07006369426751592, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.777308464050293, + "learning_rate": 2.3316251830161055e-07, + "loss": 1.1529, + "mean_token_accuracy": 0.6635075807571411, + "num_tokens": 16532650.0, + "step": 638 + }, + { + "epoch": 0.07017351197012958, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.7725026607513428, + "learning_rate": 2.3352855051244507e-07, + "loss": 1.2248, + "mean_token_accuracy": 0.6588432788848877, + "num_tokens": 16560255.0, + "step": 639 + }, + { + "epoch": 0.07028332967274324, + "ewc_loss": 1.0281801223754883e-06, + "grad_norm": 2.94869065284729, + "learning_rate": 2.3389458272327964e-07, + "loss": 1.1817, + "mean_token_accuracy": 0.6579218506813049, + "num_tokens": 16591859.0, + "step": 640 + }, + { + "epoch": 0.0703931473753569, + "ewc_loss": 1.0356307029724121e-06, + "grad_norm": 2.7798378467559814, + "learning_rate": 2.342606149341142e-07, + "loss": 1.1739, + "mean_token_accuracy": 0.6651117205619812, + "num_tokens": 16620782.0, + "step": 641 + }, + { + "epoch": 0.07050296507797058, + "ewc_loss": 1.0356307029724121e-06, + "grad_norm": 3.0157811641693115, + "learning_rate": 2.3462664714494877e-07, + "loss": 1.1814, + "mean_token_accuracy": 0.654600977897644, + "num_tokens": 16648051.0, + "step": 642 + }, + { + "epoch": 0.07061278278058424, + "ewc_loss": 1.0356307029724121e-06, + "grad_norm": 2.7533814907073975, + "learning_rate": 2.349926793557833e-07, + "loss": 1.2174, + "mean_token_accuracy": 0.6567906141281128, + "num_tokens": 16677294.0, + "step": 643 + }, + { + "epoch": 0.0707226004831979, + "ewc_loss": 1.0356307029724121e-06, + "grad_norm": 2.6964685916900635, + "learning_rate": 2.3535871156661784e-07, + "loss": 1.2508, + "mean_token_accuracy": 0.6456199884414673, + "num_tokens": 16707122.0, + "step": 644 + }, + { + "epoch": 0.07083241818581155, + "ewc_loss": 1.0505318641662598e-06, + "grad_norm": 3.28334641456604, + "learning_rate": 2.3572474377745242e-07, + "loss": 1.1535, + "mean_token_accuracy": 0.6641832590103149, + "num_tokens": 16735039.0, + "step": 645 + }, + { + "epoch": 0.07094223588842521, + "ewc_loss": 1.0356307029724121e-06, + "grad_norm": 3.321802854537964, + "learning_rate": 2.3609077598828696e-07, + "loss": 1.0955, + "mean_token_accuracy": 0.6858232021331787, + "num_tokens": 16759433.0, + "step": 646 + }, + { + "epoch": 0.07105205359103887, + "ewc_loss": 1.043081283569336e-06, + "grad_norm": 2.9402098655700684, + "learning_rate": 2.3645680819912151e-07, + "loss": 1.1038, + "mean_token_accuracy": 0.6768377423286438, + "num_tokens": 16785559.0, + "step": 647 + }, + { + "epoch": 0.07116187129365253, + "ewc_loss": 1.043081283569336e-06, + "grad_norm": 3.4432880878448486, + "learning_rate": 2.3682284040995606e-07, + "loss": 1.0879, + "mean_token_accuracy": 0.6834025382995605, + "num_tokens": 16805432.0, + "step": 648 + }, + { + "epoch": 0.0712716889962662, + "ewc_loss": 1.043081283569336e-06, + "grad_norm": 3.1155903339385986, + "learning_rate": 2.3718887262079064e-07, + "loss": 1.1102, + "mean_token_accuracy": 0.6787206530570984, + "num_tokens": 16828225.0, + "step": 649 + }, + { + "epoch": 0.07138150669887985, + "ewc_loss": 1.0505318641662598e-06, + "grad_norm": 2.996037721633911, + "learning_rate": 2.375549048316252e-07, + "loss": 1.1483, + "mean_token_accuracy": 0.6746612191200256, + "num_tokens": 16853283.0, + "step": 650 + }, + { + "epoch": 0.07149132440149351, + "ewc_loss": 1.043081283569336e-06, + "grad_norm": 3.0480902194976807, + "learning_rate": 2.379209370424597e-07, + "loss": 1.1652, + "mean_token_accuracy": 0.6654667258262634, + "num_tokens": 16880661.0, + "step": 651 + }, + { + "epoch": 0.07160114210410719, + "ewc_loss": 1.043081283569336e-06, + "grad_norm": 2.9784514904022217, + "learning_rate": 2.3828696925329429e-07, + "loss": 1.204, + "mean_token_accuracy": 0.6546894311904907, + "num_tokens": 16905765.0, + "step": 652 + }, + { + "epoch": 0.07171095980672085, + "ewc_loss": 1.0505318641662598e-06, + "grad_norm": 3.298203706741333, + "learning_rate": 2.3865300146412886e-07, + "loss": 1.1365, + "mean_token_accuracy": 0.6730803847312927, + "num_tokens": 16926151.0, + "step": 653 + }, + { + "epoch": 0.07182077750933451, + "ewc_loss": 1.0505318641662598e-06, + "grad_norm": 2.784052610397339, + "learning_rate": 2.390190336749634e-07, + "loss": 1.159, + "mean_token_accuracy": 0.6673016548156738, + "num_tokens": 16955098.0, + "step": 654 + }, + { + "epoch": 0.07193059521194817, + "ewc_loss": 1.0505318641662598e-06, + "grad_norm": 3.439918041229248, + "learning_rate": 2.393850658857979e-07, + "loss": 1.0881, + "mean_token_accuracy": 0.6834281086921692, + "num_tokens": 16975049.0, + "step": 655 + }, + { + "epoch": 0.07204041291456183, + "ewc_loss": 1.0505318641662598e-06, + "grad_norm": 3.6733217239379883, + "learning_rate": 2.397510980966325e-07, + "loss": 1.0127, + "mean_token_accuracy": 0.7013222575187683, + "num_tokens": 16993119.0, + "step": 656 + }, + { + "epoch": 0.07215023061717549, + "ewc_loss": 1.0579824447631836e-06, + "grad_norm": 3.0724568367004395, + "learning_rate": 2.4011713030746706e-07, + "loss": 1.091, + "mean_token_accuracy": 0.6819170713424683, + "num_tokens": 17016760.0, + "step": 657 + }, + { + "epoch": 0.07226004831978915, + "ewc_loss": 1.0579824447631836e-06, + "grad_norm": 2.5138983726501465, + "learning_rate": 2.404831625183016e-07, + "loss": 1.1528, + "mean_token_accuracy": 0.664276123046875, + "num_tokens": 17046841.0, + "step": 658 + }, + { + "epoch": 0.0723698660224028, + "ewc_loss": 1.0728836059570312e-06, + "grad_norm": 2.909092664718628, + "learning_rate": 2.4084919472913616e-07, + "loss": 1.2384, + "mean_token_accuracy": 0.6474459171295166, + "num_tokens": 17073782.0, + "step": 659 + }, + { + "epoch": 0.07247968372501647, + "ewc_loss": 1.0579824447631836e-06, + "grad_norm": 2.871877431869507, + "learning_rate": 2.412152269399707e-07, + "loss": 1.1748, + "mean_token_accuracy": 0.6667231917381287, + "num_tokens": 17105514.0, + "step": 660 + }, + { + "epoch": 0.07258950142763014, + "ewc_loss": 1.0579824447631836e-06, + "grad_norm": 3.004518985748291, + "learning_rate": 2.4158125915080525e-07, + "loss": 1.1175, + "mean_token_accuracy": 0.6756011843681335, + "num_tokens": 17128183.0, + "step": 661 + }, + { + "epoch": 0.0726993191302438, + "ewc_loss": 1.0654330253601074e-06, + "grad_norm": 3.1760413646698, + "learning_rate": 2.419472913616398e-07, + "loss": 1.1352, + "mean_token_accuracy": 0.6689202189445496, + "num_tokens": 17149575.0, + "step": 662 + }, + { + "epoch": 0.07280913683285746, + "ewc_loss": 1.0654330253601074e-06, + "grad_norm": 2.8513026237487793, + "learning_rate": 2.4231332357247435e-07, + "loss": 1.0885, + "mean_token_accuracy": 0.6864286661148071, + "num_tokens": 17175446.0, + "step": 663 + }, + { + "epoch": 0.07291895453547112, + "ewc_loss": 1.0728836059570312e-06, + "grad_norm": 2.6683034896850586, + "learning_rate": 2.426793557833089e-07, + "loss": 1.0887, + "mean_token_accuracy": 0.6827901601791382, + "num_tokens": 17206437.0, + "step": 664 + }, + { + "epoch": 0.07302877223808478, + "ewc_loss": 1.0654330253601074e-06, + "grad_norm": 2.9488933086395264, + "learning_rate": 2.430453879941435e-07, + "loss": 1.1913, + "mean_token_accuracy": 0.6519621014595032, + "num_tokens": 17232640.0, + "step": 665 + }, + { + "epoch": 0.07313858994069844, + "ewc_loss": 1.0728836059570312e-06, + "grad_norm": 2.732487201690674, + "learning_rate": 2.4341142020497805e-07, + "loss": 1.179, + "mean_token_accuracy": 0.6618000268936157, + "num_tokens": 17260421.0, + "step": 666 + }, + { + "epoch": 0.0732484076433121, + "ewc_loss": 1.0728836059570312e-06, + "grad_norm": 2.938994884490967, + "learning_rate": 2.4377745241581255e-07, + "loss": 0.9819, + "mean_token_accuracy": 0.711733341217041, + "num_tokens": 17283253.0, + "step": 667 + }, + { + "epoch": 0.07335822534592576, + "ewc_loss": 1.0728836059570312e-06, + "grad_norm": 3.395470380783081, + "learning_rate": 2.4414348462664715e-07, + "loss": 1.1424, + "mean_token_accuracy": 0.6645792722702026, + "num_tokens": 17303932.0, + "step": 668 + }, + { + "epoch": 0.07346804304853942, + "ewc_loss": 1.080334186553955e-06, + "grad_norm": 3.1520588397979736, + "learning_rate": 2.445095168374817e-07, + "loss": 1.0246, + "mean_token_accuracy": 0.6976009607315063, + "num_tokens": 17324634.0, + "step": 669 + }, + { + "epoch": 0.07357786075115308, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 3.0363271236419678, + "learning_rate": 2.4487554904831625e-07, + "loss": 1.0001, + "mean_token_accuracy": 0.7051469683647156, + "num_tokens": 17347927.0, + "step": 670 + }, + { + "epoch": 0.07368767845376675, + "ewc_loss": 1.080334186553955e-06, + "grad_norm": 2.85443377494812, + "learning_rate": 2.452415812591508e-07, + "loss": 1.0578, + "mean_token_accuracy": 0.6948363780975342, + "num_tokens": 17374596.0, + "step": 671 + }, + { + "epoch": 0.07379749615638041, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 3.2663278579711914, + "learning_rate": 2.4560761346998535e-07, + "loss": 1.0492, + "mean_token_accuracy": 0.6838511824607849, + "num_tokens": 17397546.0, + "step": 672 + }, + { + "epoch": 0.07390731385899407, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 3.091784715652466, + "learning_rate": 2.459736456808199e-07, + "loss": 1.2327, + "mean_token_accuracy": 0.655414879322052, + "num_tokens": 17420746.0, + "step": 673 + }, + { + "epoch": 0.07401713156160773, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 3.159782886505127, + "learning_rate": 2.4633967789165444e-07, + "loss": 1.0852, + "mean_token_accuracy": 0.6902639865875244, + "num_tokens": 17441694.0, + "step": 674 + }, + { + "epoch": 0.07412694926422139, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 2.6994080543518066, + "learning_rate": 2.46705710102489e-07, + "loss": 1.1621, + "mean_token_accuracy": 0.6692777872085571, + "num_tokens": 17468898.0, + "step": 675 + }, + { + "epoch": 0.07423676696683505, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 3.281284809112549, + "learning_rate": 2.4707174231332354e-07, + "loss": 1.0439, + "mean_token_accuracy": 0.6925181150436401, + "num_tokens": 17488143.0, + "step": 676 + }, + { + "epoch": 0.07434658466944871, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 2.9296000003814697, + "learning_rate": 2.4743777452415815e-07, + "loss": 1.0652, + "mean_token_accuracy": 0.6879631280899048, + "num_tokens": 17514042.0, + "step": 677 + }, + { + "epoch": 0.07445640237206237, + "ewc_loss": 1.087784767150879e-06, + "grad_norm": 3.017298936843872, + "learning_rate": 2.478038067349927e-07, + "loss": 1.0528, + "mean_token_accuracy": 0.689277172088623, + "num_tokens": 17538766.0, + "step": 678 + }, + { + "epoch": 0.07456622007467603, + "ewc_loss": 1.0952353477478027e-06, + "grad_norm": 2.5658767223358154, + "learning_rate": 2.481698389458272e-07, + "loss": 1.1673, + "mean_token_accuracy": 0.6791503429412842, + "num_tokens": 17567824.0, + "step": 679 + }, + { + "epoch": 0.0746760377772897, + "ewc_loss": 1.0952353477478027e-06, + "grad_norm": 2.608788013458252, + "learning_rate": 2.485358711566618e-07, + "loss": 1.1901, + "mean_token_accuracy": 0.6602576971054077, + "num_tokens": 17599355.0, + "step": 680 + }, + { + "epoch": 0.07478585547990337, + "ewc_loss": 1.0952353477478027e-06, + "grad_norm": 2.728842258453369, + "learning_rate": 2.4890190336749634e-07, + "loss": 1.1617, + "mean_token_accuracy": 0.66267991065979, + "num_tokens": 17625251.0, + "step": 681 + }, + { + "epoch": 0.07489567318251703, + "ewc_loss": 1.0952353477478027e-06, + "grad_norm": 2.733438491821289, + "learning_rate": 2.492679355783309e-07, + "loss": 1.1599, + "mean_token_accuracy": 0.6642677187919617, + "num_tokens": 17654107.0, + "step": 682 + }, + { + "epoch": 0.07500549088513069, + "ewc_loss": 1.1026859283447266e-06, + "grad_norm": 3.0793731212615967, + "learning_rate": 2.4963396778916544e-07, + "loss": 1.0622, + "mean_token_accuracy": 0.6945348381996155, + "num_tokens": 17677403.0, + "step": 683 + }, + { + "epoch": 0.07511530858774434, + "ewc_loss": 1.1026859283447266e-06, + "grad_norm": 2.630096912384033, + "learning_rate": 2.5e-07, + "loss": 1.1798, + "mean_token_accuracy": 0.6623175144195557, + "num_tokens": 17706274.0, + "step": 684 + }, + { + "epoch": 0.075225126290358, + "ewc_loss": 1.1026859283447266e-06, + "grad_norm": 2.83957576751709, + "learning_rate": 2.5036603221083454e-07, + "loss": 1.1529, + "mean_token_accuracy": 0.658057689666748, + "num_tokens": 17732923.0, + "step": 685 + }, + { + "epoch": 0.07533494399297166, + "ewc_loss": 1.1026859283447266e-06, + "grad_norm": 2.9244980812072754, + "learning_rate": 2.507320644216691e-07, + "loss": 1.0969, + "mean_token_accuracy": 0.6854361891746521, + "num_tokens": 17758057.0, + "step": 686 + }, + { + "epoch": 0.07544476169558532, + "ewc_loss": 1.1101365089416504e-06, + "grad_norm": 2.5548038482666016, + "learning_rate": 2.5109809663250364e-07, + "loss": 1.0962, + "mean_token_accuracy": 0.6787727475166321, + "num_tokens": 17787037.0, + "step": 687 + }, + { + "epoch": 0.07555457939819898, + "ewc_loss": 1.1101365089416504e-06, + "grad_norm": 3.340839147567749, + "learning_rate": 2.514641288433382e-07, + "loss": 1.1523, + "mean_token_accuracy": 0.6711806654930115, + "num_tokens": 17807531.0, + "step": 688 + }, + { + "epoch": 0.07566439710081264, + "ewc_loss": 1.1101365089416504e-06, + "grad_norm": 3.6799066066741943, + "learning_rate": 2.518301610541728e-07, + "loss": 1.043, + "mean_token_accuracy": 0.6904457211494446, + "num_tokens": 17825910.0, + "step": 689 + }, + { + "epoch": 0.07577421480342632, + "ewc_loss": 1.1101365089416504e-06, + "grad_norm": 2.998643636703491, + "learning_rate": 2.521961932650073e-07, + "loss": 1.2015, + "mean_token_accuracy": 0.6507375836372375, + "num_tokens": 17849674.0, + "step": 690 + }, + { + "epoch": 0.07588403250603998, + "ewc_loss": 1.1101365089416504e-06, + "grad_norm": 2.4893863201141357, + "learning_rate": 2.525622254758419e-07, + "loss": 1.2148, + "mean_token_accuracy": 0.6600173115730286, + "num_tokens": 17884281.0, + "step": 691 + }, + { + "epoch": 0.07599385020865364, + "ewc_loss": 1.1175870895385742e-06, + "grad_norm": 2.876359462738037, + "learning_rate": 2.5292825768667643e-07, + "loss": 1.1308, + "mean_token_accuracy": 0.6757559776306152, + "num_tokens": 17907327.0, + "step": 692 + }, + { + "epoch": 0.0761036679112673, + "ewc_loss": 1.1175870895385742e-06, + "grad_norm": 2.806422233581543, + "learning_rate": 2.5329428989751093e-07, + "loss": 1.1259, + "mean_token_accuracy": 0.6762319803237915, + "num_tokens": 17931212.0, + "step": 693 + }, + { + "epoch": 0.07621348561388096, + "ewc_loss": 1.1101365089416504e-06, + "grad_norm": 2.9914913177490234, + "learning_rate": 2.5366032210834553e-07, + "loss": 1.1904, + "mean_token_accuracy": 0.6594496369361877, + "num_tokens": 17957452.0, + "step": 694 + }, + { + "epoch": 0.07632330331649462, + "ewc_loss": 1.1175870895385742e-06, + "grad_norm": 3.533108711242676, + "learning_rate": 2.540263543191801e-07, + "loss": 1.068, + "mean_token_accuracy": 0.6822611093521118, + "num_tokens": 17974591.0, + "step": 695 + }, + { + "epoch": 0.07643312101910828, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 3.0201468467712402, + "learning_rate": 2.543923865300146e-07, + "loss": 1.1125, + "mean_token_accuracy": 0.6847000122070312, + "num_tokens": 18002528.0, + "step": 696 + }, + { + "epoch": 0.07654293872172194, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 3.016359567642212, + "learning_rate": 2.547584187408492e-07, + "loss": 1.1293, + "mean_token_accuracy": 0.6686645150184631, + "num_tokens": 18027367.0, + "step": 697 + }, + { + "epoch": 0.0766527564243356, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.9141547679901123, + "learning_rate": 2.5512445095168373e-07, + "loss": 1.1613, + "mean_token_accuracy": 0.6621391773223877, + "num_tokens": 18052873.0, + "step": 698 + }, + { + "epoch": 0.07676257412694927, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.696906089782715, + "learning_rate": 2.5549048316251833e-07, + "loss": 1.1481, + "mean_token_accuracy": 0.6651158332824707, + "num_tokens": 18086772.0, + "step": 699 + }, + { + "epoch": 0.07687239182956293, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.7461259365081787, + "learning_rate": 2.5585651537335283e-07, + "loss": 1.1843, + "mean_token_accuracy": 0.659534215927124, + "num_tokens": 18115267.0, + "step": 700 + }, + { + "epoch": 0.07698220953217659, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.780100107192993, + "learning_rate": 2.562225475841874e-07, + "loss": 1.1395, + "mean_token_accuracy": 0.6707535982131958, + "num_tokens": 18144689.0, + "step": 701 + }, + { + "epoch": 0.07709202723479025, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 3.072483539581299, + "learning_rate": 2.56588579795022e-07, + "loss": 1.1511, + "mean_token_accuracy": 0.6736788749694824, + "num_tokens": 18168911.0, + "step": 702 + }, + { + "epoch": 0.07720184493740391, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.9649648666381836, + "learning_rate": 2.569546120058565e-07, + "loss": 1.117, + "mean_token_accuracy": 0.6761940121650696, + "num_tokens": 18193758.0, + "step": 703 + }, + { + "epoch": 0.07731166264001757, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 3.2045443058013916, + "learning_rate": 2.57320644216691e-07, + "loss": 1.1106, + "mean_token_accuracy": 0.6733614802360535, + "num_tokens": 18216265.0, + "step": 704 + }, + { + "epoch": 0.07742148034263123, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.7573816776275635, + "learning_rate": 2.576866764275256e-07, + "loss": 1.0794, + "mean_token_accuracy": 0.6823141574859619, + "num_tokens": 18240261.0, + "step": 705 + }, + { + "epoch": 0.07753129804524489, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.640812635421753, + "learning_rate": 2.580527086383602e-07, + "loss": 1.2169, + "mean_token_accuracy": 0.6494032144546509, + "num_tokens": 18267819.0, + "step": 706 + }, + { + "epoch": 0.07764111574785855, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.775503396987915, + "learning_rate": 2.584187408491947e-07, + "loss": 1.1233, + "mean_token_accuracy": 0.6772988438606262, + "num_tokens": 18295498.0, + "step": 707 + }, + { + "epoch": 0.07775093345047221, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 2.675908327102661, + "learning_rate": 2.5878477306002927e-07, + "loss": 1.1563, + "mean_token_accuracy": 0.6615128517150879, + "num_tokens": 18324344.0, + "step": 708 + }, + { + "epoch": 0.07786075115308588, + "ewc_loss": 1.125037670135498e-06, + "grad_norm": 3.296266555786133, + "learning_rate": 2.591508052708638e-07, + "loss": 1.1232, + "mean_token_accuracy": 0.6743223667144775, + "num_tokens": 18345249.0, + "step": 709 + }, + { + "epoch": 0.07797056885569954, + "ewc_loss": 1.1324882507324219e-06, + "grad_norm": 2.8604815006256104, + "learning_rate": 2.5951683748169837e-07, + "loss": 1.227, + "mean_token_accuracy": 0.6507944464683533, + "num_tokens": 18369976.0, + "step": 710 + }, + { + "epoch": 0.0780803865583132, + "ewc_loss": 1.1399388313293457e-06, + "grad_norm": 3.0928754806518555, + "learning_rate": 2.598828696925329e-07, + "loss": 1.0578, + "mean_token_accuracy": 0.6878416538238525, + "num_tokens": 18389828.0, + "step": 711 + }, + { + "epoch": 0.07819020426092686, + "ewc_loss": 1.1548399925231934e-06, + "grad_norm": 3.537111759185791, + "learning_rate": 2.6024890190336747e-07, + "loss": 1.1333, + "mean_token_accuracy": 0.6791631579399109, + "num_tokens": 18418228.0, + "step": 712 + }, + { + "epoch": 0.07830002196354052, + "ewc_loss": 1.1548399925231934e-06, + "grad_norm": 2.8679168224334717, + "learning_rate": 2.6061493411420207e-07, + "loss": 1.1328, + "mean_token_accuracy": 0.6640797257423401, + "num_tokens": 18442229.0, + "step": 713 + }, + { + "epoch": 0.07840983966615418, + "ewc_loss": 1.1548399925231934e-06, + "grad_norm": 2.6089818477630615, + "learning_rate": 2.6098096632503657e-07, + "loss": 1.1638, + "mean_token_accuracy": 0.6642701029777527, + "num_tokens": 18469419.0, + "step": 714 + }, + { + "epoch": 0.07851965736876784, + "ewc_loss": 1.1548399925231934e-06, + "grad_norm": 2.836975574493408, + "learning_rate": 2.6134699853587117e-07, + "loss": 0.9652, + "mean_token_accuracy": 0.7119178771972656, + "num_tokens": 18493762.0, + "step": 715 + }, + { + "epoch": 0.0786294750713815, + "ewc_loss": 1.1548399925231934e-06, + "grad_norm": 3.267180919647217, + "learning_rate": 2.617130307467057e-07, + "loss": 1.1309, + "mean_token_accuracy": 0.671434760093689, + "num_tokens": 18516052.0, + "step": 716 + }, + { + "epoch": 0.07873929277399516, + "ewc_loss": 1.1548399925231934e-06, + "grad_norm": 2.596318483352661, + "learning_rate": 2.620790629575402e-07, + "loss": 1.1199, + "mean_token_accuracy": 0.6799324750900269, + "num_tokens": 18542974.0, + "step": 717 + }, + { + "epoch": 0.07884911047660884, + "ewc_loss": 1.1622905731201172e-06, + "grad_norm": 3.325484037399292, + "learning_rate": 2.624450951683748e-07, + "loss": 1.0577, + "mean_token_accuracy": 0.6868950724601746, + "num_tokens": 18561988.0, + "step": 718 + }, + { + "epoch": 0.0789589281792225, + "ewc_loss": 1.169741153717041e-06, + "grad_norm": 2.879453420639038, + "learning_rate": 2.6281112737920937e-07, + "loss": 1.1772, + "mean_token_accuracy": 0.6617428064346313, + "num_tokens": 18588546.0, + "step": 719 + }, + { + "epoch": 0.07906874588183616, + "ewc_loss": 1.169741153717041e-06, + "grad_norm": 2.868313789367676, + "learning_rate": 2.6317715959004386e-07, + "loss": 1.2044, + "mean_token_accuracy": 0.6518961191177368, + "num_tokens": 18615106.0, + "step": 720 + }, + { + "epoch": 0.07917856358444982, + "ewc_loss": 1.169741153717041e-06, + "grad_norm": 2.5591657161712646, + "learning_rate": 2.6354319180087846e-07, + "loss": 1.1206, + "mean_token_accuracy": 0.6750530004501343, + "num_tokens": 18642433.0, + "step": 721 + }, + { + "epoch": 0.07928838128706348, + "ewc_loss": 1.169741153717041e-06, + "grad_norm": 2.5749330520629883, + "learning_rate": 2.63909224011713e-07, + "loss": 1.115, + "mean_token_accuracy": 0.6728061437606812, + "num_tokens": 18670567.0, + "step": 722 + }, + { + "epoch": 0.07939819898967714, + "ewc_loss": 1.169741153717041e-06, + "grad_norm": 3.0252652168273926, + "learning_rate": 2.642752562225476e-07, + "loss": 1.1246, + "mean_token_accuracy": 0.6696712970733643, + "num_tokens": 18691814.0, + "step": 723 + }, + { + "epoch": 0.0795080166922908, + "ewc_loss": 1.1771917343139648e-06, + "grad_norm": 2.706364870071411, + "learning_rate": 2.646412884333821e-07, + "loss": 1.16, + "mean_token_accuracy": 0.6655562520027161, + "num_tokens": 18717182.0, + "step": 724 + }, + { + "epoch": 0.07961783439490445, + "ewc_loss": 1.1846423149108887e-06, + "grad_norm": 2.5382590293884277, + "learning_rate": 2.6500732064421666e-07, + "loss": 1.1397, + "mean_token_accuracy": 0.6650310754776001, + "num_tokens": 18746831.0, + "step": 725 + }, + { + "epoch": 0.07972765209751811, + "ewc_loss": 1.1846423149108887e-06, + "grad_norm": 2.5631048679351807, + "learning_rate": 2.6537335285505126e-07, + "loss": 1.0812, + "mean_token_accuracy": 0.6823241114616394, + "num_tokens": 18777233.0, + "step": 726 + }, + { + "epoch": 0.07983746980013177, + "ewc_loss": 1.1846423149108887e-06, + "grad_norm": 2.794966459274292, + "learning_rate": 2.6573938506588576e-07, + "loss": 1.1326, + "mean_token_accuracy": 0.6762831211090088, + "num_tokens": 18802479.0, + "step": 727 + }, + { + "epoch": 0.07994728750274545, + "ewc_loss": 1.1846423149108887e-06, + "grad_norm": 2.5686557292938232, + "learning_rate": 2.661054172767203e-07, + "loss": 1.1543, + "mean_token_accuracy": 0.6619802713394165, + "num_tokens": 18831630.0, + "step": 728 + }, + { + "epoch": 0.08005710520535911, + "ewc_loss": 1.1920928955078125e-06, + "grad_norm": 3.0104734897613525, + "learning_rate": 2.664714494875549e-07, + "loss": 0.9738, + "mean_token_accuracy": 0.7143491506576538, + "num_tokens": 18853499.0, + "step": 729 + }, + { + "epoch": 0.08016692290797277, + "ewc_loss": 1.1920928955078125e-06, + "grad_norm": 2.861440658569336, + "learning_rate": 2.6683748169838946e-07, + "loss": 1.189, + "mean_token_accuracy": 0.663261890411377, + "num_tokens": 18875616.0, + "step": 730 + }, + { + "epoch": 0.08027674061058643, + "ewc_loss": 1.2069940567016602e-06, + "grad_norm": 2.966501235961914, + "learning_rate": 2.67203513909224e-07, + "loss": 1.1162, + "mean_token_accuracy": 0.6772201061248779, + "num_tokens": 18895943.0, + "step": 731 + }, + { + "epoch": 0.08038655831320009, + "ewc_loss": 1.2069940567016602e-06, + "grad_norm": 2.6264569759368896, + "learning_rate": 2.6756954612005856e-07, + "loss": 1.1722, + "mean_token_accuracy": 0.6614583134651184, + "num_tokens": 18930183.0, + "step": 732 + }, + { + "epoch": 0.08049637601581375, + "ewc_loss": 1.2069940567016602e-06, + "grad_norm": 2.9744629859924316, + "learning_rate": 2.679355783308931e-07, + "loss": 1.1704, + "mean_token_accuracy": 0.6638044118881226, + "num_tokens": 18953193.0, + "step": 733 + }, + { + "epoch": 0.08060619371842741, + "ewc_loss": 1.2069940567016602e-06, + "grad_norm": 2.614504337310791, + "learning_rate": 2.6830161054172765e-07, + "loss": 1.2013, + "mean_token_accuracy": 0.6583535671234131, + "num_tokens": 18981251.0, + "step": 734 + }, + { + "epoch": 0.08071601142104107, + "ewc_loss": 1.2218952178955078e-06, + "grad_norm": 2.6058855056762695, + "learning_rate": 2.686676427525622e-07, + "loss": 1.1509, + "mean_token_accuracy": 0.6768515110015869, + "num_tokens": 19009491.0, + "step": 735 + }, + { + "epoch": 0.08082582912365473, + "ewc_loss": 1.2218952178955078e-06, + "grad_norm": 2.577690839767456, + "learning_rate": 2.6903367496339675e-07, + "loss": 1.0823, + "mean_token_accuracy": 0.6835185289382935, + "num_tokens": 19037182.0, + "step": 736 + }, + { + "epoch": 0.0809356468262684, + "ewc_loss": 1.2218952178955078e-06, + "grad_norm": 2.9682884216308594, + "learning_rate": 2.693997071742313e-07, + "loss": 1.0943, + "mean_token_accuracy": 0.6797610521316528, + "num_tokens": 19059901.0, + "step": 737 + }, + { + "epoch": 0.08104546452888206, + "ewc_loss": 1.2218952178955078e-06, + "grad_norm": 2.8102011680603027, + "learning_rate": 2.6976573938506585e-07, + "loss": 1.178, + "mean_token_accuracy": 0.6622779369354248, + "num_tokens": 19085117.0, + "step": 738 + }, + { + "epoch": 0.08115528223149572, + "ewc_loss": 1.2293457984924316e-06, + "grad_norm": 2.5829901695251465, + "learning_rate": 2.7013177159590045e-07, + "loss": 1.0855, + "mean_token_accuracy": 0.6814039945602417, + "num_tokens": 19111093.0, + "step": 739 + }, + { + "epoch": 0.08126509993410938, + "ewc_loss": 1.2293457984924316e-06, + "grad_norm": 2.863966226577759, + "learning_rate": 2.70497803806735e-07, + "loss": 1.0992, + "mean_token_accuracy": 0.6743087768554688, + "num_tokens": 19135550.0, + "step": 740 + }, + { + "epoch": 0.08137491763672304, + "ewc_loss": 1.2442469596862793e-06, + "grad_norm": 2.7028636932373047, + "learning_rate": 2.708638360175695e-07, + "loss": 1.0707, + "mean_token_accuracy": 0.6844254732131958, + "num_tokens": 19162633.0, + "step": 741 + }, + { + "epoch": 0.0814847353393367, + "ewc_loss": 1.2442469596862793e-06, + "grad_norm": 2.9051551818847656, + "learning_rate": 2.712298682284041e-07, + "loss": 1.0632, + "mean_token_accuracy": 0.6967699527740479, + "num_tokens": 19186813.0, + "step": 742 + }, + { + "epoch": 0.08159455304195036, + "ewc_loss": 1.2442469596862793e-06, + "grad_norm": 2.46110200881958, + "learning_rate": 2.7159590043923865e-07, + "loss": 1.1119, + "mean_token_accuracy": 0.6741212606430054, + "num_tokens": 19217664.0, + "step": 743 + }, + { + "epoch": 0.08170437074456402, + "ewc_loss": 1.2516975402832031e-06, + "grad_norm": 3.001224994659424, + "learning_rate": 2.7196193265007315e-07, + "loss": 1.0911, + "mean_token_accuracy": 0.6763099431991577, + "num_tokens": 19239127.0, + "step": 744 + }, + { + "epoch": 0.08181418844717768, + "ewc_loss": 1.259148120880127e-06, + "grad_norm": 2.7050888538360596, + "learning_rate": 2.7232796486090775e-07, + "loss": 1.1124, + "mean_token_accuracy": 0.6777967214584351, + "num_tokens": 19263028.0, + "step": 745 + }, + { + "epoch": 0.08192400614979134, + "ewc_loss": 1.259148120880127e-06, + "grad_norm": 2.474622964859009, + "learning_rate": 2.726939970717423e-07, + "loss": 1.1941, + "mean_token_accuracy": 0.6528751254081726, + "num_tokens": 19292808.0, + "step": 746 + }, + { + "epoch": 0.08203382385240501, + "ewc_loss": 1.259148120880127e-06, + "grad_norm": 2.968973398208618, + "learning_rate": 2.730600292825769e-07, + "loss": 1.1257, + "mean_token_accuracy": 0.6731305122375488, + "num_tokens": 19314443.0, + "step": 747 + }, + { + "epoch": 0.08214364155501867, + "ewc_loss": 1.259148120880127e-06, + "grad_norm": 2.7174322605133057, + "learning_rate": 2.734260614934114e-07, + "loss": 1.0734, + "mean_token_accuracy": 0.6901847720146179, + "num_tokens": 19339945.0, + "step": 748 + }, + { + "epoch": 0.08225345925763233, + "ewc_loss": 1.2665987014770508e-06, + "grad_norm": 2.4913313388824463, + "learning_rate": 2.7379209370424594e-07, + "loss": 1.1927, + "mean_token_accuracy": 0.6580515503883362, + "num_tokens": 19370487.0, + "step": 749 + }, + { + "epoch": 0.08236327696024599, + "ewc_loss": 1.2665987014770508e-06, + "grad_norm": 2.6341423988342285, + "learning_rate": 2.7415812591508055e-07, + "loss": 1.0941, + "mean_token_accuracy": 0.6802284717559814, + "num_tokens": 19395644.0, + "step": 750 + }, + { + "epoch": 0.08247309466285965, + "ewc_loss": 1.2665987014770508e-06, + "grad_norm": 2.318535566329956, + "learning_rate": 2.7452415812591504e-07, + "loss": 1.1749, + "mean_token_accuracy": 0.6589652895927429, + "num_tokens": 19426218.0, + "step": 751 + }, + { + "epoch": 0.08258291236547331, + "ewc_loss": 1.2665987014770508e-06, + "grad_norm": 2.658430814743042, + "learning_rate": 2.748901903367496e-07, + "loss": 1.172, + "mean_token_accuracy": 0.6655164361000061, + "num_tokens": 19451846.0, + "step": 752 + }, + { + "epoch": 0.08269273006808697, + "ewc_loss": 1.2665987014770508e-06, + "grad_norm": 2.851618766784668, + "learning_rate": 2.752562225475842e-07, + "loss": 1.0954, + "mean_token_accuracy": 0.6859294176101685, + "num_tokens": 19476296.0, + "step": 753 + }, + { + "epoch": 0.08280254777070063, + "ewc_loss": 1.2665987014770508e-06, + "grad_norm": 2.4722867012023926, + "learning_rate": 2.756222547584187e-07, + "loss": 1.1346, + "mean_token_accuracy": 0.6780863404273987, + "num_tokens": 19504200.0, + "step": 754 + }, + { + "epoch": 0.08291236547331429, + "ewc_loss": 1.2740492820739746e-06, + "grad_norm": 2.8674471378326416, + "learning_rate": 2.759882869692533e-07, + "loss": 1.2122, + "mean_token_accuracy": 0.655156135559082, + "num_tokens": 19527134.0, + "step": 755 + }, + { + "epoch": 0.08302218317592797, + "ewc_loss": 1.2740492820739746e-06, + "grad_norm": 2.493267059326172, + "learning_rate": 2.7635431918008784e-07, + "loss": 1.0956, + "mean_token_accuracy": 0.6832151412963867, + "num_tokens": 19556429.0, + "step": 756 + }, + { + "epoch": 0.08313200087854163, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 2.4029242992401123, + "learning_rate": 2.767203513909224e-07, + "loss": 1.1695, + "mean_token_accuracy": 0.6609424352645874, + "num_tokens": 19587759.0, + "step": 757 + }, + { + "epoch": 0.08324181858115529, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 3.182471990585327, + "learning_rate": 2.7708638360175694e-07, + "loss": 1.1183, + "mean_token_accuracy": 0.6815283298492432, + "num_tokens": 19607412.0, + "step": 758 + }, + { + "epoch": 0.08335163628376895, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 2.741009473800659, + "learning_rate": 2.774524158125915e-07, + "loss": 1.0149, + "mean_token_accuracy": 0.7098939418792725, + "num_tokens": 19630184.0, + "step": 759 + }, + { + "epoch": 0.0834614539863826, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 2.7005481719970703, + "learning_rate": 2.7781844802342604e-07, + "loss": 1.0972, + "mean_token_accuracy": 0.6815471053123474, + "num_tokens": 19655787.0, + "step": 760 + }, + { + "epoch": 0.08357127168899627, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 2.5623834133148193, + "learning_rate": 2.781844802342606e-07, + "loss": 1.0622, + "mean_token_accuracy": 0.6921109557151794, + "num_tokens": 19683813.0, + "step": 761 + }, + { + "epoch": 0.08368108939160993, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 2.9149649143218994, + "learning_rate": 2.7855051244509513e-07, + "loss": 1.0898, + "mean_token_accuracy": 0.6807785034179688, + "num_tokens": 19706512.0, + "step": 762 + }, + { + "epoch": 0.08379090709422359, + "ewc_loss": 1.2889504432678223e-06, + "grad_norm": 2.5997071266174316, + "learning_rate": 2.7891654465592974e-07, + "loss": 1.1596, + "mean_token_accuracy": 0.6688681840896606, + "num_tokens": 19737097.0, + "step": 763 + }, + { + "epoch": 0.08390072479683724, + "ewc_loss": 1.296401023864746e-06, + "grad_norm": 2.782137155532837, + "learning_rate": 2.792825768667643e-07, + "loss": 1.0142, + "mean_token_accuracy": 0.702065646648407, + "num_tokens": 19762815.0, + "step": 764 + }, + { + "epoch": 0.0840105424994509, + "ewc_loss": 1.296401023864746e-06, + "grad_norm": 2.941204786300659, + "learning_rate": 2.796486090775988e-07, + "loss": 1.127, + "mean_token_accuracy": 0.674709141254425, + "num_tokens": 19785444.0, + "step": 765 + }, + { + "epoch": 0.08412036020206458, + "ewc_loss": 1.296401023864746e-06, + "grad_norm": 2.8644330501556396, + "learning_rate": 2.800146412884334e-07, + "loss": 1.1349, + "mean_token_accuracy": 0.6707392334938049, + "num_tokens": 19809678.0, + "step": 766 + }, + { + "epoch": 0.08423017790467824, + "ewc_loss": 1.296401023864746e-06, + "grad_norm": 2.6750924587249756, + "learning_rate": 2.8038067349926793e-07, + "loss": 1.1088, + "mean_token_accuracy": 0.6743124127388, + "num_tokens": 19834658.0, + "step": 767 + }, + { + "epoch": 0.0843399956072919, + "ewc_loss": 1.296401023864746e-06, + "grad_norm": 2.5713818073272705, + "learning_rate": 2.8074670571010243e-07, + "loss": 1.1948, + "mean_token_accuracy": 0.659990131855011, + "num_tokens": 19860076.0, + "step": 768 + }, + { + "epoch": 0.08444981330990556, + "ewc_loss": 1.296401023864746e-06, + "grad_norm": 2.802572727203369, + "learning_rate": 2.8111273792093703e-07, + "loss": 1.1493, + "mean_token_accuracy": 0.6667723059654236, + "num_tokens": 19885069.0, + "step": 769 + }, + { + "epoch": 0.08455963101251922, + "ewc_loss": 1.3113021850585938e-06, + "grad_norm": 2.849754571914673, + "learning_rate": 2.814787701317716e-07, + "loss": 1.2104, + "mean_token_accuracy": 0.6593449115753174, + "num_tokens": 19909921.0, + "step": 770 + }, + { + "epoch": 0.08466944871513288, + "ewc_loss": 1.3113021850585938e-06, + "grad_norm": 2.6650922298431396, + "learning_rate": 2.818448023426062e-07, + "loss": 1.0651, + "mean_token_accuracy": 0.6965285539627075, + "num_tokens": 19935788.0, + "step": 771 + }, + { + "epoch": 0.08477926641774654, + "ewc_loss": 1.3187527656555176e-06, + "grad_norm": 3.2408974170684814, + "learning_rate": 2.822108345534407e-07, + "loss": 1.1422, + "mean_token_accuracy": 0.6668260097503662, + "num_tokens": 19965350.0, + "step": 772 + }, + { + "epoch": 0.0848890841203602, + "ewc_loss": 1.3187527656555176e-06, + "grad_norm": 2.8682706356048584, + "learning_rate": 2.8257686676427523e-07, + "loss": 1.0147, + "mean_token_accuracy": 0.7047832012176514, + "num_tokens": 19985991.0, + "step": 773 + }, + { + "epoch": 0.08499890182297386, + "ewc_loss": 1.3187527656555176e-06, + "grad_norm": 2.616795301437378, + "learning_rate": 2.8294289897510983e-07, + "loss": 1.0934, + "mean_token_accuracy": 0.6825340986251831, + "num_tokens": 20013789.0, + "step": 774 + }, + { + "epoch": 0.08510871952558753, + "ewc_loss": 1.3187527656555176e-06, + "grad_norm": 2.931100845336914, + "learning_rate": 2.833089311859443e-07, + "loss": 1.0341, + "mean_token_accuracy": 0.6920843124389648, + "num_tokens": 20034001.0, + "step": 775 + }, + { + "epoch": 0.08521853722820119, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.7533724308013916, + "learning_rate": 2.836749633967789e-07, + "loss": 1.1548, + "mean_token_accuracy": 0.6638762354850769, + "num_tokens": 20060222.0, + "step": 776 + }, + { + "epoch": 0.08532835493081485, + "ewc_loss": 1.3187527656555176e-06, + "grad_norm": 2.4803290367126465, + "learning_rate": 2.840409956076135e-07, + "loss": 1.0887, + "mean_token_accuracy": 0.6832543611526489, + "num_tokens": 20093804.0, + "step": 777 + }, + { + "epoch": 0.08543817263342851, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.4651899337768555, + "learning_rate": 2.8440702781844797e-07, + "loss": 1.2265, + "mean_token_accuracy": 0.6418028473854065, + "num_tokens": 20126639.0, + "step": 778 + }, + { + "epoch": 0.08554799033604217, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.586597204208374, + "learning_rate": 2.847730600292826e-07, + "loss": 1.024, + "mean_token_accuracy": 0.7043284773826599, + "num_tokens": 20154205.0, + "step": 779 + }, + { + "epoch": 0.08565780803865583, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.4334805011749268, + "learning_rate": 2.851390922401171e-07, + "loss": 1.2165, + "mean_token_accuracy": 0.6643155217170715, + "num_tokens": 20185261.0, + "step": 780 + }, + { + "epoch": 0.08576762574126949, + "ewc_loss": 1.3187527656555176e-06, + "grad_norm": 2.9094386100769043, + "learning_rate": 2.8550512445095167e-07, + "loss": 1.0185, + "mean_token_accuracy": 0.6943086385726929, + "num_tokens": 20205777.0, + "step": 781 + }, + { + "epoch": 0.08587744344388315, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.397029161453247, + "learning_rate": 2.858711566617862e-07, + "loss": 1.0625, + "mean_token_accuracy": 0.6884894371032715, + "num_tokens": 20237386.0, + "step": 782 + }, + { + "epoch": 0.08598726114649681, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.6624820232391357, + "learning_rate": 2.8623718887262077e-07, + "loss": 1.1205, + "mean_token_accuracy": 0.6774642467498779, + "num_tokens": 20261696.0, + "step": 783 + }, + { + "epoch": 0.08609707884911047, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.7459559440612793, + "learning_rate": 2.866032210834553e-07, + "loss": 1.0882, + "mean_token_accuracy": 0.6808463335037231, + "num_tokens": 20287351.0, + "step": 784 + }, + { + "epoch": 0.08620689655172414, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.788349151611328, + "learning_rate": 2.8696925329428987e-07, + "loss": 1.0084, + "mean_token_accuracy": 0.7024816870689392, + "num_tokens": 20309391.0, + "step": 785 + }, + { + "epoch": 0.0863167142543378, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 6.955178737640381, + "learning_rate": 2.873352855051244e-07, + "loss": 1.0685, + "mean_token_accuracy": 0.6813777685165405, + "num_tokens": 20332515.0, + "step": 786 + }, + { + "epoch": 0.08642653195695146, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.8898420333862305, + "learning_rate": 2.87701317715959e-07, + "loss": 1.1069, + "mean_token_accuracy": 0.6847273111343384, + "num_tokens": 20359379.0, + "step": 787 + }, + { + "epoch": 0.08653634965956512, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.8780250549316406, + "learning_rate": 2.8806734992679357e-07, + "loss": 1.1365, + "mean_token_accuracy": 0.6666578054428101, + "num_tokens": 20383436.0, + "step": 788 + }, + { + "epoch": 0.08664616736217878, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.3026325702667236, + "learning_rate": 2.8843338213762807e-07, + "loss": 1.113, + "mean_token_accuracy": 0.6757996082305908, + "num_tokens": 20414443.0, + "step": 789 + }, + { + "epoch": 0.08675598506479244, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.8381025791168213, + "learning_rate": 2.8879941434846267e-07, + "loss": 1.0742, + "mean_token_accuracy": 0.6812962293624878, + "num_tokens": 20437080.0, + "step": 790 + }, + { + "epoch": 0.0868658027674061, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.735964298248291, + "learning_rate": 2.891654465592972e-07, + "loss": 0.971, + "mean_token_accuracy": 0.7140269875526428, + "num_tokens": 20461294.0, + "step": 791 + }, + { + "epoch": 0.08697562047001976, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.649268627166748, + "learning_rate": 2.895314787701317e-07, + "loss": 1.1101, + "mean_token_accuracy": 0.6724438667297363, + "num_tokens": 20486377.0, + "step": 792 + }, + { + "epoch": 0.08708543817263342, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.835515260696411, + "learning_rate": 2.898975109809663e-07, + "loss": 1.0625, + "mean_token_accuracy": 0.6900439858436584, + "num_tokens": 20508898.0, + "step": 793 + }, + { + "epoch": 0.0871952558752471, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.4449026584625244, + "learning_rate": 2.9026354319180086e-07, + "loss": 1.1406, + "mean_token_accuracy": 0.6691406965255737, + "num_tokens": 20538441.0, + "step": 794 + }, + { + "epoch": 0.08730507357786076, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.7514045238494873, + "learning_rate": 2.9062957540263547e-07, + "loss": 1.0872, + "mean_token_accuracy": 0.6839661598205566, + "num_tokens": 20561412.0, + "step": 795 + }, + { + "epoch": 0.08741489128047442, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.757291078567505, + "learning_rate": 2.9099560761346996e-07, + "loss": 1.0875, + "mean_token_accuracy": 0.681471049785614, + "num_tokens": 20586226.0, + "step": 796 + }, + { + "epoch": 0.08752470898308808, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.3870694637298584, + "learning_rate": 2.913616398243045e-07, + "loss": 1.2103, + "mean_token_accuracy": 0.6539727449417114, + "num_tokens": 20618182.0, + "step": 797 + }, + { + "epoch": 0.08763452668570174, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.490809917449951, + "learning_rate": 2.917276720351391e-07, + "loss": 1.1746, + "mean_token_accuracy": 0.6710244417190552, + "num_tokens": 20647672.0, + "step": 798 + }, + { + "epoch": 0.0877443443883154, + "ewc_loss": 1.3262033462524414e-06, + "grad_norm": 2.656543254852295, + "learning_rate": 2.920937042459736e-07, + "loss": 1.1139, + "mean_token_accuracy": 0.6886248588562012, + "num_tokens": 20671724.0, + "step": 799 + }, + { + "epoch": 0.08785416209092906, + "ewc_loss": 1.3336539268493652e-06, + "grad_norm": 2.4208598136901855, + "learning_rate": 2.9245973645680816e-07, + "loss": 1.1546, + "mean_token_accuracy": 0.6831238269805908, + "num_tokens": 20700193.0, + "step": 800 + }, + { + "epoch": 0.08796397979354272, + "ewc_loss": 1.3336539268493652e-06, + "grad_norm": 2.667301893234253, + "learning_rate": 2.9282576866764276e-07, + "loss": 1.0802, + "mean_token_accuracy": 0.6797026991844177, + "num_tokens": 20724905.0, + "step": 801 + }, + { + "epoch": 0.08807379749615638, + "ewc_loss": 1.3336539268493652e-06, + "grad_norm": 2.7503609657287598, + "learning_rate": 2.9319180087847726e-07, + "loss": 1.1397, + "mean_token_accuracy": 0.669073760509491, + "num_tokens": 20749749.0, + "step": 802 + }, + { + "epoch": 0.08818361519877004, + "ewc_loss": 1.341104507446289e-06, + "grad_norm": 2.4330039024353027, + "learning_rate": 2.9355783308931186e-07, + "loss": 1.0124, + "mean_token_accuracy": 0.6979039311408997, + "num_tokens": 20778139.0, + "step": 803 + }, + { + "epoch": 0.08829343290138371, + "ewc_loss": 1.3485550880432129e-06, + "grad_norm": 2.7744109630584717, + "learning_rate": 2.939238653001464e-07, + "loss": 1.125, + "mean_token_accuracy": 0.6866639852523804, + "num_tokens": 20801934.0, + "step": 804 + }, + { + "epoch": 0.08840325060399737, + "ewc_loss": 1.3560056686401367e-06, + "grad_norm": 2.409921407699585, + "learning_rate": 2.9428989751098096e-07, + "loss": 1.1228, + "mean_token_accuracy": 0.6755895018577576, + "num_tokens": 20832336.0, + "step": 805 + }, + { + "epoch": 0.08851306830661103, + "ewc_loss": 1.3709068298339844e-06, + "grad_norm": 2.6031596660614014, + "learning_rate": 2.946559297218155e-07, + "loss": 1.1268, + "mean_token_accuracy": 0.6698471307754517, + "num_tokens": 20858646.0, + "step": 806 + }, + { + "epoch": 0.08862288600922469, + "ewc_loss": 1.3709068298339844e-06, + "grad_norm": 2.5057787895202637, + "learning_rate": 2.9502196193265006e-07, + "loss": 1.1518, + "mean_token_accuracy": 0.6667685508728027, + "num_tokens": 20887602.0, + "step": 807 + }, + { + "epoch": 0.08873270371183835, + "ewc_loss": 1.3709068298339844e-06, + "grad_norm": 2.733651876449585, + "learning_rate": 2.953879941434846e-07, + "loss": 1.1766, + "mean_token_accuracy": 0.6595904231071472, + "num_tokens": 20912488.0, + "step": 808 + }, + { + "epoch": 0.08884252141445201, + "ewc_loss": 1.3709068298339844e-06, + "grad_norm": 2.426698684692383, + "learning_rate": 2.9575402635431915e-07, + "loss": 1.1995, + "mean_token_accuracy": 0.6525692343711853, + "num_tokens": 20940624.0, + "step": 809 + }, + { + "epoch": 0.08895233911706567, + "ewc_loss": 1.3709068298339844e-06, + "grad_norm": 2.393739938735962, + "learning_rate": 2.961200585651537e-07, + "loss": 1.0936, + "mean_token_accuracy": 0.6830503344535828, + "num_tokens": 20970873.0, + "step": 810 + }, + { + "epoch": 0.08906215681967933, + "ewc_loss": 1.3932585716247559e-06, + "grad_norm": 2.8620197772979736, + "learning_rate": 2.964860907759883e-07, + "loss": 1.0329, + "mean_token_accuracy": 0.6933596730232239, + "num_tokens": 20993869.0, + "step": 811 + }, + { + "epoch": 0.08917197452229299, + "ewc_loss": 1.385807991027832e-06, + "grad_norm": 2.573791980743408, + "learning_rate": 2.9685212298682285e-07, + "loss": 1.0384, + "mean_token_accuracy": 0.6926065683364868, + "num_tokens": 21019651.0, + "step": 812 + }, + { + "epoch": 0.08928179222490666, + "ewc_loss": 1.385807991027832e-06, + "grad_norm": 2.3950977325439453, + "learning_rate": 2.9721815519765735e-07, + "loss": 1.0702, + "mean_token_accuracy": 0.6885907053947449, + "num_tokens": 21048736.0, + "step": 813 + }, + { + "epoch": 0.08939160992752032, + "ewc_loss": 1.385807991027832e-06, + "grad_norm": 2.8842177391052246, + "learning_rate": 2.9758418740849195e-07, + "loss": 1.2005, + "mean_token_accuracy": 0.6665027141571045, + "num_tokens": 21071505.0, + "step": 814 + }, + { + "epoch": 0.08950142763013398, + "ewc_loss": 1.385807991027832e-06, + "grad_norm": 2.8830270767211914, + "learning_rate": 2.979502196193265e-07, + "loss": 1.0329, + "mean_token_accuracy": 0.6972272396087646, + "num_tokens": 21094803.0, + "step": 815 + }, + { + "epoch": 0.08961124533274764, + "ewc_loss": 1.3932585716247559e-06, + "grad_norm": 2.7178940773010254, + "learning_rate": 2.98316251830161e-07, + "loss": 1.1079, + "mean_token_accuracy": 0.6806344985961914, + "num_tokens": 21120328.0, + "step": 816 + }, + { + "epoch": 0.0897210630353613, + "ewc_loss": 1.4081597328186035e-06, + "grad_norm": 2.541783332824707, + "learning_rate": 2.986822840409956e-07, + "loss": 1.1231, + "mean_token_accuracy": 0.6720662117004395, + "num_tokens": 21146162.0, + "step": 817 + }, + { + "epoch": 0.08983088073797496, + "ewc_loss": 1.4156103134155273e-06, + "grad_norm": 2.613617181777954, + "learning_rate": 2.9904831625183015e-07, + "loss": 1.0543, + "mean_token_accuracy": 0.6877405643463135, + "num_tokens": 21169041.0, + "step": 818 + }, + { + "epoch": 0.08994069844058862, + "ewc_loss": 1.4081597328186035e-06, + "grad_norm": 2.5702779293060303, + "learning_rate": 2.994143484626647e-07, + "loss": 1.1066, + "mean_token_accuracy": 0.681891143321991, + "num_tokens": 21195693.0, + "step": 819 + }, + { + "epoch": 0.09005051614320228, + "ewc_loss": 1.4156103134155273e-06, + "grad_norm": 2.8624744415283203, + "learning_rate": 2.9978038067349925e-07, + "loss": 1.0696, + "mean_token_accuracy": 0.6917357444763184, + "num_tokens": 21217125.0, + "step": 820 + }, + { + "epoch": 0.09016033384581594, + "ewc_loss": 1.4156103134155273e-06, + "grad_norm": 2.575934886932373, + "learning_rate": 3.001464128843338e-07, + "loss": 1.1707, + "mean_token_accuracy": 0.6595132350921631, + "num_tokens": 21243201.0, + "step": 821 + }, + { + "epoch": 0.0902701515484296, + "ewc_loss": 1.4156103134155273e-06, + "grad_norm": 2.446119785308838, + "learning_rate": 3.005124450951684e-07, + "loss": 1.0733, + "mean_token_accuracy": 0.689525842666626, + "num_tokens": 21272069.0, + "step": 822 + }, + { + "epoch": 0.09037996925104327, + "ewc_loss": 1.430511474609375e-06, + "grad_norm": 2.5241708755493164, + "learning_rate": 3.008784773060029e-07, + "loss": 1.1596, + "mean_token_accuracy": 0.6636790037155151, + "num_tokens": 21299739.0, + "step": 823 + }, + { + "epoch": 0.09048978695365693, + "ewc_loss": 1.430511474609375e-06, + "grad_norm": 3.078723669052124, + "learning_rate": 3.0124450951683744e-07, + "loss": 1.0729, + "mean_token_accuracy": 0.6766268014907837, + "num_tokens": 21318516.0, + "step": 824 + }, + { + "epoch": 0.0905996046562706, + "ewc_loss": 1.430511474609375e-06, + "grad_norm": 2.9639384746551514, + "learning_rate": 3.0161054172767204e-07, + "loss": 1.0902, + "mean_token_accuracy": 0.6736141443252563, + "num_tokens": 21337893.0, + "step": 825 + }, + { + "epoch": 0.09070942235888425, + "ewc_loss": 1.430511474609375e-06, + "grad_norm": 2.4829866886138916, + "learning_rate": 3.0197657393850654e-07, + "loss": 1.2055, + "mean_token_accuracy": 0.6478376388549805, + "num_tokens": 21366728.0, + "step": 826 + }, + { + "epoch": 0.09081924006149791, + "ewc_loss": 1.430511474609375e-06, + "grad_norm": 2.3521735668182373, + "learning_rate": 3.0234260614934114e-07, + "loss": 1.1956, + "mean_token_accuracy": 0.6507014036178589, + "num_tokens": 21398742.0, + "step": 827 + }, + { + "epoch": 0.09092905776411157, + "ewc_loss": 1.4454126358032227e-06, + "grad_norm": 2.724141836166382, + "learning_rate": 3.027086383601757e-07, + "loss": 1.0034, + "mean_token_accuracy": 0.7035799622535706, + "num_tokens": 21422256.0, + "step": 828 + }, + { + "epoch": 0.09103887546672523, + "ewc_loss": 1.4454126358032227e-06, + "grad_norm": 2.6716601848602295, + "learning_rate": 3.0307467057101024e-07, + "loss": 1.1362, + "mean_token_accuracy": 0.6708507537841797, + "num_tokens": 21445882.0, + "step": 829 + }, + { + "epoch": 0.09114869316933889, + "ewc_loss": 1.4528632164001465e-06, + "grad_norm": 2.672761917114258, + "learning_rate": 3.034407027818448e-07, + "loss": 1.1133, + "mean_token_accuracy": 0.674243688583374, + "num_tokens": 21472863.0, + "step": 830 + }, + { + "epoch": 0.09125851087195255, + "ewc_loss": 1.4454126358032227e-06, + "grad_norm": 3.362767219543457, + "learning_rate": 3.0380673499267934e-07, + "loss": 1.0508, + "mean_token_accuracy": 0.6903156042098999, + "num_tokens": 21496899.0, + "step": 831 + }, + { + "epoch": 0.09136832857456623, + "ewc_loss": 1.4454126358032227e-06, + "grad_norm": 2.902446985244751, + "learning_rate": 3.041727672035139e-07, + "loss": 1.1254, + "mean_token_accuracy": 0.6748003959655762, + "num_tokens": 21519107.0, + "step": 832 + }, + { + "epoch": 0.09147814627717989, + "ewc_loss": 1.4454126358032227e-06, + "grad_norm": 2.7542970180511475, + "learning_rate": 3.0453879941434844e-07, + "loss": 1.1605, + "mean_token_accuracy": 0.6652454137802124, + "num_tokens": 21544059.0, + "step": 833 + }, + { + "epoch": 0.09158796397979355, + "ewc_loss": 1.4528632164001465e-06, + "grad_norm": 2.710181474685669, + "learning_rate": 3.04904831625183e-07, + "loss": 1.0837, + "mean_token_accuracy": 0.6868232488632202, + "num_tokens": 21568468.0, + "step": 834 + }, + { + "epoch": 0.0916977816824072, + "ewc_loss": 1.4528632164001465e-06, + "grad_norm": 2.6928679943084717, + "learning_rate": 3.052708638360176e-07, + "loss": 1.1023, + "mean_token_accuracy": 0.6736444234848022, + "num_tokens": 21592891.0, + "step": 835 + }, + { + "epoch": 0.09180759938502087, + "ewc_loss": 1.4677643775939941e-06, + "grad_norm": 2.85624623298645, + "learning_rate": 3.056368960468521e-07, + "loss": 1.0942, + "mean_token_accuracy": 0.6831061840057373, + "num_tokens": 21614569.0, + "step": 836 + }, + { + "epoch": 0.09191741708763453, + "ewc_loss": 1.4603137969970703e-06, + "grad_norm": 2.6984596252441406, + "learning_rate": 3.0600292825768663e-07, + "loss": 1.162, + "mean_token_accuracy": 0.65671706199646, + "num_tokens": 21638089.0, + "step": 837 + }, + { + "epoch": 0.09202723479024819, + "ewc_loss": 1.4603137969970703e-06, + "grad_norm": 2.2249059677124023, + "learning_rate": 3.0636896046852124e-07, + "loss": 1.1063, + "mean_token_accuracy": 0.6868947148323059, + "num_tokens": 21668891.0, + "step": 838 + }, + { + "epoch": 0.09213705249286185, + "ewc_loss": 1.475214958190918e-06, + "grad_norm": 2.5341007709503174, + "learning_rate": 3.067349926793558e-07, + "loss": 1.1577, + "mean_token_accuracy": 0.6672496795654297, + "num_tokens": 21696713.0, + "step": 839 + }, + { + "epoch": 0.0922468701954755, + "ewc_loss": 1.4603137969970703e-06, + "grad_norm": 3.1170384883880615, + "learning_rate": 3.071010248901903e-07, + "loss": 1.0194, + "mean_token_accuracy": 0.7027544379234314, + "num_tokens": 21724588.0, + "step": 840 + }, + { + "epoch": 0.09235668789808917, + "ewc_loss": 1.475214958190918e-06, + "grad_norm": 3.1076810359954834, + "learning_rate": 3.074670571010249e-07, + "loss": 1.0824, + "mean_token_accuracy": 0.6930224299430847, + "num_tokens": 21745178.0, + "step": 841 + }, + { + "epoch": 0.09246650560070284, + "ewc_loss": 1.475214958190918e-06, + "grad_norm": 2.7095558643341064, + "learning_rate": 3.0783308931185943e-07, + "loss": 1.082, + "mean_token_accuracy": 0.6880662441253662, + "num_tokens": 21768870.0, + "step": 842 + }, + { + "epoch": 0.0925763233033165, + "ewc_loss": 1.475214958190918e-06, + "grad_norm": 2.51775860786438, + "learning_rate": 3.08199121522694e-07, + "loss": 1.068, + "mean_token_accuracy": 0.6867291927337646, + "num_tokens": 21796552.0, + "step": 843 + }, + { + "epoch": 0.09268614100593016, + "ewc_loss": 1.4901161193847656e-06, + "grad_norm": 2.4655089378356934, + "learning_rate": 3.0856515373352853e-07, + "loss": 1.1399, + "mean_token_accuracy": 0.6733516454696655, + "num_tokens": 21825989.0, + "step": 844 + }, + { + "epoch": 0.09279595870854382, + "ewc_loss": 1.4975666999816895e-06, + "grad_norm": 2.8792407512664795, + "learning_rate": 3.089311859443631e-07, + "loss": 1.0468, + "mean_token_accuracy": 0.6906084418296814, + "num_tokens": 21848075.0, + "step": 845 + }, + { + "epoch": 0.09290577641115748, + "ewc_loss": 1.4975666999816895e-06, + "grad_norm": 2.4614028930664062, + "learning_rate": 3.092972181551977e-07, + "loss": 1.1432, + "mean_token_accuracy": 0.6723535060882568, + "num_tokens": 21877666.0, + "step": 846 + }, + { + "epoch": 0.09301559411377114, + "ewc_loss": 1.4975666999816895e-06, + "grad_norm": 2.457052230834961, + "learning_rate": 3.096632503660322e-07, + "loss": 1.0926, + "mean_token_accuracy": 0.6778676509857178, + "num_tokens": 21904463.0, + "step": 847 + }, + { + "epoch": 0.0931254118163848, + "ewc_loss": 1.4975666999816895e-06, + "grad_norm": 2.746173143386841, + "learning_rate": 3.1002928257686673e-07, + "loss": 1.0682, + "mean_token_accuracy": 0.6813949346542358, + "num_tokens": 21927852.0, + "step": 848 + }, + { + "epoch": 0.09323522951899846, + "ewc_loss": 1.5050172805786133e-06, + "grad_norm": 2.4741334915161133, + "learning_rate": 3.1039531478770133e-07, + "loss": 1.0635, + "mean_token_accuracy": 0.6945946216583252, + "num_tokens": 21956317.0, + "step": 849 + }, + { + "epoch": 0.09334504722161212, + "ewc_loss": 1.5050172805786133e-06, + "grad_norm": 2.6173715591430664, + "learning_rate": 3.107613469985358e-07, + "loss": 1.0641, + "mean_token_accuracy": 0.6832833290100098, + "num_tokens": 21981920.0, + "step": 850 + }, + { + "epoch": 0.09345486492422579, + "ewc_loss": 1.5124678611755371e-06, + "grad_norm": 2.8821771144866943, + "learning_rate": 3.1112737920937043e-07, + "loss": 1.1265, + "mean_token_accuracy": 0.6713647246360779, + "num_tokens": 22006163.0, + "step": 851 + }, + { + "epoch": 0.09356468262683945, + "ewc_loss": 1.5050172805786133e-06, + "grad_norm": 2.982027769088745, + "learning_rate": 3.11493411420205e-07, + "loss": 1.0982, + "mean_token_accuracy": 0.6761182546615601, + "num_tokens": 22029906.0, + "step": 852 + }, + { + "epoch": 0.09367450032945311, + "ewc_loss": 1.5124678611755371e-06, + "grad_norm": 2.755378484725952, + "learning_rate": 3.1185944363103947e-07, + "loss": 1.1368, + "mean_token_accuracy": 0.6746012568473816, + "num_tokens": 22055611.0, + "step": 853 + }, + { + "epoch": 0.09378431803206677, + "ewc_loss": 1.5050172805786133e-06, + "grad_norm": 2.5326690673828125, + "learning_rate": 3.122254758418741e-07, + "loss": 1.0855, + "mean_token_accuracy": 0.6741873025894165, + "num_tokens": 22081614.0, + "step": 854 + }, + { + "epoch": 0.09389413573468043, + "ewc_loss": 1.519918441772461e-06, + "grad_norm": 2.822035789489746, + "learning_rate": 3.125915080527086e-07, + "loss": 1.0691, + "mean_token_accuracy": 0.6815604567527771, + "num_tokens": 22104956.0, + "step": 855 + }, + { + "epoch": 0.09400395343729409, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.348663806915283, + "learning_rate": 3.1295754026354317e-07, + "loss": 1.1163, + "mean_token_accuracy": 0.672010064125061, + "num_tokens": 22135119.0, + "step": 856 + }, + { + "epoch": 0.09411377113990775, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.8682570457458496, + "learning_rate": 3.133235724743777e-07, + "loss": 1.1317, + "mean_token_accuracy": 0.6760078072547913, + "num_tokens": 22157625.0, + "step": 857 + }, + { + "epoch": 0.09422358884252141, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.387507200241089, + "learning_rate": 3.1368960468521227e-07, + "loss": 1.105, + "mean_token_accuracy": 0.6809611916542053, + "num_tokens": 22188178.0, + "step": 858 + }, + { + "epoch": 0.09433340654513507, + "ewc_loss": 1.5348196029663086e-06, + "grad_norm": 2.507063150405884, + "learning_rate": 3.1405563689604687e-07, + "loss": 1.1076, + "mean_token_accuracy": 0.6733555197715759, + "num_tokens": 22213762.0, + "step": 859 + }, + { + "epoch": 0.09444322424774873, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.773329257965088, + "learning_rate": 3.1442166910688137e-07, + "loss": 1.115, + "mean_token_accuracy": 0.6769521832466125, + "num_tokens": 22235680.0, + "step": 860 + }, + { + "epoch": 0.0945530419503624, + "ewc_loss": 1.519918441772461e-06, + "grad_norm": 2.765474796295166, + "learning_rate": 3.147877013177159e-07, + "loss": 1.0439, + "mean_token_accuracy": 0.6919921636581421, + "num_tokens": 22258638.0, + "step": 861 + }, + { + "epoch": 0.09466285965297606, + "ewc_loss": 1.519918441772461e-06, + "grad_norm": 2.7963521480560303, + "learning_rate": 3.151537335285505e-07, + "loss": 1.087, + "mean_token_accuracy": 0.681768000125885, + "num_tokens": 22282129.0, + "step": 862 + }, + { + "epoch": 0.09477267735558972, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.3499584197998047, + "learning_rate": 3.1551976573938507e-07, + "loss": 1.0067, + "mean_token_accuracy": 0.7017791271209717, + "num_tokens": 22312222.0, + "step": 863 + }, + { + "epoch": 0.09488249505820338, + "ewc_loss": 1.519918441772461e-06, + "grad_norm": 2.885300397872925, + "learning_rate": 3.1588579795021956e-07, + "loss": 1.1007, + "mean_token_accuracy": 0.6739199161529541, + "num_tokens": 22335476.0, + "step": 864 + }, + { + "epoch": 0.09499231276081704, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.540010690689087, + "learning_rate": 3.1625183016105417e-07, + "loss": 1.1538, + "mean_token_accuracy": 0.6845405101776123, + "num_tokens": 22362628.0, + "step": 865 + }, + { + "epoch": 0.0951021304634307, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.6913630962371826, + "learning_rate": 3.166178623718887e-07, + "loss": 1.1547, + "mean_token_accuracy": 0.6685677766799927, + "num_tokens": 22388402.0, + "step": 866 + }, + { + "epoch": 0.09521194816604436, + "ewc_loss": 1.5273690223693848e-06, + "grad_norm": 2.396263837814331, + "learning_rate": 3.1698389458272327e-07, + "loss": 1.1339, + "mean_token_accuracy": 0.6763401031494141, + "num_tokens": 22417436.0, + "step": 867 + }, + { + "epoch": 0.09532176586865802, + "ewc_loss": 1.5422701835632324e-06, + "grad_norm": 2.8437469005584717, + "learning_rate": 3.173499267935578e-07, + "loss": 1.1613, + "mean_token_accuracy": 0.663306474685669, + "num_tokens": 22440356.0, + "step": 868 + }, + { + "epoch": 0.09543158357127168, + "ewc_loss": 1.5348196029663086e-06, + "grad_norm": 2.8589861392974854, + "learning_rate": 3.1771595900439236e-07, + "loss": 1.1567, + "mean_token_accuracy": 0.6724355816841125, + "num_tokens": 22462445.0, + "step": 869 + }, + { + "epoch": 0.09554140127388536, + "ewc_loss": 1.5348196029663086e-06, + "grad_norm": 2.3000967502593994, + "learning_rate": 3.1808199121522697e-07, + "loss": 1.1531, + "mean_token_accuracy": 0.6672874689102173, + "num_tokens": 22495714.0, + "step": 870 + }, + { + "epoch": 0.09565121897649902, + "ewc_loss": 1.5348196029663086e-06, + "grad_norm": 2.5619499683380127, + "learning_rate": 3.1844802342606146e-07, + "loss": 1.1375, + "mean_token_accuracy": 0.6686104536056519, + "num_tokens": 22520890.0, + "step": 871 + }, + { + "epoch": 0.09576103667911268, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.3863277435302734, + "learning_rate": 3.18814055636896e-07, + "loss": 1.1099, + "mean_token_accuracy": 0.6825809478759766, + "num_tokens": 22548125.0, + "step": 872 + }, + { + "epoch": 0.09587085438172634, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.6956698894500732, + "learning_rate": 3.191800878477306e-07, + "loss": 1.0546, + "mean_token_accuracy": 0.6957818269729614, + "num_tokens": 22571203.0, + "step": 873 + }, + { + "epoch": 0.09598067208434, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.518561601638794, + "learning_rate": 3.195461200585651e-07, + "loss": 1.0793, + "mean_token_accuracy": 0.6806507110595703, + "num_tokens": 22600078.0, + "step": 874 + }, + { + "epoch": 0.09609048978695366, + "ewc_loss": 1.5422701835632324e-06, + "grad_norm": 2.6831908226013184, + "learning_rate": 3.199121522693997e-07, + "loss": 1.12, + "mean_token_accuracy": 0.6811941266059875, + "num_tokens": 22626236.0, + "step": 875 + }, + { + "epoch": 0.09620030748956732, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.437100410461426, + "learning_rate": 3.2027818448023426e-07, + "loss": 1.1076, + "mean_token_accuracy": 0.6732855439186096, + "num_tokens": 22653233.0, + "step": 876 + }, + { + "epoch": 0.09631012519218098, + "ewc_loss": 1.5422701835632324e-06, + "grad_norm": 2.5075721740722656, + "learning_rate": 3.2064421669106876e-07, + "loss": 1.1249, + "mean_token_accuracy": 0.6693416833877563, + "num_tokens": 22679272.0, + "step": 877 + }, + { + "epoch": 0.09641994289479464, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.6547462940216064, + "learning_rate": 3.2101024890190336e-07, + "loss": 1.0623, + "mean_token_accuracy": 0.689916729927063, + "num_tokens": 22704003.0, + "step": 878 + }, + { + "epoch": 0.0965297605974083, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.5313222408294678, + "learning_rate": 3.213762811127379e-07, + "loss": 1.0486, + "mean_token_accuracy": 0.6934205293655396, + "num_tokens": 22729119.0, + "step": 879 + }, + { + "epoch": 0.09663957830002197, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.54244065284729, + "learning_rate": 3.2174231332357246e-07, + "loss": 1.1265, + "mean_token_accuracy": 0.6738234758377075, + "num_tokens": 22755678.0, + "step": 880 + }, + { + "epoch": 0.09674939600263563, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.2854676246643066, + "learning_rate": 3.22108345534407e-07, + "loss": 1.103, + "mean_token_accuracy": 0.6792607307434082, + "num_tokens": 22788600.0, + "step": 881 + }, + { + "epoch": 0.09685921370524929, + "ewc_loss": 1.5497207641601562e-06, + "grad_norm": 2.4138851165771484, + "learning_rate": 3.2247437774524155e-07, + "loss": 1.0248, + "mean_token_accuracy": 0.6969621181488037, + "num_tokens": 22814636.0, + "step": 882 + }, + { + "epoch": 0.09696903140786295, + "ewc_loss": 1.564621925354004e-06, + "grad_norm": 3.086933135986328, + "learning_rate": 3.2284040995607616e-07, + "loss": 1.1397, + "mean_token_accuracy": 0.6690537333488464, + "num_tokens": 22839289.0, + "step": 883 + }, + { + "epoch": 0.09707884911047661, + "ewc_loss": 1.55717134475708e-06, + "grad_norm": 2.59019136428833, + "learning_rate": 3.2320644216691065e-07, + "loss": 1.0896, + "mean_token_accuracy": 0.6773701906204224, + "num_tokens": 22864607.0, + "step": 884 + }, + { + "epoch": 0.09718866681309027, + "ewc_loss": 1.55717134475708e-06, + "grad_norm": 2.6274139881134033, + "learning_rate": 3.235724743777452e-07, + "loss": 1.139, + "mean_token_accuracy": 0.6818122863769531, + "num_tokens": 22892406.0, + "step": 885 + }, + { + "epoch": 0.09729848451570393, + "ewc_loss": 1.564621925354004e-06, + "grad_norm": 2.740365982055664, + "learning_rate": 3.239385065885798e-07, + "loss": 1.1132, + "mean_token_accuracy": 0.6712632179260254, + "num_tokens": 22917563.0, + "step": 886 + }, + { + "epoch": 0.09740830221831759, + "ewc_loss": 1.564621925354004e-06, + "grad_norm": 2.4781224727630615, + "learning_rate": 3.2430453879941435e-07, + "loss": 1.0888, + "mean_token_accuracy": 0.6817207336425781, + "num_tokens": 22944690.0, + "step": 887 + }, + { + "epoch": 0.09751811992093125, + "ewc_loss": 1.564621925354004e-06, + "grad_norm": 2.575669288635254, + "learning_rate": 3.2467057101024885e-07, + "loss": 1.0772, + "mean_token_accuracy": 0.6858739256858826, + "num_tokens": 22970435.0, + "step": 888 + }, + { + "epoch": 0.09762793762354492, + "ewc_loss": 1.5720725059509277e-06, + "grad_norm": 2.554760456085205, + "learning_rate": 3.2503660322108345e-07, + "loss": 1.0478, + "mean_token_accuracy": 0.6952900886535645, + "num_tokens": 22996469.0, + "step": 889 + }, + { + "epoch": 0.09773775532615858, + "ewc_loss": 1.5720725059509277e-06, + "grad_norm": 2.7166478633880615, + "learning_rate": 3.25402635431918e-07, + "loss": 1.1693, + "mean_token_accuracy": 0.6695389747619629, + "num_tokens": 23019491.0, + "step": 890 + }, + { + "epoch": 0.09784757302877224, + "ewc_loss": 1.5720725059509277e-06, + "grad_norm": 2.401074171066284, + "learning_rate": 3.2576866764275255e-07, + "loss": 1.0869, + "mean_token_accuracy": 0.6876906156539917, + "num_tokens": 23048481.0, + "step": 891 + }, + { + "epoch": 0.0979573907313859, + "ewc_loss": 1.5720725059509277e-06, + "grad_norm": 2.4847986698150635, + "learning_rate": 3.261346998535871e-07, + "loss": 1.0889, + "mean_token_accuracy": 0.6804201602935791, + "num_tokens": 23074453.0, + "step": 892 + }, + { + "epoch": 0.09806720843399956, + "ewc_loss": 1.5720725059509277e-06, + "grad_norm": 2.3569388389587402, + "learning_rate": 3.2650073206442165e-07, + "loss": 1.0003, + "mean_token_accuracy": 0.7043098211288452, + "num_tokens": 23103151.0, + "step": 893 + }, + { + "epoch": 0.09817702613661322, + "ewc_loss": 1.5720725059509277e-06, + "grad_norm": 2.3213648796081543, + "learning_rate": 3.2686676427525625e-07, + "loss": 1.1106, + "mean_token_accuracy": 0.6750556826591492, + "num_tokens": 23134289.0, + "step": 894 + }, + { + "epoch": 0.09828684383922688, + "ewc_loss": 1.5795230865478516e-06, + "grad_norm": 2.76297664642334, + "learning_rate": 3.2723279648609075e-07, + "loss": 1.0398, + "mean_token_accuracy": 0.6994891166687012, + "num_tokens": 23156101.0, + "step": 895 + }, + { + "epoch": 0.09839666154184054, + "ewc_loss": 1.5869736671447754e-06, + "grad_norm": 2.6416637897491455, + "learning_rate": 3.275988286969253e-07, + "loss": 1.0134, + "mean_token_accuracy": 0.6998195052146912, + "num_tokens": 23179178.0, + "step": 896 + }, + { + "epoch": 0.0985064792444542, + "ewc_loss": 1.5869736671447754e-06, + "grad_norm": 2.823021411895752, + "learning_rate": 3.279648609077599e-07, + "loss": 1.1142, + "mean_token_accuracy": 0.6763421297073364, + "num_tokens": 23200455.0, + "step": 897 + }, + { + "epoch": 0.09861629694706786, + "ewc_loss": 1.601874828338623e-06, + "grad_norm": 2.5804717540740967, + "learning_rate": 3.283308931185944e-07, + "loss": 1.1072, + "mean_token_accuracy": 0.6932299137115479, + "num_tokens": 23226237.0, + "step": 898 + }, + { + "epoch": 0.09872611464968153, + "ewc_loss": 1.601874828338623e-06, + "grad_norm": 2.4808008670806885, + "learning_rate": 3.28696925329429e-07, + "loss": 1.0937, + "mean_token_accuracy": 0.6866647005081177, + "num_tokens": 23251813.0, + "step": 899 + }, + { + "epoch": 0.0988359323522952, + "ewc_loss": 1.601874828338623e-06, + "grad_norm": 2.4927616119384766, + "learning_rate": 3.2906295754026354e-07, + "loss": 1.1569, + "mean_token_accuracy": 0.6580684185028076, + "num_tokens": 23278810.0, + "step": 900 + }, + { + "epoch": 0.09894575005490885, + "ewc_loss": 1.601874828338623e-06, + "grad_norm": 2.4376590251922607, + "learning_rate": 3.2942898975109804e-07, + "loss": 1.1435, + "mean_token_accuracy": 0.6731738448143005, + "num_tokens": 23307162.0, + "step": 901 + }, + { + "epoch": 0.09905556775752251, + "ewc_loss": 1.6093254089355469e-06, + "grad_norm": 2.39223313331604, + "learning_rate": 3.2979502196193264e-07, + "loss": 1.1543, + "mean_token_accuracy": 0.6657078266143799, + "num_tokens": 23335308.0, + "step": 902 + }, + { + "epoch": 0.09916538546013617, + "ewc_loss": 1.6167759895324707e-06, + "grad_norm": 2.5814013481140137, + "learning_rate": 3.301610541727672e-07, + "loss": 1.0555, + "mean_token_accuracy": 0.6929019689559937, + "num_tokens": 23359068.0, + "step": 903 + }, + { + "epoch": 0.09927520316274983, + "ewc_loss": 1.6316771507263184e-06, + "grad_norm": 2.5813405513763428, + "learning_rate": 3.3052708638360174e-07, + "loss": 1.0196, + "mean_token_accuracy": 0.70587158203125, + "num_tokens": 23381987.0, + "step": 904 + }, + { + "epoch": 0.0993850208653635, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.37391996383667, + "learning_rate": 3.308931185944363e-07, + "loss": 1.1331, + "mean_token_accuracy": 0.6707707643508911, + "num_tokens": 23420367.0, + "step": 905 + }, + { + "epoch": 0.09949483856797715, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.7696354389190674, + "learning_rate": 3.3125915080527084e-07, + "loss": 1.0733, + "mean_token_accuracy": 0.6888932585716248, + "num_tokens": 23442587.0, + "step": 906 + }, + { + "epoch": 0.09960465627059081, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.5135438442230225, + "learning_rate": 3.3162518301610544e-07, + "loss": 1.1407, + "mean_token_accuracy": 0.67292320728302, + "num_tokens": 23469268.0, + "step": 907 + }, + { + "epoch": 0.09971447397320449, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.33807373046875, + "learning_rate": 3.3199121522693994e-07, + "loss": 1.2019, + "mean_token_accuracy": 0.6471982002258301, + "num_tokens": 23500886.0, + "step": 908 + }, + { + "epoch": 0.09982429167581815, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.7452011108398438, + "learning_rate": 3.323572474377745e-07, + "loss": 1.1044, + "mean_token_accuracy": 0.6824062466621399, + "num_tokens": 23523429.0, + "step": 909 + }, + { + "epoch": 0.0999341093784318, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.639899969100952, + "learning_rate": 3.327232796486091e-07, + "loss": 1.0528, + "mean_token_accuracy": 0.6980099678039551, + "num_tokens": 23545940.0, + "step": 910 + }, + { + "epoch": 0.10004392708104547, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.6142942905426025, + "learning_rate": 3.3308931185944364e-07, + "loss": 1.1972, + "mean_token_accuracy": 0.6595012545585632, + "num_tokens": 23572000.0, + "step": 911 + }, + { + "epoch": 0.10015374478365913, + "ewc_loss": 1.6391277313232422e-06, + "grad_norm": 2.4270517826080322, + "learning_rate": 3.3345534407027813e-07, + "loss": 1.1205, + "mean_token_accuracy": 0.6774054765701294, + "num_tokens": 23600798.0, + "step": 912 + }, + { + "epoch": 0.10026356248627279, + "ewc_loss": 1.646578311920166e-06, + "grad_norm": 2.597285032272339, + "learning_rate": 3.3382137628111273e-07, + "loss": 1.0353, + "mean_token_accuracy": 0.69464510679245, + "num_tokens": 23624264.0, + "step": 913 + }, + { + "epoch": 0.10037338018888645, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.5085105895996094, + "learning_rate": 3.341874084919473e-07, + "loss": 1.0745, + "mean_token_accuracy": 0.6845740675926208, + "num_tokens": 23655171.0, + "step": 914 + }, + { + "epoch": 0.1004831978915001, + "ewc_loss": 1.646578311920166e-06, + "grad_norm": 2.4602839946746826, + "learning_rate": 3.3455344070278183e-07, + "loss": 1.1349, + "mean_token_accuracy": 0.6716212630271912, + "num_tokens": 23683903.0, + "step": 915 + }, + { + "epoch": 0.10059301559411377, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.7040727138519287, + "learning_rate": 3.349194729136164e-07, + "loss": 1.0695, + "mean_token_accuracy": 0.6923830509185791, + "num_tokens": 23706406.0, + "step": 916 + }, + { + "epoch": 0.10070283329672743, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.624464988708496, + "learning_rate": 3.3528550512445093e-07, + "loss": 1.1383, + "mean_token_accuracy": 0.6659623980522156, + "num_tokens": 23732433.0, + "step": 917 + }, + { + "epoch": 0.1008126509993411, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.4005019664764404, + "learning_rate": 3.356515373352855e-07, + "loss": 1.0432, + "mean_token_accuracy": 0.6919375658035278, + "num_tokens": 23759236.0, + "step": 918 + }, + { + "epoch": 0.10092246870195476, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.8870856761932373, + "learning_rate": 3.3601756954612003e-07, + "loss": 1.0333, + "mean_token_accuracy": 0.6969932317733765, + "num_tokens": 23784398.0, + "step": 919 + }, + { + "epoch": 0.10103228640456842, + "ewc_loss": 1.646578311920166e-06, + "grad_norm": 2.4785757064819336, + "learning_rate": 3.363836017569546e-07, + "loss": 1.0165, + "mean_token_accuracy": 0.6976145505905151, + "num_tokens": 23812650.0, + "step": 920 + }, + { + "epoch": 0.10114210410718208, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.4054181575775146, + "learning_rate": 3.367496339677892e-07, + "loss": 1.1046, + "mean_token_accuracy": 0.6795263290405273, + "num_tokens": 23841776.0, + "step": 921 + }, + { + "epoch": 0.10125192180979574, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.646782398223877, + "learning_rate": 3.371156661786237e-07, + "loss": 1.0906, + "mean_token_accuracy": 0.6779608726501465, + "num_tokens": 23865229.0, + "step": 922 + }, + { + "epoch": 0.1013617395124094, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.5878515243530273, + "learning_rate": 3.374816983894583e-07, + "loss": 1.0943, + "mean_token_accuracy": 0.67315673828125, + "num_tokens": 23889843.0, + "step": 923 + }, + { + "epoch": 0.10147155721502306, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.585052013397217, + "learning_rate": 3.3784773060029283e-07, + "loss": 1.1724, + "mean_token_accuracy": 0.6643416285514832, + "num_tokens": 23917810.0, + "step": 924 + }, + { + "epoch": 0.10158137491763672, + "ewc_loss": 1.6540288925170898e-06, + "grad_norm": 2.72701358795166, + "learning_rate": 3.382137628111273e-07, + "loss": 1.0945, + "mean_token_accuracy": 0.6786898970603943, + "num_tokens": 23940531.0, + "step": 925 + }, + { + "epoch": 0.10169119262025038, + "ewc_loss": 1.6614794731140137e-06, + "grad_norm": 2.3735246658325195, + "learning_rate": 3.385797950219619e-07, + "loss": 1.086, + "mean_token_accuracy": 0.6817498207092285, + "num_tokens": 23968313.0, + "step": 926 + }, + { + "epoch": 0.10180101032286405, + "ewc_loss": 1.6614794731140137e-06, + "grad_norm": 2.347346067428589, + "learning_rate": 3.389458272327965e-07, + "loss": 1.1291, + "mean_token_accuracy": 0.6749516725540161, + "num_tokens": 23999841.0, + "step": 927 + }, + { + "epoch": 0.10191082802547771, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.357720375061035, + "learning_rate": 3.39311859443631e-07, + "loss": 1.1511, + "mean_token_accuracy": 0.6629819273948669, + "num_tokens": 24031207.0, + "step": 928 + }, + { + "epoch": 0.10202064572809137, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.2888853549957275, + "learning_rate": 3.3967789165446557e-07, + "loss": 1.1364, + "mean_token_accuracy": 0.6713900566101074, + "num_tokens": 24060419.0, + "step": 929 + }, + { + "epoch": 0.10213046343070503, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.571558713912964, + "learning_rate": 3.400439238653001e-07, + "loss": 1.1038, + "mean_token_accuracy": 0.6849402785301208, + "num_tokens": 24086752.0, + "step": 930 + }, + { + "epoch": 0.10224028113331869, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.6351962089538574, + "learning_rate": 3.404099560761347e-07, + "loss": 1.052, + "mean_token_accuracy": 0.6860665082931519, + "num_tokens": 24110671.0, + "step": 931 + }, + { + "epoch": 0.10235009883593235, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.35817289352417, + "learning_rate": 3.407759882869692e-07, + "loss": 1.0965, + "mean_token_accuracy": 0.6781187653541565, + "num_tokens": 24142021.0, + "step": 932 + }, + { + "epoch": 0.10245991653854601, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.8657889366149902, + "learning_rate": 3.4114202049780377e-07, + "loss": 1.1008, + "mean_token_accuracy": 0.6788947582244873, + "num_tokens": 24163153.0, + "step": 933 + }, + { + "epoch": 0.10256973424115967, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.3352675437927246, + "learning_rate": 3.4150805270863837e-07, + "loss": 1.1643, + "mean_token_accuracy": 0.6616929769515991, + "num_tokens": 24194395.0, + "step": 934 + }, + { + "epoch": 0.10267955194377333, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.554675579071045, + "learning_rate": 3.4187408491947287e-07, + "loss": 1.1486, + "mean_token_accuracy": 0.6646515727043152, + "num_tokens": 24220717.0, + "step": 935 + }, + { + "epoch": 0.10278936964638699, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.582496404647827, + "learning_rate": 3.422401171303074e-07, + "loss": 1.1099, + "mean_token_accuracy": 0.6817057132720947, + "num_tokens": 24245109.0, + "step": 936 + }, + { + "epoch": 0.10289918734900066, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.9820899963378906, + "learning_rate": 3.42606149341142e-07, + "loss": 1.179, + "mean_token_accuracy": 0.6668309569358826, + "num_tokens": 24264983.0, + "step": 937 + }, + { + "epoch": 0.10300900505161432, + "ewc_loss": 1.6763806343078613e-06, + "grad_norm": 2.707554340362549, + "learning_rate": 3.4297218155197657e-07, + "loss": 1.1017, + "mean_token_accuracy": 0.6868175268173218, + "num_tokens": 24288101.0, + "step": 938 + }, + { + "epoch": 0.10311882275422798, + "ewc_loss": 1.6689300537109375e-06, + "grad_norm": 2.454665422439575, + "learning_rate": 3.433382137628111e-07, + "loss": 1.0721, + "mean_token_accuracy": 0.6850851774215698, + "num_tokens": 24314589.0, + "step": 939 + }, + { + "epoch": 0.10322864045684164, + "ewc_loss": 1.6763806343078613e-06, + "grad_norm": 2.697195529937744, + "learning_rate": 3.4370424597364567e-07, + "loss": 1.1531, + "mean_token_accuracy": 0.6741743087768555, + "num_tokens": 24337109.0, + "step": 940 + }, + { + "epoch": 0.1033384581594553, + "ewc_loss": 1.6763806343078613e-06, + "grad_norm": 2.3596646785736084, + "learning_rate": 3.440702781844802e-07, + "loss": 1.1403, + "mean_token_accuracy": 0.6737672090530396, + "num_tokens": 24367221.0, + "step": 941 + }, + { + "epoch": 0.10344827586206896, + "ewc_loss": 1.6763806343078613e-06, + "grad_norm": 2.4698727130889893, + "learning_rate": 3.4443631039531476e-07, + "loss": 1.1593, + "mean_token_accuracy": 0.6638023257255554, + "num_tokens": 24394210.0, + "step": 942 + }, + { + "epoch": 0.10355809356468262, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.49001145362854, + "learning_rate": 3.448023426061493e-07, + "loss": 1.0249, + "mean_token_accuracy": 0.6955649852752686, + "num_tokens": 24419005.0, + "step": 943 + }, + { + "epoch": 0.10366791126729628, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.367703437805176, + "learning_rate": 3.4516837481698386e-07, + "loss": 1.108, + "mean_token_accuracy": 0.6797807216644287, + "num_tokens": 24448427.0, + "step": 944 + }, + { + "epoch": 0.10377772896990994, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.2412188053131104, + "learning_rate": 3.4553440702781846e-07, + "loss": 1.2123, + "mean_token_accuracy": 0.6511927843093872, + "num_tokens": 24480209.0, + "step": 945 + }, + { + "epoch": 0.10388754667252362, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.631147861480713, + "learning_rate": 3.4590043923865296e-07, + "loss": 1.0376, + "mean_token_accuracy": 0.6971915364265442, + "num_tokens": 24503484.0, + "step": 946 + }, + { + "epoch": 0.10399736437513728, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.558112144470215, + "learning_rate": 3.4626647144948756e-07, + "loss": 1.113, + "mean_token_accuracy": 0.6735547780990601, + "num_tokens": 24528827.0, + "step": 947 + }, + { + "epoch": 0.10410718207775094, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.866147994995117, + "learning_rate": 3.466325036603221e-07, + "loss": 1.0455, + "mean_token_accuracy": 0.6972007751464844, + "num_tokens": 24550278.0, + "step": 948 + }, + { + "epoch": 0.1042169997803646, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.6436798572540283, + "learning_rate": 3.469985358711566e-07, + "loss": 1.018, + "mean_token_accuracy": 0.6965538859367371, + "num_tokens": 24573147.0, + "step": 949 + }, + { + "epoch": 0.10432681748297826, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.666837215423584, + "learning_rate": 3.473645680819912e-07, + "loss": 1.085, + "mean_token_accuracy": 0.6763901114463806, + "num_tokens": 24597279.0, + "step": 950 + }, + { + "epoch": 0.10443663518559192, + "ewc_loss": 1.6838312149047852e-06, + "grad_norm": 2.59039044380188, + "learning_rate": 3.4773060029282576e-07, + "loss": 1.1337, + "mean_token_accuracy": 0.6611204743385315, + "num_tokens": 24623328.0, + "step": 951 + }, + { + "epoch": 0.10454645288820558, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.8583719730377197, + "learning_rate": 3.480966325036603e-07, + "loss": 1.0818, + "mean_token_accuracy": 0.6931122541427612, + "num_tokens": 24644301.0, + "step": 952 + }, + { + "epoch": 0.10465627059081924, + "ewc_loss": 1.691281795501709e-06, + "grad_norm": 2.1444947719573975, + "learning_rate": 3.4846266471449486e-07, + "loss": 1.1456, + "mean_token_accuracy": 0.6687964200973511, + "num_tokens": 24678615.0, + "step": 953 + }, + { + "epoch": 0.1047660882934329, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.773209571838379, + "learning_rate": 3.488286969253294e-07, + "loss": 0.9852, + "mean_token_accuracy": 0.7061529755592346, + "num_tokens": 24701336.0, + "step": 954 + }, + { + "epoch": 0.10487590599604656, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.5274815559387207, + "learning_rate": 3.49194729136164e-07, + "loss": 1.0442, + "mean_token_accuracy": 0.6940749883651733, + "num_tokens": 24727917.0, + "step": 955 + }, + { + "epoch": 0.10498572369866023, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.8544726371765137, + "learning_rate": 3.495607613469985e-07, + "loss": 1.0379, + "mean_token_accuracy": 0.6896399855613708, + "num_tokens": 24748147.0, + "step": 956 + }, + { + "epoch": 0.10509554140127389, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.5930216312408447, + "learning_rate": 3.4992679355783305e-07, + "loss": 1.077, + "mean_token_accuracy": 0.6896052360534668, + "num_tokens": 24773384.0, + "step": 957 + }, + { + "epoch": 0.10520535910388755, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.4209330081939697, + "learning_rate": 3.5029282576866766e-07, + "loss": 1.1117, + "mean_token_accuracy": 0.6721972227096558, + "num_tokens": 24801521.0, + "step": 958 + }, + { + "epoch": 0.10531517680650121, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.5052437782287598, + "learning_rate": 3.5065885797950215e-07, + "loss": 1.1331, + "mean_token_accuracy": 0.6739596128463745, + "num_tokens": 24830015.0, + "step": 959 + }, + { + "epoch": 0.10542499450911487, + "ewc_loss": 1.6987323760986328e-06, + "grad_norm": 2.631153106689453, + "learning_rate": 3.510248901903367e-07, + "loss": 1.0948, + "mean_token_accuracy": 0.6775557994842529, + "num_tokens": 24853232.0, + "step": 960 + }, + { + "epoch": 0.10553481221172853, + "ewc_loss": 1.7061829566955566e-06, + "grad_norm": 2.560682773590088, + "learning_rate": 3.513909224011713e-07, + "loss": 1.1543, + "mean_token_accuracy": 0.6677340865135193, + "num_tokens": 24877653.0, + "step": 961 + }, + { + "epoch": 0.10564462991434219, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.6725358963012695, + "learning_rate": 3.5175695461200585e-07, + "loss": 1.1355, + "mean_token_accuracy": 0.6697906851768494, + "num_tokens": 24903434.0, + "step": 962 + }, + { + "epoch": 0.10575444761695585, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 3.546147346496582, + "learning_rate": 3.521229868228404e-07, + "loss": 1.1707, + "mean_token_accuracy": 0.6685020923614502, + "num_tokens": 24931347.0, + "step": 963 + }, + { + "epoch": 0.10586426531956951, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.7325334548950195, + "learning_rate": 3.5248901903367495e-07, + "loss": 1.071, + "mean_token_accuracy": 0.6885442733764648, + "num_tokens": 24955365.0, + "step": 964 + }, + { + "epoch": 0.10597408302218318, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.8082733154296875, + "learning_rate": 3.528550512445095e-07, + "loss": 1.1599, + "mean_token_accuracy": 0.6619488596916199, + "num_tokens": 24977249.0, + "step": 965 + }, + { + "epoch": 0.10608390072479684, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.3610243797302246, + "learning_rate": 3.5322108345534405e-07, + "loss": 1.1699, + "mean_token_accuracy": 0.6655440926551819, + "num_tokens": 25008578.0, + "step": 966 + }, + { + "epoch": 0.1061937184274105, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.860365152359009, + "learning_rate": 3.535871156661786e-07, + "loss": 1.0322, + "mean_token_accuracy": 0.6880861520767212, + "num_tokens": 25028258.0, + "step": 967 + }, + { + "epoch": 0.10630353613002416, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.6141936779022217, + "learning_rate": 3.5395314787701315e-07, + "loss": 1.0613, + "mean_token_accuracy": 0.6888270974159241, + "num_tokens": 25052108.0, + "step": 968 + }, + { + "epoch": 0.10641335383263782, + "ewc_loss": 1.7061829566955566e-06, + "grad_norm": 2.163766860961914, + "learning_rate": 3.5431918008784775e-07, + "loss": 1.1223, + "mean_token_accuracy": 0.6777603626251221, + "num_tokens": 25085379.0, + "step": 969 + }, + { + "epoch": 0.10652317153525148, + "ewc_loss": 1.7061829566955566e-06, + "grad_norm": 2.664623260498047, + "learning_rate": 3.5468521229868224e-07, + "loss": 1.1006, + "mean_token_accuracy": 0.6808484792709351, + "num_tokens": 25108319.0, + "step": 970 + }, + { + "epoch": 0.10663298923786514, + "ewc_loss": 1.7061829566955566e-06, + "grad_norm": 2.3807525634765625, + "learning_rate": 3.5505124450951685e-07, + "loss": 1.2085, + "mean_token_accuracy": 0.6573866605758667, + "num_tokens": 25139744.0, + "step": 971 + }, + { + "epoch": 0.1067428069404788, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.442575216293335, + "learning_rate": 3.554172767203514e-07, + "loss": 1.09, + "mean_token_accuracy": 0.6809488534927368, + "num_tokens": 25166647.0, + "step": 972 + }, + { + "epoch": 0.10685262464309246, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.3818869590759277, + "learning_rate": 3.557833089311859e-07, + "loss": 1.0753, + "mean_token_accuracy": 0.6869714856147766, + "num_tokens": 25195503.0, + "step": 973 + }, + { + "epoch": 0.10696244234570612, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.575819730758667, + "learning_rate": 3.561493411420205e-07, + "loss": 1.0591, + "mean_token_accuracy": 0.6907937526702881, + "num_tokens": 25220138.0, + "step": 974 + }, + { + "epoch": 0.1070722600483198, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.533452033996582, + "learning_rate": 3.5651537335285504e-07, + "loss": 1.0701, + "mean_token_accuracy": 0.683269202709198, + "num_tokens": 25244278.0, + "step": 975 + }, + { + "epoch": 0.10718207775093345, + "ewc_loss": 1.7136335372924805e-06, + "grad_norm": 2.4602603912353516, + "learning_rate": 3.5688140556368954e-07, + "loss": 1.1093, + "mean_token_accuracy": 0.6696701645851135, + "num_tokens": 25275205.0, + "step": 976 + }, + { + "epoch": 0.10729189545354711, + "ewc_loss": 1.7285346984863281e-06, + "grad_norm": 2.7560536861419678, + "learning_rate": 3.5724743777452414e-07, + "loss": 1.0451, + "mean_token_accuracy": 0.6917582750320435, + "num_tokens": 25296551.0, + "step": 977 + }, + { + "epoch": 0.10740171315616077, + "ewc_loss": 1.7285346984863281e-06, + "grad_norm": 2.747159719467163, + "learning_rate": 3.576134699853587e-07, + "loss": 0.9553, + "mean_token_accuracy": 0.717395544052124, + "num_tokens": 25319702.0, + "step": 978 + }, + { + "epoch": 0.10751153085877443, + "ewc_loss": 1.7434358596801758e-06, + "grad_norm": 2.5569121837615967, + "learning_rate": 3.579795021961933e-07, + "loss": 1.1456, + "mean_token_accuracy": 0.6664362549781799, + "num_tokens": 25345390.0, + "step": 979 + }, + { + "epoch": 0.1076213485613881, + "ewc_loss": 1.7434358596801758e-06, + "grad_norm": 2.479515790939331, + "learning_rate": 3.583455344070278e-07, + "loss": 1.0704, + "mean_token_accuracy": 0.6848717927932739, + "num_tokens": 25371403.0, + "step": 980 + }, + { + "epoch": 0.10773116626400175, + "ewc_loss": 1.7434358596801758e-06, + "grad_norm": 2.9410831928253174, + "learning_rate": 3.5871156661786234e-07, + "loss": 1.022, + "mean_token_accuracy": 0.697364091873169, + "num_tokens": 25389771.0, + "step": 981 + }, + { + "epoch": 0.10784098396661541, + "ewc_loss": 1.7583370208740234e-06, + "grad_norm": 2.7461071014404297, + "learning_rate": 3.5907759882869694e-07, + "loss": 1.0592, + "mean_token_accuracy": 0.6868317127227783, + "num_tokens": 25410123.0, + "step": 982 + }, + { + "epoch": 0.10795080166922907, + "ewc_loss": 1.7434358596801758e-06, + "grad_norm": 2.691476821899414, + "learning_rate": 3.5944363103953144e-07, + "loss": 1.0682, + "mean_token_accuracy": 0.6821449995040894, + "num_tokens": 25432542.0, + "step": 983 + }, + { + "epoch": 0.10806061937184275, + "ewc_loss": 1.7583370208740234e-06, + "grad_norm": 2.4862217903137207, + "learning_rate": 3.59809663250366e-07, + "loss": 1.1179, + "mean_token_accuracy": 0.6700556874275208, + "num_tokens": 25458481.0, + "step": 984 + }, + { + "epoch": 0.10817043707445641, + "ewc_loss": 1.7583370208740234e-06, + "grad_norm": 2.5998876094818115, + "learning_rate": 3.601756954612006e-07, + "loss": 1.0373, + "mean_token_accuracy": 0.6994715332984924, + "num_tokens": 25483261.0, + "step": 985 + }, + { + "epoch": 0.10828025477707007, + "ewc_loss": 1.7583370208740234e-06, + "grad_norm": 2.560678720474243, + "learning_rate": 3.6054172767203514e-07, + "loss": 1.0381, + "mean_token_accuracy": 0.6978949308395386, + "num_tokens": 25511391.0, + "step": 986 + }, + { + "epoch": 0.10839007247968373, + "ewc_loss": 1.7657876014709473e-06, + "grad_norm": 2.3647379875183105, + "learning_rate": 3.609077598828697e-07, + "loss": 1.1142, + "mean_token_accuracy": 0.6750856041908264, + "num_tokens": 25541179.0, + "step": 987 + }, + { + "epoch": 0.10849989018229739, + "ewc_loss": 1.7657876014709473e-06, + "grad_norm": 2.6731457710266113, + "learning_rate": 3.6127379209370423e-07, + "loss": 1.0249, + "mean_token_accuracy": 0.6997569799423218, + "num_tokens": 25564057.0, + "step": 988 + }, + { + "epoch": 0.10860970788491105, + "ewc_loss": 1.7657876014709473e-06, + "grad_norm": 2.4868128299713135, + "learning_rate": 3.616398243045388e-07, + "loss": 1.0624, + "mean_token_accuracy": 0.6942200660705566, + "num_tokens": 25590617.0, + "step": 989 + }, + { + "epoch": 0.1087195255875247, + "ewc_loss": 1.7657876014709473e-06, + "grad_norm": 3.0752980709075928, + "learning_rate": 3.6200585651537333e-07, + "loss": 1.0679, + "mean_token_accuracy": 0.6898156404495239, + "num_tokens": 25610733.0, + "step": 990 + }, + { + "epoch": 0.10882934329013837, + "ewc_loss": 1.780688762664795e-06, + "grad_norm": 2.539466381072998, + "learning_rate": 3.623718887262079e-07, + "loss": 1.097, + "mean_token_accuracy": 0.6792514324188232, + "num_tokens": 25636827.0, + "step": 991 + }, + { + "epoch": 0.10893916099275203, + "ewc_loss": 1.780688762664795e-06, + "grad_norm": 2.694546937942505, + "learning_rate": 3.6273792093704243e-07, + "loss": 1.1297, + "mean_token_accuracy": 0.6742320656776428, + "num_tokens": 25659218.0, + "step": 992 + }, + { + "epoch": 0.10904897869536569, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.7398681640625, + "learning_rate": 3.6310395314787703e-07, + "loss": 1.0837, + "mean_token_accuracy": 0.682666540145874, + "num_tokens": 25682816.0, + "step": 993 + }, + { + "epoch": 0.10915879639797936, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.350788116455078, + "learning_rate": 3.6346998535871153e-07, + "loss": 1.1031, + "mean_token_accuracy": 0.6785271167755127, + "num_tokens": 25712238.0, + "step": 994 + }, + { + "epoch": 0.10926861410059302, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.505795478820801, + "learning_rate": 3.6383601756954613e-07, + "loss": 0.9975, + "mean_token_accuracy": 0.6975138187408447, + "num_tokens": 25737964.0, + "step": 995 + }, + { + "epoch": 0.10937843180320668, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.4064552783966064, + "learning_rate": 3.642020497803807e-07, + "loss": 1.0238, + "mean_token_accuracy": 0.6978960037231445, + "num_tokens": 25765993.0, + "step": 996 + }, + { + "epoch": 0.10948824950582034, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.9664463996887207, + "learning_rate": 3.645680819912152e-07, + "loss": 1.0684, + "mean_token_accuracy": 0.6893216371536255, + "num_tokens": 25789007.0, + "step": 997 + }, + { + "epoch": 0.109598067208434, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.547736406326294, + "learning_rate": 3.649341142020498e-07, + "loss": 0.9985, + "mean_token_accuracy": 0.7060402631759644, + "num_tokens": 25814004.0, + "step": 998 + }, + { + "epoch": 0.10970788491104766, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.7780892848968506, + "learning_rate": 3.653001464128843e-07, + "loss": 1.0727, + "mean_token_accuracy": 0.6871623992919922, + "num_tokens": 25835515.0, + "step": 999 + }, + { + "epoch": 0.10981770261366132, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.6932640075683594, + "learning_rate": 3.656661786237188e-07, + "loss": 1.1134, + "mean_token_accuracy": 0.6803106665611267, + "num_tokens": 25859300.0, + "step": 1000 + }, + { + "epoch": 0.10992752031627498, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.5871520042419434, + "learning_rate": 3.660322108345534e-07, + "loss": 1.0508, + "mean_token_accuracy": 0.6894925236701965, + "num_tokens": 25888053.0, + "step": 1001 + }, + { + "epoch": 0.11003733801888864, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.550781488418579, + "learning_rate": 3.66398243045388e-07, + "loss": 1.1662, + "mean_token_accuracy": 0.660053014755249, + "num_tokens": 25913789.0, + "step": 1002 + }, + { + "epoch": 0.11014715572150231, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.699885368347168, + "learning_rate": 3.667642752562226e-07, + "loss": 1.1764, + "mean_token_accuracy": 0.6686364412307739, + "num_tokens": 25936587.0, + "step": 1003 + }, + { + "epoch": 0.11025697342411597, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.348752498626709, + "learning_rate": 3.6713030746705707e-07, + "loss": 1.0603, + "mean_token_accuracy": 0.6889822483062744, + "num_tokens": 25965940.0, + "step": 1004 + }, + { + "epoch": 0.11036679112672963, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.6509692668914795, + "learning_rate": 3.674963396778916e-07, + "loss": 1.0018, + "mean_token_accuracy": 0.7064557671546936, + "num_tokens": 25993188.0, + "step": 1005 + }, + { + "epoch": 0.11047660882934329, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.5547611713409424, + "learning_rate": 3.678623718887262e-07, + "loss": 1.0392, + "mean_token_accuracy": 0.692260205745697, + "num_tokens": 26016160.0, + "step": 1006 + }, + { + "epoch": 0.11058642653195695, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.6644554138183594, + "learning_rate": 3.682284040995607e-07, + "loss": 1.0494, + "mean_token_accuracy": 0.6891720294952393, + "num_tokens": 26039892.0, + "step": 1007 + }, + { + "epoch": 0.11069624423457061, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.7017393112182617, + "learning_rate": 3.6859443631039527e-07, + "loss": 1.0799, + "mean_token_accuracy": 0.6817066669464111, + "num_tokens": 26061680.0, + "step": 1008 + }, + { + "epoch": 0.11080606193718427, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.505732297897339, + "learning_rate": 3.6896046852122987e-07, + "loss": 1.1755, + "mean_token_accuracy": 0.6576499938964844, + "num_tokens": 26088811.0, + "step": 1009 + }, + { + "epoch": 0.11091587963979793, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.635509729385376, + "learning_rate": 3.693265007320644e-07, + "loss": 1.1362, + "mean_token_accuracy": 0.6695331335067749, + "num_tokens": 26113462.0, + "step": 1010 + }, + { + "epoch": 0.11102569734241159, + "ewc_loss": 1.8104910850524902e-06, + "grad_norm": 2.450144052505493, + "learning_rate": 3.6969253294289897e-07, + "loss": 1.1625, + "mean_token_accuracy": 0.663495659828186, + "num_tokens": 26140867.0, + "step": 1011 + }, + { + "epoch": 0.11113551504502525, + "ewc_loss": 1.8030405044555664e-06, + "grad_norm": 2.457430362701416, + "learning_rate": 3.700585651537335e-07, + "loss": 1.0826, + "mean_token_accuracy": 0.6775895357131958, + "num_tokens": 26168121.0, + "step": 1012 + }, + { + "epoch": 0.11124533274763893, + "ewc_loss": 1.8104910850524902e-06, + "grad_norm": 2.7435364723205566, + "learning_rate": 3.7042459736456807e-07, + "loss": 1.0412, + "mean_token_accuracy": 0.6906453371047974, + "num_tokens": 26189626.0, + "step": 1013 + }, + { + "epoch": 0.11135515045025259, + "ewc_loss": 1.8104910850524902e-06, + "grad_norm": 2.7111587524414062, + "learning_rate": 3.707906295754026e-07, + "loss": 1.1111, + "mean_token_accuracy": 0.6748253703117371, + "num_tokens": 26213658.0, + "step": 1014 + }, + { + "epoch": 0.11146496815286625, + "ewc_loss": 1.8104910850524902e-06, + "grad_norm": 2.266930341720581, + "learning_rate": 3.7115666178623716e-07, + "loss": 1.0507, + "mean_token_accuracy": 0.6899991035461426, + "num_tokens": 26243731.0, + "step": 1015 + }, + { + "epoch": 0.1115747858554799, + "ewc_loss": 1.817941665649414e-06, + "grad_norm": 2.7615184783935547, + "learning_rate": 3.715226939970717e-07, + "loss": 1.0728, + "mean_token_accuracy": 0.6831750869750977, + "num_tokens": 26266656.0, + "step": 1016 + }, + { + "epoch": 0.11168460355809356, + "ewc_loss": 1.817941665649414e-06, + "grad_norm": 3.0646071434020996, + "learning_rate": 3.7188872620790626e-07, + "loss": 1.0425, + "mean_token_accuracy": 0.6925250291824341, + "num_tokens": 26285162.0, + "step": 1017 + }, + { + "epoch": 0.11179442126070722, + "ewc_loss": 1.817941665649414e-06, + "grad_norm": 2.7489676475524902, + "learning_rate": 3.722547584187408e-07, + "loss": 1.0533, + "mean_token_accuracy": 0.6902198195457458, + "num_tokens": 26309229.0, + "step": 1018 + }, + { + "epoch": 0.11190423896332088, + "ewc_loss": 1.8253922462463379e-06, + "grad_norm": 2.605833053588867, + "learning_rate": 3.726207906295754e-07, + "loss": 1.1144, + "mean_token_accuracy": 0.6779667139053345, + "num_tokens": 26334702.0, + "step": 1019 + }, + { + "epoch": 0.11201405666593454, + "ewc_loss": 1.8253922462463379e-06, + "grad_norm": 2.792872428894043, + "learning_rate": 3.7298682284040996e-07, + "loss": 1.0753, + "mean_token_accuracy": 0.6941183805465698, + "num_tokens": 26359954.0, + "step": 1020 + }, + { + "epoch": 0.1121238743685482, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.608522415161133, + "learning_rate": 3.7335285505124446e-07, + "loss": 1.0329, + "mean_token_accuracy": 0.6967594027519226, + "num_tokens": 26383642.0, + "step": 1021 + }, + { + "epoch": 0.11223369207116188, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.3815577030181885, + "learning_rate": 3.7371888726207906e-07, + "loss": 1.045, + "mean_token_accuracy": 0.6975876092910767, + "num_tokens": 26413718.0, + "step": 1022 + }, + { + "epoch": 0.11234350977377554, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.1723830699920654, + "learning_rate": 3.740849194729136e-07, + "loss": 1.1605, + "mean_token_accuracy": 0.665530800819397, + "num_tokens": 26448070.0, + "step": 1023 + }, + { + "epoch": 0.1124533274763892, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 3.0587611198425293, + "learning_rate": 3.744509516837481e-07, + "loss": 1.1682, + "mean_token_accuracy": 0.6635569334030151, + "num_tokens": 26477876.0, + "step": 1024 + }, + { + "epoch": 0.11256314517900286, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.586576461791992, + "learning_rate": 3.748169838945827e-07, + "loss": 1.1779, + "mean_token_accuracy": 0.6535383462905884, + "num_tokens": 26503463.0, + "step": 1025 + }, + { + "epoch": 0.11267296288161652, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.4890758991241455, + "learning_rate": 3.7518301610541726e-07, + "loss": 1.0758, + "mean_token_accuracy": 0.6854788064956665, + "num_tokens": 26530910.0, + "step": 1026 + }, + { + "epoch": 0.11278278058423018, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.5738117694854736, + "learning_rate": 3.7554904831625186e-07, + "loss": 1.14, + "mean_token_accuracy": 0.6646485328674316, + "num_tokens": 26556511.0, + "step": 1027 + }, + { + "epoch": 0.11289259828684384, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.9527924060821533, + "learning_rate": 3.7591508052708636e-07, + "loss": 1.0975, + "mean_token_accuracy": 0.6744831800460815, + "num_tokens": 26577702.0, + "step": 1028 + }, + { + "epoch": 0.1130024159894575, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.6445999145507812, + "learning_rate": 3.762811127379209e-07, + "loss": 1.168, + "mean_token_accuracy": 0.6684356927871704, + "num_tokens": 26602465.0, + "step": 1029 + }, + { + "epoch": 0.11311223369207116, + "ewc_loss": 1.8328428268432617e-06, + "grad_norm": 2.8536596298217773, + "learning_rate": 3.766471449487555e-07, + "loss": 1.0675, + "mean_token_accuracy": 0.6830675601959229, + "num_tokens": 26622436.0, + "step": 1030 + }, + { + "epoch": 0.11322205139468482, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 2.559370517730713, + "learning_rate": 3.7701317715959e-07, + "loss": 1.0783, + "mean_token_accuracy": 0.6853772401809692, + "num_tokens": 26646386.0, + "step": 1031 + }, + { + "epoch": 0.11333186909729849, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 2.517425537109375, + "learning_rate": 3.7737920937042455e-07, + "loss": 1.1746, + "mean_token_accuracy": 0.659778356552124, + "num_tokens": 26674495.0, + "step": 1032 + }, + { + "epoch": 0.11344168679991215, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 2.467764377593994, + "learning_rate": 3.7774524158125915e-07, + "loss": 1.0523, + "mean_token_accuracy": 0.6916862726211548, + "num_tokens": 26701395.0, + "step": 1033 + }, + { + "epoch": 0.11355150450252581, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 2.696324348449707, + "learning_rate": 3.781112737920937e-07, + "loss": 1.1399, + "mean_token_accuracy": 0.6754515171051025, + "num_tokens": 26724537.0, + "step": 1034 + }, + { + "epoch": 0.11366132220513947, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 2.72658371925354, + "learning_rate": 3.7847730600292825e-07, + "loss": 1.1153, + "mean_token_accuracy": 0.6779115796089172, + "num_tokens": 26747889.0, + "step": 1035 + }, + { + "epoch": 0.11377113990775313, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 3.028163194656372, + "learning_rate": 3.788433382137628e-07, + "loss": 1.0235, + "mean_token_accuracy": 0.6971884965896606, + "num_tokens": 26767707.0, + "step": 1036 + }, + { + "epoch": 0.11388095761036679, + "ewc_loss": 1.8402934074401855e-06, + "grad_norm": 2.487380027770996, + "learning_rate": 3.7920937042459735e-07, + "loss": 1.1406, + "mean_token_accuracy": 0.6817249059677124, + "num_tokens": 26793032.0, + "step": 1037 + }, + { + "epoch": 0.11399077531298045, + "ewc_loss": 1.8477439880371094e-06, + "grad_norm": 2.512549877166748, + "learning_rate": 3.795754026354319e-07, + "loss": 1.0506, + "mean_token_accuracy": 0.689734935760498, + "num_tokens": 26819051.0, + "step": 1038 + }, + { + "epoch": 0.11410059301559411, + "ewc_loss": 1.8477439880371094e-06, + "grad_norm": 2.6809194087982178, + "learning_rate": 3.7994143484626645e-07, + "loss": 1.0561, + "mean_token_accuracy": 0.6918587684631348, + "num_tokens": 26842616.0, + "step": 1039 + }, + { + "epoch": 0.11421041071820777, + "ewc_loss": 1.8477439880371094e-06, + "grad_norm": 2.724506139755249, + "learning_rate": 3.80307467057101e-07, + "loss": 1.1262, + "mean_token_accuracy": 0.6815096139907837, + "num_tokens": 26865703.0, + "step": 1040 + }, + { + "epoch": 0.11432022842082144, + "ewc_loss": 1.8477439880371094e-06, + "grad_norm": 3.1238832473754883, + "learning_rate": 3.8067349926793555e-07, + "loss": 1.117, + "mean_token_accuracy": 0.6675956845283508, + "num_tokens": 26884984.0, + "step": 1041 + }, + { + "epoch": 0.1144300461234351, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.542207717895508, + "learning_rate": 3.810395314787701e-07, + "loss": 1.0947, + "mean_token_accuracy": 0.6825364828109741, + "num_tokens": 26912097.0, + "step": 1042 + }, + { + "epoch": 0.11453986382604876, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.7740695476531982, + "learning_rate": 3.814055636896047e-07, + "loss": 1.0291, + "mean_token_accuracy": 0.6986544132232666, + "num_tokens": 26932033.0, + "step": 1043 + }, + { + "epoch": 0.11464968152866242, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.652003288269043, + "learning_rate": 3.8177159590043925e-07, + "loss": 1.1131, + "mean_token_accuracy": 0.6747097373008728, + "num_tokens": 26954418.0, + "step": 1044 + }, + { + "epoch": 0.11475949923127608, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.430419921875, + "learning_rate": 3.8213762811127374e-07, + "loss": 1.0782, + "mean_token_accuracy": 0.6859344244003296, + "num_tokens": 26984406.0, + "step": 1045 + }, + { + "epoch": 0.11486931693388974, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.6404879093170166, + "learning_rate": 3.8250366032210835e-07, + "loss": 1.0787, + "mean_token_accuracy": 0.688662052154541, + "num_tokens": 27012181.0, + "step": 1046 + }, + { + "epoch": 0.1149791346365034, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.531170606613159, + "learning_rate": 3.828696925329429e-07, + "loss": 1.1034, + "mean_token_accuracy": 0.6775604486465454, + "num_tokens": 27040817.0, + "step": 1047 + }, + { + "epoch": 0.11508895233911706, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.6552557945251465, + "learning_rate": 3.832357247437774e-07, + "loss": 1.0153, + "mean_token_accuracy": 0.710334062576294, + "num_tokens": 27063531.0, + "step": 1048 + }, + { + "epoch": 0.11519877004173072, + "ewc_loss": 1.8775463104248047e-06, + "grad_norm": 2.8901538848876953, + "learning_rate": 3.83601756954612e-07, + "loss": 1.0529, + "mean_token_accuracy": 0.6850557327270508, + "num_tokens": 27085193.0, + "step": 1049 + }, + { + "epoch": 0.11530858774434438, + "ewc_loss": 1.8700957298278809e-06, + "grad_norm": 2.490525484085083, + "learning_rate": 3.8396778916544654e-07, + "loss": 1.0966, + "mean_token_accuracy": 0.6746451258659363, + "num_tokens": 27112026.0, + "step": 1050 + }, + { + "epoch": 0.11541840544695806, + "ewc_loss": 1.8700957298278809e-06, + "grad_norm": 2.7034828662872314, + "learning_rate": 3.8433382137628114e-07, + "loss": 1.0922, + "mean_token_accuracy": 0.6845065951347351, + "num_tokens": 27134736.0, + "step": 1051 + }, + { + "epoch": 0.11552822314957172, + "ewc_loss": 1.8700957298278809e-06, + "grad_norm": 2.240363836288452, + "learning_rate": 3.8469985358711564e-07, + "loss": 1.204, + "mean_token_accuracy": 0.6501563787460327, + "num_tokens": 27167154.0, + "step": 1052 + }, + { + "epoch": 0.11563804085218538, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.4666266441345215, + "learning_rate": 3.850658857979502e-07, + "loss": 1.0638, + "mean_token_accuracy": 0.6926844120025635, + "num_tokens": 27191779.0, + "step": 1053 + }, + { + "epoch": 0.11574785855479904, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.805861234664917, + "learning_rate": 3.854319180087848e-07, + "loss": 1.0801, + "mean_token_accuracy": 0.6798173189163208, + "num_tokens": 27214207.0, + "step": 1054 + }, + { + "epoch": 0.1158576762574127, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.476106643676758, + "learning_rate": 3.857979502196193e-07, + "loss": 1.1138, + "mean_token_accuracy": 0.67973792552948, + "num_tokens": 27240519.0, + "step": 1055 + }, + { + "epoch": 0.11596749396002635, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.6097989082336426, + "learning_rate": 3.8616398243045384e-07, + "loss": 1.0405, + "mean_token_accuracy": 0.6972341537475586, + "num_tokens": 27265773.0, + "step": 1056 + }, + { + "epoch": 0.11607731166264001, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.359248638153076, + "learning_rate": 3.8653001464128844e-07, + "loss": 1.1389, + "mean_token_accuracy": 0.6664254665374756, + "num_tokens": 27294062.0, + "step": 1057 + }, + { + "epoch": 0.11618712936525367, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.6523027420043945, + "learning_rate": 3.8689604685212293e-07, + "loss": 1.0728, + "mean_token_accuracy": 0.6847143769264221, + "num_tokens": 27317200.0, + "step": 1058 + }, + { + "epoch": 0.11629694706786733, + "ewc_loss": 1.862645149230957e-06, + "grad_norm": 2.4103219509124756, + "learning_rate": 3.8726207906295754e-07, + "loss": 1.1281, + "mean_token_accuracy": 0.672534704208374, + "num_tokens": 27345129.0, + "step": 1059 + }, + { + "epoch": 0.11640676477048101, + "ewc_loss": 1.8551945686340332e-06, + "grad_norm": 2.328585386276245, + "learning_rate": 3.876281112737921e-07, + "loss": 1.1429, + "mean_token_accuracy": 0.6684240698814392, + "num_tokens": 27376500.0, + "step": 1060 + }, + { + "epoch": 0.11651658247309467, + "ewc_loss": 1.8775463104248047e-06, + "grad_norm": 2.700481653213501, + "learning_rate": 3.8799414348462663e-07, + "loss": 1.1249, + "mean_token_accuracy": 0.6676461696624756, + "num_tokens": 27405136.0, + "step": 1061 + }, + { + "epoch": 0.11662640017570833, + "ewc_loss": 1.8998980522155762e-06, + "grad_norm": 2.5311007499694824, + "learning_rate": 3.883601756954612e-07, + "loss": 1.0802, + "mean_token_accuracy": 0.6798485517501831, + "num_tokens": 27430922.0, + "step": 1062 + }, + { + "epoch": 0.11673621787832199, + "ewc_loss": 1.8998980522155762e-06, + "grad_norm": 2.806986093521118, + "learning_rate": 3.8872620790629573e-07, + "loss": 1.0483, + "mean_token_accuracy": 0.6929945945739746, + "num_tokens": 27452160.0, + "step": 1063 + }, + { + "epoch": 0.11684603558093565, + "ewc_loss": 1.8998980522155762e-06, + "grad_norm": 2.588648796081543, + "learning_rate": 3.890922401171303e-07, + "loss": 1.1093, + "mean_token_accuracy": 0.6791097521781921, + "num_tokens": 27477265.0, + "step": 1064 + }, + { + "epoch": 0.11695585328354931, + "ewc_loss": 1.8998980522155762e-06, + "grad_norm": 2.575958013534546, + "learning_rate": 3.8945827232796483e-07, + "loss": 1.1405, + "mean_token_accuracy": 0.6695111393928528, + "num_tokens": 27503052.0, + "step": 1065 + }, + { + "epoch": 0.11706567098616297, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.6839303970336914, + "learning_rate": 3.898243045387994e-07, + "loss": 1.0652, + "mean_token_accuracy": 0.6928037405014038, + "num_tokens": 27526173.0, + "step": 1066 + }, + { + "epoch": 0.11717548868877663, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.304993152618408, + "learning_rate": 3.90190336749634e-07, + "loss": 1.1098, + "mean_token_accuracy": 0.685994029045105, + "num_tokens": 27556183.0, + "step": 1067 + }, + { + "epoch": 0.11728530639139029, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.5172178745269775, + "learning_rate": 3.9055636896046853e-07, + "loss": 1.0832, + "mean_token_accuracy": 0.6841873526573181, + "num_tokens": 27583306.0, + "step": 1068 + }, + { + "epoch": 0.11739512409400395, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.5554463863372803, + "learning_rate": 3.9092240117130303e-07, + "loss": 1.0424, + "mean_token_accuracy": 0.6923055648803711, + "num_tokens": 27607080.0, + "step": 1069 + }, + { + "epoch": 0.11750494179661762, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.2424139976501465, + "learning_rate": 3.9128843338213763e-07, + "loss": 1.1437, + "mean_token_accuracy": 0.6682911515235901, + "num_tokens": 27639131.0, + "step": 1070 + }, + { + "epoch": 0.11761475949923128, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.7650561332702637, + "learning_rate": 3.916544655929722e-07, + "loss": 1.0698, + "mean_token_accuracy": 0.6820962429046631, + "num_tokens": 27660915.0, + "step": 1071 + }, + { + "epoch": 0.11772457720184494, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.2409260272979736, + "learning_rate": 3.920204978038067e-07, + "loss": 1.1199, + "mean_token_accuracy": 0.6729480028152466, + "num_tokens": 27693707.0, + "step": 1072 + }, + { + "epoch": 0.1178343949044586, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.4747300148010254, + "learning_rate": 3.923865300146413e-07, + "loss": 1.2001, + "mean_token_accuracy": 0.6562448143959045, + "num_tokens": 27721535.0, + "step": 1073 + }, + { + "epoch": 0.11794421260707226, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.6165411472320557, + "learning_rate": 3.927525622254758e-07, + "loss": 0.9999, + "mean_token_accuracy": 0.7003992199897766, + "num_tokens": 27744277.0, + "step": 1074 + }, + { + "epoch": 0.11805403030968592, + "ewc_loss": 1.9073486328125e-06, + "grad_norm": 2.3640122413635254, + "learning_rate": 3.9311859443631043e-07, + "loss": 1.116, + "mean_token_accuracy": 0.6750466823577881, + "num_tokens": 27773010.0, + "step": 1075 + }, + { + "epoch": 0.11816384801229958, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.80232572555542, + "learning_rate": 3.934846266471449e-07, + "loss": 1.1412, + "mean_token_accuracy": 0.6814056634902954, + "num_tokens": 27794162.0, + "step": 1076 + }, + { + "epoch": 0.11827366571491324, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.6461610794067383, + "learning_rate": 3.9385065885797947e-07, + "loss": 1.0982, + "mean_token_accuracy": 0.678185224533081, + "num_tokens": 27817334.0, + "step": 1077 + }, + { + "epoch": 0.1183834834175269, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.67307710647583, + "learning_rate": 3.942166910688141e-07, + "loss": 1.1624, + "mean_token_accuracy": 0.6591452360153198, + "num_tokens": 27841083.0, + "step": 1078 + }, + { + "epoch": 0.11849330112014057, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.841846227645874, + "learning_rate": 3.9458272327964857e-07, + "loss": 1.027, + "mean_token_accuracy": 0.689906656742096, + "num_tokens": 27861503.0, + "step": 1079 + }, + { + "epoch": 0.11860311882275423, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.4842276573181152, + "learning_rate": 3.949487554904831e-07, + "loss": 1.069, + "mean_token_accuracy": 0.6945875883102417, + "num_tokens": 27888130.0, + "step": 1080 + }, + { + "epoch": 0.1187129365253679, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.4402048587799072, + "learning_rate": 3.953147877013177e-07, + "loss": 1.0767, + "mean_token_accuracy": 0.6806212663650513, + "num_tokens": 27913796.0, + "step": 1081 + }, + { + "epoch": 0.11882275422798155, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.4633514881134033, + "learning_rate": 3.956808199121522e-07, + "loss": 1.1026, + "mean_token_accuracy": 0.6800616979598999, + "num_tokens": 27938790.0, + "step": 1082 + }, + { + "epoch": 0.11893257193059521, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.426873207092285, + "learning_rate": 3.960468521229868e-07, + "loss": 1.0743, + "mean_token_accuracy": 0.6838198304176331, + "num_tokens": 27964042.0, + "step": 1083 + }, + { + "epoch": 0.11904238963320887, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.5025932788848877, + "learning_rate": 3.9641288433382137e-07, + "loss": 1.0923, + "mean_token_accuracy": 0.6867702007293701, + "num_tokens": 27990871.0, + "step": 1084 + }, + { + "epoch": 0.11915220733582253, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.3640449047088623, + "learning_rate": 3.967789165446559e-07, + "loss": 1.0924, + "mean_token_accuracy": 0.6718018054962158, + "num_tokens": 28019249.0, + "step": 1085 + }, + { + "epoch": 0.11926202503843619, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.427250623703003, + "learning_rate": 3.9714494875549047e-07, + "loss": 1.0925, + "mean_token_accuracy": 0.6739042401313782, + "num_tokens": 28047209.0, + "step": 1086 + }, + { + "epoch": 0.11937184274104985, + "ewc_loss": 1.9222497940063477e-06, + "grad_norm": 2.3265981674194336, + "learning_rate": 3.97510980966325e-07, + "loss": 1.1708, + "mean_token_accuracy": 0.6700524091720581, + "num_tokens": 28076909.0, + "step": 1087 + }, + { + "epoch": 0.11948166044366351, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.4631195068359375, + "learning_rate": 3.9787701317715957e-07, + "loss": 1.1733, + "mean_token_accuracy": 0.6605854630470276, + "num_tokens": 28105066.0, + "step": 1088 + }, + { + "epoch": 0.11959147814627719, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.8501105308532715, + "learning_rate": 3.982430453879941e-07, + "loss": 1.0764, + "mean_token_accuracy": 0.6777508854866028, + "num_tokens": 28126388.0, + "step": 1089 + }, + { + "epoch": 0.11970129584889085, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 3.0538065433502197, + "learning_rate": 3.9860907759882866e-07, + "loss": 1.0326, + "mean_token_accuracy": 0.6911900639533997, + "num_tokens": 28152525.0, + "step": 1090 + }, + { + "epoch": 0.1198111135515045, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.6202731132507324, + "learning_rate": 3.9897510980966327e-07, + "loss": 1.006, + "mean_token_accuracy": 0.7044124007225037, + "num_tokens": 28176649.0, + "step": 1091 + }, + { + "epoch": 0.11992093125411817, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.9969258308410645, + "learning_rate": 3.993411420204978e-07, + "loss": 0.9904, + "mean_token_accuracy": 0.715130627155304, + "num_tokens": 28194986.0, + "step": 1092 + }, + { + "epoch": 0.12003074895673183, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.3194503784179688, + "learning_rate": 3.997071742313323e-07, + "loss": 1.0829, + "mean_token_accuracy": 0.6829824447631836, + "num_tokens": 28225155.0, + "step": 1093 + }, + { + "epoch": 0.12014056665934549, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.326199531555176, + "learning_rate": 4.000732064421669e-07, + "loss": 1.0752, + "mean_token_accuracy": 0.6805174350738525, + "num_tokens": 28254599.0, + "step": 1094 + }, + { + "epoch": 0.12025038436195915, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.9446985721588135, + "learning_rate": 4.0043923865300146e-07, + "loss": 1.0202, + "mean_token_accuracy": 0.6947224140167236, + "num_tokens": 28277178.0, + "step": 1095 + }, + { + "epoch": 0.1203602020645728, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.5347518920898438, + "learning_rate": 4.0080527086383596e-07, + "loss": 1.0872, + "mean_token_accuracy": 0.6837224960327148, + "num_tokens": 28304301.0, + "step": 1096 + }, + { + "epoch": 0.12047001976718646, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.4995570182800293, + "learning_rate": 4.0117130307467056e-07, + "loss": 0.974, + "mean_token_accuracy": 0.7106835842132568, + "num_tokens": 28328572.0, + "step": 1097 + }, + { + "epoch": 0.12057983746980014, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.6619505882263184, + "learning_rate": 4.015373352855051e-07, + "loss": 1.0659, + "mean_token_accuracy": 0.6865452527999878, + "num_tokens": 28352705.0, + "step": 1098 + }, + { + "epoch": 0.1206896551724138, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.4393622875213623, + "learning_rate": 4.0190336749633966e-07, + "loss": 1.1167, + "mean_token_accuracy": 0.6778508424758911, + "num_tokens": 28379284.0, + "step": 1099 + }, + { + "epoch": 0.12079947287502746, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.0992283821105957, + "learning_rate": 4.022693997071742e-07, + "loss": 1.1119, + "mean_token_accuracy": 0.6773478984832764, + "num_tokens": 28412910.0, + "step": 1100 + }, + { + "epoch": 0.12090929057764112, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.3790996074676514, + "learning_rate": 4.0263543191800876e-07, + "loss": 1.0293, + "mean_token_accuracy": 0.7034351229667664, + "num_tokens": 28439611.0, + "step": 1101 + }, + { + "epoch": 0.12101910828025478, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.559525489807129, + "learning_rate": 4.0300146412884336e-07, + "loss": 1.1374, + "mean_token_accuracy": 0.6728231906890869, + "num_tokens": 28464781.0, + "step": 1102 + }, + { + "epoch": 0.12112892598286844, + "ewc_loss": 1.9371509552001953e-06, + "grad_norm": 2.779179334640503, + "learning_rate": 4.0336749633967785e-07, + "loss": 1.0439, + "mean_token_accuracy": 0.6910112500190735, + "num_tokens": 28486031.0, + "step": 1103 + }, + { + "epoch": 0.1212387436854821, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.6491260528564453, + "learning_rate": 4.037335285505124e-07, + "loss": 0.9885, + "mean_token_accuracy": 0.7042901515960693, + "num_tokens": 28507282.0, + "step": 1104 + }, + { + "epoch": 0.12134856138809576, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.632836103439331, + "learning_rate": 4.04099560761347e-07, + "loss": 1.0701, + "mean_token_accuracy": 0.6810010671615601, + "num_tokens": 28532491.0, + "step": 1105 + }, + { + "epoch": 0.12145837909070942, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.3466131687164307, + "learning_rate": 4.044655929721815e-07, + "loss": 1.0285, + "mean_token_accuracy": 0.7058090567588806, + "num_tokens": 28559439.0, + "step": 1106 + }, + { + "epoch": 0.12156819679332308, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.4952375888824463, + "learning_rate": 4.048316251830161e-07, + "loss": 1.1273, + "mean_token_accuracy": 0.6713650226593018, + "num_tokens": 28584914.0, + "step": 1107 + }, + { + "epoch": 0.12167801449593675, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.41035795211792, + "learning_rate": 4.0519765739385065e-07, + "loss": 1.0915, + "mean_token_accuracy": 0.6843543648719788, + "num_tokens": 28613358.0, + "step": 1108 + }, + { + "epoch": 0.12178783219855041, + "ewc_loss": 1.9669532775878906e-06, + "grad_norm": 2.6088814735412598, + "learning_rate": 4.055636896046852e-07, + "loss": 1.1234, + "mean_token_accuracy": 0.6856136322021484, + "num_tokens": 28637827.0, + "step": 1109 + }, + { + "epoch": 0.12189764990116407, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.4866065979003906, + "learning_rate": 4.0592972181551975e-07, + "loss": 1.1756, + "mean_token_accuracy": 0.6627068519592285, + "num_tokens": 28663784.0, + "step": 1110 + }, + { + "epoch": 0.12200746760377773, + "ewc_loss": 1.952052116394043e-06, + "grad_norm": 2.563394784927368, + "learning_rate": 4.062957540263543e-07, + "loss": 1.1376, + "mean_token_accuracy": 0.6685918569564819, + "num_tokens": 28686562.0, + "step": 1111 + }, + { + "epoch": 0.12211728530639139, + "ewc_loss": 1.9669532775878906e-06, + "grad_norm": 2.7045860290527344, + "learning_rate": 4.0666178623718885e-07, + "loss": 1.1338, + "mean_token_accuracy": 0.6778908967971802, + "num_tokens": 28713361.0, + "step": 1112 + }, + { + "epoch": 0.12222710300900505, + "ewc_loss": 1.9669532775878906e-06, + "grad_norm": 2.6847267150878906, + "learning_rate": 4.070278184480234e-07, + "loss": 1.0537, + "mean_token_accuracy": 0.6884559392929077, + "num_tokens": 28736221.0, + "step": 1113 + }, + { + "epoch": 0.12233692071161871, + "ewc_loss": 1.9669532775878906e-06, + "grad_norm": 2.203547239303589, + "learning_rate": 4.0739385065885795e-07, + "loss": 1.0341, + "mean_token_accuracy": 0.6936799883842468, + "num_tokens": 28767480.0, + "step": 1114 + }, + { + "epoch": 0.12244673841423237, + "ewc_loss": 1.9669532775878906e-06, + "grad_norm": 2.7180774211883545, + "learning_rate": 4.0775988286969255e-07, + "loss": 1.0523, + "mean_token_accuracy": 0.6931810975074768, + "num_tokens": 28790622.0, + "step": 1115 + }, + { + "epoch": 0.12255655611684603, + "ewc_loss": 1.9669532775878906e-06, + "grad_norm": 2.460984468460083, + "learning_rate": 4.081259150805271e-07, + "loss": 1.0937, + "mean_token_accuracy": 0.6891516447067261, + "num_tokens": 28816316.0, + "step": 1116 + }, + { + "epoch": 0.1226663738194597, + "ewc_loss": 1.9818544387817383e-06, + "grad_norm": 2.6152613162994385, + "learning_rate": 4.084919472913616e-07, + "loss": 1.0638, + "mean_token_accuracy": 0.6868100166320801, + "num_tokens": 28838421.0, + "step": 1117 + }, + { + "epoch": 0.12277619152207336, + "ewc_loss": 1.996755599975586e-06, + "grad_norm": 2.7607831954956055, + "learning_rate": 4.088579795021962e-07, + "loss": 1.1805, + "mean_token_accuracy": 0.6605367064476013, + "num_tokens": 28863527.0, + "step": 1118 + }, + { + "epoch": 0.12288600922468702, + "ewc_loss": 1.996755599975586e-06, + "grad_norm": 2.8591268062591553, + "learning_rate": 4.0922401171303075e-07, + "loss": 0.9287, + "mean_token_accuracy": 0.72350013256073, + "num_tokens": 28882731.0, + "step": 1119 + }, + { + "epoch": 0.12299582692730068, + "ewc_loss": 1.9818544387817383e-06, + "grad_norm": 2.645897626876831, + "learning_rate": 4.0959004392386524e-07, + "loss": 1.0972, + "mean_token_accuracy": 0.6771144270896912, + "num_tokens": 28906336.0, + "step": 1120 + }, + { + "epoch": 0.12310564462991434, + "ewc_loss": 2.0116567611694336e-06, + "grad_norm": 2.7017996311187744, + "learning_rate": 4.0995607613469984e-07, + "loss": 1.0037, + "mean_token_accuracy": 0.7052395343780518, + "num_tokens": 28927718.0, + "step": 1121 + }, + { + "epoch": 0.123215462332528, + "ewc_loss": 2.0116567611694336e-06, + "grad_norm": 2.417703151702881, + "learning_rate": 4.103221083455344e-07, + "loss": 1.1638, + "mean_token_accuracy": 0.6606659889221191, + "num_tokens": 28956062.0, + "step": 1122 + }, + { + "epoch": 0.12332528003514166, + "ewc_loss": 2.0116567611694336e-06, + "grad_norm": 2.597444772720337, + "learning_rate": 4.1068814055636894e-07, + "loss": 1.1057, + "mean_token_accuracy": 0.6808242201805115, + "num_tokens": 28980636.0, + "step": 1123 + }, + { + "epoch": 0.12343509773775532, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 2.563466787338257, + "learning_rate": 4.110541727672035e-07, + "loss": 1.1345, + "mean_token_accuracy": 0.6623113751411438, + "num_tokens": 29005838.0, + "step": 1124 + }, + { + "epoch": 0.12354491544036898, + "ewc_loss": 2.0116567611694336e-06, + "grad_norm": 2.4881787300109863, + "learning_rate": 4.1142020497803804e-07, + "loss": 1.1495, + "mean_token_accuracy": 0.6631456613540649, + "num_tokens": 29032584.0, + "step": 1125 + }, + { + "epoch": 0.12365473314298264, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 2.658134698867798, + "learning_rate": 4.1178623718887264e-07, + "loss": 1.0838, + "mean_token_accuracy": 0.6822656989097595, + "num_tokens": 29054249.0, + "step": 1126 + }, + { + "epoch": 0.12376455084559632, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 2.3446247577667236, + "learning_rate": 4.1215226939970714e-07, + "loss": 1.1011, + "mean_token_accuracy": 0.6763274669647217, + "num_tokens": 29083825.0, + "step": 1127 + }, + { + "epoch": 0.12387436854820998, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 2.284111976623535, + "learning_rate": 4.125183016105417e-07, + "loss": 1.1078, + "mean_token_accuracy": 0.6857732534408569, + "num_tokens": 29113619.0, + "step": 1128 + }, + { + "epoch": 0.12398418625082364, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 3.1978447437286377, + "learning_rate": 4.128843338213763e-07, + "loss": 1.1485, + "mean_token_accuracy": 0.6707633137702942, + "num_tokens": 29141354.0, + "step": 1129 + }, + { + "epoch": 0.1240940039534373, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 2.699327230453491, + "learning_rate": 4.132503660322108e-07, + "loss": 1.0885, + "mean_token_accuracy": 0.6771151423454285, + "num_tokens": 29163113.0, + "step": 1130 + }, + { + "epoch": 0.12420382165605096, + "ewc_loss": 2.0265579223632812e-06, + "grad_norm": 2.396864891052246, + "learning_rate": 4.136163982430454e-07, + "loss": 1.044, + "mean_token_accuracy": 0.6935330629348755, + "num_tokens": 29190986.0, + "step": 1131 + }, + { + "epoch": 0.12431363935866462, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.3375394344329834, + "learning_rate": 4.1398243045387994e-07, + "loss": 0.9764, + "mean_token_accuracy": 0.7103258371353149, + "num_tokens": 29218126.0, + "step": 1132 + }, + { + "epoch": 0.12442345706127828, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.4764668941497803, + "learning_rate": 4.143484626647145e-07, + "loss": 1.0959, + "mean_token_accuracy": 0.6768537759780884, + "num_tokens": 29246177.0, + "step": 1133 + }, + { + "epoch": 0.12453327476389194, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.563077926635742, + "learning_rate": 4.1471449487554904e-07, + "loss": 1.0618, + "mean_token_accuracy": 0.6945103406906128, + "num_tokens": 29270132.0, + "step": 1134 + }, + { + "epoch": 0.1246430924665056, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.3791391849517822, + "learning_rate": 4.150805270863836e-07, + "loss": 1.1315, + "mean_token_accuracy": 0.6671429872512817, + "num_tokens": 29298680.0, + "step": 1135 + }, + { + "epoch": 0.12475291016911927, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.75785493850708, + "learning_rate": 4.1544655929721813e-07, + "loss": 1.0352, + "mean_token_accuracy": 0.695551335811615, + "num_tokens": 29319341.0, + "step": 1136 + }, + { + "epoch": 0.12486272787173293, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.4508931636810303, + "learning_rate": 4.158125915080527e-07, + "loss": 1.0782, + "mean_token_accuracy": 0.697014570236206, + "num_tokens": 29344645.0, + "step": 1137 + }, + { + "epoch": 0.12497254557434659, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.3701424598693848, + "learning_rate": 4.1617862371888723e-07, + "loss": 1.0504, + "mean_token_accuracy": 0.6911747455596924, + "num_tokens": 29370995.0, + "step": 1138 + }, + { + "epoch": 0.12508236327696023, + "ewc_loss": 2.041459083557129e-06, + "grad_norm": 2.5101914405822754, + "learning_rate": 4.1654465592972183e-07, + "loss": 1.1613, + "mean_token_accuracy": 0.6653017997741699, + "num_tokens": 29396692.0, + "step": 1139 + }, + { + "epoch": 0.1251921809795739, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.450969696044922, + "learning_rate": 4.1691068814055633e-07, + "loss": 1.1102, + "mean_token_accuracy": 0.6709421873092651, + "num_tokens": 29426669.0, + "step": 1140 + }, + { + "epoch": 0.12530199868218758, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.536015033721924, + "learning_rate": 4.172767203513909e-07, + "loss": 1.0181, + "mean_token_accuracy": 0.6983998417854309, + "num_tokens": 29451681.0, + "step": 1141 + }, + { + "epoch": 0.12541181638480123, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.1950159072875977, + "learning_rate": 4.176427525622255e-07, + "loss": 1.1449, + "mean_token_accuracy": 0.6716936230659485, + "num_tokens": 29484702.0, + "step": 1142 + }, + { + "epoch": 0.1255216340874149, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 3.0949831008911133, + "learning_rate": 4.1800878477306003e-07, + "loss": 0.9935, + "mean_token_accuracy": 0.7046188116073608, + "num_tokens": 29504341.0, + "step": 1143 + }, + { + "epoch": 0.12563145179002855, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.400425434112549, + "learning_rate": 4.183748169838945e-07, + "loss": 1.0223, + "mean_token_accuracy": 0.6913341283798218, + "num_tokens": 29530130.0, + "step": 1144 + }, + { + "epoch": 0.12574126949264222, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.345775842666626, + "learning_rate": 4.1874084919472913e-07, + "loss": 1.1222, + "mean_token_accuracy": 0.6725404262542725, + "num_tokens": 29559283.0, + "step": 1145 + }, + { + "epoch": 0.12585108719525587, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.5677621364593506, + "learning_rate": 4.191068814055637e-07, + "loss": 1.1156, + "mean_token_accuracy": 0.6753799915313721, + "num_tokens": 29584911.0, + "step": 1146 + }, + { + "epoch": 0.12596090489786954, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.5215003490448, + "learning_rate": 4.194729136163982e-07, + "loss": 1.0016, + "mean_token_accuracy": 0.7010422348976135, + "num_tokens": 29610399.0, + "step": 1147 + }, + { + "epoch": 0.1260707226004832, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.630452871322632, + "learning_rate": 4.198389458272328e-07, + "loss": 0.9716, + "mean_token_accuracy": 0.7067018747329712, + "num_tokens": 29633032.0, + "step": 1148 + }, + { + "epoch": 0.12618054030309686, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.5660977363586426, + "learning_rate": 4.202049780380673e-07, + "loss": 1.0808, + "mean_token_accuracy": 0.6807981133460999, + "num_tokens": 29657214.0, + "step": 1149 + }, + { + "epoch": 0.1262903580057105, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.2809860706329346, + "learning_rate": 4.205710102489019e-07, + "loss": 1.088, + "mean_token_accuracy": 0.6816182136535645, + "num_tokens": 29688013.0, + "step": 1150 + }, + { + "epoch": 0.12640017570832418, + "ewc_loss": 2.086162567138672e-06, + "grad_norm": 2.7261106967926025, + "learning_rate": 4.209370424597364e-07, + "loss": 1.027, + "mean_token_accuracy": 0.6937025785446167, + "num_tokens": 29710451.0, + "step": 1151 + }, + { + "epoch": 0.12650999341093785, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.7018914222717285, + "learning_rate": 4.2130307467057097e-07, + "loss": 1.101, + "mean_token_accuracy": 0.6764256954193115, + "num_tokens": 29735654.0, + "step": 1152 + }, + { + "epoch": 0.1266198111135515, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.534836530685425, + "learning_rate": 4.2166910688140557e-07, + "loss": 1.1494, + "mean_token_accuracy": 0.6620609760284424, + "num_tokens": 29764641.0, + "step": 1153 + }, + { + "epoch": 0.12672962881616517, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.5539557933807373, + "learning_rate": 4.2203513909224007e-07, + "loss": 1.1593, + "mean_token_accuracy": 0.6683144569396973, + "num_tokens": 29790458.0, + "step": 1154 + }, + { + "epoch": 0.12683944651877882, + "ewc_loss": 2.0563602447509766e-06, + "grad_norm": 2.5858521461486816, + "learning_rate": 4.2240117130307467e-07, + "loss": 1.0659, + "mean_token_accuracy": 0.6862736344337463, + "num_tokens": 29816208.0, + "step": 1155 + }, + { + "epoch": 0.1269492642213925, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.554473876953125, + "learning_rate": 4.227672035139092e-07, + "loss": 1.0977, + "mean_token_accuracy": 0.687232255935669, + "num_tokens": 29840722.0, + "step": 1156 + }, + { + "epoch": 0.12705908192400614, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.4466660022735596, + "learning_rate": 4.231332357247437e-07, + "loss": 1.1891, + "mean_token_accuracy": 0.6565048694610596, + "num_tokens": 29869138.0, + "step": 1157 + }, + { + "epoch": 0.1271688996266198, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.6407458782196045, + "learning_rate": 4.234992679355783e-07, + "loss": 1.0598, + "mean_token_accuracy": 0.6841599941253662, + "num_tokens": 29891804.0, + "step": 1158 + }, + { + "epoch": 0.12727871732923346, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.4059886932373047, + "learning_rate": 4.2386530014641287e-07, + "loss": 1.1462, + "mean_token_accuracy": 0.6735758781433105, + "num_tokens": 29918368.0, + "step": 1159 + }, + { + "epoch": 0.12738853503184713, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.372243642807007, + "learning_rate": 4.242313323572474e-07, + "loss": 1.1076, + "mean_token_accuracy": 0.6814891695976257, + "num_tokens": 29947243.0, + "step": 1160 + }, + { + "epoch": 0.1274983527344608, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.321668863296509, + "learning_rate": 4.2459736456808197e-07, + "loss": 1.1387, + "mean_token_accuracy": 0.670540452003479, + "num_tokens": 29976185.0, + "step": 1161 + }, + { + "epoch": 0.12760817043707445, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.419572353363037, + "learning_rate": 4.249633967789165e-07, + "loss": 1.106, + "mean_token_accuracy": 0.674712061882019, + "num_tokens": 30004893.0, + "step": 1162 + }, + { + "epoch": 0.12771798813968813, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.5228209495544434, + "learning_rate": 4.253294289897511e-07, + "loss": 1.1975, + "mean_token_accuracy": 0.6565080285072327, + "num_tokens": 30033645.0, + "step": 1163 + }, + { + "epoch": 0.12782780584230177, + "ewc_loss": 2.0712614059448242e-06, + "grad_norm": 2.656765937805176, + "learning_rate": 4.256954612005856e-07, + "loss": 1.0981, + "mean_token_accuracy": 0.6847410202026367, + "num_tokens": 30055734.0, + "step": 1164 + }, + { + "epoch": 0.12793762354491545, + "ewc_loss": 2.086162567138672e-06, + "grad_norm": 2.844794511795044, + "learning_rate": 4.2606149341142016e-07, + "loss": 1.0191, + "mean_token_accuracy": 0.6935898661613464, + "num_tokens": 30076609.0, + "step": 1165 + }, + { + "epoch": 0.1280474412475291, + "ewc_loss": 2.086162567138672e-06, + "grad_norm": 2.7908239364624023, + "learning_rate": 4.2642752562225476e-07, + "loss": 1.1365, + "mean_token_accuracy": 0.662526547908783, + "num_tokens": 30101695.0, + "step": 1166 + }, + { + "epoch": 0.12815725895014277, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.577817678451538, + "learning_rate": 4.267935578330893e-07, + "loss": 1.143, + "mean_token_accuracy": 0.6727045178413391, + "num_tokens": 30128385.0, + "step": 1167 + }, + { + "epoch": 0.1282670766527564, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 3.13857364654541, + "learning_rate": 4.271595900439238e-07, + "loss": 0.9239, + "mean_token_accuracy": 0.726658046245575, + "num_tokens": 30149434.0, + "step": 1168 + }, + { + "epoch": 0.12837689435537009, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.445997953414917, + "learning_rate": 4.275256222547584e-07, + "loss": 1.0346, + "mean_token_accuracy": 0.6957895755767822, + "num_tokens": 30175072.0, + "step": 1169 + }, + { + "epoch": 0.12848671205798376, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.433748483657837, + "learning_rate": 4.2789165446559296e-07, + "loss": 1.093, + "mean_token_accuracy": 0.6826236248016357, + "num_tokens": 30204221.0, + "step": 1170 + }, + { + "epoch": 0.1285965297605974, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.317833185195923, + "learning_rate": 4.282576866764275e-07, + "loss": 1.1148, + "mean_token_accuracy": 0.6760677099227905, + "num_tokens": 30232624.0, + "step": 1171 + }, + { + "epoch": 0.12870634746321108, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.4013731479644775, + "learning_rate": 4.2862371888726206e-07, + "loss": 1.0982, + "mean_token_accuracy": 0.685143768787384, + "num_tokens": 30259309.0, + "step": 1172 + }, + { + "epoch": 0.12881616516582473, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.3825128078460693, + "learning_rate": 4.289897510980966e-07, + "loss": 1.0932, + "mean_token_accuracy": 0.6847867965698242, + "num_tokens": 30286533.0, + "step": 1173 + }, + { + "epoch": 0.1289259828684384, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.6582422256469727, + "learning_rate": 4.293557833089312e-07, + "loss": 1.0542, + "mean_token_accuracy": 0.6879417300224304, + "num_tokens": 30311150.0, + "step": 1174 + }, + { + "epoch": 0.12903580057105205, + "ewc_loss": 2.115964889526367e-06, + "grad_norm": 2.5626749992370605, + "learning_rate": 4.297218155197657e-07, + "loss": 1.02, + "mean_token_accuracy": 0.6949633359909058, + "num_tokens": 30334858.0, + "step": 1175 + }, + { + "epoch": 0.12914561827366572, + "ewc_loss": 2.130866050720215e-06, + "grad_norm": 2.6711316108703613, + "learning_rate": 4.3008784773060026e-07, + "loss": 0.9974, + "mean_token_accuracy": 0.7100471258163452, + "num_tokens": 30358502.0, + "step": 1176 + }, + { + "epoch": 0.12925543597627936, + "ewc_loss": 2.130866050720215e-06, + "grad_norm": 2.492790460586548, + "learning_rate": 4.3045387994143486e-07, + "loss": 1.1284, + "mean_token_accuracy": 0.6700063943862915, + "num_tokens": 30384783.0, + "step": 1177 + }, + { + "epoch": 0.12936525367889304, + "ewc_loss": 2.130866050720215e-06, + "grad_norm": 2.565830707550049, + "learning_rate": 4.3081991215226935e-07, + "loss": 1.0985, + "mean_token_accuracy": 0.6889243721961975, + "num_tokens": 30408913.0, + "step": 1178 + }, + { + "epoch": 0.1294750713815067, + "ewc_loss": 2.130866050720215e-06, + "grad_norm": 2.769623279571533, + "learning_rate": 4.3118594436310396e-07, + "loss": 1.0528, + "mean_token_accuracy": 0.6948105692863464, + "num_tokens": 30433900.0, + "step": 1179 + }, + { + "epoch": 0.12958488908412036, + "ewc_loss": 2.130866050720215e-06, + "grad_norm": 2.7060258388519287, + "learning_rate": 4.315519765739385e-07, + "loss": 1.0163, + "mean_token_accuracy": 0.7023412585258484, + "num_tokens": 30461190.0, + "step": 1180 + }, + { + "epoch": 0.12969470678673403, + "ewc_loss": 2.130866050720215e-06, + "grad_norm": 2.557631254196167, + "learning_rate": 4.31918008784773e-07, + "loss": 1.1162, + "mean_token_accuracy": 0.6744120121002197, + "num_tokens": 30489651.0, + "step": 1181 + }, + { + "epoch": 0.12980452448934768, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.5252790451049805, + "learning_rate": 4.322840409956076e-07, + "loss": 1.0922, + "mean_token_accuracy": 0.693905770778656, + "num_tokens": 30519052.0, + "step": 1182 + }, + { + "epoch": 0.12991434219196135, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.395524263381958, + "learning_rate": 4.3265007320644215e-07, + "loss": 0.9994, + "mean_token_accuracy": 0.7032973766326904, + "num_tokens": 30543441.0, + "step": 1183 + }, + { + "epoch": 0.130024159894575, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.6431472301483154, + "learning_rate": 4.330161054172767e-07, + "loss": 1.148, + "mean_token_accuracy": 0.668258786201477, + "num_tokens": 30567401.0, + "step": 1184 + }, + { + "epoch": 0.13013397759718867, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.528191089630127, + "learning_rate": 4.3338213762811125e-07, + "loss": 1.1463, + "mean_token_accuracy": 0.666115403175354, + "num_tokens": 30594937.0, + "step": 1185 + }, + { + "epoch": 0.13024379529980232, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.389143943786621, + "learning_rate": 4.337481698389458e-07, + "loss": 0.9994, + "mean_token_accuracy": 0.7075477242469788, + "num_tokens": 30623647.0, + "step": 1186 + }, + { + "epoch": 0.130353613002416, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.494504928588867, + "learning_rate": 4.341142020497804e-07, + "loss": 1.1082, + "mean_token_accuracy": 0.6816840767860413, + "num_tokens": 30650386.0, + "step": 1187 + }, + { + "epoch": 0.13046343070502964, + "ewc_loss": 2.1457672119140625e-06, + "grad_norm": 2.725665807723999, + "learning_rate": 4.344802342606149e-07, + "loss": 1.0825, + "mean_token_accuracy": 0.6809501647949219, + "num_tokens": 30673298.0, + "step": 1188 + }, + { + "epoch": 0.1305732484076433, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.324030637741089, + "learning_rate": 4.3484626647144945e-07, + "loss": 1.1069, + "mean_token_accuracy": 0.68405681848526, + "num_tokens": 30702162.0, + "step": 1189 + }, + { + "epoch": 0.13068306611025698, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.5450501441955566, + "learning_rate": 4.3521229868228405e-07, + "loss": 1.1367, + "mean_token_accuracy": 0.6747273802757263, + "num_tokens": 30728125.0, + "step": 1190 + }, + { + "epoch": 0.13079288381287063, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.433805465698242, + "learning_rate": 4.355783308931186e-07, + "loss": 1.0164, + "mean_token_accuracy": 0.7002696990966797, + "num_tokens": 30753050.0, + "step": 1191 + }, + { + "epoch": 0.1309027015154843, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.5627241134643555, + "learning_rate": 4.359443631039531e-07, + "loss": 1.0664, + "mean_token_accuracy": 0.6861664056777954, + "num_tokens": 30778703.0, + "step": 1192 + }, + { + "epoch": 0.13101251921809795, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.315375328063965, + "learning_rate": 4.363103953147877e-07, + "loss": 1.0619, + "mean_token_accuracy": 0.6865348219871521, + "num_tokens": 30810814.0, + "step": 1193 + }, + { + "epoch": 0.13112233692071162, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.5229785442352295, + "learning_rate": 4.3667642752562224e-07, + "loss": 1.0729, + "mean_token_accuracy": 0.6974503993988037, + "num_tokens": 30836427.0, + "step": 1194 + }, + { + "epoch": 0.13123215462332527, + "ewc_loss": 2.175569534301758e-06, + "grad_norm": 2.74843692779541, + "learning_rate": 4.370424597364568e-07, + "loss": 1.1384, + "mean_token_accuracy": 0.6679970622062683, + "num_tokens": 30865532.0, + "step": 1195 + }, + { + "epoch": 0.13134197232593894, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.835171937942505, + "learning_rate": 4.3740849194729134e-07, + "loss": 1.0609, + "mean_token_accuracy": 0.6850931644439697, + "num_tokens": 30885416.0, + "step": 1196 + }, + { + "epoch": 0.1314517900285526, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.507782459259033, + "learning_rate": 4.377745241581259e-07, + "loss": 1.0341, + "mean_token_accuracy": 0.6909513473510742, + "num_tokens": 30909752.0, + "step": 1197 + }, + { + "epoch": 0.13156160773116626, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 3.0360682010650635, + "learning_rate": 4.381405563689605e-07, + "loss": 1.017, + "mean_token_accuracy": 0.7005009651184082, + "num_tokens": 30928463.0, + "step": 1198 + }, + { + "epoch": 0.13167142543377994, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.805553913116455, + "learning_rate": 4.38506588579795e-07, + "loss": 1.0593, + "mean_token_accuracy": 0.6886870265007019, + "num_tokens": 30950612.0, + "step": 1199 + }, + { + "epoch": 0.13178124313639358, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.3956990242004395, + "learning_rate": 4.3887262079062954e-07, + "loss": 1.0564, + "mean_token_accuracy": 0.6970747709274292, + "num_tokens": 30975408.0, + "step": 1200 + }, + { + "epoch": 0.13189106083900726, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.626758098602295, + "learning_rate": 4.3923865300146414e-07, + "loss": 0.9648, + "mean_token_accuracy": 0.7116629481315613, + "num_tokens": 30998382.0, + "step": 1201 + }, + { + "epoch": 0.1320008785416209, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.403014659881592, + "learning_rate": 4.3960468521229864e-07, + "loss": 1.0185, + "mean_token_accuracy": 0.7017443776130676, + "num_tokens": 31024645.0, + "step": 1202 + }, + { + "epoch": 0.13211069624423458, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.5080699920654297, + "learning_rate": 4.3997071742313324e-07, + "loss": 1.0754, + "mean_token_accuracy": 0.6924759149551392, + "num_tokens": 31050392.0, + "step": 1203 + }, + { + "epoch": 0.13222051394684822, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.6286520957946777, + "learning_rate": 4.403367496339678e-07, + "loss": 1.0111, + "mean_token_accuracy": 0.7003214359283447, + "num_tokens": 31072099.0, + "step": 1204 + }, + { + "epoch": 0.1323303316494619, + "ewc_loss": 2.16066837310791e-06, + "grad_norm": 2.416381597518921, + "learning_rate": 4.407027818448023e-07, + "loss": 1.0454, + "mean_token_accuracy": 0.6893038749694824, + "num_tokens": 31099348.0, + "step": 1205 + }, + { + "epoch": 0.13244014935207554, + "ewc_loss": 2.1904706954956055e-06, + "grad_norm": 2.7723779678344727, + "learning_rate": 4.410688140556369e-07, + "loss": 1.0448, + "mean_token_accuracy": 0.690679669380188, + "num_tokens": 31127674.0, + "step": 1206 + }, + { + "epoch": 0.13254996705468922, + "ewc_loss": 2.1904706954956055e-06, + "grad_norm": 2.8875060081481934, + "learning_rate": 4.4143484626647144e-07, + "loss": 1.0382, + "mean_token_accuracy": 0.7015697956085205, + "num_tokens": 31147826.0, + "step": 1207 + }, + { + "epoch": 0.1326597847573029, + "ewc_loss": 2.205371856689453e-06, + "grad_norm": 2.9229297637939453, + "learning_rate": 4.41800878477306e-07, + "loss": 0.986, + "mean_token_accuracy": 0.7049826383590698, + "num_tokens": 31168192.0, + "step": 1208 + }, + { + "epoch": 0.13276960245991654, + "ewc_loss": 2.205371856689453e-06, + "grad_norm": 3.1886544227600098, + "learning_rate": 4.4216691068814053e-07, + "loss": 0.8785, + "mean_token_accuracy": 0.7228952646255493, + "num_tokens": 31186404.0, + "step": 1209 + }, + { + "epoch": 0.1328794201625302, + "ewc_loss": 2.205371856689453e-06, + "grad_norm": 2.76485538482666, + "learning_rate": 4.425329428989751e-07, + "loss": 1.1337, + "mean_token_accuracy": 0.6643761396408081, + "num_tokens": 31216377.0, + "step": 1210 + }, + { + "epoch": 0.13298923786514386, + "ewc_loss": 2.205371856689453e-06, + "grad_norm": 2.6521854400634766, + "learning_rate": 4.428989751098097e-07, + "loss": 1.085, + "mean_token_accuracy": 0.6819456815719604, + "num_tokens": 31241083.0, + "step": 1211 + }, + { + "epoch": 0.13309905556775753, + "ewc_loss": 2.205371856689453e-06, + "grad_norm": 2.5627963542938232, + "learning_rate": 4.432650073206442e-07, + "loss": 1.1193, + "mean_token_accuracy": 0.6751137375831604, + "num_tokens": 31269362.0, + "step": 1212 + }, + { + "epoch": 0.13320887327037118, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.379329204559326, + "learning_rate": 4.4363103953147873e-07, + "loss": 1.0217, + "mean_token_accuracy": 0.6960593461990356, + "num_tokens": 31301121.0, + "step": 1213 + }, + { + "epoch": 0.13331869097298485, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.7348105907440186, + "learning_rate": 4.4399707174231333e-07, + "loss": 1.1208, + "mean_token_accuracy": 0.6674745082855225, + "num_tokens": 31323850.0, + "step": 1214 + }, + { + "epoch": 0.1334285086755985, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.498084306716919, + "learning_rate": 4.443631039531479e-07, + "loss": 1.1052, + "mean_token_accuracy": 0.6837121248245239, + "num_tokens": 31349231.0, + "step": 1215 + }, + { + "epoch": 0.13353832637821217, + "ewc_loss": 2.205371856689453e-06, + "grad_norm": 2.9173355102539062, + "learning_rate": 4.447291361639824e-07, + "loss": 1.1159, + "mean_token_accuracy": 0.6758848428726196, + "num_tokens": 31368491.0, + "step": 1216 + }, + { + "epoch": 0.13364814408082584, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.483999729156494, + "learning_rate": 4.45095168374817e-07, + "loss": 0.983, + "mean_token_accuracy": 0.7041502594947815, + "num_tokens": 31392465.0, + "step": 1217 + }, + { + "epoch": 0.1337579617834395, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.6990649700164795, + "learning_rate": 4.4546120058565153e-07, + "loss": 1.0876, + "mean_token_accuracy": 0.6767308712005615, + "num_tokens": 31417587.0, + "step": 1218 + }, + { + "epoch": 0.13386777948605316, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.305950403213501, + "learning_rate": 4.458272327964861e-07, + "loss": 1.0419, + "mean_token_accuracy": 0.6943243741989136, + "num_tokens": 31448313.0, + "step": 1219 + }, + { + "epoch": 0.1339775971886668, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.582153797149658, + "learning_rate": 4.4619326500732063e-07, + "loss": 1.0448, + "mean_token_accuracy": 0.688705563545227, + "num_tokens": 31472179.0, + "step": 1220 + }, + { + "epoch": 0.13408741489128048, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.3219478130340576, + "learning_rate": 4.465592972181552e-07, + "loss": 1.1052, + "mean_token_accuracy": 0.6850653886795044, + "num_tokens": 31501837.0, + "step": 1221 + }, + { + "epoch": 0.13419723259389413, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.6691558361053467, + "learning_rate": 4.469253294289897e-07, + "loss": 1.0273, + "mean_token_accuracy": 0.6973628997802734, + "num_tokens": 31523276.0, + "step": 1222 + }, + { + "epoch": 0.1343070502965078, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.678565740585327, + "learning_rate": 4.472913616398243e-07, + "loss": 1.1746, + "mean_token_accuracy": 0.6639101505279541, + "num_tokens": 31547514.0, + "step": 1223 + }, + { + "epoch": 0.13441686799912145, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.4493188858032227, + "learning_rate": 4.476573938506588e-07, + "loss": 1.0081, + "mean_token_accuracy": 0.6984409093856812, + "num_tokens": 31575007.0, + "step": 1224 + }, + { + "epoch": 0.13452668570173512, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.56583571434021, + "learning_rate": 4.480234260614934e-07, + "loss": 1.0529, + "mean_token_accuracy": 0.6867201328277588, + "num_tokens": 31600483.0, + "step": 1225 + }, + { + "epoch": 0.13463650340434877, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.6629326343536377, + "learning_rate": 4.483894582723279e-07, + "loss": 1.0407, + "mean_token_accuracy": 0.6980476379394531, + "num_tokens": 31622639.0, + "step": 1226 + }, + { + "epoch": 0.13474632110696244, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.6106131076812744, + "learning_rate": 4.487554904831625e-07, + "loss": 1.0607, + "mean_token_accuracy": 0.6983325481414795, + "num_tokens": 31647351.0, + "step": 1227 + }, + { + "epoch": 0.13485613880957611, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.8800294399261475, + "learning_rate": 4.4912152269399707e-07, + "loss": 1.1212, + "mean_token_accuracy": 0.67473304271698, + "num_tokens": 31667220.0, + "step": 1228 + }, + { + "epoch": 0.13496595651218976, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.692182779312134, + "learning_rate": 4.4948755490483157e-07, + "loss": 1.1309, + "mean_token_accuracy": 0.6787645220756531, + "num_tokens": 31689044.0, + "step": 1229 + }, + { + "epoch": 0.13507577421480343, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.5102856159210205, + "learning_rate": 4.4985358711566617e-07, + "loss": 1.2633, + "mean_token_accuracy": 0.6450693607330322, + "num_tokens": 31717794.0, + "step": 1230 + }, + { + "epoch": 0.13518559191741708, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.5410215854644775, + "learning_rate": 4.502196193265007e-07, + "loss": 1.1282, + "mean_token_accuracy": 0.6717593669891357, + "num_tokens": 31743349.0, + "step": 1231 + }, + { + "epoch": 0.13529540962003075, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.9554102420806885, + "learning_rate": 4.5058565153733527e-07, + "loss": 1.0108, + "mean_token_accuracy": 0.6954622268676758, + "num_tokens": 31761698.0, + "step": 1232 + }, + { + "epoch": 0.1354052273226444, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 3.052840232849121, + "learning_rate": 4.509516837481698e-07, + "loss": 1.0051, + "mean_token_accuracy": 0.7061142921447754, + "num_tokens": 31780060.0, + "step": 1233 + }, + { + "epoch": 0.13551504502525807, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.5460398197174072, + "learning_rate": 4.5131771595900437e-07, + "loss": 1.0202, + "mean_token_accuracy": 0.7010465264320374, + "num_tokens": 31810081.0, + "step": 1234 + }, + { + "epoch": 0.13562486272787172, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.3219614028930664, + "learning_rate": 4.5168374816983897e-07, + "loss": 1.0873, + "mean_token_accuracy": 0.6805665493011475, + "num_tokens": 31838983.0, + "step": 1235 + }, + { + "epoch": 0.1357346804304854, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.401139497756958, + "learning_rate": 4.5204978038067347e-07, + "loss": 1.0954, + "mean_token_accuracy": 0.6873452067375183, + "num_tokens": 31864058.0, + "step": 1236 + }, + { + "epoch": 0.13584449813309907, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.4951159954071045, + "learning_rate": 4.52415812591508e-07, + "loss": 1.0737, + "mean_token_accuracy": 0.6835799217224121, + "num_tokens": 31889614.0, + "step": 1237 + }, + { + "epoch": 0.1359543158357127, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.525599241256714, + "learning_rate": 4.527818448023426e-07, + "loss": 1.1782, + "mean_token_accuracy": 0.6691774725914001, + "num_tokens": 31915897.0, + "step": 1238 + }, + { + "epoch": 0.1360641335383264, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.338819980621338, + "learning_rate": 4.531478770131771e-07, + "loss": 1.0308, + "mean_token_accuracy": 0.6910585761070251, + "num_tokens": 31943588.0, + "step": 1239 + }, + { + "epoch": 0.13617395124094003, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.508396863937378, + "learning_rate": 4.5351390922401166e-07, + "loss": 1.0926, + "mean_token_accuracy": 0.6837136745452881, + "num_tokens": 31970532.0, + "step": 1240 + }, + { + "epoch": 0.1362837689435537, + "ewc_loss": 2.2202730178833008e-06, + "grad_norm": 2.462196111679077, + "learning_rate": 4.5387994143484626e-07, + "loss": 1.0577, + "mean_token_accuracy": 0.6955980062484741, + "num_tokens": 31996697.0, + "step": 1241 + }, + { + "epoch": 0.13639358664616735, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.7663159370422363, + "learning_rate": 4.542459736456808e-07, + "loss": 1.0108, + "mean_token_accuracy": 0.7052263021469116, + "num_tokens": 32018303.0, + "step": 1242 + }, + { + "epoch": 0.13650340434878103, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.3387346267700195, + "learning_rate": 4.5461200585651536e-07, + "loss": 1.1334, + "mean_token_accuracy": 0.666737973690033, + "num_tokens": 32046571.0, + "step": 1243 + }, + { + "epoch": 0.13661322205139467, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.5852131843566895, + "learning_rate": 4.549780380673499e-07, + "loss": 1.1573, + "mean_token_accuracy": 0.6679955720901489, + "num_tokens": 32072281.0, + "step": 1244 + }, + { + "epoch": 0.13672303975400835, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.5460264682769775, + "learning_rate": 4.5534407027818446e-07, + "loss": 1.194, + "mean_token_accuracy": 0.6591952443122864, + "num_tokens": 32099451.0, + "step": 1245 + }, + { + "epoch": 0.13683285745662202, + "ewc_loss": 2.2351741790771484e-06, + "grad_norm": 2.4915213584899902, + "learning_rate": 4.55710102489019e-07, + "loss": 1.0588, + "mean_token_accuracy": 0.6912966370582581, + "num_tokens": 32124080.0, + "step": 1246 + }, + { + "epoch": 0.13694267515923567, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.3917970657348633, + "learning_rate": 4.5607613469985356e-07, + "loss": 1.1609, + "mean_token_accuracy": 0.6617399454116821, + "num_tokens": 32153434.0, + "step": 1247 + }, + { + "epoch": 0.13705249286184934, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.675631284713745, + "learning_rate": 4.564421669106881e-07, + "loss": 1.0363, + "mean_token_accuracy": 0.6959346532821655, + "num_tokens": 32176510.0, + "step": 1248 + }, + { + "epoch": 0.13716231056446299, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.85341739654541, + "learning_rate": 4.568081991215227e-07, + "loss": 1.0895, + "mean_token_accuracy": 0.6778663992881775, + "num_tokens": 32206030.0, + "step": 1249 + }, + { + "epoch": 0.13727212826707666, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.457876443862915, + "learning_rate": 4.571742313323572e-07, + "loss": 1.0669, + "mean_token_accuracy": 0.6909143924713135, + "num_tokens": 32233255.0, + "step": 1250 + }, + { + "epoch": 0.1373819459696903, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.501814603805542, + "learning_rate": 4.575402635431918e-07, + "loss": 1.0553, + "mean_token_accuracy": 0.6934974193572998, + "num_tokens": 32257818.0, + "step": 1251 + }, + { + "epoch": 0.13749176367230398, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.4041709899902344, + "learning_rate": 4.5790629575402636e-07, + "loss": 1.0605, + "mean_token_accuracy": 0.6881811618804932, + "num_tokens": 32286826.0, + "step": 1252 + }, + { + "epoch": 0.13760158137491763, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.793849229812622, + "learning_rate": 4.5827232796486085e-07, + "loss": 1.0686, + "mean_token_accuracy": 0.6851341724395752, + "num_tokens": 32310447.0, + "step": 1253 + }, + { + "epoch": 0.1377113990775313, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.584986925125122, + "learning_rate": 4.5863836017569545e-07, + "loss": 1.1125, + "mean_token_accuracy": 0.6899474859237671, + "num_tokens": 32335713.0, + "step": 1254 + }, + { + "epoch": 0.13782121678014497, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.508181095123291, + "learning_rate": 4.5900439238653e-07, + "loss": 1.0794, + "mean_token_accuracy": 0.6877830028533936, + "num_tokens": 32361259.0, + "step": 1255 + }, + { + "epoch": 0.13793103448275862, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.8571910858154297, + "learning_rate": 4.593704245973645e-07, + "loss": 1.064, + "mean_token_accuracy": 0.685820996761322, + "num_tokens": 32383138.0, + "step": 1256 + }, + { + "epoch": 0.1380408521853723, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.371929883956909, + "learning_rate": 4.597364568081991e-07, + "loss": 1.0646, + "mean_token_accuracy": 0.6864646077156067, + "num_tokens": 32409879.0, + "step": 1257 + }, + { + "epoch": 0.13815066988798594, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.958805561065674, + "learning_rate": 4.6010248901903365e-07, + "loss": 1.0448, + "mean_token_accuracy": 0.6898446679115295, + "num_tokens": 32430039.0, + "step": 1258 + }, + { + "epoch": 0.1382604875905996, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.423967123031616, + "learning_rate": 4.6046852122986825e-07, + "loss": 1.0474, + "mean_token_accuracy": 0.6892445683479309, + "num_tokens": 32456043.0, + "step": 1259 + }, + { + "epoch": 0.13837030529321326, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.2191410064697266, + "learning_rate": 4.6083455344070275e-07, + "loss": 1.1564, + "mean_token_accuracy": 0.6640454530715942, + "num_tokens": 32489703.0, + "step": 1260 + }, + { + "epoch": 0.13848012299582693, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.1264872550964355, + "learning_rate": 4.612005856515373e-07, + "loss": 1.1104, + "mean_token_accuracy": 0.6746249198913574, + "num_tokens": 32523481.0, + "step": 1261 + }, + { + "epoch": 0.13858994069844058, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.810495138168335, + "learning_rate": 4.615666178623719e-07, + "loss": 1.0467, + "mean_token_accuracy": 0.6936300992965698, + "num_tokens": 32544230.0, + "step": 1262 + }, + { + "epoch": 0.13869975840105425, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.7880618572235107, + "learning_rate": 4.619326500732064e-07, + "loss": 1.0187, + "mean_token_accuracy": 0.6955009698867798, + "num_tokens": 32563592.0, + "step": 1263 + }, + { + "epoch": 0.1388095761036679, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.4041855335235596, + "learning_rate": 4.6229868228404095e-07, + "loss": 1.1746, + "mean_token_accuracy": 0.6710079908370972, + "num_tokens": 32592096.0, + "step": 1264 + }, + { + "epoch": 0.13891939380628157, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.5235602855682373, + "learning_rate": 4.6266471449487555e-07, + "loss": 1.0426, + "mean_token_accuracy": 0.689624547958374, + "num_tokens": 32615477.0, + "step": 1265 + }, + { + "epoch": 0.13902921150889525, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 3.0415940284729004, + "learning_rate": 4.630307467057101e-07, + "loss": 1.0393, + "mean_token_accuracy": 0.6916304230690002, + "num_tokens": 32637654.0, + "step": 1266 + }, + { + "epoch": 0.1391390292115089, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.4936866760253906, + "learning_rate": 4.6339677891654465e-07, + "loss": 1.0148, + "mean_token_accuracy": 0.6980936527252197, + "num_tokens": 32661199.0, + "step": 1267 + }, + { + "epoch": 0.13924884691412256, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.6838760375976562, + "learning_rate": 4.637628111273792e-07, + "loss": 1.1098, + "mean_token_accuracy": 0.6753883361816406, + "num_tokens": 32684302.0, + "step": 1268 + }, + { + "epoch": 0.1393586646167362, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.3972649574279785, + "learning_rate": 4.6412884333821374e-07, + "loss": 1.1455, + "mean_token_accuracy": 0.6666993498802185, + "num_tokens": 32711546.0, + "step": 1269 + }, + { + "epoch": 0.13946848231934988, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.2787346839904785, + "learning_rate": 4.644948755490483e-07, + "loss": 1.1602, + "mean_token_accuracy": 0.6644284725189209, + "num_tokens": 32741165.0, + "step": 1270 + }, + { + "epoch": 0.13957830002196353, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.287672281265259, + "learning_rate": 4.6486090775988284e-07, + "loss": 1.0734, + "mean_token_accuracy": 0.686006486415863, + "num_tokens": 32774701.0, + "step": 1271 + }, + { + "epoch": 0.1396881177245772, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.628626585006714, + "learning_rate": 4.652269399707174e-07, + "loss": 1.1594, + "mean_token_accuracy": 0.6651763916015625, + "num_tokens": 32798129.0, + "step": 1272 + }, + { + "epoch": 0.13979793542719085, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.232854127883911, + "learning_rate": 4.65592972181552e-07, + "loss": 1.0767, + "mean_token_accuracy": 0.6894116401672363, + "num_tokens": 32829323.0, + "step": 1273 + }, + { + "epoch": 0.13990775312980452, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.4303267002105713, + "learning_rate": 4.659590043923865e-07, + "loss": 1.0015, + "mean_token_accuracy": 0.704585075378418, + "num_tokens": 32856609.0, + "step": 1274 + }, + { + "epoch": 0.1400175708324182, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.54705548286438, + "learning_rate": 4.663250366032211e-07, + "loss": 1.0556, + "mean_token_accuracy": 0.6852672100067139, + "num_tokens": 32881060.0, + "step": 1275 + }, + { + "epoch": 0.14012738853503184, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.5060606002807617, + "learning_rate": 4.6669106881405564e-07, + "loss": 1.02, + "mean_token_accuracy": 0.7120616436004639, + "num_tokens": 32904678.0, + "step": 1276 + }, + { + "epoch": 0.14023720623764552, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.6444973945617676, + "learning_rate": 4.6705710102489014e-07, + "loss": 1.0305, + "mean_token_accuracy": 0.6909171342849731, + "num_tokens": 32928285.0, + "step": 1277 + }, + { + "epoch": 0.14034702394025916, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.62241792678833, + "learning_rate": 4.6742313323572474e-07, + "loss": 1.0619, + "mean_token_accuracy": 0.6903878450393677, + "num_tokens": 32953717.0, + "step": 1278 + }, + { + "epoch": 0.14045684164287284, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.149670124053955, + "learning_rate": 4.677891654465593e-07, + "loss": 1.0389, + "mean_token_accuracy": 0.6952957510948181, + "num_tokens": 32985152.0, + "step": 1279 + }, + { + "epoch": 0.14056665934548648, + "ewc_loss": 2.250075340270996e-06, + "grad_norm": 2.5722620487213135, + "learning_rate": 4.681551976573938e-07, + "loss": 1.0545, + "mean_token_accuracy": 0.6899592876434326, + "num_tokens": 33009420.0, + "step": 1280 + }, + { + "epoch": 0.14067647704810016, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.483698844909668, + "learning_rate": 4.685212298682284e-07, + "loss": 1.1833, + "mean_token_accuracy": 0.6541085243225098, + "num_tokens": 33042302.0, + "step": 1281 + }, + { + "epoch": 0.1407862947507138, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.2900550365448, + "learning_rate": 4.6888726207906293e-07, + "loss": 1.0752, + "mean_token_accuracy": 0.6823545694351196, + "num_tokens": 33072431.0, + "step": 1282 + }, + { + "epoch": 0.14089611245332748, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.4232797622680664, + "learning_rate": 4.6925329428989754e-07, + "loss": 1.1201, + "mean_token_accuracy": 0.6689980626106262, + "num_tokens": 33098031.0, + "step": 1283 + }, + { + "epoch": 0.14100593015594115, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.561245918273926, + "learning_rate": 4.6961932650073203e-07, + "loss": 1.1038, + "mean_token_accuracy": 0.684256911277771, + "num_tokens": 33121574.0, + "step": 1284 + }, + { + "epoch": 0.1411157478585548, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.6313018798828125, + "learning_rate": 4.699853587115666e-07, + "loss": 1.1175, + "mean_token_accuracy": 0.6814731359481812, + "num_tokens": 33145414.0, + "step": 1285 + }, + { + "epoch": 0.14122556556116847, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.7177541255950928, + "learning_rate": 4.703513909224012e-07, + "loss": 1.0141, + "mean_token_accuracy": 0.6985379457473755, + "num_tokens": 33166555.0, + "step": 1286 + }, + { + "epoch": 0.14133538326378212, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.258530855178833, + "learning_rate": 4.707174231332357e-07, + "loss": 1.1404, + "mean_token_accuracy": 0.6686983108520508, + "num_tokens": 33200074.0, + "step": 1287 + }, + { + "epoch": 0.1414452009663958, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.819610357284546, + "learning_rate": 4.7108345534407023e-07, + "loss": 0.9149, + "mean_token_accuracy": 0.728095293045044, + "num_tokens": 33220090.0, + "step": 1288 + }, + { + "epoch": 0.14155501866900944, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.5557823181152344, + "learning_rate": 4.7144948755490483e-07, + "loss": 1.0333, + "mean_token_accuracy": 0.6883938312530518, + "num_tokens": 33247080.0, + "step": 1289 + }, + { + "epoch": 0.1416648363716231, + "ewc_loss": 2.2649765014648438e-06, + "grad_norm": 2.302750825881958, + "learning_rate": 4.718155197657394e-07, + "loss": 1.1312, + "mean_token_accuracy": 0.6774462461471558, + "num_tokens": 33276588.0, + "step": 1290 + }, + { + "epoch": 0.14177465407423676, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.8767306804656982, + "learning_rate": 4.7218155197657393e-07, + "loss": 0.9865, + "mean_token_accuracy": 0.7050602436065674, + "num_tokens": 33296291.0, + "step": 1291 + }, + { + "epoch": 0.14188447177685043, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.329561471939087, + "learning_rate": 4.725475841874085e-07, + "loss": 1.1372, + "mean_token_accuracy": 0.6678591370582581, + "num_tokens": 33327544.0, + "step": 1292 + }, + { + "epoch": 0.1419942894794641, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.177764892578125, + "learning_rate": 4.7291361639824303e-07, + "loss": 1.0636, + "mean_token_accuracy": 0.690792441368103, + "num_tokens": 33360733.0, + "step": 1293 + }, + { + "epoch": 0.14210410718207775, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.393307685852051, + "learning_rate": 4.732796486090776e-07, + "loss": 1.0327, + "mean_token_accuracy": 0.696483850479126, + "num_tokens": 33386974.0, + "step": 1294 + }, + { + "epoch": 0.14221392488469142, + "ewc_loss": 2.2798776626586914e-06, + "grad_norm": 2.5895607471466064, + "learning_rate": 4.736456808199121e-07, + "loss": 1.0616, + "mean_token_accuracy": 0.6979262828826904, + "num_tokens": 33410360.0, + "step": 1295 + }, + { + "epoch": 0.14232374258730507, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.7946648597717285, + "learning_rate": 4.740117130307467e-07, + "loss": 1.0613, + "mean_token_accuracy": 0.687209963798523, + "num_tokens": 33435322.0, + "step": 1296 + }, + { + "epoch": 0.14243356028991874, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.480137586593628, + "learning_rate": 4.743777452415813e-07, + "loss": 1.2219, + "mean_token_accuracy": 0.6536271572113037, + "num_tokens": 33464229.0, + "step": 1297 + }, + { + "epoch": 0.1425433779925324, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.650911808013916, + "learning_rate": 4.7474377745241577e-07, + "loss": 0.9679, + "mean_token_accuracy": 0.7180019617080688, + "num_tokens": 33485248.0, + "step": 1298 + }, + { + "epoch": 0.14265319569514606, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.3639135360717773, + "learning_rate": 4.751098096632504e-07, + "loss": 1.0702, + "mean_token_accuracy": 0.6859645247459412, + "num_tokens": 33516065.0, + "step": 1299 + }, + { + "epoch": 0.1427630133977597, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.2319960594177246, + "learning_rate": 4.754758418740849e-07, + "loss": 1.0122, + "mean_token_accuracy": 0.7056030035018921, + "num_tokens": 33548790.0, + "step": 1300 + }, + { + "epoch": 0.14287283110037338, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.4010190963745117, + "learning_rate": 4.758418740849194e-07, + "loss": 1.082, + "mean_token_accuracy": 0.6873431205749512, + "num_tokens": 33577167.0, + "step": 1301 + }, + { + "epoch": 0.14298264880298703, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.723184585571289, + "learning_rate": 4.76207906295754e-07, + "loss": 1.0544, + "mean_token_accuracy": 0.6967132091522217, + "num_tokens": 33599933.0, + "step": 1302 + }, + { + "epoch": 0.1430924665056007, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.914813756942749, + "learning_rate": 4.7657393850658857e-07, + "loss": 0.9671, + "mean_token_accuracy": 0.7079356908798218, + "num_tokens": 33618648.0, + "step": 1303 + }, + { + "epoch": 0.14320228420821438, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.478175163269043, + "learning_rate": 4.769399707174231e-07, + "loss": 1.0388, + "mean_token_accuracy": 0.6929572224617004, + "num_tokens": 33643320.0, + "step": 1304 + }, + { + "epoch": 0.14331210191082802, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.563979387283325, + "learning_rate": 4.773060029282577e-07, + "loss": 1.1253, + "mean_token_accuracy": 0.6682852506637573, + "num_tokens": 33667215.0, + "step": 1305 + }, + { + "epoch": 0.1434219196134417, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.533174514770508, + "learning_rate": 4.776720351390922e-07, + "loss": 1.1341, + "mean_token_accuracy": 0.6671179533004761, + "num_tokens": 33691314.0, + "step": 1306 + }, + { + "epoch": 0.14353173731605534, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.5830485820770264, + "learning_rate": 4.780380673499268e-07, + "loss": 1.1063, + "mean_token_accuracy": 0.6815258860588074, + "num_tokens": 33716231.0, + "step": 1307 + }, + { + "epoch": 0.14364155501866901, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.3010330200195312, + "learning_rate": 4.784040995607613e-07, + "loss": 1.0813, + "mean_token_accuracy": 0.6812323331832886, + "num_tokens": 33750223.0, + "step": 1308 + }, + { + "epoch": 0.14375137272128266, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.652256488800049, + "learning_rate": 4.787701317715958e-07, + "loss": 1.0117, + "mean_token_accuracy": 0.7044488787651062, + "num_tokens": 33772320.0, + "step": 1309 + }, + { + "epoch": 0.14386119042389633, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.7562255859375, + "learning_rate": 4.791361639824304e-07, + "loss": 1.0609, + "mean_token_accuracy": 0.6885166168212891, + "num_tokens": 33793342.0, + "step": 1310 + }, + { + "epoch": 0.14397100812650998, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.7154009342193604, + "learning_rate": 4.79502196193265e-07, + "loss": 0.9779, + "mean_token_accuracy": 0.7036738991737366, + "num_tokens": 33813186.0, + "step": 1311 + }, + { + "epoch": 0.14408082582912365, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.4808356761932373, + "learning_rate": 4.798682284040995e-07, + "loss": 1.0268, + "mean_token_accuracy": 0.6903846263885498, + "num_tokens": 33840214.0, + "step": 1312 + }, + { + "epoch": 0.14419064353173733, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.7145426273345947, + "learning_rate": 4.802342606149341e-07, + "loss": 1.0448, + "mean_token_accuracy": 0.6882784366607666, + "num_tokens": 33863647.0, + "step": 1313 + }, + { + "epoch": 0.14430046123435097, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.2184691429138184, + "learning_rate": 4.806002928257686e-07, + "loss": 1.1173, + "mean_token_accuracy": 0.677255392074585, + "num_tokens": 33896827.0, + "step": 1314 + }, + { + "epoch": 0.14441027893696465, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.6733973026275635, + "learning_rate": 4.809663250366032e-07, + "loss": 1.0776, + "mean_token_accuracy": 0.6808122396469116, + "num_tokens": 33920697.0, + "step": 1315 + }, + { + "epoch": 0.1445200966395783, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.4314708709716797, + "learning_rate": 4.813323572474377e-07, + "loss": 1.0351, + "mean_token_accuracy": 0.6988286972045898, + "num_tokens": 33945998.0, + "step": 1316 + }, + { + "epoch": 0.14462991434219197, + "ewc_loss": 2.294778823852539e-06, + "grad_norm": 2.6002650260925293, + "learning_rate": 4.816983894582723e-07, + "loss": 1.0533, + "mean_token_accuracy": 0.6957069039344788, + "num_tokens": 33969471.0, + "step": 1317 + }, + { + "epoch": 0.1447397320448056, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.4034085273742676, + "learning_rate": 4.820644216691069e-07, + "loss": 1.0871, + "mean_token_accuracy": 0.6771621704101562, + "num_tokens": 33995373.0, + "step": 1318 + }, + { + "epoch": 0.1448495497474193, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.5645265579223633, + "learning_rate": 4.824304538799414e-07, + "loss": 0.9888, + "mean_token_accuracy": 0.706964910030365, + "num_tokens": 34019470.0, + "step": 1319 + }, + { + "epoch": 0.14495936745003293, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.4921224117279053, + "learning_rate": 4.827964860907759e-07, + "loss": 1.1096, + "mean_token_accuracy": 0.6776554584503174, + "num_tokens": 34046792.0, + "step": 1320 + }, + { + "epoch": 0.1450691851526466, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.4105961322784424, + "learning_rate": 4.831625183016105e-07, + "loss": 1.0864, + "mean_token_accuracy": 0.6790288686752319, + "num_tokens": 34073908.0, + "step": 1321 + }, + { + "epoch": 0.14517900285526028, + "ewc_loss": 2.3096799850463867e-06, + "grad_norm": 2.3223209381103516, + "learning_rate": 4.835285505124451e-07, + "loss": 0.9475, + "mean_token_accuracy": 0.7142386436462402, + "num_tokens": 34101665.0, + "step": 1322 + }, + { + "epoch": 0.14528882055787393, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.862473487854004, + "learning_rate": 4.838945827232796e-07, + "loss": 0.975, + "mean_token_accuracy": 0.7098947763442993, + "num_tokens": 34123530.0, + "step": 1323 + }, + { + "epoch": 0.1453986382604876, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.647045373916626, + "learning_rate": 4.842606149341142e-07, + "loss": 1.0675, + "mean_token_accuracy": 0.676479697227478, + "num_tokens": 34148038.0, + "step": 1324 + }, + { + "epoch": 0.14550845596310125, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.5846846103668213, + "learning_rate": 4.846266471449487e-07, + "loss": 1.0718, + "mean_token_accuracy": 0.6807249188423157, + "num_tokens": 34171039.0, + "step": 1325 + }, + { + "epoch": 0.14561827366571492, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.5076396465301514, + "learning_rate": 4.849926793557833e-07, + "loss": 1.0746, + "mean_token_accuracy": 0.6827881336212158, + "num_tokens": 34197557.0, + "step": 1326 + }, + { + "epoch": 0.14572809136832857, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.6390185356140137, + "learning_rate": 4.853587115666178e-07, + "loss": 1.0139, + "mean_token_accuracy": 0.6957784295082092, + "num_tokens": 34219659.0, + "step": 1327 + }, + { + "epoch": 0.14583790907094224, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.763279914855957, + "learning_rate": 4.857247437774524e-07, + "loss": 1.0781, + "mean_token_accuracy": 0.6814442873001099, + "num_tokens": 34241179.0, + "step": 1328 + }, + { + "epoch": 0.14594772677355589, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.768519163131714, + "learning_rate": 4.86090775988287e-07, + "loss": 0.9588, + "mean_token_accuracy": 0.7151703834533691, + "num_tokens": 34261000.0, + "step": 1329 + }, + { + "epoch": 0.14605754447616956, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.7001101970672607, + "learning_rate": 4.864568081991215e-07, + "loss": 1.0646, + "mean_token_accuracy": 0.6849561929702759, + "num_tokens": 34285817.0, + "step": 1330 + }, + { + "epoch": 0.14616736217878323, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.443354368209839, + "learning_rate": 4.868228404099561e-07, + "loss": 1.1489, + "mean_token_accuracy": 0.6650068163871765, + "num_tokens": 34315537.0, + "step": 1331 + }, + { + "epoch": 0.14627717988139688, + "ewc_loss": 2.3245811462402344e-06, + "grad_norm": 2.55609393119812, + "learning_rate": 4.871888726207906e-07, + "loss": 1.0729, + "mean_token_accuracy": 0.6844680309295654, + "num_tokens": 34338741.0, + "step": 1332 + }, + { + "epoch": 0.14638699758401055, + "ewc_loss": 2.339482307434082e-06, + "grad_norm": 2.60331130027771, + "learning_rate": 4.875549048316251e-07, + "loss": 1.0527, + "mean_token_accuracy": 0.6917620897293091, + "num_tokens": 34362546.0, + "step": 1333 + }, + { + "epoch": 0.1464968152866242, + "ewc_loss": 2.339482307434082e-06, + "grad_norm": 2.2052299976348877, + "learning_rate": 4.879209370424597e-07, + "loss": 1.1442, + "mean_token_accuracy": 0.665809154510498, + "num_tokens": 34394885.0, + "step": 1334 + }, + { + "epoch": 0.14660663298923787, + "ewc_loss": 2.339482307434082e-06, + "grad_norm": 2.3518834114074707, + "learning_rate": 4.882869692532943e-07, + "loss": 1.0945, + "mean_token_accuracy": 0.6723960638046265, + "num_tokens": 34425299.0, + "step": 1335 + }, + { + "epoch": 0.14671645069185152, + "ewc_loss": 2.339482307434082e-06, + "grad_norm": 2.6807737350463867, + "learning_rate": 4.886530014641288e-07, + "loss": 1.0522, + "mean_token_accuracy": 0.6931209564208984, + "num_tokens": 34448680.0, + "step": 1336 + }, + { + "epoch": 0.1468262683944652, + "ewc_loss": 2.3543834686279297e-06, + "grad_norm": 2.4413042068481445, + "learning_rate": 4.890190336749634e-07, + "loss": 1.1013, + "mean_token_accuracy": 0.6815142035484314, + "num_tokens": 34474831.0, + "step": 1337 + }, + { + "epoch": 0.14693608609707884, + "ewc_loss": 2.3543834686279297e-06, + "grad_norm": 2.4904210567474365, + "learning_rate": 4.893850658857979e-07, + "loss": 0.9345, + "mean_token_accuracy": 0.7195410132408142, + "num_tokens": 34496472.0, + "step": 1338 + }, + { + "epoch": 0.1470459037996925, + "ewc_loss": 2.3543834686279297e-06, + "grad_norm": 2.661785125732422, + "learning_rate": 4.897510980966325e-07, + "loss": 1.0033, + "mean_token_accuracy": 0.713570773601532, + "num_tokens": 34519372.0, + "step": 1339 + }, + { + "epoch": 0.14715572150230616, + "ewc_loss": 2.3543834686279297e-06, + "grad_norm": 2.4635818004608154, + "learning_rate": 4.90117130307467e-07, + "loss": 1.0964, + "mean_token_accuracy": 0.6907554864883423, + "num_tokens": 34545661.0, + "step": 1340 + }, + { + "epoch": 0.14726553920491983, + "ewc_loss": 2.3543834686279297e-06, + "grad_norm": 2.4484689235687256, + "learning_rate": 4.904831625183016e-07, + "loss": 1.1794, + "mean_token_accuracy": 0.6629171967506409, + "num_tokens": 34576126.0, + "step": 1341 + }, + { + "epoch": 0.1473753569075335, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.6931350231170654, + "learning_rate": 4.908491947291362e-07, + "loss": 1.1006, + "mean_token_accuracy": 0.6789532899856567, + "num_tokens": 34600086.0, + "step": 1342 + }, + { + "epoch": 0.14748517461014715, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.523465871810913, + "learning_rate": 4.912152269399707e-07, + "loss": 1.0125, + "mean_token_accuracy": 0.7040127515792847, + "num_tokens": 34626401.0, + "step": 1343 + }, + { + "epoch": 0.14759499231276083, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.543665885925293, + "learning_rate": 4.915812591508052e-07, + "loss": 1.1086, + "mean_token_accuracy": 0.6751536130905151, + "num_tokens": 34653201.0, + "step": 1344 + }, + { + "epoch": 0.14770481001537447, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.43159818649292, + "learning_rate": 4.919472913616398e-07, + "loss": 1.0909, + "mean_token_accuracy": 0.677271842956543, + "num_tokens": 34680307.0, + "step": 1345 + }, + { + "epoch": 0.14781462771798815, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.63649845123291, + "learning_rate": 4.923133235724744e-07, + "loss": 1.0478, + "mean_token_accuracy": 0.6961060762405396, + "num_tokens": 34702531.0, + "step": 1346 + }, + { + "epoch": 0.1479244454206018, + "ewc_loss": 2.384185791015625e-06, + "grad_norm": 2.332787275314331, + "learning_rate": 4.926793557833089e-07, + "loss": 1.0772, + "mean_token_accuracy": 0.6829843521118164, + "num_tokens": 34733708.0, + "step": 1347 + }, + { + "epoch": 0.14803426312321546, + "ewc_loss": 2.384185791015625e-06, + "grad_norm": 2.5523085594177246, + "learning_rate": 4.930453879941435e-07, + "loss": 1.1075, + "mean_token_accuracy": 0.6762590408325195, + "num_tokens": 34758625.0, + "step": 1348 + }, + { + "epoch": 0.1481440808258291, + "ewc_loss": 2.384185791015625e-06, + "grad_norm": 2.407076835632324, + "learning_rate": 4.93411420204978e-07, + "loss": 1.0714, + "mean_token_accuracy": 0.6938130855560303, + "num_tokens": 34786952.0, + "step": 1349 + }, + { + "epoch": 0.14825389852844278, + "ewc_loss": 2.3543834686279297e-06, + "grad_norm": 2.378549337387085, + "learning_rate": 4.937774524158126e-07, + "loss": 1.1316, + "mean_token_accuracy": 0.6712225675582886, + "num_tokens": 34813181.0, + "step": 1350 + }, + { + "epoch": 0.14836371623105646, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.2597899436950684, + "learning_rate": 4.941434846266471e-07, + "loss": 1.0264, + "mean_token_accuracy": 0.693908154964447, + "num_tokens": 34845742.0, + "step": 1351 + }, + { + "epoch": 0.1484735339336701, + "ewc_loss": 2.3692846298217773e-06, + "grad_norm": 2.385265588760376, + "learning_rate": 4.945095168374817e-07, + "loss": 1.0711, + "mean_token_accuracy": 0.6838217377662659, + "num_tokens": 34872515.0, + "step": 1352 + }, + { + "epoch": 0.14858335163628378, + "ewc_loss": 2.384185791015625e-06, + "grad_norm": 2.5177979469299316, + "learning_rate": 4.948755490483163e-07, + "loss": 1.1274, + "mean_token_accuracy": 0.668665885925293, + "num_tokens": 34899778.0, + "step": 1353 + }, + { + "epoch": 0.14869316933889742, + "ewc_loss": 2.384185791015625e-06, + "grad_norm": 2.4405200481414795, + "learning_rate": 4.952415812591508e-07, + "loss": 1.0121, + "mean_token_accuracy": 0.6948229670524597, + "num_tokens": 34926103.0, + "step": 1354 + }, + { + "epoch": 0.1488029870415111, + "ewc_loss": 2.3990869522094727e-06, + "grad_norm": 2.33100962638855, + "learning_rate": 4.956076134699854e-07, + "loss": 0.9818, + "mean_token_accuracy": 0.7060462236404419, + "num_tokens": 34952471.0, + "step": 1355 + }, + { + "epoch": 0.14891280474412474, + "ewc_loss": 2.4139881134033203e-06, + "grad_norm": 2.3725392818450928, + "learning_rate": 4.959736456808199e-07, + "loss": 1.1115, + "mean_token_accuracy": 0.670109748840332, + "num_tokens": 34981653.0, + "step": 1356 + }, + { + "epoch": 0.14902262244673842, + "ewc_loss": 2.4139881134033203e-06, + "grad_norm": 2.6642260551452637, + "learning_rate": 4.963396778916544e-07, + "loss": 0.9961, + "mean_token_accuracy": 0.7083134651184082, + "num_tokens": 35004160.0, + "step": 1357 + }, + { + "epoch": 0.14913244014935206, + "ewc_loss": 2.428889274597168e-06, + "grad_norm": 2.38297176361084, + "learning_rate": 4.96705710102489e-07, + "loss": 1.0931, + "mean_token_accuracy": 0.6842519640922546, + "num_tokens": 35033695.0, + "step": 1358 + }, + { + "epoch": 0.14924225785196574, + "ewc_loss": 2.428889274597168e-06, + "grad_norm": 2.454883337020874, + "learning_rate": 4.970717423133236e-07, + "loss": 1.0238, + "mean_token_accuracy": 0.6896829605102539, + "num_tokens": 35058960.0, + "step": 1359 + }, + { + "epoch": 0.1493520755545794, + "ewc_loss": 2.428889274597168e-06, + "grad_norm": 2.499603033065796, + "learning_rate": 4.974377745241581e-07, + "loss": 1.0461, + "mean_token_accuracy": 0.691497266292572, + "num_tokens": 35082458.0, + "step": 1360 + }, + { + "epoch": 0.14946189325719306, + "ewc_loss": 2.428889274597168e-06, + "grad_norm": 2.4433562755584717, + "learning_rate": 4.978038067349927e-07, + "loss": 1.0871, + "mean_token_accuracy": 0.689429521560669, + "num_tokens": 35111320.0, + "step": 1361 + }, + { + "epoch": 0.14957171095980673, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.959536552429199, + "learning_rate": 4.981698389458272e-07, + "loss": 1.1144, + "mean_token_accuracy": 0.6672778129577637, + "num_tokens": 35136034.0, + "step": 1362 + }, + { + "epoch": 0.14968152866242038, + "ewc_loss": 2.428889274597168e-06, + "grad_norm": 2.5809364318847656, + "learning_rate": 4.985358711566618e-07, + "loss": 1.0579, + "mean_token_accuracy": 0.6908245086669922, + "num_tokens": 35160498.0, + "step": 1363 + }, + { + "epoch": 0.14979134636503405, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.222705602645874, + "learning_rate": 4.989019033674963e-07, + "loss": 1.0955, + "mean_token_accuracy": 0.682675302028656, + "num_tokens": 35191074.0, + "step": 1364 + }, + { + "epoch": 0.1499011640676477, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.5496573448181152, + "learning_rate": 4.992679355783309e-07, + "loss": 1.0671, + "mean_token_accuracy": 0.691595733165741, + "num_tokens": 35216079.0, + "step": 1365 + }, + { + "epoch": 0.15001098177026137, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.6665329933166504, + "learning_rate": 4.996339677891655e-07, + "loss": 1.0703, + "mean_token_accuracy": 0.6934361457824707, + "num_tokens": 35239149.0, + "step": 1366 + }, + { + "epoch": 0.15012079947287502, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.3609087467193604, + "learning_rate": 5e-07, + "loss": 1.1189, + "mean_token_accuracy": 0.6781884431838989, + "num_tokens": 35267266.0, + "step": 1367 + }, + { + "epoch": 0.1502306171754887, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.337230920791626, + "learning_rate": 5.003660322108345e-07, + "loss": 1.1262, + "mean_token_accuracy": 0.6750568151473999, + "num_tokens": 35295774.0, + "step": 1368 + }, + { + "epoch": 0.15034043487810236, + "ewc_loss": 2.4437904357910156e-06, + "grad_norm": 2.841157913208008, + "learning_rate": 5.007320644216691e-07, + "loss": 1.0534, + "mean_token_accuracy": 0.6875234842300415, + "num_tokens": 35315866.0, + "step": 1369 + }, + { + "epoch": 0.150450252580716, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.638953924179077, + "learning_rate": 5.010980966325037e-07, + "loss": 1.0881, + "mean_token_accuracy": 0.6859794855117798, + "num_tokens": 35340461.0, + "step": 1370 + }, + { + "epoch": 0.15056007028332968, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.2483325004577637, + "learning_rate": 5.014641288433382e-07, + "loss": 0.9991, + "mean_token_accuracy": 0.7028948664665222, + "num_tokens": 35367551.0, + "step": 1371 + }, + { + "epoch": 0.15066988798594333, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.3728396892547607, + "learning_rate": 5.018301610541728e-07, + "loss": 1.1287, + "mean_token_accuracy": 0.6647330522537231, + "num_tokens": 35397737.0, + "step": 1372 + }, + { + "epoch": 0.150779705688557, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.6507813930511475, + "learning_rate": 5.021961932650073e-07, + "loss": 1.0455, + "mean_token_accuracy": 0.6928078532218933, + "num_tokens": 35422047.0, + "step": 1373 + }, + { + "epoch": 0.15088952339117065, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.3216259479522705, + "learning_rate": 5.025622254758418e-07, + "loss": 1.0659, + "mean_token_accuracy": 0.6871399879455566, + "num_tokens": 35450974.0, + "step": 1374 + }, + { + "epoch": 0.15099934109378432, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.5734596252441406, + "learning_rate": 5.029282576866764e-07, + "loss": 1.0019, + "mean_token_accuracy": 0.704353928565979, + "num_tokens": 35478196.0, + "step": 1375 + }, + { + "epoch": 0.15110915879639797, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.2767229080200195, + "learning_rate": 5.03294289897511e-07, + "loss": 1.1303, + "mean_token_accuracy": 0.668136715888977, + "num_tokens": 35510077.0, + "step": 1376 + }, + { + "epoch": 0.15121897649901164, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.496352195739746, + "learning_rate": 5.036603221083456e-07, + "loss": 1.0785, + "mean_token_accuracy": 0.6831573247909546, + "num_tokens": 35538238.0, + "step": 1377 + }, + { + "epoch": 0.1513287942016253, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.302431106567383, + "learning_rate": 5.040263543191801e-07, + "loss": 1.1253, + "mean_token_accuracy": 0.6679090261459351, + "num_tokens": 35569282.0, + "step": 1378 + }, + { + "epoch": 0.15143861190423896, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.5613491535186768, + "learning_rate": 5.043923865300146e-07, + "loss": 1.1677, + "mean_token_accuracy": 0.6697646379470825, + "num_tokens": 35595162.0, + "step": 1379 + }, + { + "epoch": 0.15154842960685264, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.551145553588867, + "learning_rate": 5.047584187408492e-07, + "loss": 1.0102, + "mean_token_accuracy": 0.6980856657028198, + "num_tokens": 35620173.0, + "step": 1380 + }, + { + "epoch": 0.15165824730946628, + "ewc_loss": 2.4586915969848633e-06, + "grad_norm": 2.7684237957000732, + "learning_rate": 5.051244509516838e-07, + "loss": 1.0014, + "mean_token_accuracy": 0.70256108045578, + "num_tokens": 35641699.0, + "step": 1381 + }, + { + "epoch": 0.15176806501207996, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.4373741149902344, + "learning_rate": 5.054904831625183e-07, + "loss": 1.1096, + "mean_token_accuracy": 0.6737090945243835, + "num_tokens": 35672048.0, + "step": 1382 + }, + { + "epoch": 0.1518778827146936, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.4695489406585693, + "learning_rate": 5.058565153733529e-07, + "loss": 1.1056, + "mean_token_accuracy": 0.6774070858955383, + "num_tokens": 35697483.0, + "step": 1383 + }, + { + "epoch": 0.15198770041730728, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.4317283630371094, + "learning_rate": 5.062225475841874e-07, + "loss": 1.0751, + "mean_token_accuracy": 0.6962506771087646, + "num_tokens": 35723408.0, + "step": 1384 + }, + { + "epoch": 0.15209751811992092, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.51332426071167, + "learning_rate": 5.065885797950219e-07, + "loss": 1.0457, + "mean_token_accuracy": 0.6903512477874756, + "num_tokens": 35747189.0, + "step": 1385 + }, + { + "epoch": 0.1522073358225346, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.373323917388916, + "learning_rate": 5.069546120058566e-07, + "loss": 1.0746, + "mean_token_accuracy": 0.6811755299568176, + "num_tokens": 35777084.0, + "step": 1386 + }, + { + "epoch": 0.15231715352514824, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.6475894451141357, + "learning_rate": 5.073206442166911e-07, + "loss": 0.9884, + "mean_token_accuracy": 0.7080281376838684, + "num_tokens": 35801318.0, + "step": 1387 + }, + { + "epoch": 0.15242697122776191, + "ewc_loss": 2.473592758178711e-06, + "grad_norm": 2.9707212448120117, + "learning_rate": 5.076866764275256e-07, + "loss": 1.0074, + "mean_token_accuracy": 0.7036387920379639, + "num_tokens": 35820852.0, + "step": 1388 + }, + { + "epoch": 0.1525367889303756, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 20.646934509277344, + "learning_rate": 5.080527086383602e-07, + "loss": 1.0979, + "mean_token_accuracy": 0.6750994324684143, + "num_tokens": 35846653.0, + "step": 1389 + }, + { + "epoch": 0.15264660663298923, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.312872886657715, + "learning_rate": 5.084187408491947e-07, + "loss": 1.1051, + "mean_token_accuracy": 0.6746540069580078, + "num_tokens": 35877674.0, + "step": 1390 + }, + { + "epoch": 0.1527564243356029, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.2987096309661865, + "learning_rate": 5.087847730600292e-07, + "loss": 1.0972, + "mean_token_accuracy": 0.6791633367538452, + "num_tokens": 35908450.0, + "step": 1391 + }, + { + "epoch": 0.15286624203821655, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.4965460300445557, + "learning_rate": 5.091508052708639e-07, + "loss": 1.0614, + "mean_token_accuracy": 0.6916478872299194, + "num_tokens": 35934858.0, + "step": 1392 + }, + { + "epoch": 0.15297605974083023, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.4268147945404053, + "learning_rate": 5.095168374816984e-07, + "loss": 1.065, + "mean_token_accuracy": 0.6838470101356506, + "num_tokens": 35962276.0, + "step": 1393 + }, + { + "epoch": 0.15308587744344387, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.646977663040161, + "learning_rate": 5.09882869692533e-07, + "loss": 1.1886, + "mean_token_accuracy": 0.6608401536941528, + "num_tokens": 35987873.0, + "step": 1394 + }, + { + "epoch": 0.15319569514605755, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.5794475078582764, + "learning_rate": 5.102489019033675e-07, + "loss": 1.063, + "mean_token_accuracy": 0.6852099895477295, + "num_tokens": 36013354.0, + "step": 1395 + }, + { + "epoch": 0.1533055128486712, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.470139741897583, + "learning_rate": 5.10614934114202e-07, + "loss": 1.0789, + "mean_token_accuracy": 0.6823989152908325, + "num_tokens": 36040211.0, + "step": 1396 + }, + { + "epoch": 0.15341533055128487, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.4517781734466553, + "learning_rate": 5.109809663250367e-07, + "loss": 1.1073, + "mean_token_accuracy": 0.6761339902877808, + "num_tokens": 36066374.0, + "step": 1397 + }, + { + "epoch": 0.15352514825389854, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.490320920944214, + "learning_rate": 5.113469985358712e-07, + "loss": 1.0061, + "mean_token_accuracy": 0.6998056173324585, + "num_tokens": 36092114.0, + "step": 1398 + }, + { + "epoch": 0.1536349659565122, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.627815008163452, + "learning_rate": 5.117130307467057e-07, + "loss": 1.0691, + "mean_token_accuracy": 0.6889735460281372, + "num_tokens": 36117150.0, + "step": 1399 + }, + { + "epoch": 0.15374478365912586, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.372453451156616, + "learning_rate": 5.120790629575403e-07, + "loss": 1.1625, + "mean_token_accuracy": 0.6655094623565674, + "num_tokens": 36146676.0, + "step": 1400 + }, + { + "epoch": 0.1538546013617395, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.5060088634490967, + "learning_rate": 5.124450951683748e-07, + "loss": 1.1164, + "mean_token_accuracy": 0.6722222566604614, + "num_tokens": 36173244.0, + "step": 1401 + }, + { + "epoch": 0.15396441906435318, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.3258190155029297, + "learning_rate": 5.128111273792094e-07, + "loss": 1.1188, + "mean_token_accuracy": 0.6775875687599182, + "num_tokens": 36203197.0, + "step": 1402 + }, + { + "epoch": 0.15407423676696683, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.6759166717529297, + "learning_rate": 5.13177159590044e-07, + "loss": 1.1193, + "mean_token_accuracy": 0.6736017465591431, + "num_tokens": 36231276.0, + "step": 1403 + }, + { + "epoch": 0.1541840544695805, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.2022578716278076, + "learning_rate": 5.135431918008785e-07, + "loss": 1.0842, + "mean_token_accuracy": 0.6900675892829895, + "num_tokens": 36261923.0, + "step": 1404 + }, + { + "epoch": 0.15429387217219415, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.6015255451202393, + "learning_rate": 5.13909224011713e-07, + "loss": 0.9987, + "mean_token_accuracy": 0.7101227045059204, + "num_tokens": 36285200.0, + "step": 1405 + }, + { + "epoch": 0.15440368987480782, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.417923927307129, + "learning_rate": 5.142752562225476e-07, + "loss": 1.0857, + "mean_token_accuracy": 0.6793569922447205, + "num_tokens": 36310732.0, + "step": 1406 + }, + { + "epoch": 0.1545135075774215, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.6738884449005127, + "learning_rate": 5.14641288433382e-07, + "loss": 1.0135, + "mean_token_accuracy": 0.7003214359283447, + "num_tokens": 36333833.0, + "step": 1407 + }, + { + "epoch": 0.15462332528003514, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.5590524673461914, + "learning_rate": 5.150073206442168e-07, + "loss": 1.0239, + "mean_token_accuracy": 0.7022621631622314, + "num_tokens": 36357548.0, + "step": 1408 + }, + { + "epoch": 0.1547331429826488, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.376589775085449, + "learning_rate": 5.153733528550513e-07, + "loss": 0.9811, + "mean_token_accuracy": 0.7069408893585205, + "num_tokens": 36386854.0, + "step": 1409 + }, + { + "epoch": 0.15484296068526246, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.2187693119049072, + "learning_rate": 5.157393850658857e-07, + "loss": 1.1762, + "mean_token_accuracy": 0.6612781882286072, + "num_tokens": 36425311.0, + "step": 1410 + }, + { + "epoch": 0.15495277838787613, + "ewc_loss": 2.5480985641479492e-06, + "grad_norm": 2.2817912101745605, + "learning_rate": 5.161054172767203e-07, + "loss": 0.9915, + "mean_token_accuracy": 0.7008216381072998, + "num_tokens": 36453543.0, + "step": 1411 + }, + { + "epoch": 0.15506259609048978, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.531979560852051, + "learning_rate": 5.164714494875548e-07, + "loss": 1.1021, + "mean_token_accuracy": 0.687190055847168, + "num_tokens": 36479897.0, + "step": 1412 + }, + { + "epoch": 0.15517241379310345, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.4715259075164795, + "learning_rate": 5.168374816983894e-07, + "loss": 1.0945, + "mean_token_accuracy": 0.6800373792648315, + "num_tokens": 36508596.0, + "step": 1413 + }, + { + "epoch": 0.1552822314957171, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.7149672508239746, + "learning_rate": 5.17203513909224e-07, + "loss": 1.0912, + "mean_token_accuracy": 0.6782618761062622, + "num_tokens": 36532995.0, + "step": 1414 + }, + { + "epoch": 0.15539204919833077, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.24442195892334, + "learning_rate": 5.175695461200585e-07, + "loss": 1.1, + "mean_token_accuracy": 0.6778879761695862, + "num_tokens": 36565082.0, + "step": 1415 + }, + { + "epoch": 0.15550186690094442, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.394304037094116, + "learning_rate": 5.17935578330893e-07, + "loss": 1.178, + "mean_token_accuracy": 0.6540803909301758, + "num_tokens": 36595531.0, + "step": 1416 + }, + { + "epoch": 0.1556116846035581, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.5228137969970703, + "learning_rate": 5.183016105417276e-07, + "loss": 1.033, + "mean_token_accuracy": 0.6975818276405334, + "num_tokens": 36618337.0, + "step": 1417 + }, + { + "epoch": 0.15572150230617177, + "ewc_loss": 2.562999725341797e-06, + "grad_norm": 2.849180221557617, + "learning_rate": 5.186676427525622e-07, + "loss": 1.1298, + "mean_token_accuracy": 0.6684250831604004, + "num_tokens": 36641123.0, + "step": 1418 + }, + { + "epoch": 0.1558313200087854, + "ewc_loss": 2.6226043701171875e-06, + "grad_norm": 8.796476364135742, + "learning_rate": 5.190336749633967e-07, + "loss": 1.0769, + "mean_token_accuracy": 0.6899842023849487, + "num_tokens": 36667870.0, + "step": 1419 + }, + { + "epoch": 0.15594113771139909, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.391516923904419, + "learning_rate": 5.193997071742313e-07, + "loss": 1.0647, + "mean_token_accuracy": 0.6903964877128601, + "num_tokens": 36697458.0, + "step": 1420 + }, + { + "epoch": 0.15605095541401273, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.3951475620269775, + "learning_rate": 5.197657393850658e-07, + "loss": 1.1178, + "mean_token_accuracy": 0.6779218912124634, + "num_tokens": 36728999.0, + "step": 1421 + }, + { + "epoch": 0.1561607731166264, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.920792579650879, + "learning_rate": 5.201317715959003e-07, + "loss": 1.1546, + "mean_token_accuracy": 0.6672343015670776, + "num_tokens": 36751360.0, + "step": 1422 + }, + { + "epoch": 0.15627059081924005, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.4945068359375, + "learning_rate": 5.204978038067349e-07, + "loss": 1.1129, + "mean_token_accuracy": 0.6805871725082397, + "num_tokens": 36778609.0, + "step": 1423 + }, + { + "epoch": 0.15638040852185373, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.594143867492676, + "learning_rate": 5.208638360175695e-07, + "loss": 1.036, + "mean_token_accuracy": 0.692102313041687, + "num_tokens": 36801124.0, + "step": 1424 + }, + { + "epoch": 0.15649022622446737, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.5712947845458984, + "learning_rate": 5.212298682284041e-07, + "loss": 1.1144, + "mean_token_accuracy": 0.6706631779670715, + "num_tokens": 36826030.0, + "step": 1425 + }, + { + "epoch": 0.15660004392708105, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.3050355911254883, + "learning_rate": 5.215959004392386e-07, + "loss": 1.1389, + "mean_token_accuracy": 0.6684393286705017, + "num_tokens": 36857373.0, + "step": 1426 + }, + { + "epoch": 0.15670986162969472, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.465702533721924, + "learning_rate": 5.219619326500731e-07, + "loss": 1.0397, + "mean_token_accuracy": 0.6960639357566833, + "num_tokens": 36882743.0, + "step": 1427 + }, + { + "epoch": 0.15681967933230836, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.4346768856048584, + "learning_rate": 5.223279648609077e-07, + "loss": 1.0537, + "mean_token_accuracy": 0.6898795366287231, + "num_tokens": 36908317.0, + "step": 1428 + }, + { + "epoch": 0.15692949703492204, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.672381639480591, + "learning_rate": 5.226939970717423e-07, + "loss": 1.0856, + "mean_token_accuracy": 0.6775422096252441, + "num_tokens": 36933934.0, + "step": 1429 + }, + { + "epoch": 0.15703931473753568, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.528325080871582, + "learning_rate": 5.230600292825768e-07, + "loss": 1.0237, + "mean_token_accuracy": 0.6954628825187683, + "num_tokens": 36961658.0, + "step": 1430 + }, + { + "epoch": 0.15714913244014936, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.502633810043335, + "learning_rate": 5.234260614934114e-07, + "loss": 1.0523, + "mean_token_accuracy": 0.6852378845214844, + "num_tokens": 36989890.0, + "step": 1431 + }, + { + "epoch": 0.157258950142763, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.736053228378296, + "learning_rate": 5.237920937042459e-07, + "loss": 1.1451, + "mean_token_accuracy": 0.6692034006118774, + "num_tokens": 37014359.0, + "step": 1432 + }, + { + "epoch": 0.15736876784537668, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.663618326187134, + "learning_rate": 5.241581259150804e-07, + "loss": 1.0515, + "mean_token_accuracy": 0.6922943592071533, + "num_tokens": 37036466.0, + "step": 1433 + }, + { + "epoch": 0.15747858554799032, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.898911476135254, + "learning_rate": 5.245241581259151e-07, + "loss": 0.9984, + "mean_token_accuracy": 0.7033485174179077, + "num_tokens": 37056557.0, + "step": 1434 + }, + { + "epoch": 0.157588403250604, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.6169850826263428, + "learning_rate": 5.248901903367496e-07, + "loss": 1.1129, + "mean_token_accuracy": 0.6727210283279419, + "num_tokens": 37079546.0, + "step": 1435 + }, + { + "epoch": 0.15769822095321767, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.795379161834717, + "learning_rate": 5.252562225475841e-07, + "loss": 1.14, + "mean_token_accuracy": 0.667387068271637, + "num_tokens": 37103143.0, + "step": 1436 + }, + { + "epoch": 0.15780803865583132, + "ewc_loss": 2.5779008865356445e-06, + "grad_norm": 2.472705602645874, + "learning_rate": 5.256222547584187e-07, + "loss": 1.0041, + "mean_token_accuracy": 0.719074010848999, + "num_tokens": 37128958.0, + "step": 1437 + }, + { + "epoch": 0.157917856358445, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.7835402488708496, + "learning_rate": 5.259882869692532e-07, + "loss": 0.998, + "mean_token_accuracy": 0.7064732313156128, + "num_tokens": 37148862.0, + "step": 1438 + }, + { + "epoch": 0.15802767406105864, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.7132792472839355, + "learning_rate": 5.263543191800877e-07, + "loss": 0.9583, + "mean_token_accuracy": 0.7111896872520447, + "num_tokens": 37168500.0, + "step": 1439 + }, + { + "epoch": 0.1581374917636723, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.6638994216918945, + "learning_rate": 5.267203513909224e-07, + "loss": 1.0827, + "mean_token_accuracy": 0.6799057722091675, + "num_tokens": 37191202.0, + "step": 1440 + }, + { + "epoch": 0.15824730946628596, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.601888656616211, + "learning_rate": 5.270863836017569e-07, + "loss": 1.1389, + "mean_token_accuracy": 0.6633559465408325, + "num_tokens": 37215939.0, + "step": 1441 + }, + { + "epoch": 0.15835712716889963, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.363375663757324, + "learning_rate": 5.274524158125915e-07, + "loss": 1.2561, + "mean_token_accuracy": 0.6605101823806763, + "num_tokens": 37246502.0, + "step": 1442 + }, + { + "epoch": 0.15846694487151328, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.619359016418457, + "learning_rate": 5.27818448023426e-07, + "loss": 0.9618, + "mean_token_accuracy": 0.7125669121742249, + "num_tokens": 37269519.0, + "step": 1443 + }, + { + "epoch": 0.15857676257412695, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.421994924545288, + "learning_rate": 5.281844802342605e-07, + "loss": 1.0302, + "mean_token_accuracy": 0.6985598206520081, + "num_tokens": 37295472.0, + "step": 1444 + }, + { + "epoch": 0.15868658027674062, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.4268627166748047, + "learning_rate": 5.285505124450952e-07, + "loss": 1.1097, + "mean_token_accuracy": 0.6774346232414246, + "num_tokens": 37323089.0, + "step": 1445 + }, + { + "epoch": 0.15879639797935427, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.6108386516571045, + "learning_rate": 5.289165446559297e-07, + "loss": 1.1174, + "mean_token_accuracy": 0.6691844463348389, + "num_tokens": 37349083.0, + "step": 1446 + }, + { + "epoch": 0.15890621568196794, + "ewc_loss": 2.592802047729492e-06, + "grad_norm": 2.780517101287842, + "learning_rate": 5.292825768667642e-07, + "loss": 0.9326, + "mean_token_accuracy": 0.7178113460540771, + "num_tokens": 37370736.0, + "step": 1447 + }, + { + "epoch": 0.1590160333845816, + "ewc_loss": 2.60770320892334e-06, + "grad_norm": 2.6765248775482178, + "learning_rate": 5.296486090775988e-07, + "loss": 1.0937, + "mean_token_accuracy": 0.6815215945243835, + "num_tokens": 37395554.0, + "step": 1448 + }, + { + "epoch": 0.15912585108719526, + "ewc_loss": 2.637505531311035e-06, + "grad_norm": 2.5943968296051025, + "learning_rate": 5.300146412884333e-07, + "loss": 1.0597, + "mean_token_accuracy": 0.6874808073043823, + "num_tokens": 37418863.0, + "step": 1449 + }, + { + "epoch": 0.1592356687898089, + "ewc_loss": 2.637505531311035e-06, + "grad_norm": 3.1525967121124268, + "learning_rate": 5.303806734992679e-07, + "loss": 1.0511, + "mean_token_accuracy": 0.6835231781005859, + "num_tokens": 37435803.0, + "step": 1450 + }, + { + "epoch": 0.15934548649242258, + "ewc_loss": 2.637505531311035e-06, + "grad_norm": 2.506889581680298, + "learning_rate": 5.307467057101025e-07, + "loss": 1.1684, + "mean_token_accuracy": 0.6565335988998413, + "num_tokens": 37465254.0, + "step": 1451 + }, + { + "epoch": 0.15945530419503623, + "ewc_loss": 2.637505531311035e-06, + "grad_norm": 2.4593241214752197, + "learning_rate": 5.31112737920937e-07, + "loss": 1.0075, + "mean_token_accuracy": 0.7061479091644287, + "num_tokens": 37490392.0, + "step": 1452 + }, + { + "epoch": 0.1595651218976499, + "ewc_loss": 2.652406692504883e-06, + "grad_norm": 2.508796453475952, + "learning_rate": 5.314787701317715e-07, + "loss": 1.0626, + "mean_token_accuracy": 0.6859419345855713, + "num_tokens": 37515868.0, + "step": 1453 + }, + { + "epoch": 0.15967493960026355, + "ewc_loss": 2.652406692504883e-06, + "grad_norm": 2.7151894569396973, + "learning_rate": 5.318448023426061e-07, + "loss": 1.0089, + "mean_token_accuracy": 0.7028732299804688, + "num_tokens": 37539607.0, + "step": 1454 + }, + { + "epoch": 0.15978475730287722, + "ewc_loss": 2.652406692504883e-06, + "grad_norm": 3.115355968475342, + "learning_rate": 5.322108345534406e-07, + "loss": 0.9403, + "mean_token_accuracy": 0.7139208316802979, + "num_tokens": 37557033.0, + "step": 1455 + }, + { + "epoch": 0.1598945750054909, + "ewc_loss": 2.652406692504883e-06, + "grad_norm": 2.4859328269958496, + "learning_rate": 5.325768667642752e-07, + "loss": 0.9974, + "mean_token_accuracy": 0.6997157335281372, + "num_tokens": 37581118.0, + "step": 1456 + }, + { + "epoch": 0.16000439270810454, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.5146329402923584, + "learning_rate": 5.329428989751098e-07, + "loss": 1.1495, + "mean_token_accuracy": 0.6667271852493286, + "num_tokens": 37609075.0, + "step": 1457 + }, + { + "epoch": 0.16011421041071822, + "ewc_loss": 2.682209014892578e-06, + "grad_norm": 2.32045316696167, + "learning_rate": 5.333089311859443e-07, + "loss": 1.0596, + "mean_token_accuracy": 0.6874954700469971, + "num_tokens": 37636878.0, + "step": 1458 + }, + { + "epoch": 0.16022402811333186, + "ewc_loss": 2.682209014892578e-06, + "grad_norm": 2.3253707885742188, + "learning_rate": 5.336749633967789e-07, + "loss": 1.1173, + "mean_token_accuracy": 0.6676127910614014, + "num_tokens": 37667267.0, + "step": 1459 + }, + { + "epoch": 0.16033384581594554, + "ewc_loss": 2.682209014892578e-06, + "grad_norm": 2.930811643600464, + "learning_rate": 5.340409956076134e-07, + "loss": 1.0265, + "mean_token_accuracy": 0.698307991027832, + "num_tokens": 37688240.0, + "step": 1460 + }, + { + "epoch": 0.16044366351855918, + "ewc_loss": 2.682209014892578e-06, + "grad_norm": 2.833984136581421, + "learning_rate": 5.34407027818448e-07, + "loss": 0.9467, + "mean_token_accuracy": 0.71248459815979, + "num_tokens": 37707416.0, + "step": 1461 + }, + { + "epoch": 0.16055348122117286, + "ewc_loss": 2.682209014892578e-06, + "grad_norm": 2.2792327404022217, + "learning_rate": 5.347730600292826e-07, + "loss": 1.0955, + "mean_token_accuracy": 0.6828501224517822, + "num_tokens": 37740039.0, + "step": 1462 + }, + { + "epoch": 0.1606632989237865, + "ewc_loss": 2.682209014892578e-06, + "grad_norm": 2.2151870727539062, + "learning_rate": 5.351390922401171e-07, + "loss": 1.1544, + "mean_token_accuracy": 0.6676994562149048, + "num_tokens": 37773476.0, + "step": 1463 + }, + { + "epoch": 0.16077311662640018, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.3973889350891113, + "learning_rate": 5.355051244509516e-07, + "loss": 1.0765, + "mean_token_accuracy": 0.6885712146759033, + "num_tokens": 37802625.0, + "step": 1464 + }, + { + "epoch": 0.16088293432901385, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 3.03605580329895, + "learning_rate": 5.358711566617862e-07, + "loss": 0.9798, + "mean_token_accuracy": 0.7089778780937195, + "num_tokens": 37822336.0, + "step": 1465 + }, + { + "epoch": 0.1609927520316275, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.528489828109741, + "learning_rate": 5.362371888726208e-07, + "loss": 1.0155, + "mean_token_accuracy": 0.6976297497749329, + "num_tokens": 37845801.0, + "step": 1466 + }, + { + "epoch": 0.16110256973424117, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.5519542694091797, + "learning_rate": 5.366032210834553e-07, + "loss": 1.1264, + "mean_token_accuracy": 0.6763547658920288, + "num_tokens": 37872498.0, + "step": 1467 + }, + { + "epoch": 0.16121238743685481, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.298821449279785, + "learning_rate": 5.369692532942899e-07, + "loss": 1.0674, + "mean_token_accuracy": 0.6869387626647949, + "num_tokens": 37904152.0, + "step": 1468 + }, + { + "epoch": 0.1613222051394685, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.297395944595337, + "learning_rate": 5.373352855051244e-07, + "loss": 0.9926, + "mean_token_accuracy": 0.7105991244316101, + "num_tokens": 37931150.0, + "step": 1469 + }, + { + "epoch": 0.16143202284208213, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.592435836791992, + "learning_rate": 5.377013177159589e-07, + "loss": 1.0679, + "mean_token_accuracy": 0.684264600276947, + "num_tokens": 37955011.0, + "step": 1470 + }, + { + "epoch": 0.1615418405446958, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.645087480545044, + "learning_rate": 5.380673499267935e-07, + "loss": 1.1336, + "mean_token_accuracy": 0.6706168055534363, + "num_tokens": 37981518.0, + "step": 1471 + }, + { + "epoch": 0.16165165824730945, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.508427619934082, + "learning_rate": 5.384333821376281e-07, + "loss": 1.1, + "mean_token_accuracy": 0.6800264120101929, + "num_tokens": 38008938.0, + "step": 1472 + }, + { + "epoch": 0.16176147594992313, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.50714373588562, + "learning_rate": 5.387994143484626e-07, + "loss": 1.057, + "mean_token_accuracy": 0.6858639717102051, + "num_tokens": 38036626.0, + "step": 1473 + }, + { + "epoch": 0.1618712936525368, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.1904659271240234, + "learning_rate": 5.391654465592972e-07, + "loss": 1.1148, + "mean_token_accuracy": 0.6732478141784668, + "num_tokens": 38070141.0, + "step": 1474 + }, + { + "epoch": 0.16198111135515045, + "ewc_loss": 2.6673078536987305e-06, + "grad_norm": 2.7954204082489014, + "learning_rate": 5.395314787701317e-07, + "loss": 1.0325, + "mean_token_accuracy": 0.6926717758178711, + "num_tokens": 38092134.0, + "step": 1475 + }, + { + "epoch": 0.16209092905776412, + "ewc_loss": 2.6971101760864258e-06, + "grad_norm": 3.571324348449707, + "learning_rate": 5.398975109809663e-07, + "loss": 0.9973, + "mean_token_accuracy": 0.6999035477638245, + "num_tokens": 38111775.0, + "step": 1476 + }, + { + "epoch": 0.16220074676037777, + "ewc_loss": 2.6971101760864258e-06, + "grad_norm": 2.861355781555176, + "learning_rate": 5.402635431918009e-07, + "loss": 1.1069, + "mean_token_accuracy": 0.6759517192840576, + "num_tokens": 38134366.0, + "step": 1477 + }, + { + "epoch": 0.16231056446299144, + "ewc_loss": 2.6971101760864258e-06, + "grad_norm": 2.619075298309326, + "learning_rate": 5.406295754026354e-07, + "loss": 1.0962, + "mean_token_accuracy": 0.6764318943023682, + "num_tokens": 38159731.0, + "step": 1478 + }, + { + "epoch": 0.1624203821656051, + "ewc_loss": 2.7120113372802734e-06, + "grad_norm": 2.486194133758545, + "learning_rate": 5.4099560761347e-07, + "loss": 1.0728, + "mean_token_accuracy": 0.6793676614761353, + "num_tokens": 38189414.0, + "step": 1479 + }, + { + "epoch": 0.16253019986821876, + "ewc_loss": 2.7120113372802734e-06, + "grad_norm": 2.6005308628082275, + "learning_rate": 5.413616398243045e-07, + "loss": 1.0413, + "mean_token_accuracy": 0.6905560493469238, + "num_tokens": 38213059.0, + "step": 1480 + }, + { + "epoch": 0.1626400175708324, + "ewc_loss": 2.7120113372802734e-06, + "grad_norm": 2.44386887550354, + "learning_rate": 5.41727672035139e-07, + "loss": 1.1326, + "mean_token_accuracy": 0.6662691831588745, + "num_tokens": 38242367.0, + "step": 1481 + }, + { + "epoch": 0.16274983527344608, + "ewc_loss": 2.7120113372802734e-06, + "grad_norm": 2.37910532951355, + "learning_rate": 5.420937042459737e-07, + "loss": 1.1475, + "mean_token_accuracy": 0.6610085964202881, + "num_tokens": 38269270.0, + "step": 1482 + }, + { + "epoch": 0.16285965297605975, + "ewc_loss": 2.7120113372802734e-06, + "grad_norm": 2.423428773880005, + "learning_rate": 5.424597364568082e-07, + "loss": 1.11, + "mean_token_accuracy": 0.6864285469055176, + "num_tokens": 38297171.0, + "step": 1483 + }, + { + "epoch": 0.1629694706786734, + "ewc_loss": 2.6971101760864258e-06, + "grad_norm": 3.679067373275757, + "learning_rate": 5.428257686676427e-07, + "loss": 1.11, + "mean_token_accuracy": 0.6727390289306641, + "num_tokens": 38319751.0, + "step": 1484 + }, + { + "epoch": 0.16307928838128707, + "ewc_loss": 2.7120113372802734e-06, + "grad_norm": 2.6244699954986572, + "learning_rate": 5.431918008784773e-07, + "loss": 1.0414, + "mean_token_accuracy": 0.6913816332817078, + "num_tokens": 38344590.0, + "step": 1485 + }, + { + "epoch": 0.16318910608390072, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.5965659618377686, + "learning_rate": 5.435578330893118e-07, + "loss": 1.0209, + "mean_token_accuracy": 0.7011120319366455, + "num_tokens": 38369984.0, + "step": 1486 + }, + { + "epoch": 0.1632989237865144, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.707644462585449, + "learning_rate": 5.439238653001463e-07, + "loss": 1.0178, + "mean_token_accuracy": 0.6989149451255798, + "num_tokens": 38392266.0, + "step": 1487 + }, + { + "epoch": 0.16340874148912804, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.177626848220825, + "learning_rate": 5.44289897510981e-07, + "loss": 1.1879, + "mean_token_accuracy": 0.6583385467529297, + "num_tokens": 38426150.0, + "step": 1488 + }, + { + "epoch": 0.1635185591917417, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.21036958694458, + "learning_rate": 5.446559297218155e-07, + "loss": 1.0277, + "mean_token_accuracy": 0.7026016712188721, + "num_tokens": 38458153.0, + "step": 1489 + }, + { + "epoch": 0.16362837689435536, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 3.198282480239868, + "learning_rate": 5.4502196193265e-07, + "loss": 1.0607, + "mean_token_accuracy": 0.6946844458580017, + "num_tokens": 38478408.0, + "step": 1490 + }, + { + "epoch": 0.16373819459696903, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.3617825508117676, + "learning_rate": 5.453879941434846e-07, + "loss": 0.9794, + "mean_token_accuracy": 0.7038341164588928, + "num_tokens": 38507099.0, + "step": 1491 + }, + { + "epoch": 0.16384801229958268, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.3563249111175537, + "learning_rate": 5.457540263543191e-07, + "loss": 1.0505, + "mean_token_accuracy": 0.6909745335578918, + "num_tokens": 38532923.0, + "step": 1492 + }, + { + "epoch": 0.16395783000219635, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.493039608001709, + "learning_rate": 5.461200585651538e-07, + "loss": 1.1233, + "mean_token_accuracy": 0.6733568906784058, + "num_tokens": 38558137.0, + "step": 1493 + }, + { + "epoch": 0.16406764770481003, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.4233975410461426, + "learning_rate": 5.464860907759883e-07, + "loss": 1.0634, + "mean_token_accuracy": 0.6879614591598511, + "num_tokens": 38584890.0, + "step": 1494 + }, + { + "epoch": 0.16417746540742367, + "ewc_loss": 2.726912498474121e-06, + "grad_norm": 2.5572938919067383, + "learning_rate": 5.468521229868228e-07, + "loss": 1.0671, + "mean_token_accuracy": 0.6853164434432983, + "num_tokens": 38611310.0, + "step": 1495 + }, + { + "epoch": 0.16428728311003735, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.4798102378845215, + "learning_rate": 5.472181551976574e-07, + "loss": 1.1154, + "mean_token_accuracy": 0.6756832599639893, + "num_tokens": 38638118.0, + "step": 1496 + }, + { + "epoch": 0.164397100812651, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.7936580181121826, + "learning_rate": 5.475841874084919e-07, + "loss": 0.9784, + "mean_token_accuracy": 0.7076689600944519, + "num_tokens": 38660015.0, + "step": 1497 + }, + { + "epoch": 0.16450691851526467, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.483268976211548, + "learning_rate": 5.479502196193265e-07, + "loss": 1.0885, + "mean_token_accuracy": 0.6797440052032471, + "num_tokens": 38688580.0, + "step": 1498 + }, + { + "epoch": 0.1646167362178783, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.5069663524627686, + "learning_rate": 5.483162518301611e-07, + "loss": 1.099, + "mean_token_accuracy": 0.6787666082382202, + "num_tokens": 38714193.0, + "step": 1499 + }, + { + "epoch": 0.16472655392049199, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.4015462398529053, + "learning_rate": 5.486822840409956e-07, + "loss": 1.0884, + "mean_token_accuracy": 0.6787887811660767, + "num_tokens": 38742448.0, + "step": 1500 + }, + { + "epoch": 0.16483637162310563, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.7823266983032227, + "learning_rate": 5.490483162518301e-07, + "loss": 0.9955, + "mean_token_accuracy": 0.7070558667182922, + "num_tokens": 38765620.0, + "step": 1501 + }, + { + "epoch": 0.1649461893257193, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.5707144737243652, + "learning_rate": 5.494143484626647e-07, + "loss": 0.9711, + "mean_token_accuracy": 0.7112984657287598, + "num_tokens": 38788077.0, + "step": 1502 + }, + { + "epoch": 0.16505600702833298, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.6536645889282227, + "learning_rate": 5.497803806734992e-07, + "loss": 1.0333, + "mean_token_accuracy": 0.6979369521141052, + "num_tokens": 38811822.0, + "step": 1503 + }, + { + "epoch": 0.16516582473094663, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.532141923904419, + "learning_rate": 5.501464128843338e-07, + "loss": 1.0514, + "mean_token_accuracy": 0.6884059906005859, + "num_tokens": 38835722.0, + "step": 1504 + }, + { + "epoch": 0.1652756424335603, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.441377878189087, + "learning_rate": 5.505124450951684e-07, + "loss": 1.0069, + "mean_token_accuracy": 0.7001391649246216, + "num_tokens": 38859889.0, + "step": 1505 + }, + { + "epoch": 0.16538546013617395, + "ewc_loss": 2.7418136596679688e-06, + "grad_norm": 2.302013397216797, + "learning_rate": 5.508784773060029e-07, + "loss": 1.0073, + "mean_token_accuracy": 0.7140339612960815, + "num_tokens": 38890641.0, + "step": 1506 + }, + { + "epoch": 0.16549527783878762, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.700157880783081, + "learning_rate": 5.512445095168374e-07, + "loss": 1.1146, + "mean_token_accuracy": 0.6776021718978882, + "num_tokens": 38914684.0, + "step": 1507 + }, + { + "epoch": 0.16560509554140126, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.7780709266662598, + "learning_rate": 5.51610541727672e-07, + "loss": 1.0154, + "mean_token_accuracy": 0.7029469013214111, + "num_tokens": 38935499.0, + "step": 1508 + }, + { + "epoch": 0.16571491324401494, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.5710484981536865, + "learning_rate": 5.519765739385066e-07, + "loss": 1.0549, + "mean_token_accuracy": 0.6861419081687927, + "num_tokens": 38960393.0, + "step": 1509 + }, + { + "epoch": 0.16582473094662858, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.7378060817718506, + "learning_rate": 5.523426061493412e-07, + "loss": 1.0845, + "mean_token_accuracy": 0.6896432042121887, + "num_tokens": 38983330.0, + "step": 1510 + }, + { + "epoch": 0.16593454864924226, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.3130075931549072, + "learning_rate": 5.527086383601757e-07, + "loss": 1.1455, + "mean_token_accuracy": 0.6675930619239807, + "num_tokens": 39011686.0, + "step": 1511 + }, + { + "epoch": 0.16604436635185593, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.487351894378662, + "learning_rate": 5.530746705710102e-07, + "loss": 1.0722, + "mean_token_accuracy": 0.6868326663970947, + "num_tokens": 39037527.0, + "step": 1512 + }, + { + "epoch": 0.16615418405446958, + "ewc_loss": 2.7567148208618164e-06, + "grad_norm": 2.749279737472534, + "learning_rate": 5.534407027818448e-07, + "loss": 1.1173, + "mean_token_accuracy": 0.671414852142334, + "num_tokens": 39061584.0, + "step": 1513 + }, + { + "epoch": 0.16626400175708325, + "ewc_loss": 2.7865171432495117e-06, + "grad_norm": 2.719109058380127, + "learning_rate": 5.538067349926794e-07, + "loss": 1.0323, + "mean_token_accuracy": 0.6942468881607056, + "num_tokens": 39085730.0, + "step": 1514 + }, + { + "epoch": 0.1663738194596969, + "ewc_loss": 2.7865171432495117e-06, + "grad_norm": 3.0287435054779053, + "learning_rate": 5.541727672035139e-07, + "loss": 1.0435, + "mean_token_accuracy": 0.6971535086631775, + "num_tokens": 39103747.0, + "step": 1515 + }, + { + "epoch": 0.16648363716231057, + "ewc_loss": 2.7865171432495117e-06, + "grad_norm": 3.502547264099121, + "learning_rate": 5.545387994143485e-07, + "loss": 0.9994, + "mean_token_accuracy": 0.6988981366157532, + "num_tokens": 39124300.0, + "step": 1516 + }, + { + "epoch": 0.16659345486492422, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.597113847732544, + "learning_rate": 5.54904831625183e-07, + "loss": 0.9672, + "mean_token_accuracy": 0.7068397998809814, + "num_tokens": 39148700.0, + "step": 1517 + }, + { + "epoch": 0.1667032725675379, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.837911367416382, + "learning_rate": 5.552708638360175e-07, + "loss": 1.1274, + "mean_token_accuracy": 0.666939377784729, + "num_tokens": 39172508.0, + "step": 1518 + }, + { + "epoch": 0.16681309027015154, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.7075068950653076, + "learning_rate": 5.556368960468521e-07, + "loss": 0.9731, + "mean_token_accuracy": 0.7096099853515625, + "num_tokens": 39197782.0, + "step": 1519 + }, + { + "epoch": 0.1669229079727652, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.4275710582733154, + "learning_rate": 5.560029282576867e-07, + "loss": 1.0244, + "mean_token_accuracy": 0.6965091228485107, + "num_tokens": 39224616.0, + "step": 1520 + }, + { + "epoch": 0.16703272567537888, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.45424747467041, + "learning_rate": 5.563689604685212e-07, + "loss": 1.0132, + "mean_token_accuracy": 0.6982669830322266, + "num_tokens": 39249771.0, + "step": 1521 + }, + { + "epoch": 0.16714254337799253, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.660874128341675, + "learning_rate": 5.567349926793558e-07, + "loss": 0.9434, + "mean_token_accuracy": 0.7130613923072815, + "num_tokens": 39271524.0, + "step": 1522 + }, + { + "epoch": 0.1672523610806062, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.4118809700012207, + "learning_rate": 5.571010248901903e-07, + "loss": 1.0141, + "mean_token_accuracy": 0.6990441083908081, + "num_tokens": 39299086.0, + "step": 1523 + }, + { + "epoch": 0.16736217878321985, + "ewc_loss": 2.7865171432495117e-06, + "grad_norm": 2.2778923511505127, + "learning_rate": 5.574670571010248e-07, + "loss": 1.0248, + "mean_token_accuracy": 0.7041472792625427, + "num_tokens": 39331423.0, + "step": 1524 + }, + { + "epoch": 0.16747199648583352, + "ewc_loss": 2.7865171432495117e-06, + "grad_norm": 2.755121946334839, + "learning_rate": 5.578330893118595e-07, + "loss": 1.0797, + "mean_token_accuracy": 0.6820632219314575, + "num_tokens": 39352747.0, + "step": 1525 + }, + { + "epoch": 0.16758181418844717, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.6037981510162354, + "learning_rate": 5.58199121522694e-07, + "loss": 1.0936, + "mean_token_accuracy": 0.681583821773529, + "num_tokens": 39377552.0, + "step": 1526 + }, + { + "epoch": 0.16769163189106084, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.633192777633667, + "learning_rate": 5.585651537335286e-07, + "loss": 1.0688, + "mean_token_accuracy": 0.6845273971557617, + "num_tokens": 39401174.0, + "step": 1527 + }, + { + "epoch": 0.1678014495936745, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.3766636848449707, + "learning_rate": 5.589311859443631e-07, + "loss": 0.9717, + "mean_token_accuracy": 0.709598958492279, + "num_tokens": 39427179.0, + "step": 1528 + }, + { + "epoch": 0.16791126729628816, + "ewc_loss": 2.7865171432495117e-06, + "grad_norm": 2.6666698455810547, + "learning_rate": 5.592972181551976e-07, + "loss": 0.9645, + "mean_token_accuracy": 0.7066785097122192, + "num_tokens": 39446853.0, + "step": 1529 + }, + { + "epoch": 0.1680210849989018, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.682957887649536, + "learning_rate": 5.596632503660323e-07, + "loss": 0.9763, + "mean_token_accuracy": 0.7100669145584106, + "num_tokens": 39467826.0, + "step": 1530 + }, + { + "epoch": 0.16813090270151548, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 16.70714569091797, + "learning_rate": 5.600292825768668e-07, + "loss": 0.934, + "mean_token_accuracy": 0.7170472741127014, + "num_tokens": 39492608.0, + "step": 1531 + }, + { + "epoch": 0.16824072040412916, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.986797571182251, + "learning_rate": 5.603953147877013e-07, + "loss": 1.0785, + "mean_token_accuracy": 0.6881169080734253, + "num_tokens": 39513984.0, + "step": 1532 + }, + { + "epoch": 0.1683505381067428, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.440434217453003, + "learning_rate": 5.607613469985359e-07, + "loss": 0.9961, + "mean_token_accuracy": 0.7036008834838867, + "num_tokens": 39541191.0, + "step": 1533 + }, + { + "epoch": 0.16846035580935648, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.621342420578003, + "learning_rate": 5.611273792093704e-07, + "loss": 1.0052, + "mean_token_accuracy": 0.7143714427947998, + "num_tokens": 39565937.0, + "step": 1534 + }, + { + "epoch": 0.16857017351197012, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.8513190746307373, + "learning_rate": 5.614934114202049e-07, + "loss": 1.0457, + "mean_token_accuracy": 0.6943168640136719, + "num_tokens": 39585586.0, + "step": 1535 + }, + { + "epoch": 0.1686799912145838, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.6313164234161377, + "learning_rate": 5.618594436310396e-07, + "loss": 1.1601, + "mean_token_accuracy": 0.6637606620788574, + "num_tokens": 39621116.0, + "step": 1536 + }, + { + "epoch": 0.16878980891719744, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.6284608840942383, + "learning_rate": 5.622254758418741e-07, + "loss": 1.0925, + "mean_token_accuracy": 0.6826357841491699, + "num_tokens": 39645545.0, + "step": 1537 + }, + { + "epoch": 0.16889962661981112, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.5600900650024414, + "learning_rate": 5.625915080527086e-07, + "loss": 1.0918, + "mean_token_accuracy": 0.6853343844413757, + "num_tokens": 39671881.0, + "step": 1538 + }, + { + "epoch": 0.16900944432242476, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.684398651123047, + "learning_rate": 5.629575402635432e-07, + "loss": 1.0354, + "mean_token_accuracy": 0.6940061450004578, + "num_tokens": 39695290.0, + "step": 1539 + }, + { + "epoch": 0.16911926202503844, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.4587864875793457, + "learning_rate": 5.633235724743777e-07, + "loss": 1.0408, + "mean_token_accuracy": 0.6996798515319824, + "num_tokens": 39722135.0, + "step": 1540 + }, + { + "epoch": 0.1692290797276521, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.4502522945404053, + "learning_rate": 5.636896046852124e-07, + "loss": 1.1699, + "mean_token_accuracy": 0.6604301333427429, + "num_tokens": 39752087.0, + "step": 1541 + }, + { + "epoch": 0.16933889743026576, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.335397243499756, + "learning_rate": 5.640556368960469e-07, + "loss": 1.0446, + "mean_token_accuracy": 0.6933793425559998, + "num_tokens": 39782108.0, + "step": 1542 + }, + { + "epoch": 0.16944871513287943, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.3871376514434814, + "learning_rate": 5.644216691068814e-07, + "loss": 1.0918, + "mean_token_accuracy": 0.6749311089515686, + "num_tokens": 39810707.0, + "step": 1543 + }, + { + "epoch": 0.16955853283549308, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.493884801864624, + "learning_rate": 5.64787701317716e-07, + "loss": 1.0073, + "mean_token_accuracy": 0.6985870599746704, + "num_tokens": 39836838.0, + "step": 1544 + }, + { + "epoch": 0.16966835053810675, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.5983619689941406, + "learning_rate": 5.651537335285505e-07, + "loss": 0.9876, + "mean_token_accuracy": 0.7102533578872681, + "num_tokens": 39860798.0, + "step": 1545 + }, + { + "epoch": 0.1697781682407204, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.609163999557495, + "learning_rate": 5.655197657393851e-07, + "loss": 1.1398, + "mean_token_accuracy": 0.6703338623046875, + "num_tokens": 39886558.0, + "step": 1546 + }, + { + "epoch": 0.16988798594333407, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.7949981689453125, + "learning_rate": 5.658857979502197e-07, + "loss": 0.9669, + "mean_token_accuracy": 0.7053533792495728, + "num_tokens": 39907445.0, + "step": 1547 + }, + { + "epoch": 0.16999780364594771, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.315898895263672, + "learning_rate": 5.662518301610542e-07, + "loss": 0.9656, + "mean_token_accuracy": 0.7170801162719727, + "num_tokens": 39937347.0, + "step": 1548 + }, + { + "epoch": 0.1701076213485614, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.4310855865478516, + "learning_rate": 5.666178623718887e-07, + "loss": 1.0039, + "mean_token_accuracy": 0.7046077847480774, + "num_tokens": 39966512.0, + "step": 1549 + }, + { + "epoch": 0.17021743905117506, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.545819044113159, + "learning_rate": 5.669838945827233e-07, + "loss": 1.0883, + "mean_token_accuracy": 0.6787731647491455, + "num_tokens": 39993941.0, + "step": 1550 + }, + { + "epoch": 0.1703272567537887, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.9648635387420654, + "learning_rate": 5.673499267935577e-07, + "loss": 0.9067, + "mean_token_accuracy": 0.730489194393158, + "num_tokens": 40012552.0, + "step": 1551 + }, + { + "epoch": 0.17043707445640238, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.2949745655059814, + "learning_rate": 5.677159590043924e-07, + "loss": 1.0775, + "mean_token_accuracy": 0.6861027479171753, + "num_tokens": 40043398.0, + "step": 1552 + }, + { + "epoch": 0.17054689215901603, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.5463948249816895, + "learning_rate": 5.68081991215227e-07, + "loss": 1.1131, + "mean_token_accuracy": 0.6765391230583191, + "num_tokens": 40070386.0, + "step": 1553 + }, + { + "epoch": 0.1706567098616297, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.649190902709961, + "learning_rate": 5.684480234260615e-07, + "loss": 1.0415, + "mean_token_accuracy": 0.6950928568840027, + "num_tokens": 40094339.0, + "step": 1554 + }, + { + "epoch": 0.17076652756424335, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.461422920227051, + "learning_rate": 5.688140556368959e-07, + "loss": 0.9545, + "mean_token_accuracy": 0.7194868326187134, + "num_tokens": 40118938.0, + "step": 1555 + }, + { + "epoch": 0.17087634526685702, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.442594051361084, + "learning_rate": 5.691800878477305e-07, + "loss": 1.1304, + "mean_token_accuracy": 0.666756272315979, + "num_tokens": 40148530.0, + "step": 1556 + }, + { + "epoch": 0.17098616296947067, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.4379868507385254, + "learning_rate": 5.695461200585652e-07, + "loss": 1.0849, + "mean_token_accuracy": 0.6868032217025757, + "num_tokens": 40175048.0, + "step": 1557 + }, + { + "epoch": 0.17109598067208434, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.648909330368042, + "learning_rate": 5.699121522693998e-07, + "loss": 1.043, + "mean_token_accuracy": 0.6922048330307007, + "num_tokens": 40198815.0, + "step": 1558 + }, + { + "epoch": 0.17120579837469801, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.6338112354278564, + "learning_rate": 5.702781844802342e-07, + "loss": 1.0154, + "mean_token_accuracy": 0.706497848033905, + "num_tokens": 40221652.0, + "step": 1559 + }, + { + "epoch": 0.17131561607731166, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.569525718688965, + "learning_rate": 5.706442166910687e-07, + "loss": 1.0308, + "mean_token_accuracy": 0.6952993869781494, + "num_tokens": 40248573.0, + "step": 1560 + }, + { + "epoch": 0.17142543377992533, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.7924559116363525, + "learning_rate": 5.710102489019033e-07, + "loss": 1.1058, + "mean_token_accuracy": 0.6786457300186157, + "num_tokens": 40273242.0, + "step": 1561 + }, + { + "epoch": 0.17153525148253898, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 2.6578361988067627, + "learning_rate": 5.71376281112738e-07, + "loss": 1.0303, + "mean_token_accuracy": 0.69501793384552, + "num_tokens": 40295379.0, + "step": 1562 + }, + { + "epoch": 0.17164506918515265, + "ewc_loss": 2.8014183044433594e-06, + "grad_norm": 3.0185325145721436, + "learning_rate": 5.717423133235724e-07, + "loss": 1.0344, + "mean_token_accuracy": 0.6932237148284912, + "num_tokens": 40313918.0, + "step": 1563 + }, + { + "epoch": 0.1717548868877663, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.529045581817627, + "learning_rate": 5.72108345534407e-07, + "loss": 1.0516, + "mean_token_accuracy": 0.690239429473877, + "num_tokens": 40338811.0, + "step": 1564 + }, + { + "epoch": 0.17186470459037997, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.70546293258667, + "learning_rate": 5.724743777452415e-07, + "loss": 1.0585, + "mean_token_accuracy": 0.688822865486145, + "num_tokens": 40361823.0, + "step": 1565 + }, + { + "epoch": 0.17197452229299362, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.6366748809814453, + "learning_rate": 5.72840409956076e-07, + "loss": 1.0931, + "mean_token_accuracy": 0.6886813640594482, + "num_tokens": 40386526.0, + "step": 1566 + }, + { + "epoch": 0.1720843399956073, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.4891409873962402, + "learning_rate": 5.732064421669106e-07, + "loss": 1.0188, + "mean_token_accuracy": 0.7007985711097717, + "num_tokens": 40412670.0, + "step": 1567 + }, + { + "epoch": 0.17219415769822094, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.484132766723633, + "learning_rate": 5.735724743777452e-07, + "loss": 1.0389, + "mean_token_accuracy": 0.6950249671936035, + "num_tokens": 40439359.0, + "step": 1568 + }, + { + "epoch": 0.1723039754008346, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.355456829071045, + "learning_rate": 5.739385065885797e-07, + "loss": 1.0328, + "mean_token_accuracy": 0.6984506249427795, + "num_tokens": 40466494.0, + "step": 1569 + }, + { + "epoch": 0.1724137931034483, + "ewc_loss": 2.816319465637207e-06, + "grad_norm": 2.781970500946045, + "learning_rate": 5.743045387994143e-07, + "loss": 0.9904, + "mean_token_accuracy": 0.7027437686920166, + "num_tokens": 40488412.0, + "step": 1570 + }, + { + "epoch": 0.17252361080606193, + "ewc_loss": 2.8312206268310547e-06, + "grad_norm": 2.6122841835021973, + "learning_rate": 5.746705710102488e-07, + "loss": 1.0856, + "mean_token_accuracy": 0.6815703511238098, + "num_tokens": 40511460.0, + "step": 1571 + }, + { + "epoch": 0.1726334285086756, + "ewc_loss": 2.8312206268310547e-06, + "grad_norm": 2.362157106399536, + "learning_rate": 5.750366032210833e-07, + "loss": 1.0035, + "mean_token_accuracy": 0.7060770988464355, + "num_tokens": 40537397.0, + "step": 1572 + }, + { + "epoch": 0.17274324621128925, + "ewc_loss": 2.8461217880249023e-06, + "grad_norm": 3.053466796875, + "learning_rate": 5.75402635431918e-07, + "loss": 1.0121, + "mean_token_accuracy": 0.7002522945404053, + "num_tokens": 40556838.0, + "step": 1573 + }, + { + "epoch": 0.17285306391390293, + "ewc_loss": 2.8312206268310547e-06, + "grad_norm": 2.973149538040161, + "learning_rate": 5.757686676427525e-07, + "loss": 1.0224, + "mean_token_accuracy": 0.6944022178649902, + "num_tokens": 40575543.0, + "step": 1574 + }, + { + "epoch": 0.17296288161651657, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.340524911880493, + "learning_rate": 5.761346998535871e-07, + "loss": 1.0759, + "mean_token_accuracy": 0.6867619156837463, + "num_tokens": 40604152.0, + "step": 1575 + }, + { + "epoch": 0.17307269931913025, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.5692639350891113, + "learning_rate": 5.765007320644216e-07, + "loss": 1.0693, + "mean_token_accuracy": 0.6857622861862183, + "num_tokens": 40627574.0, + "step": 1576 + }, + { + "epoch": 0.1731825170217439, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.58907413482666, + "learning_rate": 5.768667642752561e-07, + "loss": 1.0108, + "mean_token_accuracy": 0.6987965106964111, + "num_tokens": 40651057.0, + "step": 1577 + }, + { + "epoch": 0.17329233472435757, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.469935655593872, + "learning_rate": 5.772327964860908e-07, + "loss": 1.0455, + "mean_token_accuracy": 0.6886382102966309, + "num_tokens": 40678662.0, + "step": 1578 + }, + { + "epoch": 0.17340215242697124, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.631850004196167, + "learning_rate": 5.775988286969253e-07, + "loss": 1.1087, + "mean_token_accuracy": 0.6829572916030884, + "num_tokens": 40704905.0, + "step": 1579 + }, + { + "epoch": 0.17351197012958489, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.797675132751465, + "learning_rate": 5.779648609077598e-07, + "loss": 0.9732, + "mean_token_accuracy": 0.7124413251876831, + "num_tokens": 40725916.0, + "step": 1580 + }, + { + "epoch": 0.17362178783219856, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.729539632797241, + "learning_rate": 5.783308931185944e-07, + "loss": 1.0716, + "mean_token_accuracy": 0.6842572093009949, + "num_tokens": 40746748.0, + "step": 1581 + }, + { + "epoch": 0.1737316055348122, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.3211135864257812, + "learning_rate": 5.786969253294289e-07, + "loss": 1.1423, + "mean_token_accuracy": 0.6848962306976318, + "num_tokens": 40779091.0, + "step": 1582 + }, + { + "epoch": 0.17384142323742588, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.25596022605896, + "learning_rate": 5.790629575402634e-07, + "loss": 1.122, + "mean_token_accuracy": 0.669992208480835, + "num_tokens": 40809761.0, + "step": 1583 + }, + { + "epoch": 0.17395124094003953, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.3337810039520264, + "learning_rate": 5.794289897510981e-07, + "loss": 1.0873, + "mean_token_accuracy": 0.6845816373825073, + "num_tokens": 40841486.0, + "step": 1584 + }, + { + "epoch": 0.1740610586426532, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.2433512210845947, + "learning_rate": 5.797950219619326e-07, + "loss": 1.1034, + "mean_token_accuracy": 0.681702733039856, + "num_tokens": 40870417.0, + "step": 1585 + }, + { + "epoch": 0.17417087634526685, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.4361515045166016, + "learning_rate": 5.801610541727671e-07, + "loss": 1.0325, + "mean_token_accuracy": 0.7017472386360168, + "num_tokens": 40898422.0, + "step": 1586 + }, + { + "epoch": 0.17428069404788052, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.8394808769226074, + "learning_rate": 5.805270863836017e-07, + "loss": 1.1466, + "mean_token_accuracy": 0.6678264141082764, + "num_tokens": 40921246.0, + "step": 1587 + }, + { + "epoch": 0.1743905117504942, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.4153456687927246, + "learning_rate": 5.808931185944362e-07, + "loss": 1.0339, + "mean_token_accuracy": 0.6897264122962952, + "num_tokens": 40947697.0, + "step": 1588 + }, + { + "epoch": 0.17450032945310784, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.595957040786743, + "learning_rate": 5.812591508052709e-07, + "loss": 0.9598, + "mean_token_accuracy": 0.7159906625747681, + "num_tokens": 40971954.0, + "step": 1589 + }, + { + "epoch": 0.1746101471557215, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.368894338607788, + "learning_rate": 5.816251830161054e-07, + "loss": 1.0766, + "mean_token_accuracy": 0.6804496049880981, + "num_tokens": 41003026.0, + "step": 1590 + }, + { + "epoch": 0.17471996485833516, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.572024345397949, + "learning_rate": 5.819912152269399e-07, + "loss": 1.084, + "mean_token_accuracy": 0.680292546749115, + "num_tokens": 41027889.0, + "step": 1591 + }, + { + "epoch": 0.17482978256094883, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.813654661178589, + "learning_rate": 5.823572474377745e-07, + "loss": 1.0523, + "mean_token_accuracy": 0.6884948015213013, + "num_tokens": 41048983.0, + "step": 1592 + }, + { + "epoch": 0.17493960026356248, + "ewc_loss": 2.86102294921875e-06, + "grad_norm": 2.564664125442505, + "learning_rate": 5.82723279648609e-07, + "loss": 1.105, + "mean_token_accuracy": 0.6709978580474854, + "num_tokens": 41074025.0, + "step": 1593 + }, + { + "epoch": 0.17504941796617615, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.3940718173980713, + "learning_rate": 5.830893118594436e-07, + "loss": 1.1038, + "mean_token_accuracy": 0.6762522459030151, + "num_tokens": 41104142.0, + "step": 1594 + }, + { + "epoch": 0.1751592356687898, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.624232530593872, + "learning_rate": 5.834553440702782e-07, + "loss": 1.0704, + "mean_token_accuracy": 0.6869714260101318, + "num_tokens": 41128610.0, + "step": 1595 + }, + { + "epoch": 0.17526905337140347, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.4940271377563477, + "learning_rate": 5.838213762811127e-07, + "loss": 1.011, + "mean_token_accuracy": 0.7020330429077148, + "num_tokens": 41154215.0, + "step": 1596 + }, + { + "epoch": 0.17537887107401715, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.3809845447540283, + "learning_rate": 5.841874084919472e-07, + "loss": 1.101, + "mean_token_accuracy": 0.6728259325027466, + "num_tokens": 41186683.0, + "step": 1597 + }, + { + "epoch": 0.1754886887766308, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.564112901687622, + "learning_rate": 5.845534407027818e-07, + "loss": 1.0201, + "mean_token_accuracy": 0.6939969062805176, + "num_tokens": 41211138.0, + "step": 1598 + }, + { + "epoch": 0.17559850647924446, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 3.048501968383789, + "learning_rate": 5.849194729136163e-07, + "loss": 0.9477, + "mean_token_accuracy": 0.7160730361938477, + "num_tokens": 41229133.0, + "step": 1599 + }, + { + "epoch": 0.1757083241818581, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.657076835632324, + "learning_rate": 5.852855051244509e-07, + "loss": 1.0807, + "mean_token_accuracy": 0.6868948936462402, + "num_tokens": 41254358.0, + "step": 1600 + }, + { + "epoch": 0.17581814188447178, + "ewc_loss": 2.8908252716064453e-06, + "grad_norm": 2.4165780544281006, + "learning_rate": 5.856515373352855e-07, + "loss": 1.107, + "mean_token_accuracy": 0.6842125058174133, + "num_tokens": 41282152.0, + "step": 1601 + }, + { + "epoch": 0.17592795958708543, + "ewc_loss": 2.905726432800293e-06, + "grad_norm": 2.486022472381592, + "learning_rate": 5.8601756954612e-07, + "loss": 0.9831, + "mean_token_accuracy": 0.7130230069160461, + "num_tokens": 41307760.0, + "step": 1602 + }, + { + "epoch": 0.1760377772896991, + "ewc_loss": 2.905726432800293e-06, + "grad_norm": 2.4954442977905273, + "learning_rate": 5.863836017569545e-07, + "loss": 1.0019, + "mean_token_accuracy": 0.7019009590148926, + "num_tokens": 41333522.0, + "step": 1603 + }, + { + "epoch": 0.17614759499231275, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.5451345443725586, + "learning_rate": 5.867496339677891e-07, + "loss": 1.037, + "mean_token_accuracy": 0.6957181692123413, + "num_tokens": 41357614.0, + "step": 1604 + }, + { + "epoch": 0.17625741269492642, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.4655609130859375, + "learning_rate": 5.871156661786237e-07, + "loss": 0.991, + "mean_token_accuracy": 0.709450364112854, + "num_tokens": 41383591.0, + "step": 1605 + }, + { + "epoch": 0.17636723039754007, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.524016857147217, + "learning_rate": 5.874816983894583e-07, + "loss": 1.1172, + "mean_token_accuracy": 0.6727821826934814, + "num_tokens": 41410821.0, + "step": 1606 + }, + { + "epoch": 0.17647704810015374, + "ewc_loss": 2.9206275939941406e-06, + "grad_norm": 2.6756701469421387, + "learning_rate": 5.878477306002928e-07, + "loss": 1.0889, + "mean_token_accuracy": 0.6842179298400879, + "num_tokens": 41433212.0, + "step": 1607 + }, + { + "epoch": 0.17658686580276742, + "ewc_loss": 2.9206275939941406e-06, + "grad_norm": 2.420332193374634, + "learning_rate": 5.882137628111273e-07, + "loss": 1.1647, + "mean_token_accuracy": 0.6581733226776123, + "num_tokens": 41461618.0, + "step": 1608 + }, + { + "epoch": 0.17669668350538106, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.332998275756836, + "learning_rate": 5.885797950219619e-07, + "loss": 1.0758, + "mean_token_accuracy": 0.6808621883392334, + "num_tokens": 41490369.0, + "step": 1609 + }, + { + "epoch": 0.17680650120799474, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.489532709121704, + "learning_rate": 5.889458272327965e-07, + "loss": 1.2307, + "mean_token_accuracy": 0.6470752954483032, + "num_tokens": 41518493.0, + "step": 1610 + }, + { + "epoch": 0.17691631891060838, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.4877817630767822, + "learning_rate": 5.89311859443631e-07, + "loss": 1.1397, + "mean_token_accuracy": 0.673457682132721, + "num_tokens": 41545826.0, + "step": 1611 + }, + { + "epoch": 0.17702613661322206, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.626561164855957, + "learning_rate": 5.896778916544656e-07, + "loss": 1.0336, + "mean_token_accuracy": 0.6902704834938049, + "num_tokens": 41568877.0, + "step": 1612 + }, + { + "epoch": 0.1771359543158357, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.494407892227173, + "learning_rate": 5.900439238653001e-07, + "loss": 1.0026, + "mean_token_accuracy": 0.7027072310447693, + "num_tokens": 41594228.0, + "step": 1613 + }, + { + "epoch": 0.17724577201844938, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.599519729614258, + "learning_rate": 5.904099560761346e-07, + "loss": 1.1109, + "mean_token_accuracy": 0.6782333254814148, + "num_tokens": 41620861.0, + "step": 1614 + }, + { + "epoch": 0.17735558972106302, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.4643332958221436, + "learning_rate": 5.907759882869692e-07, + "loss": 1.1093, + "mean_token_accuracy": 0.679897129535675, + "num_tokens": 41646859.0, + "step": 1615 + }, + { + "epoch": 0.1774654074236767, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.5424554347991943, + "learning_rate": 5.911420204978038e-07, + "loss": 1.0699, + "mean_token_accuracy": 0.6858869791030884, + "num_tokens": 41670100.0, + "step": 1616 + }, + { + "epoch": 0.17757522512629037, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.726036548614502, + "learning_rate": 5.915080527086383e-07, + "loss": 1.0109, + "mean_token_accuracy": 0.7043467164039612, + "num_tokens": 41690417.0, + "step": 1617 + }, + { + "epoch": 0.17768504282890402, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.5440704822540283, + "learning_rate": 5.918740849194729e-07, + "loss": 0.9757, + "mean_token_accuracy": 0.7192330360412598, + "num_tokens": 41723702.0, + "step": 1618 + }, + { + "epoch": 0.1777948605315177, + "ewc_loss": 2.9355287551879883e-06, + "grad_norm": 2.599290370941162, + "learning_rate": 5.922401171303074e-07, + "loss": 0.9919, + "mean_token_accuracy": 0.7032043933868408, + "num_tokens": 41747292.0, + "step": 1619 + }, + { + "epoch": 0.17790467823413134, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.7326409816741943, + "learning_rate": 5.926061493411419e-07, + "loss": 1.0944, + "mean_token_accuracy": 0.6800866723060608, + "num_tokens": 41771664.0, + "step": 1620 + }, + { + "epoch": 0.178014495936745, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.440568447113037, + "learning_rate": 5.929721815519766e-07, + "loss": 1.105, + "mean_token_accuracy": 0.6765694618225098, + "num_tokens": 41802272.0, + "step": 1621 + }, + { + "epoch": 0.17812431363935866, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.472064733505249, + "learning_rate": 5.933382137628111e-07, + "loss": 1.0468, + "mean_token_accuracy": 0.6879183650016785, + "num_tokens": 41826323.0, + "step": 1622 + }, + { + "epoch": 0.17823413134197233, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.3512792587280273, + "learning_rate": 5.937042459736457e-07, + "loss": 1.1564, + "mean_token_accuracy": 0.6647728085517883, + "num_tokens": 41859043.0, + "step": 1623 + }, + { + "epoch": 0.17834394904458598, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.705500841140747, + "learning_rate": 5.940702781844802e-07, + "loss": 1.0204, + "mean_token_accuracy": 0.6990989446640015, + "num_tokens": 41880055.0, + "step": 1624 + }, + { + "epoch": 0.17845376674719965, + "ewc_loss": 2.950429916381836e-06, + "grad_norm": 2.4493839740753174, + "learning_rate": 5.944363103953147e-07, + "loss": 1.0855, + "mean_token_accuracy": 0.6784408092498779, + "num_tokens": 41906987.0, + "step": 1625 + }, + { + "epoch": 0.17856358444981332, + "ewc_loss": 2.950429916381836e-06, + "grad_norm": 2.685511350631714, + "learning_rate": 5.948023426061494e-07, + "loss": 1.0609, + "mean_token_accuracy": 0.6996237635612488, + "num_tokens": 41929196.0, + "step": 1626 + }, + { + "epoch": 0.17867340215242697, + "ewc_loss": 2.950429916381836e-06, + "grad_norm": 2.931117296218872, + "learning_rate": 5.951683748169839e-07, + "loss": 1.1682, + "mean_token_accuracy": 0.6658121943473816, + "num_tokens": 41955354.0, + "step": 1627 + }, + { + "epoch": 0.17878321985504064, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.5311121940612793, + "learning_rate": 5.955344070278184e-07, + "loss": 1.1673, + "mean_token_accuracy": 0.6571204662322998, + "num_tokens": 41982518.0, + "step": 1628 + }, + { + "epoch": 0.1788930375576543, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.544487714767456, + "learning_rate": 5.95900439238653e-07, + "loss": 1.0183, + "mean_token_accuracy": 0.6974310278892517, + "num_tokens": 42006184.0, + "step": 1629 + }, + { + "epoch": 0.17900285526026796, + "ewc_loss": 2.950429916381836e-06, + "grad_norm": 2.861506462097168, + "learning_rate": 5.962664714494875e-07, + "loss": 1.012, + "mean_token_accuracy": 0.7000073194503784, + "num_tokens": 42025392.0, + "step": 1630 + }, + { + "epoch": 0.1791126729628816, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.371899366378784, + "learning_rate": 5.96632503660322e-07, + "loss": 1.1951, + "mean_token_accuracy": 0.6596313714981079, + "num_tokens": 42055514.0, + "step": 1631 + }, + { + "epoch": 0.17922249066549528, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.5050859451293945, + "learning_rate": 5.969985358711567e-07, + "loss": 1.0904, + "mean_token_accuracy": 0.6817895770072937, + "num_tokens": 42081033.0, + "step": 1632 + }, + { + "epoch": 0.17933230836810893, + "ewc_loss": 2.950429916381836e-06, + "grad_norm": 2.5865743160247803, + "learning_rate": 5.973645680819912e-07, + "loss": 0.997, + "mean_token_accuracy": 0.7088583707809448, + "num_tokens": 42104399.0, + "step": 1633 + }, + { + "epoch": 0.1794421260707226, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.565467596054077, + "learning_rate": 5.977306002928257e-07, + "loss": 1.075, + "mean_token_accuracy": 0.6821130514144897, + "num_tokens": 42128912.0, + "step": 1634 + }, + { + "epoch": 0.17955194377333628, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.8356690406799316, + "learning_rate": 5.980966325036603e-07, + "loss": 1.1116, + "mean_token_accuracy": 0.6750312447547913, + "num_tokens": 42150892.0, + "step": 1635 + }, + { + "epoch": 0.17966176147594992, + "ewc_loss": 2.9802322387695312e-06, + "grad_norm": 2.7466042041778564, + "learning_rate": 5.984626647144948e-07, + "loss": 0.9956, + "mean_token_accuracy": 0.7110235095024109, + "num_tokens": 42172510.0, + "step": 1636 + }, + { + "epoch": 0.1797715791785636, + "ewc_loss": 2.9802322387695312e-06, + "grad_norm": 2.4686036109924316, + "learning_rate": 5.988286969253294e-07, + "loss": 1.0717, + "mean_token_accuracy": 0.6839044094085693, + "num_tokens": 42199310.0, + "step": 1637 + }, + { + "epoch": 0.17988139688117724, + "ewc_loss": 2.9802322387695312e-06, + "grad_norm": 2.334550380706787, + "learning_rate": 5.99194729136164e-07, + "loss": 1.1079, + "mean_token_accuracy": 0.6731339693069458, + "num_tokens": 42229990.0, + "step": 1638 + }, + { + "epoch": 0.17999121458379091, + "ewc_loss": 2.9653310775756836e-06, + "grad_norm": 2.8208816051483154, + "learning_rate": 5.995607613469985e-07, + "loss": 0.9777, + "mean_token_accuracy": 0.7116377353668213, + "num_tokens": 42249858.0, + "step": 1639 + }, + { + "epoch": 0.18010103228640456, + "ewc_loss": 2.995133399963379e-06, + "grad_norm": 2.8669469356536865, + "learning_rate": 5.999267935578331e-07, + "loss": 0.9738, + "mean_token_accuracy": 0.7073010206222534, + "num_tokens": 42271355.0, + "step": 1640 + }, + { + "epoch": 0.18021084998901823, + "ewc_loss": 3.0100345611572266e-06, + "grad_norm": 2.5465614795684814, + "learning_rate": 6.002928257686676e-07, + "loss": 1.0438, + "mean_token_accuracy": 0.6974177956581116, + "num_tokens": 42297624.0, + "step": 1641 + }, + { + "epoch": 0.18032066769163188, + "ewc_loss": 3.0100345611572266e-06, + "grad_norm": 2.5202198028564453, + "learning_rate": 6.006588579795022e-07, + "loss": 0.9837, + "mean_token_accuracy": 0.7053536176681519, + "num_tokens": 42322395.0, + "step": 1642 + }, + { + "epoch": 0.18043048539424555, + "ewc_loss": 3.0100345611572266e-06, + "grad_norm": 2.8394412994384766, + "learning_rate": 6.010248901903368e-07, + "loss": 1.0371, + "mean_token_accuracy": 0.7075306177139282, + "num_tokens": 42344634.0, + "step": 1643 + }, + { + "epoch": 0.1805403030968592, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.648582696914673, + "learning_rate": 6.013909224011713e-07, + "loss": 1.1183, + "mean_token_accuracy": 0.6760274171829224, + "num_tokens": 42370501.0, + "step": 1644 + }, + { + "epoch": 0.18065012079947287, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.424619436264038, + "learning_rate": 6.017569546120058e-07, + "loss": 1.0672, + "mean_token_accuracy": 0.6907824277877808, + "num_tokens": 42396966.0, + "step": 1645 + }, + { + "epoch": 0.18075993850208655, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.4601547718048096, + "learning_rate": 6.021229868228404e-07, + "loss": 1.1144, + "mean_token_accuracy": 0.6851463317871094, + "num_tokens": 42423033.0, + "step": 1646 + }, + { + "epoch": 0.1808697562047002, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.1953859329223633, + "learning_rate": 6.024890190336749e-07, + "loss": 1.0632, + "mean_token_accuracy": 0.6929745674133301, + "num_tokens": 42456449.0, + "step": 1647 + }, + { + "epoch": 0.18097957390731387, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.68426775932312, + "learning_rate": 6.028550512445095e-07, + "loss": 1.0282, + "mean_token_accuracy": 0.6999751925468445, + "num_tokens": 42480035.0, + "step": 1648 + }, + { + "epoch": 0.1810893916099275, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.5962181091308594, + "learning_rate": 6.032210834553441e-07, + "loss": 1.0374, + "mean_token_accuracy": 0.6991217732429504, + "num_tokens": 42503308.0, + "step": 1649 + }, + { + "epoch": 0.1811992093125412, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 7.938530445098877, + "learning_rate": 6.035871156661786e-07, + "loss": 1.1036, + "mean_token_accuracy": 0.6772769093513489, + "num_tokens": 42532075.0, + "step": 1650 + }, + { + "epoch": 0.18130902701515483, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.596435308456421, + "learning_rate": 6.039531478770131e-07, + "loss": 1.0151, + "mean_token_accuracy": 0.7017651796340942, + "num_tokens": 42556184.0, + "step": 1651 + }, + { + "epoch": 0.1814188447177685, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.7196106910705566, + "learning_rate": 6.043191800878477e-07, + "loss": 1.1144, + "mean_token_accuracy": 0.6793912649154663, + "num_tokens": 42578435.0, + "step": 1652 + }, + { + "epoch": 0.18152866242038215, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.553696870803833, + "learning_rate": 6.046852122986823e-07, + "loss": 1.0699, + "mean_token_accuracy": 0.6865024566650391, + "num_tokens": 42604295.0, + "step": 1653 + }, + { + "epoch": 0.18163848012299583, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.6016414165496826, + "learning_rate": 6.050512445095168e-07, + "loss": 1.1418, + "mean_token_accuracy": 0.6799820065498352, + "num_tokens": 42632476.0, + "step": 1654 + }, + { + "epoch": 0.1817482978256095, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.7187576293945312, + "learning_rate": 6.054172767203514e-07, + "loss": 1.1199, + "mean_token_accuracy": 0.673485279083252, + "num_tokens": 42657921.0, + "step": 1655 + }, + { + "epoch": 0.18185811552822315, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.6769745349884033, + "learning_rate": 6.057833089311859e-07, + "loss": 1.1621, + "mean_token_accuracy": 0.6634977459907532, + "num_tokens": 42683798.0, + "step": 1656 + }, + { + "epoch": 0.18196793323083682, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.5480644702911377, + "learning_rate": 6.061493411420205e-07, + "loss": 1.0148, + "mean_token_accuracy": 0.7011920213699341, + "num_tokens": 42707333.0, + "step": 1657 + }, + { + "epoch": 0.18207775093345047, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.376542806625366, + "learning_rate": 6.065153733528551e-07, + "loss": 1.0676, + "mean_token_accuracy": 0.694180965423584, + "num_tokens": 42733776.0, + "step": 1658 + }, + { + "epoch": 0.18218756863606414, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.1564128398895264, + "learning_rate": 6.068814055636896e-07, + "loss": 1.1072, + "mean_token_accuracy": 0.6768898963928223, + "num_tokens": 42769503.0, + "step": 1659 + }, + { + "epoch": 0.18229738633867779, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.8486437797546387, + "learning_rate": 6.072474377745242e-07, + "loss": 1.0259, + "mean_token_accuracy": 0.7032131552696228, + "num_tokens": 42791718.0, + "step": 1660 + }, + { + "epoch": 0.18240720404129146, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.6975181102752686, + "learning_rate": 6.076134699853587e-07, + "loss": 1.052, + "mean_token_accuracy": 0.6911462545394897, + "num_tokens": 42816722.0, + "step": 1661 + }, + { + "epoch": 0.1825170217439051, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.453562021255493, + "learning_rate": 6.079795021961932e-07, + "loss": 1.003, + "mean_token_accuracy": 0.6990269422531128, + "num_tokens": 42844961.0, + "step": 1662 + }, + { + "epoch": 0.18262683944651878, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.450234889984131, + "learning_rate": 6.083455344070278e-07, + "loss": 0.9601, + "mean_token_accuracy": 0.7091425061225891, + "num_tokens": 42870511.0, + "step": 1663 + }, + { + "epoch": 0.18273665714913245, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.3657243251800537, + "learning_rate": 6.087115666178624e-07, + "loss": 0.9866, + "mean_token_accuracy": 0.708132803440094, + "num_tokens": 42897649.0, + "step": 1664 + }, + { + "epoch": 0.1828464748517461, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.3280630111694336, + "learning_rate": 6.090775988286969e-07, + "loss": 0.9969, + "mean_token_accuracy": 0.703006386756897, + "num_tokens": 42926306.0, + "step": 1665 + }, + { + "epoch": 0.18295629255435977, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 3.2573139667510986, + "learning_rate": 6.094436310395315e-07, + "loss": 1.0371, + "mean_token_accuracy": 0.6917297840118408, + "num_tokens": 42951568.0, + "step": 1666 + }, + { + "epoch": 0.18306611025697342, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.6486549377441406, + "learning_rate": 6.09809663250366e-07, + "loss": 1.1175, + "mean_token_accuracy": 0.6855819225311279, + "num_tokens": 42975783.0, + "step": 1667 + }, + { + "epoch": 0.1831759279595871, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.4400289058685303, + "learning_rate": 6.101756954612005e-07, + "loss": 1.1404, + "mean_token_accuracy": 0.6691521406173706, + "num_tokens": 43004209.0, + "step": 1668 + }, + { + "epoch": 0.18328574566220074, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.5989389419555664, + "learning_rate": 6.105417276720352e-07, + "loss": 1.0529, + "mean_token_accuracy": 0.6856951713562012, + "num_tokens": 43028852.0, + "step": 1669 + }, + { + "epoch": 0.1833955633648144, + "ewc_loss": 3.0249357223510742e-06, + "grad_norm": 2.521514654159546, + "learning_rate": 6.109077598828697e-07, + "loss": 1.0254, + "mean_token_accuracy": 0.6915217638015747, + "num_tokens": 43051679.0, + "step": 1670 + }, + { + "epoch": 0.18350538106742806, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.6026628017425537, + "learning_rate": 6.112737920937042e-07, + "loss": 1.1146, + "mean_token_accuracy": 0.6760408878326416, + "num_tokens": 43079437.0, + "step": 1671 + }, + { + "epoch": 0.18361519877004173, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.2858309745788574, + "learning_rate": 6.116398243045388e-07, + "loss": 1.0853, + "mean_token_accuracy": 0.6800129413604736, + "num_tokens": 43111882.0, + "step": 1672 + }, + { + "epoch": 0.1837250164726554, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.5031936168670654, + "learning_rate": 6.120058565153733e-07, + "loss": 1.0899, + "mean_token_accuracy": 0.6899569630622864, + "num_tokens": 43138332.0, + "step": 1673 + }, + { + "epoch": 0.18383483417526905, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.539400577545166, + "learning_rate": 6.12371888726208e-07, + "loss": 1.0862, + "mean_token_accuracy": 0.6855359673500061, + "num_tokens": 43163166.0, + "step": 1674 + }, + { + "epoch": 0.18394465187788273, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.26664137840271, + "learning_rate": 6.127379209370425e-07, + "loss": 1.0395, + "mean_token_accuracy": 0.6963599920272827, + "num_tokens": 43192052.0, + "step": 1675 + }, + { + "epoch": 0.18405446958049637, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.3834571838378906, + "learning_rate": 6.13103953147877e-07, + "loss": 0.9978, + "mean_token_accuracy": 0.7001214027404785, + "num_tokens": 43218645.0, + "step": 1676 + }, + { + "epoch": 0.18416428728311005, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.6608564853668213, + "learning_rate": 6.134699853587116e-07, + "loss": 1.096, + "mean_token_accuracy": 0.6733343005180359, + "num_tokens": 43241509.0, + "step": 1677 + }, + { + "epoch": 0.1842741049857237, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.7975125312805176, + "learning_rate": 6.138360175695461e-07, + "loss": 1.0357, + "mean_token_accuracy": 0.6949355602264404, + "num_tokens": 43264698.0, + "step": 1678 + }, + { + "epoch": 0.18438392268833736, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.9559996128082275, + "learning_rate": 6.142020497803806e-07, + "loss": 1.0087, + "mean_token_accuracy": 0.7074198722839355, + "num_tokens": 43287337.0, + "step": 1679 + }, + { + "epoch": 0.184493740390951, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.620089530944824, + "learning_rate": 6.145680819912153e-07, + "loss": 1.0945, + "mean_token_accuracy": 0.6820334196090698, + "num_tokens": 43313430.0, + "step": 1680 + }, + { + "epoch": 0.18460355809356468, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.6454761028289795, + "learning_rate": 6.149341142020498e-07, + "loss": 1.0691, + "mean_token_accuracy": 0.6871265172958374, + "num_tokens": 43337787.0, + "step": 1681 + }, + { + "epoch": 0.18471337579617833, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.230534791946411, + "learning_rate": 6.153001464128843e-07, + "loss": 1.1198, + "mean_token_accuracy": 0.6729018688201904, + "num_tokens": 43373378.0, + "step": 1682 + }, + { + "epoch": 0.184823193498792, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.2628512382507324, + "learning_rate": 6.156661786237189e-07, + "loss": 0.9782, + "mean_token_accuracy": 0.708638608455658, + "num_tokens": 43404353.0, + "step": 1683 + }, + { + "epoch": 0.18493301120140568, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.3930535316467285, + "learning_rate": 6.160322108345534e-07, + "loss": 1.1102, + "mean_token_accuracy": 0.6770429611206055, + "num_tokens": 43433934.0, + "step": 1684 + }, + { + "epoch": 0.18504282890401932, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.2380919456481934, + "learning_rate": 6.16398243045388e-07, + "loss": 1.0062, + "mean_token_accuracy": 0.6974568963050842, + "num_tokens": 43466980.0, + "step": 1685 + }, + { + "epoch": 0.185152646606633, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.3060462474823, + "learning_rate": 6.167642752562226e-07, + "loss": 1.0371, + "mean_token_accuracy": 0.7018808126449585, + "num_tokens": 43494439.0, + "step": 1686 + }, + { + "epoch": 0.18526246430924664, + "ewc_loss": 3.039836883544922e-06, + "grad_norm": 2.408395290374756, + "learning_rate": 6.171303074670571e-07, + "loss": 1.1531, + "mean_token_accuracy": 0.6701831817626953, + "num_tokens": 43521827.0, + "step": 1687 + }, + { + "epoch": 0.18537228201186032, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.613035202026367, + "learning_rate": 6.174963396778916e-07, + "loss": 1.0994, + "mean_token_accuracy": 0.672541618347168, + "num_tokens": 43546844.0, + "step": 1688 + }, + { + "epoch": 0.18548209971447396, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.4825236797332764, + "learning_rate": 6.178623718887262e-07, + "loss": 1.1291, + "mean_token_accuracy": 0.6664943099021912, + "num_tokens": 43579098.0, + "step": 1689 + }, + { + "epoch": 0.18559191741708764, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 2.9280784130096436, + "learning_rate": 6.182284040995608e-07, + "loss": 0.9259, + "mean_token_accuracy": 0.7240391969680786, + "num_tokens": 43599155.0, + "step": 1690 + }, + { + "epoch": 0.18570173511970128, + "ewc_loss": 3.0547380447387695e-06, + "grad_norm": 3.3393020629882812, + "learning_rate": 6.185944363103954e-07, + "loss": 1.031, + "mean_token_accuracy": 0.6965122222900391, + "num_tokens": 43622669.0, + "step": 1691 + }, + { + "epoch": 0.18581155282231496, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.6300241947174072, + "learning_rate": 6.189604685212299e-07, + "loss": 1.0205, + "mean_token_accuracy": 0.6974279284477234, + "num_tokens": 43646422.0, + "step": 1692 + }, + { + "epoch": 0.18592137052492863, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.981447696685791, + "learning_rate": 6.193265007320644e-07, + "loss": 1.0219, + "mean_token_accuracy": 0.6990832686424255, + "num_tokens": 43676400.0, + "step": 1693 + }, + { + "epoch": 0.18603118822754228, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.7979581356048584, + "learning_rate": 6.19692532942899e-07, + "loss": 1.0292, + "mean_token_accuracy": 0.697556734085083, + "num_tokens": 43697193.0, + "step": 1694 + }, + { + "epoch": 0.18614100593015595, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.7385690212249756, + "learning_rate": 6.200585651537335e-07, + "loss": 1.0685, + "mean_token_accuracy": 0.6873914003372192, + "num_tokens": 43721079.0, + "step": 1695 + }, + { + "epoch": 0.1862508236327696, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.5841786861419678, + "learning_rate": 6.204245973645681e-07, + "loss": 1.0612, + "mean_token_accuracy": 0.6934053301811218, + "num_tokens": 43747631.0, + "step": 1696 + }, + { + "epoch": 0.18636064133538327, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.3629372119903564, + "learning_rate": 6.207906295754027e-07, + "loss": 1.0312, + "mean_token_accuracy": 0.6941488981246948, + "num_tokens": 43778341.0, + "step": 1697 + }, + { + "epoch": 0.18647045903799692, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.619260787963867, + "learning_rate": 6.211566617862372e-07, + "loss": 1.0283, + "mean_token_accuracy": 0.6964471936225891, + "num_tokens": 43802076.0, + "step": 1698 + }, + { + "epoch": 0.1865802767406106, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.4437406063079834, + "learning_rate": 6.215226939970716e-07, + "loss": 1.104, + "mean_token_accuracy": 0.6762474775314331, + "num_tokens": 43829357.0, + "step": 1699 + }, + { + "epoch": 0.18669009444322424, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.61080002784729, + "learning_rate": 6.218887262079063e-07, + "loss": 0.978, + "mean_token_accuracy": 0.7025152444839478, + "num_tokens": 43852687.0, + "step": 1700 + }, + { + "epoch": 0.1867999121458379, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.618583917617798, + "learning_rate": 6.222547584187409e-07, + "loss": 1.0751, + "mean_token_accuracy": 0.6895791292190552, + "num_tokens": 43877297.0, + "step": 1701 + }, + { + "epoch": 0.18690972984845158, + "ewc_loss": 3.084540367126465e-06, + "grad_norm": 2.7678637504577637, + "learning_rate": 6.226207906295753e-07, + "loss": 1.0175, + "mean_token_accuracy": 0.6944229602813721, + "num_tokens": 43899048.0, + "step": 1702 + }, + { + "epoch": 0.18701954755106523, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.754392147064209, + "learning_rate": 6.2298682284041e-07, + "loss": 0.949, + "mean_token_accuracy": 0.7161177396774292, + "num_tokens": 43921618.0, + "step": 1703 + }, + { + "epoch": 0.1871293652536789, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.263596534729004, + "learning_rate": 6.233528550512444e-07, + "loss": 1.1017, + "mean_token_accuracy": 0.6883978247642517, + "num_tokens": 43952213.0, + "step": 1704 + }, + { + "epoch": 0.18723918295629255, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.301727294921875, + "learning_rate": 6.237188872620789e-07, + "loss": 1.0294, + "mean_token_accuracy": 0.7010643482208252, + "num_tokens": 43979710.0, + "step": 1705 + }, + { + "epoch": 0.18734900065890622, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.446881055831909, + "learning_rate": 6.240849194729137e-07, + "loss": 1.0313, + "mean_token_accuracy": 0.6948770880699158, + "num_tokens": 44005736.0, + "step": 1706 + }, + { + "epoch": 0.18745881836151987, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.3315937519073486, + "learning_rate": 6.244509516837481e-07, + "loss": 1.1304, + "mean_token_accuracy": 0.6775261759757996, + "num_tokens": 44036907.0, + "step": 1707 + }, + { + "epoch": 0.18756863606413354, + "ewc_loss": 3.069639205932617e-06, + "grad_norm": 2.263190269470215, + "learning_rate": 6.248169838945828e-07, + "loss": 1.0871, + "mean_token_accuracy": 0.6790582537651062, + "num_tokens": 44067574.0, + "step": 1708 + }, + { + "epoch": 0.1876784537667472, + "ewc_loss": 3.084540367126465e-06, + "grad_norm": 2.655547618865967, + "learning_rate": 6.251830161054172e-07, + "loss": 1.1209, + "mean_token_accuracy": 0.6733550429344177, + "num_tokens": 44092129.0, + "step": 1709 + }, + { + "epoch": 0.18778827146936086, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.3916919231414795, + "learning_rate": 6.255490483162517e-07, + "loss": 1.1418, + "mean_token_accuracy": 0.672545313835144, + "num_tokens": 44124520.0, + "step": 1710 + }, + { + "epoch": 0.18789808917197454, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.5138747692108154, + "learning_rate": 6.259150805270863e-07, + "loss": 1.092, + "mean_token_accuracy": 0.6883278489112854, + "num_tokens": 44150189.0, + "step": 1711 + }, + { + "epoch": 0.18800790687458818, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.730719566345215, + "learning_rate": 6.262811127379209e-07, + "loss": 1.0057, + "mean_token_accuracy": 0.7031267881393433, + "num_tokens": 44171622.0, + "step": 1712 + }, + { + "epoch": 0.18811772457720186, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.5458619594573975, + "learning_rate": 6.266471449487554e-07, + "loss": 1.0958, + "mean_token_accuracy": 0.6892498731613159, + "num_tokens": 44196722.0, + "step": 1713 + }, + { + "epoch": 0.1882275422798155, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.2886266708374023, + "learning_rate": 6.2701317715959e-07, + "loss": 1.1332, + "mean_token_accuracy": 0.6679106950759888, + "num_tokens": 44228788.0, + "step": 1714 + }, + { + "epoch": 0.18833735998242918, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.706369400024414, + "learning_rate": 6.273792093704245e-07, + "loss": 1.0789, + "mean_token_accuracy": 0.6834816336631775, + "num_tokens": 44253946.0, + "step": 1715 + }, + { + "epoch": 0.18844717768504282, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.6147043704986572, + "learning_rate": 6.27745241581259e-07, + "loss": 1.0841, + "mean_token_accuracy": 0.6821408271789551, + "num_tokens": 44277666.0, + "step": 1716 + }, + { + "epoch": 0.1885569953876565, + "ewc_loss": 3.0994415283203125e-06, + "grad_norm": 2.342200517654419, + "learning_rate": 6.281112737920937e-07, + "loss": 1.0154, + "mean_token_accuracy": 0.6997746229171753, + "num_tokens": 44306333.0, + "step": 1717 + }, + { + "epoch": 0.18866681309027014, + "ewc_loss": 3.11434268951416e-06, + "grad_norm": 2.5479624271392822, + "learning_rate": 6.284773060029282e-07, + "loss": 1.1013, + "mean_token_accuracy": 0.6757168769836426, + "num_tokens": 44333933.0, + "step": 1718 + }, + { + "epoch": 0.18877663079288381, + "ewc_loss": 3.11434268951416e-06, + "grad_norm": 2.6224498748779297, + "learning_rate": 6.288433382137627e-07, + "loss": 1.0137, + "mean_token_accuracy": 0.6959103345870972, + "num_tokens": 44358642.0, + "step": 1719 + }, + { + "epoch": 0.18888644849549746, + "ewc_loss": 3.11434268951416e-06, + "grad_norm": 2.7381913661956787, + "learning_rate": 6.292093704245973e-07, + "loss": 1.0462, + "mean_token_accuracy": 0.6875061988830566, + "num_tokens": 44381083.0, + "step": 1720 + }, + { + "epoch": 0.18899626619811113, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.645080327987671, + "learning_rate": 6.295754026354318e-07, + "loss": 1.0984, + "mean_token_accuracy": 0.6840682029724121, + "num_tokens": 44406847.0, + "step": 1721 + }, + { + "epoch": 0.1891060839007248, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.721686601638794, + "learning_rate": 6.299414348462665e-07, + "loss": 1.0846, + "mean_token_accuracy": 0.6766037940979004, + "num_tokens": 44433809.0, + "step": 1722 + }, + { + "epoch": 0.18921590160333845, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.6806411743164062, + "learning_rate": 6.30307467057101e-07, + "loss": 1.0755, + "mean_token_accuracy": 0.6828282475471497, + "num_tokens": 44458042.0, + "step": 1723 + }, + { + "epoch": 0.18932571930595213, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 3.050680637359619, + "learning_rate": 6.306734992679355e-07, + "loss": 1.0285, + "mean_token_accuracy": 0.6957530975341797, + "num_tokens": 44478035.0, + "step": 1724 + }, + { + "epoch": 0.18943553700856577, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.2014434337615967, + "learning_rate": 6.310395314787701e-07, + "loss": 1.1311, + "mean_token_accuracy": 0.6661993265151978, + "num_tokens": 44513865.0, + "step": 1725 + }, + { + "epoch": 0.18954535471117945, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.898653030395508, + "learning_rate": 6.314055636896046e-07, + "loss": 1.0512, + "mean_token_accuracy": 0.6976968050003052, + "num_tokens": 44534733.0, + "step": 1726 + }, + { + "epoch": 0.1896551724137931, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.433715343475342, + "learning_rate": 6.317715959004391e-07, + "loss": 1.0021, + "mean_token_accuracy": 0.7021336555480957, + "num_tokens": 44560374.0, + "step": 1727 + }, + { + "epoch": 0.18976499011640677, + "ewc_loss": 3.129243850708008e-06, + "grad_norm": 2.8378379344940186, + "learning_rate": 6.321376281112738e-07, + "loss": 1.0411, + "mean_token_accuracy": 0.7076354026794434, + "num_tokens": 44579550.0, + "step": 1728 + }, + { + "epoch": 0.1898748078190204, + "ewc_loss": 3.159046173095703e-06, + "grad_norm": 2.4631576538085938, + "learning_rate": 6.325036603221083e-07, + "loss": 1.0792, + "mean_token_accuracy": 0.6786831617355347, + "num_tokens": 44605935.0, + "step": 1729 + }, + { + "epoch": 0.1899846255216341, + "ewc_loss": 3.159046173095703e-06, + "grad_norm": 2.7155048847198486, + "learning_rate": 6.328696925329428e-07, + "loss": 1.0451, + "mean_token_accuracy": 0.6937711238861084, + "num_tokens": 44628919.0, + "step": 1730 + }, + { + "epoch": 0.19009444322424776, + "ewc_loss": 3.159046173095703e-06, + "grad_norm": 2.6485824584960938, + "learning_rate": 6.332357247437774e-07, + "loss": 1.0744, + "mean_token_accuracy": 0.6803997755050659, + "num_tokens": 44656105.0, + "step": 1731 + }, + { + "epoch": 0.1902042609268614, + "ewc_loss": 3.159046173095703e-06, + "grad_norm": 2.3587894439697266, + "learning_rate": 6.336017569546119e-07, + "loss": 1.0823, + "mean_token_accuracy": 0.6779462695121765, + "num_tokens": 44686900.0, + "step": 1732 + }, + { + "epoch": 0.19031407862947508, + "ewc_loss": 3.159046173095703e-06, + "grad_norm": 2.8866970539093018, + "learning_rate": 6.339677891654465e-07, + "loss": 1.0266, + "mean_token_accuracy": 0.6963053941726685, + "num_tokens": 44710054.0, + "step": 1733 + }, + { + "epoch": 0.19042389633208873, + "ewc_loss": 3.1739473342895508e-06, + "grad_norm": 3.0346405506134033, + "learning_rate": 6.343338213762811e-07, + "loss": 1.0104, + "mean_token_accuracy": 0.7033700942993164, + "num_tokens": 44730033.0, + "step": 1734 + }, + { + "epoch": 0.1905337140347024, + "ewc_loss": 3.1739473342895508e-06, + "grad_norm": 2.6902148723602295, + "learning_rate": 6.346998535871156e-07, + "loss": 0.976, + "mean_token_accuracy": 0.7072324156761169, + "num_tokens": 44756831.0, + "step": 1735 + }, + { + "epoch": 0.19064353173731605, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.486239433288574, + "learning_rate": 6.350658857979501e-07, + "loss": 1.0554, + "mean_token_accuracy": 0.6860166788101196, + "num_tokens": 44782982.0, + "step": 1736 + }, + { + "epoch": 0.19075334943992972, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.6332883834838867, + "learning_rate": 6.354319180087847e-07, + "loss": 1.1782, + "mean_token_accuracy": 0.6644657254219055, + "num_tokens": 44811212.0, + "step": 1737 + }, + { + "epoch": 0.19086316714254337, + "ewc_loss": 3.159046173095703e-06, + "grad_norm": 2.5701980590820312, + "learning_rate": 6.357979502196193e-07, + "loss": 1.0303, + "mean_token_accuracy": 0.6979964971542358, + "num_tokens": 44833233.0, + "step": 1738 + }, + { + "epoch": 0.19097298484515704, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.3399274349212646, + "learning_rate": 6.361639824304539e-07, + "loss": 1.0785, + "mean_token_accuracy": 0.6864607930183411, + "num_tokens": 44861753.0, + "step": 1739 + }, + { + "epoch": 0.1910828025477707, + "ewc_loss": 3.1739473342895508e-06, + "grad_norm": 2.4108757972717285, + "learning_rate": 6.365300146412884e-07, + "loss": 1.06, + "mean_token_accuracy": 0.6971811056137085, + "num_tokens": 44887824.0, + "step": 1740 + }, + { + "epoch": 0.19119262025038436, + "ewc_loss": 3.1739473342895508e-06, + "grad_norm": 2.494577169418335, + "learning_rate": 6.368960468521229e-07, + "loss": 1.0465, + "mean_token_accuracy": 0.684429407119751, + "num_tokens": 44914371.0, + "step": 1741 + }, + { + "epoch": 0.19130243795299803, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.4473717212677, + "learning_rate": 6.372620790629575e-07, + "loss": 1.1135, + "mean_token_accuracy": 0.673790693283081, + "num_tokens": 44943379.0, + "step": 1742 + }, + { + "epoch": 0.19141225565561168, + "ewc_loss": 3.1739473342895508e-06, + "grad_norm": 2.5790064334869385, + "learning_rate": 6.37628111273792e-07, + "loss": 1.1306, + "mean_token_accuracy": 0.6792083382606506, + "num_tokens": 44967204.0, + "step": 1743 + }, + { + "epoch": 0.19152207335822535, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.4241983890533447, + "learning_rate": 6.379941434846266e-07, + "loss": 1.0147, + "mean_token_accuracy": 0.6965177655220032, + "num_tokens": 44993218.0, + "step": 1744 + }, + { + "epoch": 0.191631891060839, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.6179652214050293, + "learning_rate": 6.383601756954612e-07, + "loss": 1.0374, + "mean_token_accuracy": 0.6925176382064819, + "num_tokens": 45016771.0, + "step": 1745 + }, + { + "epoch": 0.19174170876345267, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.683485507965088, + "learning_rate": 6.387262079062957e-07, + "loss": 1.0782, + "mean_token_accuracy": 0.6843976378440857, + "num_tokens": 45041696.0, + "step": 1746 + }, + { + "epoch": 0.19185152646606632, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.4391846656799316, + "learning_rate": 6.390922401171302e-07, + "loss": 1.0905, + "mean_token_accuracy": 0.6832122802734375, + "num_tokens": 45067615.0, + "step": 1747 + }, + { + "epoch": 0.19196134416868, + "ewc_loss": 3.1888484954833984e-06, + "grad_norm": 2.608492136001587, + "learning_rate": 6.394582723279648e-07, + "loss": 1.0901, + "mean_token_accuracy": 0.6911807060241699, + "num_tokens": 45092065.0, + "step": 1748 + }, + { + "epoch": 0.19207116187129367, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.59183669090271, + "learning_rate": 6.398243045387994e-07, + "loss": 1.1073, + "mean_token_accuracy": 0.6800568103790283, + "num_tokens": 45116555.0, + "step": 1749 + }, + { + "epoch": 0.1921809795739073, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.4729506969451904, + "learning_rate": 6.401903367496339e-07, + "loss": 1.1169, + "mean_token_accuracy": 0.6787590980529785, + "num_tokens": 45142968.0, + "step": 1750 + }, + { + "epoch": 0.19229079727652099, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.887732982635498, + "learning_rate": 6.405563689604685e-07, + "loss": 1.0023, + "mean_token_accuracy": 0.699073314666748, + "num_tokens": 45162421.0, + "step": 1751 + }, + { + "epoch": 0.19240061497913463, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.4823710918426514, + "learning_rate": 6.40922401171303e-07, + "loss": 0.972, + "mean_token_accuracy": 0.7100237607955933, + "num_tokens": 45188853.0, + "step": 1752 + }, + { + "epoch": 0.1925104326817483, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.548999786376953, + "learning_rate": 6.412884333821375e-07, + "loss": 1.1121, + "mean_token_accuracy": 0.6950236558914185, + "num_tokens": 45211908.0, + "step": 1753 + }, + { + "epoch": 0.19262025038436195, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.4621989727020264, + "learning_rate": 6.416544655929722e-07, + "loss": 0.974, + "mean_token_accuracy": 0.7105863094329834, + "num_tokens": 45235897.0, + "step": 1754 + }, + { + "epoch": 0.19273006808697563, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.4796457290649414, + "learning_rate": 6.420204978038067e-07, + "loss": 1.2113, + "mean_token_accuracy": 0.6526737213134766, + "num_tokens": 45266668.0, + "step": 1755 + }, + { + "epoch": 0.19283988578958927, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.605480670928955, + "learning_rate": 6.423865300146413e-07, + "loss": 1.1081, + "mean_token_accuracy": 0.6732752919197083, + "num_tokens": 45291961.0, + "step": 1756 + }, + { + "epoch": 0.19294970349220295, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.406526565551758, + "learning_rate": 6.427525622254758e-07, + "loss": 1.0717, + "mean_token_accuracy": 0.6872354745864868, + "num_tokens": 45321224.0, + "step": 1757 + }, + { + "epoch": 0.1930595211948166, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.446244955062866, + "learning_rate": 6.431185944363103e-07, + "loss": 1.0631, + "mean_token_accuracy": 0.6913452744483948, + "num_tokens": 45349120.0, + "step": 1758 + }, + { + "epoch": 0.19316933889743026, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.699479341506958, + "learning_rate": 6.434846266471449e-07, + "loss": 0.9813, + "mean_token_accuracy": 0.7086092829704285, + "num_tokens": 45370609.0, + "step": 1759 + }, + { + "epoch": 0.19327915660004394, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.2585580348968506, + "learning_rate": 6.438506588579795e-07, + "loss": 1.0424, + "mean_token_accuracy": 0.6949788331985474, + "num_tokens": 45402004.0, + "step": 1760 + }, + { + "epoch": 0.19338897430265758, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.8301823139190674, + "learning_rate": 6.44216691068814e-07, + "loss": 0.9927, + "mean_token_accuracy": 0.7047998905181885, + "num_tokens": 45423421.0, + "step": 1761 + }, + { + "epoch": 0.19349879200527126, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.4173989295959473, + "learning_rate": 6.445827232796486e-07, + "loss": 0.9976, + "mean_token_accuracy": 0.7079315185546875, + "num_tokens": 45449405.0, + "step": 1762 + }, + { + "epoch": 0.1936086097078849, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.7623279094696045, + "learning_rate": 6.449487554904831e-07, + "loss": 1.0079, + "mean_token_accuracy": 0.6926711797714233, + "num_tokens": 45470264.0, + "step": 1763 + }, + { + "epoch": 0.19371842741049858, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.239086151123047, + "learning_rate": 6.453147877013176e-07, + "loss": 1.0052, + "mean_token_accuracy": 0.709327220916748, + "num_tokens": 45501935.0, + "step": 1764 + }, + { + "epoch": 0.19382824511311222, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.4078922271728516, + "learning_rate": 6.456808199121523e-07, + "loss": 1.1011, + "mean_token_accuracy": 0.680382490158081, + "num_tokens": 45528933.0, + "step": 1765 + }, + { + "epoch": 0.1939380628157259, + "ewc_loss": 3.203749656677246e-06, + "grad_norm": 2.5884792804718018, + "learning_rate": 6.460468521229868e-07, + "loss": 0.9623, + "mean_token_accuracy": 0.7125195264816284, + "num_tokens": 45551417.0, + "step": 1766 + }, + { + "epoch": 0.19404788051833954, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.287783622741699, + "learning_rate": 6.464128843338213e-07, + "loss": 1.0557, + "mean_token_accuracy": 0.6906673908233643, + "num_tokens": 45582175.0, + "step": 1767 + }, + { + "epoch": 0.19415769822095322, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.562352180480957, + "learning_rate": 6.467789165446559e-07, + "loss": 1.1232, + "mean_token_accuracy": 0.6699574589729309, + "num_tokens": 45606661.0, + "step": 1768 + }, + { + "epoch": 0.1942675159235669, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.6540982723236084, + "learning_rate": 6.471449487554904e-07, + "loss": 1.084, + "mean_token_accuracy": 0.6800537109375, + "num_tokens": 45632245.0, + "step": 1769 + }, + { + "epoch": 0.19437733362618054, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.1681106090545654, + "learning_rate": 6.475109809663251e-07, + "loss": 1.1, + "mean_token_accuracy": 0.6823816895484924, + "num_tokens": 45666175.0, + "step": 1770 + }, + { + "epoch": 0.1944871513287942, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.519456386566162, + "learning_rate": 6.478770131771596e-07, + "loss": 1.0629, + "mean_token_accuracy": 0.697192907333374, + "num_tokens": 45692819.0, + "step": 1771 + }, + { + "epoch": 0.19459696903140786, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.391456127166748, + "learning_rate": 6.482430453879941e-07, + "loss": 1.0612, + "mean_token_accuracy": 0.6928821802139282, + "num_tokens": 45721537.0, + "step": 1772 + }, + { + "epoch": 0.19470678673402153, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.637434244155884, + "learning_rate": 6.486090775988287e-07, + "loss": 1.102, + "mean_token_accuracy": 0.686416506767273, + "num_tokens": 45745690.0, + "step": 1773 + }, + { + "epoch": 0.19481660443663518, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.5909006595611572, + "learning_rate": 6.489751098096632e-07, + "loss": 1.1285, + "mean_token_accuracy": 0.6727509498596191, + "num_tokens": 45771981.0, + "step": 1774 + }, + { + "epoch": 0.19492642213924885, + "ewc_loss": 3.2186508178710938e-06, + "grad_norm": 2.690112352371216, + "learning_rate": 6.493411420204977e-07, + "loss": 1.0595, + "mean_token_accuracy": 0.6933057904243469, + "num_tokens": 45796644.0, + "step": 1775 + }, + { + "epoch": 0.1950362398418625, + "ewc_loss": 3.2335519790649414e-06, + "grad_norm": 2.9770238399505615, + "learning_rate": 6.497071742313324e-07, + "loss": 1.0994, + "mean_token_accuracy": 0.676020085811615, + "num_tokens": 45815886.0, + "step": 1776 + }, + { + "epoch": 0.19514605754447617, + "ewc_loss": 3.2335519790649414e-06, + "grad_norm": 2.513352870941162, + "learning_rate": 6.500732064421669e-07, + "loss": 1.0712, + "mean_token_accuracy": 0.6861430406570435, + "num_tokens": 45843574.0, + "step": 1777 + }, + { + "epoch": 0.19525587524708984, + "ewc_loss": 3.2335519790649414e-06, + "grad_norm": 2.436798095703125, + "learning_rate": 6.504392386530014e-07, + "loss": 1.0384, + "mean_token_accuracy": 0.6951904296875, + "num_tokens": 45870615.0, + "step": 1778 + }, + { + "epoch": 0.1953656929497035, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.721998929977417, + "learning_rate": 6.50805270863836e-07, + "loss": 1.0471, + "mean_token_accuracy": 0.6941872835159302, + "num_tokens": 45892574.0, + "step": 1779 + }, + { + "epoch": 0.19547551065231716, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.588904619216919, + "learning_rate": 6.511713030746705e-07, + "loss": 1.1154, + "mean_token_accuracy": 0.6749621629714966, + "num_tokens": 45917428.0, + "step": 1780 + }, + { + "epoch": 0.1955853283549308, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.49690580368042, + "learning_rate": 6.515373352855051e-07, + "loss": 1.0417, + "mean_token_accuracy": 0.6895562410354614, + "num_tokens": 45947191.0, + "step": 1781 + }, + { + "epoch": 0.19569514605754448, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.433971405029297, + "learning_rate": 6.519033674963397e-07, + "loss": 1.0171, + "mean_token_accuracy": 0.6991990804672241, + "num_tokens": 45973562.0, + "step": 1782 + }, + { + "epoch": 0.19580496376015813, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.541189670562744, + "learning_rate": 6.522693997071742e-07, + "loss": 1.026, + "mean_token_accuracy": 0.7008837461471558, + "num_tokens": 46000779.0, + "step": 1783 + }, + { + "epoch": 0.1959147814627718, + "ewc_loss": 3.2633543014526367e-06, + "grad_norm": 2.536099910736084, + "learning_rate": 6.526354319180087e-07, + "loss": 1.0576, + "mean_token_accuracy": 0.6916029453277588, + "num_tokens": 46025924.0, + "step": 1784 + }, + { + "epoch": 0.19602459916538545, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.4528987407684326, + "learning_rate": 6.530014641288433e-07, + "loss": 1.0683, + "mean_token_accuracy": 0.6854180097579956, + "num_tokens": 46052335.0, + "step": 1785 + }, + { + "epoch": 0.19613441686799912, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.5842978954315186, + "learning_rate": 6.533674963396779e-07, + "loss": 0.9938, + "mean_token_accuracy": 0.7061383128166199, + "num_tokens": 46075495.0, + "step": 1786 + }, + { + "epoch": 0.1962442345706128, + "ewc_loss": 3.248453140258789e-06, + "grad_norm": 2.8860976696014404, + "learning_rate": 6.537335285505125e-07, + "loss": 1.0139, + "mean_token_accuracy": 0.6968564987182617, + "num_tokens": 46094772.0, + "step": 1787 + }, + { + "epoch": 0.19635405227322644, + "ewc_loss": 3.2633543014526367e-06, + "grad_norm": 2.563089370727539, + "learning_rate": 6.54099560761347e-07, + "loss": 1.1215, + "mean_token_accuracy": 0.6726027727127075, + "num_tokens": 46123044.0, + "step": 1788 + }, + { + "epoch": 0.19646386997584012, + "ewc_loss": 3.2782554626464844e-06, + "grad_norm": 2.6122190952301025, + "learning_rate": 6.544655929721815e-07, + "loss": 1.055, + "mean_token_accuracy": 0.6909582614898682, + "num_tokens": 46146887.0, + "step": 1789 + }, + { + "epoch": 0.19657368767845376, + "ewc_loss": 3.2782554626464844e-06, + "grad_norm": 2.7399444580078125, + "learning_rate": 6.548316251830161e-07, + "loss": 1.0478, + "mean_token_accuracy": 0.6985856294631958, + "num_tokens": 46168260.0, + "step": 1790 + }, + { + "epoch": 0.19668350538106744, + "ewc_loss": 3.2782554626464844e-06, + "grad_norm": 2.7619283199310303, + "learning_rate": 6.551976573938506e-07, + "loss": 1.1076, + "mean_token_accuracy": 0.6826929450035095, + "num_tokens": 46190266.0, + "step": 1791 + }, + { + "epoch": 0.19679332308368108, + "ewc_loss": 3.2782554626464844e-06, + "grad_norm": 2.5380990505218506, + "learning_rate": 6.555636896046852e-07, + "loss": 1.0742, + "mean_token_accuracy": 0.6851435303688049, + "num_tokens": 46214139.0, + "step": 1792 + }, + { + "epoch": 0.19690314078629476, + "ewc_loss": 3.2782554626464844e-06, + "grad_norm": 2.9082775115966797, + "learning_rate": 6.559297218155198e-07, + "loss": 1.0775, + "mean_token_accuracy": 0.6892987489700317, + "num_tokens": 46235396.0, + "step": 1793 + }, + { + "epoch": 0.1970129584889084, + "ewc_loss": 3.293156623840332e-06, + "grad_norm": 2.510444164276123, + "learning_rate": 6.562957540263543e-07, + "loss": 1.0389, + "mean_token_accuracy": 0.6938443779945374, + "num_tokens": 46261539.0, + "step": 1794 + }, + { + "epoch": 0.19712277619152208, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.570796251296997, + "learning_rate": 6.566617862371888e-07, + "loss": 1.0422, + "mean_token_accuracy": 0.6936718225479126, + "num_tokens": 46286833.0, + "step": 1795 + }, + { + "epoch": 0.19723259389413572, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.403686761856079, + "learning_rate": 6.570278184480234e-07, + "loss": 1.0886, + "mean_token_accuracy": 0.6820732951164246, + "num_tokens": 46313733.0, + "step": 1796 + }, + { + "epoch": 0.1973424115967494, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.621375799179077, + "learning_rate": 6.57393850658858e-07, + "loss": 1.0898, + "mean_token_accuracy": 0.6827249526977539, + "num_tokens": 46337087.0, + "step": 1797 + }, + { + "epoch": 0.19745222929936307, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 3.0217649936676025, + "learning_rate": 6.577598828696925e-07, + "loss": 0.9883, + "mean_token_accuracy": 0.7006375193595886, + "num_tokens": 46356648.0, + "step": 1798 + }, + { + "epoch": 0.19756204700197671, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.572136878967285, + "learning_rate": 6.581259150805271e-07, + "loss": 0.9568, + "mean_token_accuracy": 0.7107794284820557, + "num_tokens": 46378939.0, + "step": 1799 + }, + { + "epoch": 0.1976718647045904, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.4300341606140137, + "learning_rate": 6.584919472913616e-07, + "loss": 1.0605, + "mean_token_accuracy": 0.6848133206367493, + "num_tokens": 46408171.0, + "step": 1800 + }, + { + "epoch": 0.19778168240720403, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.69618558883667, + "learning_rate": 6.588579795021961e-07, + "loss": 1.0167, + "mean_token_accuracy": 0.6999887824058533, + "num_tokens": 46431563.0, + "step": 1801 + }, + { + "epoch": 0.1978915001098177, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.695038080215454, + "learning_rate": 6.592240117130308e-07, + "loss": 1.0467, + "mean_token_accuracy": 0.7006235122680664, + "num_tokens": 46455897.0, + "step": 1802 + }, + { + "epoch": 0.19800131781243135, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.390850782394409, + "learning_rate": 6.595900439238653e-07, + "loss": 1.1066, + "mean_token_accuracy": 0.6738502979278564, + "num_tokens": 46485421.0, + "step": 1803 + }, + { + "epoch": 0.19811113551504503, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.887699842453003, + "learning_rate": 6.599560761346999e-07, + "loss": 1.0213, + "mean_token_accuracy": 0.6959772109985352, + "num_tokens": 46504392.0, + "step": 1804 + }, + { + "epoch": 0.19822095321765867, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.7791061401367188, + "learning_rate": 6.603221083455344e-07, + "loss": 0.9704, + "mean_token_accuracy": 0.7135303020477295, + "num_tokens": 46526987.0, + "step": 1805 + }, + { + "epoch": 0.19833077092027235, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.589773654937744, + "learning_rate": 6.606881405563689e-07, + "loss": 1.0348, + "mean_token_accuracy": 0.6973881721496582, + "num_tokens": 46552167.0, + "step": 1806 + }, + { + "epoch": 0.19844058862288602, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.5683395862579346, + "learning_rate": 6.610541727672035e-07, + "loss": 1.0544, + "mean_token_accuracy": 0.6852658987045288, + "num_tokens": 46578431.0, + "step": 1807 + }, + { + "epoch": 0.19855040632549967, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.649108409881592, + "learning_rate": 6.614202049780381e-07, + "loss": 0.9361, + "mean_token_accuracy": 0.7174975872039795, + "num_tokens": 46597882.0, + "step": 1808 + }, + { + "epoch": 0.19866022402811334, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.311544895172119, + "learning_rate": 6.617862371888726e-07, + "loss": 1.1107, + "mean_token_accuracy": 0.690278172492981, + "num_tokens": 46627173.0, + "step": 1809 + }, + { + "epoch": 0.198770041730727, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.6305248737335205, + "learning_rate": 6.621522693997072e-07, + "loss": 1.067, + "mean_token_accuracy": 0.6925770044326782, + "num_tokens": 46652241.0, + "step": 1810 + }, + { + "epoch": 0.19887985943334066, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.4649951457977295, + "learning_rate": 6.625183016105417e-07, + "loss": 1.1091, + "mean_token_accuracy": 0.6739874482154846, + "num_tokens": 46681172.0, + "step": 1811 + }, + { + "epoch": 0.1989896771359543, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.372915267944336, + "learning_rate": 6.628843338213762e-07, + "loss": 1.0917, + "mean_token_accuracy": 0.6820138096809387, + "num_tokens": 46711892.0, + "step": 1812 + }, + { + "epoch": 0.19909949483856798, + "ewc_loss": 3.3080577850341797e-06, + "grad_norm": 2.234470844268799, + "learning_rate": 6.632503660322109e-07, + "loss": 1.0928, + "mean_token_accuracy": 0.6778671145439148, + "num_tokens": 46744839.0, + "step": 1813 + }, + { + "epoch": 0.19920931254118163, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.7300658226013184, + "learning_rate": 6.636163982430454e-07, + "loss": 0.9573, + "mean_token_accuracy": 0.7029592394828796, + "num_tokens": 46768706.0, + "step": 1814 + }, + { + "epoch": 0.1993191302437953, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.8401670455932617, + "learning_rate": 6.639824304538799e-07, + "loss": 1.0131, + "mean_token_accuracy": 0.699434757232666, + "num_tokens": 46789029.0, + "step": 1815 + }, + { + "epoch": 0.19942894794640897, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.351628303527832, + "learning_rate": 6.643484626647145e-07, + "loss": 1.0928, + "mean_token_accuracy": 0.6810581088066101, + "num_tokens": 46819385.0, + "step": 1816 + }, + { + "epoch": 0.19953876564902262, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.493837356567383, + "learning_rate": 6.64714494875549e-07, + "loss": 1.0746, + "mean_token_accuracy": 0.6845322251319885, + "num_tokens": 46845617.0, + "step": 1817 + }, + { + "epoch": 0.1996485833516363, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.6785247325897217, + "learning_rate": 6.650805270863836e-07, + "loss": 1.001, + "mean_token_accuracy": 0.7014223337173462, + "num_tokens": 46867168.0, + "step": 1818 + }, + { + "epoch": 0.19975840105424994, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.674506664276123, + "learning_rate": 6.654465592972182e-07, + "loss": 1.003, + "mean_token_accuracy": 0.7042679786682129, + "num_tokens": 46893199.0, + "step": 1819 + }, + { + "epoch": 0.1998682187568636, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 3.060561418533325, + "learning_rate": 6.658125915080527e-07, + "loss": 1.0428, + "mean_token_accuracy": 0.6934864521026611, + "num_tokens": 46912484.0, + "step": 1820 + }, + { + "epoch": 0.19997803645947726, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.257655620574951, + "learning_rate": 6.661786237188873e-07, + "loss": 1.1234, + "mean_token_accuracy": 0.6834499835968018, + "num_tokens": 46949071.0, + "step": 1821 + }, + { + "epoch": 0.20008785416209093, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.457793951034546, + "learning_rate": 6.665446559297218e-07, + "loss": 1.0123, + "mean_token_accuracy": 0.6960776448249817, + "num_tokens": 46976660.0, + "step": 1822 + }, + { + "epoch": 0.20019767186470458, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.4516313076019287, + "learning_rate": 6.669106881405563e-07, + "loss": 0.9349, + "mean_token_accuracy": 0.7258780598640442, + "num_tokens": 47003264.0, + "step": 1823 + }, + { + "epoch": 0.20030748956731825, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.197889804840088, + "learning_rate": 6.67276720351391e-07, + "loss": 1.0559, + "mean_token_accuracy": 0.6850039958953857, + "num_tokens": 47036317.0, + "step": 1824 + }, + { + "epoch": 0.2004173072699319, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.6794509887695312, + "learning_rate": 6.676427525622255e-07, + "loss": 1.0842, + "mean_token_accuracy": 0.6827518939971924, + "num_tokens": 47059959.0, + "step": 1825 + }, + { + "epoch": 0.20052712497254557, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.604398250579834, + "learning_rate": 6.6800878477306e-07, + "loss": 1.0323, + "mean_token_accuracy": 0.6953749060630798, + "num_tokens": 47083996.0, + "step": 1826 + }, + { + "epoch": 0.20063694267515925, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.397395372390747, + "learning_rate": 6.683748169838946e-07, + "loss": 1.0368, + "mean_token_accuracy": 0.7159387469291687, + "num_tokens": 47111746.0, + "step": 1827 + }, + { + "epoch": 0.2007467603777729, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.553718328475952, + "learning_rate": 6.687408491947291e-07, + "loss": 1.1121, + "mean_token_accuracy": 0.6839041709899902, + "num_tokens": 47137621.0, + "step": 1828 + }, + { + "epoch": 0.20085657808038657, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.54719614982605, + "learning_rate": 6.691068814055637e-07, + "loss": 1.0107, + "mean_token_accuracy": 0.6978538036346436, + "num_tokens": 47162400.0, + "step": 1829 + }, + { + "epoch": 0.2009663957830002, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.8103761672973633, + "learning_rate": 6.694729136163983e-07, + "loss": 1.0077, + "mean_token_accuracy": 0.7008969187736511, + "num_tokens": 47186479.0, + "step": 1830 + }, + { + "epoch": 0.20107621348561389, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.1848881244659424, + "learning_rate": 6.698389458272328e-07, + "loss": 1.1019, + "mean_token_accuracy": 0.6876657009124756, + "num_tokens": 47217322.0, + "step": 1831 + }, + { + "epoch": 0.20118603118822753, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.238572359085083, + "learning_rate": 6.702049780380673e-07, + "loss": 1.0438, + "mean_token_accuracy": 0.6876427531242371, + "num_tokens": 47247815.0, + "step": 1832 + }, + { + "epoch": 0.2012958488908412, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.7452919483184814, + "learning_rate": 6.705710102489019e-07, + "loss": 1.037, + "mean_token_accuracy": 0.6953979134559631, + "num_tokens": 47274328.0, + "step": 1833 + }, + { + "epoch": 0.20140566659345485, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.5036144256591797, + "learning_rate": 6.709370424597365e-07, + "loss": 1.0537, + "mean_token_accuracy": 0.68459153175354, + "num_tokens": 47303590.0, + "step": 1834 + }, + { + "epoch": 0.20151548429606853, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.4444375038146973, + "learning_rate": 6.71303074670571e-07, + "loss": 0.92, + "mean_token_accuracy": 0.729908287525177, + "num_tokens": 47331291.0, + "step": 1835 + }, + { + "epoch": 0.2016253019986822, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.5225250720977783, + "learning_rate": 6.716691068814056e-07, + "loss": 1.0303, + "mean_token_accuracy": 0.6949514150619507, + "num_tokens": 47356358.0, + "step": 1836 + }, + { + "epoch": 0.20173511970129585, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.4849750995635986, + "learning_rate": 6.720351390922401e-07, + "loss": 1.0209, + "mean_token_accuracy": 0.6995525360107422, + "num_tokens": 47383864.0, + "step": 1837 + }, + { + "epoch": 0.20184493740390952, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.483438014984131, + "learning_rate": 6.724011713030747e-07, + "loss": 1.052, + "mean_token_accuracy": 0.6892917156219482, + "num_tokens": 47408405.0, + "step": 1838 + }, + { + "epoch": 0.20195475510652316, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.5134215354919434, + "learning_rate": 6.727672035139092e-07, + "loss": 1.0211, + "mean_token_accuracy": 0.7020599246025085, + "num_tokens": 47432899.0, + "step": 1839 + }, + { + "epoch": 0.20206457280913684, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.6510322093963623, + "learning_rate": 6.731332357247438e-07, + "loss": 1.1476, + "mean_token_accuracy": 0.6697674989700317, + "num_tokens": 47459200.0, + "step": 1840 + }, + { + "epoch": 0.20217439051175048, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.359065294265747, + "learning_rate": 6.734992679355784e-07, + "loss": 1.1384, + "mean_token_accuracy": 0.6717801094055176, + "num_tokens": 47489755.0, + "step": 1841 + }, + { + "epoch": 0.20228420821436416, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.380812883377075, + "learning_rate": 6.738653001464129e-07, + "loss": 1.0621, + "mean_token_accuracy": 0.6873679757118225, + "num_tokens": 47521733.0, + "step": 1842 + }, + { + "epoch": 0.2023940259169778, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.7909584045410156, + "learning_rate": 6.742313323572474e-07, + "loss": 0.9775, + "mean_token_accuracy": 0.7107123136520386, + "num_tokens": 47541998.0, + "step": 1843 + }, + { + "epoch": 0.20250384361959148, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.409055233001709, + "learning_rate": 6.74597364568082e-07, + "loss": 1.0526, + "mean_token_accuracy": 0.6863391399383545, + "num_tokens": 47569865.0, + "step": 1844 + }, + { + "epoch": 0.20261366132220515, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.5246071815490723, + "learning_rate": 6.749633967789166e-07, + "loss": 1.14, + "mean_token_accuracy": 0.6721853017807007, + "num_tokens": 47598795.0, + "step": 1845 + }, + { + "epoch": 0.2027234790248188, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.4603869915008545, + "learning_rate": 6.753294289897511e-07, + "loss": 1.139, + "mean_token_accuracy": 0.670785665512085, + "num_tokens": 47627078.0, + "step": 1846 + }, + { + "epoch": 0.20283329672743247, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.1894171237945557, + "learning_rate": 6.756954612005857e-07, + "loss": 1.046, + "mean_token_accuracy": 0.6952427625656128, + "num_tokens": 47660542.0, + "step": 1847 + }, + { + "epoch": 0.20294311443004612, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.486212968826294, + "learning_rate": 6.760614934114202e-07, + "loss": 1.1042, + "mean_token_accuracy": 0.6789975166320801, + "num_tokens": 47687250.0, + "step": 1848 + }, + { + "epoch": 0.2030529321326598, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.687120199203491, + "learning_rate": 6.764275256222546e-07, + "loss": 0.9995, + "mean_token_accuracy": 0.6951888203620911, + "num_tokens": 47708672.0, + "step": 1849 + }, + { + "epoch": 0.20316274983527344, + "ewc_loss": 3.3229589462280273e-06, + "grad_norm": 2.8509342670440674, + "learning_rate": 6.767935578330894e-07, + "loss": 0.9823, + "mean_token_accuracy": 0.7126774787902832, + "num_tokens": 47730052.0, + "step": 1850 + }, + { + "epoch": 0.2032725675378871, + "ewc_loss": 3.3527612686157227e-06, + "grad_norm": 2.5265398025512695, + "learning_rate": 6.771595900439239e-07, + "loss": 1.1183, + "mean_token_accuracy": 0.6722545623779297, + "num_tokens": 47757372.0, + "step": 1851 + }, + { + "epoch": 0.20338238524050076, + "ewc_loss": 3.3527612686157227e-06, + "grad_norm": 2.642679214477539, + "learning_rate": 6.775256222547583e-07, + "loss": 1.0765, + "mean_token_accuracy": 0.6776921153068542, + "num_tokens": 47784543.0, + "step": 1852 + }, + { + "epoch": 0.20349220294311443, + "ewc_loss": 3.3527612686157227e-06, + "grad_norm": 2.909877300262451, + "learning_rate": 6.77891654465593e-07, + "loss": 1.0688, + "mean_token_accuracy": 0.685402512550354, + "num_tokens": 47813373.0, + "step": 1853 + }, + { + "epoch": 0.2036020206457281, + "ewc_loss": 3.3527612686157227e-06, + "grad_norm": 2.531451940536499, + "learning_rate": 6.782576866764274e-07, + "loss": 1.0859, + "mean_token_accuracy": 0.6837927103042603, + "num_tokens": 47837686.0, + "step": 1854 + }, + { + "epoch": 0.20371183834834175, + "ewc_loss": 3.3527612686157227e-06, + "grad_norm": 2.226144790649414, + "learning_rate": 6.78623718887262e-07, + "loss": 1.0918, + "mean_token_accuracy": 0.6887607574462891, + "num_tokens": 47870281.0, + "step": 1855 + }, + { + "epoch": 0.20382165605095542, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.377607822418213, + "learning_rate": 6.789897510980966e-07, + "loss": 1.1111, + "mean_token_accuracy": 0.6731311082839966, + "num_tokens": 47899080.0, + "step": 1856 + }, + { + "epoch": 0.20393147375356907, + "ewc_loss": 3.3527612686157227e-06, + "grad_norm": 2.41640043258667, + "learning_rate": 6.793557833089311e-07, + "loss": 1.0874, + "mean_token_accuracy": 0.6851160526275635, + "num_tokens": 47923965.0, + "step": 1857 + }, + { + "epoch": 0.20404129145618274, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.799720287322998, + "learning_rate": 6.797218155197657e-07, + "loss": 1.0158, + "mean_token_accuracy": 0.7000848650932312, + "num_tokens": 47945835.0, + "step": 1858 + }, + { + "epoch": 0.2041511091587964, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.6187171936035156, + "learning_rate": 6.800878477306002e-07, + "loss": 1.0546, + "mean_token_accuracy": 0.687341570854187, + "num_tokens": 47971101.0, + "step": 1859 + }, + { + "epoch": 0.20426092686141006, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.636929512023926, + "learning_rate": 6.804538799414347e-07, + "loss": 1.1131, + "mean_token_accuracy": 0.6919430494308472, + "num_tokens": 47995297.0, + "step": 1860 + }, + { + "epoch": 0.2043707445640237, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.4523024559020996, + "learning_rate": 6.808199121522694e-07, + "loss": 1.0991, + "mean_token_accuracy": 0.6777017116546631, + "num_tokens": 48028033.0, + "step": 1861 + }, + { + "epoch": 0.20448056226663738, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.764686107635498, + "learning_rate": 6.811859443631039e-07, + "loss": 1.0663, + "mean_token_accuracy": 0.6840710639953613, + "num_tokens": 48050471.0, + "step": 1862 + }, + { + "epoch": 0.20459037996925103, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.7046966552734375, + "learning_rate": 6.815519765739384e-07, + "loss": 1.0591, + "mean_token_accuracy": 0.688176155090332, + "num_tokens": 48073222.0, + "step": 1863 + }, + { + "epoch": 0.2047001976718647, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.8008062839508057, + "learning_rate": 6.81918008784773e-07, + "loss": 1.0489, + "mean_token_accuracy": 0.6911953687667847, + "num_tokens": 48097610.0, + "step": 1864 + }, + { + "epoch": 0.20481001537447838, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.803487539291382, + "learning_rate": 6.822840409956075e-07, + "loss": 1.0229, + "mean_token_accuracy": 0.6992942690849304, + "num_tokens": 48119076.0, + "step": 1865 + }, + { + "epoch": 0.20491983307709202, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.4622292518615723, + "learning_rate": 6.826500732064421e-07, + "loss": 1.0118, + "mean_token_accuracy": 0.699101448059082, + "num_tokens": 48145067.0, + "step": 1866 + }, + { + "epoch": 0.2050296507797057, + "ewc_loss": 3.3676624298095703e-06, + "grad_norm": 2.5950865745544434, + "learning_rate": 6.830161054172767e-07, + "loss": 1.0756, + "mean_token_accuracy": 0.6762288808822632, + "num_tokens": 48169280.0, + "step": 1867 + }, + { + "epoch": 0.20513946848231934, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.315892219543457, + "learning_rate": 6.833821376281112e-07, + "loss": 0.987, + "mean_token_accuracy": 0.7029207348823547, + "num_tokens": 48201385.0, + "step": 1868 + }, + { + "epoch": 0.20524928618493302, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.645265817642212, + "learning_rate": 6.837481698389457e-07, + "loss": 1.0215, + "mean_token_accuracy": 0.6950916051864624, + "num_tokens": 48227893.0, + "step": 1869 + }, + { + "epoch": 0.20535910388754666, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.652939558029175, + "learning_rate": 6.841142020497803e-07, + "loss": 1.0825, + "mean_token_accuracy": 0.6845203042030334, + "num_tokens": 48255573.0, + "step": 1870 + }, + { + "epoch": 0.20546892159016034, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.6894965171813965, + "learning_rate": 6.844802342606148e-07, + "loss": 0.9655, + "mean_token_accuracy": 0.7193821668624878, + "num_tokens": 48276990.0, + "step": 1871 + }, + { + "epoch": 0.20557873929277398, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.8087408542633057, + "learning_rate": 6.848462664714495e-07, + "loss": 1.054, + "mean_token_accuracy": 0.6930273771286011, + "num_tokens": 48298559.0, + "step": 1872 + }, + { + "epoch": 0.20568855699538766, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.58394193649292, + "learning_rate": 6.85212298682284e-07, + "loss": 1.0976, + "mean_token_accuracy": 0.6728967428207397, + "num_tokens": 48325901.0, + "step": 1873 + }, + { + "epoch": 0.20579837469800133, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.872689962387085, + "learning_rate": 6.855783308931185e-07, + "loss": 1.1086, + "mean_token_accuracy": 0.6836861371994019, + "num_tokens": 48347441.0, + "step": 1874 + }, + { + "epoch": 0.20590819240061498, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 3.250502586364746, + "learning_rate": 6.859443631039531e-07, + "loss": 1.0535, + "mean_token_accuracy": 0.6913511157035828, + "num_tokens": 48365276.0, + "step": 1875 + }, + { + "epoch": 0.20601801010322865, + "ewc_loss": 3.3974647521972656e-06, + "grad_norm": 2.787323474884033, + "learning_rate": 6.863103953147876e-07, + "loss": 0.9613, + "mean_token_accuracy": 0.7146371006965637, + "num_tokens": 48385953.0, + "step": 1876 + }, + { + "epoch": 0.2061278278058423, + "ewc_loss": 3.4123659133911133e-06, + "grad_norm": 2.6232316493988037, + "learning_rate": 6.866764275256222e-07, + "loss": 1.0103, + "mean_token_accuracy": 0.700941801071167, + "num_tokens": 48410901.0, + "step": 1877 + }, + { + "epoch": 0.20623764550845597, + "ewc_loss": 3.4123659133911133e-06, + "grad_norm": 2.4192357063293457, + "learning_rate": 6.870424597364568e-07, + "loss": 1.1045, + "mean_token_accuracy": 0.6792477369308472, + "num_tokens": 48438408.0, + "step": 1878 + }, + { + "epoch": 0.20634746321106961, + "ewc_loss": 3.4123659133911133e-06, + "grad_norm": 3.033268451690674, + "learning_rate": 6.874084919472913e-07, + "loss": 1.0233, + "mean_token_accuracy": 0.6987470388412476, + "num_tokens": 48457518.0, + "step": 1879 + }, + { + "epoch": 0.2064572809136833, + "ewc_loss": 3.4123659133911133e-06, + "grad_norm": 2.512451171875, + "learning_rate": 6.877745241581258e-07, + "loss": 1.0103, + "mean_token_accuracy": 0.7065882682800293, + "num_tokens": 48486392.0, + "step": 1880 + }, + { + "epoch": 0.20656709861629693, + "ewc_loss": 3.427267074584961e-06, + "grad_norm": 2.362138271331787, + "learning_rate": 6.881405563689604e-07, + "loss": 1.1453, + "mean_token_accuracy": 0.6761223673820496, + "num_tokens": 48516058.0, + "step": 1881 + }, + { + "epoch": 0.2066769163189106, + "ewc_loss": 3.4123659133911133e-06, + "grad_norm": 2.393760919570923, + "learning_rate": 6.88506588579795e-07, + "loss": 1.097, + "mean_token_accuracy": 0.6770950555801392, + "num_tokens": 48546658.0, + "step": 1882 + }, + { + "epoch": 0.20678673402152428, + "ewc_loss": 3.4123659133911133e-06, + "grad_norm": 2.632972240447998, + "learning_rate": 6.888726207906295e-07, + "loss": 1.0805, + "mean_token_accuracy": 0.6796995997428894, + "num_tokens": 48569941.0, + "step": 1883 + }, + { + "epoch": 0.20689655172413793, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.9583237171173096, + "learning_rate": 6.892386530014641e-07, + "loss": 1.0229, + "mean_token_accuracy": 0.6973749399185181, + "num_tokens": 48590774.0, + "step": 1884 + }, + { + "epoch": 0.2070063694267516, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.499871253967285, + "learning_rate": 6.896046852122986e-07, + "loss": 1.084, + "mean_token_accuracy": 0.6785259246826172, + "num_tokens": 48617652.0, + "step": 1885 + }, + { + "epoch": 0.20711618712936525, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.535628318786621, + "learning_rate": 6.899707174231331e-07, + "loss": 1.0294, + "mean_token_accuracy": 0.6901748180389404, + "num_tokens": 48642381.0, + "step": 1886 + }, + { + "epoch": 0.20722600483197892, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.3423964977264404, + "learning_rate": 6.903367496339677e-07, + "loss": 1.0229, + "mean_token_accuracy": 0.7021385431289673, + "num_tokens": 48671585.0, + "step": 1887 + }, + { + "epoch": 0.20733582253459257, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.3444912433624268, + "learning_rate": 6.907027818448023e-07, + "loss": 1.0912, + "mean_token_accuracy": 0.6782346367835999, + "num_tokens": 48703218.0, + "step": 1888 + }, + { + "epoch": 0.20744564023720624, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.489133834838867, + "learning_rate": 6.910688140556369e-07, + "loss": 1.0941, + "mean_token_accuracy": 0.6787711977958679, + "num_tokens": 48733166.0, + "step": 1889 + }, + { + "epoch": 0.2075554579398199, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.763669729232788, + "learning_rate": 6.914348462664714e-07, + "loss": 1.0506, + "mean_token_accuracy": 0.6940119862556458, + "num_tokens": 48756462.0, + "step": 1890 + }, + { + "epoch": 0.20766527564243356, + "ewc_loss": 3.4570693969726562e-06, + "grad_norm": 2.5000686645507812, + "learning_rate": 6.918008784773059e-07, + "loss": 1.106, + "mean_token_accuracy": 0.6885442137718201, + "num_tokens": 48785002.0, + "step": 1891 + }, + { + "epoch": 0.20777509334504723, + "ewc_loss": 3.4421682357788086e-06, + "grad_norm": 2.4173502922058105, + "learning_rate": 6.921669106881405e-07, + "loss": 0.9969, + "mean_token_accuracy": 0.7083761692047119, + "num_tokens": 48810763.0, + "step": 1892 + }, + { + "epoch": 0.20788491104766088, + "ewc_loss": 3.4570693969726562e-06, + "grad_norm": 2.70995831489563, + "learning_rate": 6.925329428989751e-07, + "loss": 1.0247, + "mean_token_accuracy": 0.6959061622619629, + "num_tokens": 48835009.0, + "step": 1893 + }, + { + "epoch": 0.20799472875027455, + "ewc_loss": 3.4570693969726562e-06, + "grad_norm": 2.7127671241760254, + "learning_rate": 6.928989751098096e-07, + "loss": 1.0215, + "mean_token_accuracy": 0.6969238519668579, + "num_tokens": 48858156.0, + "step": 1894 + }, + { + "epoch": 0.2081045464528882, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.7238082885742188, + "learning_rate": 6.932650073206442e-07, + "loss": 1.0884, + "mean_token_accuracy": 0.6798794269561768, + "num_tokens": 48882263.0, + "step": 1895 + }, + { + "epoch": 0.20821436415550187, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.7846779823303223, + "learning_rate": 6.936310395314787e-07, + "loss": 1.0815, + "mean_token_accuracy": 0.685515284538269, + "num_tokens": 48905461.0, + "step": 1896 + }, + { + "epoch": 0.20832418185811552, + "ewc_loss": 3.4868717193603516e-06, + "grad_norm": 2.5614993572235107, + "learning_rate": 6.939970717423132e-07, + "loss": 1.0526, + "mean_token_accuracy": 0.6885393857955933, + "num_tokens": 48931619.0, + "step": 1897 + }, + { + "epoch": 0.2084339995607292, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.534092903137207, + "learning_rate": 6.943631039531479e-07, + "loss": 1.0914, + "mean_token_accuracy": 0.6798776984214783, + "num_tokens": 48957444.0, + "step": 1898 + }, + { + "epoch": 0.20854381726334284, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.4518368244171143, + "learning_rate": 6.947291361639824e-07, + "loss": 0.9908, + "mean_token_accuracy": 0.7021256685256958, + "num_tokens": 48986415.0, + "step": 1899 + }, + { + "epoch": 0.2086536349659565, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.7559854984283447, + "learning_rate": 6.950951683748169e-07, + "loss": 1.0224, + "mean_token_accuracy": 0.6981226205825806, + "num_tokens": 49010061.0, + "step": 1900 + }, + { + "epoch": 0.20876345266857016, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.7327332496643066, + "learning_rate": 6.954612005856515e-07, + "loss": 0.977, + "mean_token_accuracy": 0.7116939425468445, + "num_tokens": 49033376.0, + "step": 1901 + }, + { + "epoch": 0.20887327037118383, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.6757700443267822, + "learning_rate": 6.95827232796486e-07, + "loss": 1.0439, + "mean_token_accuracy": 0.6931476593017578, + "num_tokens": 49057563.0, + "step": 1902 + }, + { + "epoch": 0.2089830880737975, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.563476324081421, + "learning_rate": 6.961932650073206e-07, + "loss": 1.1242, + "mean_token_accuracy": 0.6679829359054565, + "num_tokens": 49085030.0, + "step": 1903 + }, + { + "epoch": 0.20909290577641115, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.783445358276367, + "learning_rate": 6.965592972181552e-07, + "loss": 0.9606, + "mean_token_accuracy": 0.7089179158210754, + "num_tokens": 49104353.0, + "step": 1904 + }, + { + "epoch": 0.20920272347902483, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.409102201461792, + "learning_rate": 6.969253294289897e-07, + "loss": 1.0649, + "mean_token_accuracy": 0.694527268409729, + "num_tokens": 49134413.0, + "step": 1905 + }, + { + "epoch": 0.20931254118163847, + "ewc_loss": 3.4868717193603516e-06, + "grad_norm": 2.441796064376831, + "learning_rate": 6.972913616398243e-07, + "loss": 1.065, + "mean_token_accuracy": 0.6895418167114258, + "num_tokens": 49164137.0, + "step": 1906 + }, + { + "epoch": 0.20942235888425215, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.392151117324829, + "learning_rate": 6.976573938506588e-07, + "loss": 1.0476, + "mean_token_accuracy": 0.6871166229248047, + "num_tokens": 49192401.0, + "step": 1907 + }, + { + "epoch": 0.2095321765868658, + "ewc_loss": 3.471970558166504e-06, + "grad_norm": 2.936180353164673, + "learning_rate": 6.980234260614933e-07, + "loss": 1.0512, + "mean_token_accuracy": 0.6905162930488586, + "num_tokens": 49218341.0, + "step": 1908 + }, + { + "epoch": 0.20964199428947947, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.764564037322998, + "learning_rate": 6.98389458272328e-07, + "loss": 1.0345, + "mean_token_accuracy": 0.6970235109329224, + "num_tokens": 49239163.0, + "step": 1909 + }, + { + "epoch": 0.2097518119920931, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.4378037452697754, + "learning_rate": 6.987554904831625e-07, + "loss": 1.0584, + "mean_token_accuracy": 0.6909608840942383, + "num_tokens": 49268705.0, + "step": 1910 + }, + { + "epoch": 0.20986162969470679, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.562868356704712, + "learning_rate": 6.99121522693997e-07, + "loss": 0.9856, + "mean_token_accuracy": 0.707987368106842, + "num_tokens": 49293437.0, + "step": 1911 + }, + { + "epoch": 0.20997144739732046, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.562260627746582, + "learning_rate": 6.994875549048316e-07, + "loss": 1.1196, + "mean_token_accuracy": 0.6686387658119202, + "num_tokens": 49317640.0, + "step": 1912 + }, + { + "epoch": 0.2100812650999341, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.37267804145813, + "learning_rate": 6.998535871156661e-07, + "loss": 1.0601, + "mean_token_accuracy": 0.689638078212738, + "num_tokens": 49348204.0, + "step": 1913 + }, + { + "epoch": 0.21019108280254778, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.453112840652466, + "learning_rate": 7.002196193265007e-07, + "loss": 1.0469, + "mean_token_accuracy": 0.6981696486473083, + "num_tokens": 49376892.0, + "step": 1914 + }, + { + "epoch": 0.21030090050516143, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.459270715713501, + "learning_rate": 7.005856515373353e-07, + "loss": 1.0188, + "mean_token_accuracy": 0.6898207664489746, + "num_tokens": 49405788.0, + "step": 1915 + }, + { + "epoch": 0.2104107182077751, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.6480884552001953, + "learning_rate": 7.009516837481698e-07, + "loss": 1.078, + "mean_token_accuracy": 0.68352210521698, + "num_tokens": 49428328.0, + "step": 1916 + }, + { + "epoch": 0.21052053591038875, + "ewc_loss": 3.4868717193603516e-06, + "grad_norm": 2.4126834869384766, + "learning_rate": 7.013177159590043e-07, + "loss": 1.0972, + "mean_token_accuracy": 0.6897695064544678, + "num_tokens": 49456756.0, + "step": 1917 + }, + { + "epoch": 0.21063035361300242, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.674013614654541, + "learning_rate": 7.016837481698389e-07, + "loss": 1.0485, + "mean_token_accuracy": 0.6927304267883301, + "num_tokens": 49481855.0, + "step": 1918 + }, + { + "epoch": 0.21074017131561606, + "ewc_loss": 3.4868717193603516e-06, + "grad_norm": 2.587307929992676, + "learning_rate": 7.020497803806734e-07, + "loss": 1.0794, + "mean_token_accuracy": 0.684664785861969, + "num_tokens": 49507527.0, + "step": 1919 + }, + { + "epoch": 0.21084998901822974, + "ewc_loss": 3.4868717193603516e-06, + "grad_norm": 2.4396591186523438, + "learning_rate": 7.024158125915081e-07, + "loss": 1.0862, + "mean_token_accuracy": 0.6761928796768188, + "num_tokens": 49536120.0, + "step": 1920 + }, + { + "epoch": 0.2109598067208434, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.6520426273345947, + "learning_rate": 7.027818448023426e-07, + "loss": 1.0181, + "mean_token_accuracy": 0.6980679035186768, + "num_tokens": 49557633.0, + "step": 1921 + }, + { + "epoch": 0.21106962442345706, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.5633349418640137, + "learning_rate": 7.031478770131771e-07, + "loss": 1.0143, + "mean_token_accuracy": 0.702221691608429, + "num_tokens": 49581184.0, + "step": 1922 + }, + { + "epoch": 0.21117944212607073, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.322648525238037, + "learning_rate": 7.035139092240117e-07, + "loss": 1.078, + "mean_token_accuracy": 0.6845590472221375, + "num_tokens": 49612774.0, + "step": 1923 + }, + { + "epoch": 0.21128925982868438, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.8834307193756104, + "learning_rate": 7.038799414348462e-07, + "loss": 1.0136, + "mean_token_accuracy": 0.6985783576965332, + "num_tokens": 49632537.0, + "step": 1924 + }, + { + "epoch": 0.21139907753129805, + "ewc_loss": 3.5017728805541992e-06, + "grad_norm": 2.2991790771484375, + "learning_rate": 7.042459736456808e-07, + "loss": 0.9677, + "mean_token_accuracy": 0.7204364538192749, + "num_tokens": 49661533.0, + "step": 1925 + }, + { + "epoch": 0.2115088952339117, + "ewc_loss": 3.516674041748047e-06, + "grad_norm": 2.3134701251983643, + "learning_rate": 7.046120058565154e-07, + "loss": 1.016, + "mean_token_accuracy": 0.6973155736923218, + "num_tokens": 49691168.0, + "step": 1926 + }, + { + "epoch": 0.21161871293652537, + "ewc_loss": 3.516674041748047e-06, + "grad_norm": 2.4655661582946777, + "learning_rate": 7.049780380673499e-07, + "loss": 1.1504, + "mean_token_accuracy": 0.6699862480163574, + "num_tokens": 49718846.0, + "step": 1927 + }, + { + "epoch": 0.21172853063913902, + "ewc_loss": 3.516674041748047e-06, + "grad_norm": 2.989725112915039, + "learning_rate": 7.053440702781844e-07, + "loss": 1.048, + "mean_token_accuracy": 0.6947684288024902, + "num_tokens": 49738425.0, + "step": 1928 + }, + { + "epoch": 0.2118383483417527, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.6353251934051514, + "learning_rate": 7.05710102489019e-07, + "loss": 0.9942, + "mean_token_accuracy": 0.7073920369148254, + "num_tokens": 49762843.0, + "step": 1929 + }, + { + "epoch": 0.21194816604436637, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.4673311710357666, + "learning_rate": 7.060761346998536e-07, + "loss": 0.942, + "mean_token_accuracy": 0.7235240936279297, + "num_tokens": 49787535.0, + "step": 1930 + }, + { + "epoch": 0.21205798374698, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.7974631786346436, + "learning_rate": 7.064421669106881e-07, + "loss": 0.9744, + "mean_token_accuracy": 0.7153034806251526, + "num_tokens": 49808954.0, + "step": 1931 + }, + { + "epoch": 0.21216780144959368, + "ewc_loss": 3.546476364135742e-06, + "grad_norm": 2.569305419921875, + "learning_rate": 7.068081991215227e-07, + "loss": 0.9854, + "mean_token_accuracy": 0.712841808795929, + "num_tokens": 49833932.0, + "step": 1932 + }, + { + "epoch": 0.21227761915220733, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.5883162021636963, + "learning_rate": 7.071742313323572e-07, + "loss": 1.0708, + "mean_token_accuracy": 0.6871428489685059, + "num_tokens": 49861910.0, + "step": 1933 + }, + { + "epoch": 0.212387436854821, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.663968563079834, + "learning_rate": 7.075402635431917e-07, + "loss": 1.0752, + "mean_token_accuracy": 0.685951828956604, + "num_tokens": 49885140.0, + "step": 1934 + }, + { + "epoch": 0.21249725455743465, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.5406577587127686, + "learning_rate": 7.079062957540263e-07, + "loss": 0.9705, + "mean_token_accuracy": 0.7105024456977844, + "num_tokens": 49909653.0, + "step": 1935 + }, + { + "epoch": 0.21260707226004832, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.5372629165649414, + "learning_rate": 7.082723279648609e-07, + "loss": 1.0817, + "mean_token_accuracy": 0.6844508647918701, + "num_tokens": 49936584.0, + "step": 1936 + }, + { + "epoch": 0.21271688996266197, + "ewc_loss": 3.5315752029418945e-06, + "grad_norm": 2.8579728603363037, + "learning_rate": 7.086383601756955e-07, + "loss": 1.0454, + "mean_token_accuracy": 0.6896653175354004, + "num_tokens": 49956702.0, + "step": 1937 + }, + { + "epoch": 0.21282670766527564, + "ewc_loss": 3.546476364135742e-06, + "grad_norm": 2.760298490524292, + "learning_rate": 7.0900439238653e-07, + "loss": 0.9976, + "mean_token_accuracy": 0.6938092112541199, + "num_tokens": 49977896.0, + "step": 1938 + }, + { + "epoch": 0.2129365253678893, + "ewc_loss": 3.546476364135742e-06, + "grad_norm": 2.5069544315338135, + "learning_rate": 7.093704245973645e-07, + "loss": 0.9839, + "mean_token_accuracy": 0.7117176055908203, + "num_tokens": 50003208.0, + "step": 1939 + }, + { + "epoch": 0.21304634307050296, + "ewc_loss": 3.546476364135742e-06, + "grad_norm": 2.774298906326294, + "learning_rate": 7.097364568081991e-07, + "loss": 1.0508, + "mean_token_accuracy": 0.6907224655151367, + "num_tokens": 50024940.0, + "step": 1940 + }, + { + "epoch": 0.21315616077311664, + "ewc_loss": 3.56137752532959e-06, + "grad_norm": 2.359884738922119, + "learning_rate": 7.101024890190337e-07, + "loss": 0.9379, + "mean_token_accuracy": 0.722618818283081, + "num_tokens": 50051034.0, + "step": 1941 + }, + { + "epoch": 0.21326597847573028, + "ewc_loss": 3.56137752532959e-06, + "grad_norm": 2.4715001583099365, + "learning_rate": 7.104685212298682e-07, + "loss": 1.0703, + "mean_token_accuracy": 0.6851897239685059, + "num_tokens": 50077175.0, + "step": 1942 + }, + { + "epoch": 0.21337579617834396, + "ewc_loss": 3.56137752532959e-06, + "grad_norm": 2.1464974880218506, + "learning_rate": 7.108345534407028e-07, + "loss": 1.1861, + "mean_token_accuracy": 0.6528418660163879, + "num_tokens": 50115357.0, + "step": 1943 + }, + { + "epoch": 0.2134856138809576, + "ewc_loss": 3.56137752532959e-06, + "grad_norm": 2.3201136589050293, + "learning_rate": 7.112005856515373e-07, + "loss": 0.9752, + "mean_token_accuracy": 0.7116224765777588, + "num_tokens": 50143631.0, + "step": 1944 + }, + { + "epoch": 0.21359543158357128, + "ewc_loss": 3.56137752532959e-06, + "grad_norm": 2.6138389110565186, + "learning_rate": 7.115666178623718e-07, + "loss": 1.0117, + "mean_token_accuracy": 0.7062091827392578, + "num_tokens": 50167265.0, + "step": 1945 + }, + { + "epoch": 0.21370524928618492, + "ewc_loss": 3.591179847717285e-06, + "grad_norm": 2.979487895965576, + "learning_rate": 7.119326500732065e-07, + "loss": 1.1411, + "mean_token_accuracy": 0.6666300296783447, + "num_tokens": 50191169.0, + "step": 1946 + }, + { + "epoch": 0.2138150669887986, + "ewc_loss": 3.5762786865234375e-06, + "grad_norm": 2.3843584060668945, + "learning_rate": 7.12298682284041e-07, + "loss": 1.0575, + "mean_token_accuracy": 0.6873126029968262, + "num_tokens": 50218920.0, + "step": 1947 + }, + { + "epoch": 0.21392488469141224, + "ewc_loss": 3.56137752532959e-06, + "grad_norm": 2.5968260765075684, + "learning_rate": 7.126647144948755e-07, + "loss": 1.1544, + "mean_token_accuracy": 0.6642801761627197, + "num_tokens": 50244790.0, + "step": 1948 + }, + { + "epoch": 0.21403470239402592, + "ewc_loss": 3.5762786865234375e-06, + "grad_norm": 2.47412371635437, + "learning_rate": 7.130307467057101e-07, + "loss": 1.02, + "mean_token_accuracy": 0.6999154090881348, + "num_tokens": 50270316.0, + "step": 1949 + }, + { + "epoch": 0.2141445200966396, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.439039945602417, + "learning_rate": 7.133967789165446e-07, + "loss": 1.0328, + "mean_token_accuracy": 0.6957584023475647, + "num_tokens": 50296368.0, + "step": 1950 + }, + { + "epoch": 0.21425433779925324, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.8811447620391846, + "learning_rate": 7.137628111273791e-07, + "loss": 1.0146, + "mean_token_accuracy": 0.6934046149253845, + "num_tokens": 50317194.0, + "step": 1951 + }, + { + "epoch": 0.2143641555018669, + "ewc_loss": 3.5762786865234375e-06, + "grad_norm": 2.652431011199951, + "learning_rate": 7.141288433382138e-07, + "loss": 0.9816, + "mean_token_accuracy": 0.7081747651100159, + "num_tokens": 50338475.0, + "step": 1952 + }, + { + "epoch": 0.21447397320448056, + "ewc_loss": 3.5762786865234375e-06, + "grad_norm": 2.549037218093872, + "learning_rate": 7.144948755490483e-07, + "loss": 1.0931, + "mean_token_accuracy": 0.6762378215789795, + "num_tokens": 50363753.0, + "step": 1953 + }, + { + "epoch": 0.21458379090709423, + "ewc_loss": 3.591179847717285e-06, + "grad_norm": 2.534187078475952, + "learning_rate": 7.148609077598829e-07, + "loss": 0.992, + "mean_token_accuracy": 0.7031913995742798, + "num_tokens": 50387871.0, + "step": 1954 + }, + { + "epoch": 0.21469360860970788, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.755025625228882, + "learning_rate": 7.152269399707174e-07, + "loss": 1.0364, + "mean_token_accuracy": 0.6960204839706421, + "num_tokens": 50408307.0, + "step": 1955 + }, + { + "epoch": 0.21480342631232155, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.6010191440582275, + "learning_rate": 7.155929721815519e-07, + "loss": 1.1095, + "mean_token_accuracy": 0.688962459564209, + "num_tokens": 50434392.0, + "step": 1956 + }, + { + "epoch": 0.2149132440149352, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.2820749282836914, + "learning_rate": 7.159590043923866e-07, + "loss": 1.0618, + "mean_token_accuracy": 0.686681866645813, + "num_tokens": 50464382.0, + "step": 1957 + }, + { + "epoch": 0.21502306171754887, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.8337156772613525, + "learning_rate": 7.163250366032211e-07, + "loss": 1.1627, + "mean_token_accuracy": 0.6559673547744751, + "num_tokens": 50486233.0, + "step": 1958 + }, + { + "epoch": 0.21513287942016254, + "ewc_loss": 3.606081008911133e-06, + "grad_norm": 2.382664918899536, + "learning_rate": 7.166910688140556e-07, + "loss": 1.1056, + "mean_token_accuracy": 0.671788215637207, + "num_tokens": 50516822.0, + "step": 1959 + }, + { + "epoch": 0.2152426971227762, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 2.7190518379211426, + "learning_rate": 7.170571010248902e-07, + "loss": 1.1339, + "mean_token_accuracy": 0.6738255023956299, + "num_tokens": 50541270.0, + "step": 1960 + }, + { + "epoch": 0.21535251482538986, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 2.5092225074768066, + "learning_rate": 7.174231332357247e-07, + "loss": 1.0048, + "mean_token_accuracy": 0.6995059847831726, + "num_tokens": 50566736.0, + "step": 1961 + }, + { + "epoch": 0.2154623325280035, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 2.6046152114868164, + "learning_rate": 7.177891654465593e-07, + "loss": 1.0437, + "mean_token_accuracy": 0.6908263564109802, + "num_tokens": 50591212.0, + "step": 1962 + }, + { + "epoch": 0.21557215023061718, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 2.765362501144409, + "learning_rate": 7.181551976573939e-07, + "loss": 0.9678, + "mean_token_accuracy": 0.7119083404541016, + "num_tokens": 50613334.0, + "step": 1963 + }, + { + "epoch": 0.21568196793323083, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 2.337245464324951, + "learning_rate": 7.185212298682284e-07, + "loss": 0.9737, + "mean_token_accuracy": 0.7108324766159058, + "num_tokens": 50642793.0, + "step": 1964 + }, + { + "epoch": 0.2157917856358445, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 2.4927258491516113, + "learning_rate": 7.188872620790629e-07, + "loss": 1.1435, + "mean_token_accuracy": 0.6765679121017456, + "num_tokens": 50670184.0, + "step": 1965 + }, + { + "epoch": 0.21590160333845815, + "ewc_loss": 3.6209821701049805e-06, + "grad_norm": 3.088484764099121, + "learning_rate": 7.192532942898975e-07, + "loss": 0.9674, + "mean_token_accuracy": 0.707207202911377, + "num_tokens": 50688140.0, + "step": 1966 + }, + { + "epoch": 0.21601142104107182, + "ewc_loss": 3.635883331298828e-06, + "grad_norm": 2.5483973026275635, + "learning_rate": 7.19619326500732e-07, + "loss": 1.07, + "mean_token_accuracy": 0.6822307109832764, + "num_tokens": 50712549.0, + "step": 1967 + }, + { + "epoch": 0.2161212387436855, + "ewc_loss": 3.635883331298828e-06, + "grad_norm": 2.630640983581543, + "learning_rate": 7.199853587115667e-07, + "loss": 1.1553, + "mean_token_accuracy": 0.6635974645614624, + "num_tokens": 50738520.0, + "step": 1968 + }, + { + "epoch": 0.21623105644629914, + "ewc_loss": 3.635883331298828e-06, + "grad_norm": 2.50713849067688, + "learning_rate": 7.203513909224012e-07, + "loss": 0.9941, + "mean_token_accuracy": 0.7136585116386414, + "num_tokens": 50763757.0, + "step": 1969 + }, + { + "epoch": 0.21634087414891282, + "ewc_loss": 3.635883331298828e-06, + "grad_norm": 2.7103095054626465, + "learning_rate": 7.207174231332357e-07, + "loss": 1.0525, + "mean_token_accuracy": 0.7012801170349121, + "num_tokens": 50793832.0, + "step": 1970 + }, + { + "epoch": 0.21645069185152646, + "ewc_loss": 3.635883331298828e-06, + "grad_norm": 2.8039817810058594, + "learning_rate": 7.210834553440703e-07, + "loss": 0.9643, + "mean_token_accuracy": 0.708588182926178, + "num_tokens": 50814416.0, + "step": 1971 + }, + { + "epoch": 0.21656050955414013, + "ewc_loss": 3.635883331298828e-06, + "grad_norm": 3.0131418704986572, + "learning_rate": 7.214494875549048e-07, + "loss": 0.9706, + "mean_token_accuracy": 0.7052932977676392, + "num_tokens": 50833389.0, + "step": 1972 + }, + { + "epoch": 0.21667032725675378, + "ewc_loss": 3.6507844924926758e-06, + "grad_norm": 2.4130351543426514, + "learning_rate": 7.218155197657394e-07, + "loss": 1.1603, + "mean_token_accuracy": 0.6683118343353271, + "num_tokens": 50863979.0, + "step": 1973 + }, + { + "epoch": 0.21678014495936745, + "ewc_loss": 3.6507844924926758e-06, + "grad_norm": 2.5668694972991943, + "learning_rate": 7.22181551976574e-07, + "loss": 0.9841, + "mean_token_accuracy": 0.70401930809021, + "num_tokens": 50886208.0, + "step": 1974 + }, + { + "epoch": 0.2168899626619811, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.6596570014953613, + "learning_rate": 7.225475841874085e-07, + "loss": 1.0579, + "mean_token_accuracy": 0.6866642236709595, + "num_tokens": 50908463.0, + "step": 1975 + }, + { + "epoch": 0.21699978036459477, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.7453229427337646, + "learning_rate": 7.22913616398243e-07, + "loss": 1.0086, + "mean_token_accuracy": 0.7087165117263794, + "num_tokens": 50931728.0, + "step": 1976 + }, + { + "epoch": 0.21710959806720842, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.48584246635437, + "learning_rate": 7.232796486090776e-07, + "loss": 1.0855, + "mean_token_accuracy": 0.6801497936248779, + "num_tokens": 50956836.0, + "step": 1977 + }, + { + "epoch": 0.2172194157698221, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.3540432453155518, + "learning_rate": 7.236456808199122e-07, + "loss": 1.1666, + "mean_token_accuracy": 0.6590887308120728, + "num_tokens": 50988874.0, + "step": 1978 + }, + { + "epoch": 0.21732923347243577, + "ewc_loss": 3.6507844924926758e-06, + "grad_norm": 2.4872937202453613, + "learning_rate": 7.240117130307467e-07, + "loss": 1.1367, + "mean_token_accuracy": 0.6811974048614502, + "num_tokens": 51017825.0, + "step": 1979 + }, + { + "epoch": 0.2174390511750494, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.3767993450164795, + "learning_rate": 7.243777452415813e-07, + "loss": 1.018, + "mean_token_accuracy": 0.6994250416755676, + "num_tokens": 51045955.0, + "step": 1980 + }, + { + "epoch": 0.2175488688776631, + "ewc_loss": 3.6507844924926758e-06, + "grad_norm": 2.686825752258301, + "learning_rate": 7.247437774524158e-07, + "loss": 1.0656, + "mean_token_accuracy": 0.6784927248954773, + "num_tokens": 51069330.0, + "step": 1981 + }, + { + "epoch": 0.21765868658027673, + "ewc_loss": 3.6507844924926758e-06, + "grad_norm": 2.7333874702453613, + "learning_rate": 7.251098096632503e-07, + "loss": 1.0951, + "mean_token_accuracy": 0.6910849809646606, + "num_tokens": 51092178.0, + "step": 1982 + }, + { + "epoch": 0.2177685042828904, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.5370030403137207, + "learning_rate": 7.254758418740849e-07, + "loss": 1.0185, + "mean_token_accuracy": 0.6919636726379395, + "num_tokens": 51115554.0, + "step": 1983 + }, + { + "epoch": 0.21787832198550405, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.7461326122283936, + "learning_rate": 7.258418740849195e-07, + "loss": 1.0054, + "mean_token_accuracy": 0.7018174529075623, + "num_tokens": 51135508.0, + "step": 1984 + }, + { + "epoch": 0.21798813968811773, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.516695499420166, + "learning_rate": 7.262079062957541e-07, + "loss": 1.0047, + "mean_token_accuracy": 0.7043335437774658, + "num_tokens": 51159938.0, + "step": 1985 + }, + { + "epoch": 0.21809795739073137, + "ewc_loss": 3.680586814880371e-06, + "grad_norm": 2.642230987548828, + "learning_rate": 7.265739385065886e-07, + "loss": 1.1507, + "mean_token_accuracy": 0.6678062677383423, + "num_tokens": 51189162.0, + "step": 1986 + }, + { + "epoch": 0.21820777509334505, + "ewc_loss": 3.6954879760742188e-06, + "grad_norm": 2.7251129150390625, + "learning_rate": 7.269399707174231e-07, + "loss": 1.0404, + "mean_token_accuracy": 0.6910030841827393, + "num_tokens": 51213141.0, + "step": 1987 + }, + { + "epoch": 0.21831759279595872, + "ewc_loss": 3.6954879760742188e-06, + "grad_norm": 2.453198194503784, + "learning_rate": 7.273060029282577e-07, + "loss": 1.0631, + "mean_token_accuracy": 0.6874226331710815, + "num_tokens": 51243740.0, + "step": 1988 + }, + { + "epoch": 0.21842741049857237, + "ewc_loss": 3.6656856536865234e-06, + "grad_norm": 2.8247203826904297, + "learning_rate": 7.276720351390923e-07, + "loss": 1.0548, + "mean_token_accuracy": 0.6988620162010193, + "num_tokens": 51267656.0, + "step": 1989 + }, + { + "epoch": 0.21853722820118604, + "ewc_loss": 3.6954879760742188e-06, + "grad_norm": 2.650344133377075, + "learning_rate": 7.280380673499268e-07, + "loss": 0.9696, + "mean_token_accuracy": 0.7107890248298645, + "num_tokens": 51291884.0, + "step": 1990 + }, + { + "epoch": 0.21864704590379969, + "ewc_loss": 3.6954879760742188e-06, + "grad_norm": 2.826470375061035, + "learning_rate": 7.284040995607614e-07, + "loss": 1.0739, + "mean_token_accuracy": 0.6842236518859863, + "num_tokens": 51313957.0, + "step": 1991 + }, + { + "epoch": 0.21875686360641336, + "ewc_loss": 3.725290298461914e-06, + "grad_norm": 6.882008075714111, + "learning_rate": 7.287701317715959e-07, + "loss": 1.0132, + "mean_token_accuracy": 0.6989148855209351, + "num_tokens": 51333146.0, + "step": 1992 + }, + { + "epoch": 0.218866681309027, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 3.0125513076782227, + "learning_rate": 7.291361639824304e-07, + "loss": 1.0997, + "mean_token_accuracy": 0.6794861555099487, + "num_tokens": 51354676.0, + "step": 1993 + }, + { + "epoch": 0.21897649901164068, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.598590135574341, + "learning_rate": 7.295021961932651e-07, + "loss": 1.0252, + "mean_token_accuracy": 0.6988868117332458, + "num_tokens": 51377314.0, + "step": 1994 + }, + { + "epoch": 0.21908631671425433, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.519695520401001, + "learning_rate": 7.298682284040996e-07, + "loss": 1.0788, + "mean_token_accuracy": 0.6852335333824158, + "num_tokens": 51404483.0, + "step": 1995 + }, + { + "epoch": 0.219196134416868, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.584280252456665, + "learning_rate": 7.30234260614934e-07, + "loss": 1.0293, + "mean_token_accuracy": 0.6961787343025208, + "num_tokens": 51427270.0, + "step": 1996 + }, + { + "epoch": 0.21930595211948167, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.7485709190368652, + "learning_rate": 7.306002928257687e-07, + "loss": 1.0944, + "mean_token_accuracy": 0.6849486231803894, + "num_tokens": 51451033.0, + "step": 1997 + }, + { + "epoch": 0.21941576982209532, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.441197395324707, + "learning_rate": 7.309663250366031e-07, + "loss": 1.0175, + "mean_token_accuracy": 0.7006028890609741, + "num_tokens": 51477625.0, + "step": 1998 + }, + { + "epoch": 0.219525587524709, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.7469091415405273, + "learning_rate": 7.313323572474376e-07, + "loss": 1.0693, + "mean_token_accuracy": 0.6861942410469055, + "num_tokens": 51499313.0, + "step": 1999 + }, + { + "epoch": 0.21963540522732264, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.6776177883148193, + "learning_rate": 7.316983894582724e-07, + "loss": 1.0331, + "mean_token_accuracy": 0.7029784321784973, + "num_tokens": 51523584.0, + "step": 2000 + }, + { + "epoch": 0.2197452229299363, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.905590057373047, + "learning_rate": 7.320644216691068e-07, + "loss": 1.0944, + "mean_token_accuracy": 0.6808141469955444, + "num_tokens": 51544804.0, + "step": 2001 + }, + { + "epoch": 0.21985504063254996, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.6939501762390137, + "learning_rate": 7.324304538799415e-07, + "loss": 1.0003, + "mean_token_accuracy": 0.70130854845047, + "num_tokens": 51568871.0, + "step": 2002 + }, + { + "epoch": 0.21996485833516363, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.439305543899536, + "learning_rate": 7.32796486090776e-07, + "loss": 1.0426, + "mean_token_accuracy": 0.7029373049736023, + "num_tokens": 51596345.0, + "step": 2003 + }, + { + "epoch": 0.22007467603777728, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.47662091255188, + "learning_rate": 7.331625183016104e-07, + "loss": 1.0875, + "mean_token_accuracy": 0.6849671006202698, + "num_tokens": 51624644.0, + "step": 2004 + }, + { + "epoch": 0.22018449374039095, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.6491644382476807, + "learning_rate": 7.335285505124452e-07, + "loss": 0.956, + "mean_token_accuracy": 0.7140289545059204, + "num_tokens": 51647338.0, + "step": 2005 + }, + { + "epoch": 0.22029431144300463, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.7194695472717285, + "learning_rate": 7.338945827232796e-07, + "loss": 1.0658, + "mean_token_accuracy": 0.6882050037384033, + "num_tokens": 51670588.0, + "step": 2006 + }, + { + "epoch": 0.22040412914561827, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.501831531524658, + "learning_rate": 7.342606149341141e-07, + "loss": 1.0245, + "mean_token_accuracy": 0.7082641124725342, + "num_tokens": 51694915.0, + "step": 2007 + }, + { + "epoch": 0.22051394684823195, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.3639395236968994, + "learning_rate": 7.346266471449487e-07, + "loss": 1.1274, + "mean_token_accuracy": 0.6691006422042847, + "num_tokens": 51722461.0, + "step": 2008 + }, + { + "epoch": 0.2206237645508456, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.387773275375366, + "learning_rate": 7.349926793557832e-07, + "loss": 1.0171, + "mean_token_accuracy": 0.6971898078918457, + "num_tokens": 51750114.0, + "step": 2009 + }, + { + "epoch": 0.22073358225345927, + "ewc_loss": 3.7103891372680664e-06, + "grad_norm": 2.8041324615478516, + "learning_rate": 7.353587115666178e-07, + "loss": 0.9993, + "mean_token_accuracy": 0.7027890086174011, + "num_tokens": 51769519.0, + "step": 2010 + }, + { + "epoch": 0.2208433999560729, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.7981503009796143, + "learning_rate": 7.357247437774524e-07, + "loss": 0.982, + "mean_token_accuracy": 0.7121356725692749, + "num_tokens": 51790952.0, + "step": 2011 + }, + { + "epoch": 0.22095321765868658, + "ewc_loss": 3.725290298461914e-06, + "grad_norm": 2.3551154136657715, + "learning_rate": 7.360907759882869e-07, + "loss": 1.04, + "mean_token_accuracy": 0.6993298530578613, + "num_tokens": 51820766.0, + "step": 2012 + }, + { + "epoch": 0.22106303536130023, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.5034003257751465, + "learning_rate": 7.364568081991214e-07, + "loss": 1.1046, + "mean_token_accuracy": 0.679539144039154, + "num_tokens": 51848779.0, + "step": 2013 + }, + { + "epoch": 0.2211728530639139, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.7353579998016357, + "learning_rate": 7.36822840409956e-07, + "loss": 1.0783, + "mean_token_accuracy": 0.6856464147567749, + "num_tokens": 51872683.0, + "step": 2014 + }, + { + "epoch": 0.22128267076652755, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.7149667739868164, + "learning_rate": 7.371888726207905e-07, + "loss": 1.0085, + "mean_token_accuracy": 0.7019472718238831, + "num_tokens": 51892998.0, + "step": 2015 + }, + { + "epoch": 0.22139248846914122, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.6369051933288574, + "learning_rate": 7.375549048316251e-07, + "loss": 1.0233, + "mean_token_accuracy": 0.6948120594024658, + "num_tokens": 51916800.0, + "step": 2016 + }, + { + "epoch": 0.2215023061717549, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.8300328254699707, + "learning_rate": 7.379209370424597e-07, + "loss": 0.998, + "mean_token_accuracy": 0.7053383588790894, + "num_tokens": 51938004.0, + "step": 2017 + }, + { + "epoch": 0.22161212387436854, + "ewc_loss": 3.7401914596557617e-06, + "grad_norm": 2.8811168670654297, + "learning_rate": 7.382869692532942e-07, + "loss": 1.0147, + "mean_token_accuracy": 0.6975032091140747, + "num_tokens": 51957847.0, + "step": 2018 + }, + { + "epoch": 0.22172194157698222, + "ewc_loss": 3.7550926208496094e-06, + "grad_norm": 2.709105968475342, + "learning_rate": 7.386530014641288e-07, + "loss": 1.0039, + "mean_token_accuracy": 0.7030110359191895, + "num_tokens": 51980724.0, + "step": 2019 + }, + { + "epoch": 0.22183175927959586, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.612595558166504, + "learning_rate": 7.390190336749633e-07, + "loss": 0.9805, + "mean_token_accuracy": 0.703872561454773, + "num_tokens": 52004838.0, + "step": 2020 + }, + { + "epoch": 0.22194157698220954, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 9.6788969039917, + "learning_rate": 7.393850658857979e-07, + "loss": 1.1172, + "mean_token_accuracy": 0.6852753758430481, + "num_tokens": 52031951.0, + "step": 2021 + }, + { + "epoch": 0.22205139468482318, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.7470462322235107, + "learning_rate": 7.397510980966325e-07, + "loss": 1.0648, + "mean_token_accuracy": 0.6833213567733765, + "num_tokens": 52055321.0, + "step": 2022 + }, + { + "epoch": 0.22216121238743686, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.6300737857818604, + "learning_rate": 7.40117130307467e-07, + "loss": 1.1175, + "mean_token_accuracy": 0.6733402609825134, + "num_tokens": 52081626.0, + "step": 2023 + }, + { + "epoch": 0.2222710300900505, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.201078176498413, + "learning_rate": 7.404831625183015e-07, + "loss": 1.086, + "mean_token_accuracy": 0.6831626892089844, + "num_tokens": 52116812.0, + "step": 2024 + }, + { + "epoch": 0.22238084779266418, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.4308619499206543, + "learning_rate": 7.408491947291361e-07, + "loss": 1.0893, + "mean_token_accuracy": 0.6916499137878418, + "num_tokens": 52146142.0, + "step": 2025 + }, + { + "epoch": 0.22249066549527785, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.5197958946228027, + "learning_rate": 7.412152269399707e-07, + "loss": 1.0757, + "mean_token_accuracy": 0.6850438117980957, + "num_tokens": 52171070.0, + "step": 2026 + }, + { + "epoch": 0.2226004831978915, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.4781720638275146, + "learning_rate": 7.415812591508052e-07, + "loss": 0.9689, + "mean_token_accuracy": 0.7092692852020264, + "num_tokens": 52198273.0, + "step": 2027 + }, + { + "epoch": 0.22271030090050517, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.6851913928985596, + "learning_rate": 7.419472913616398e-07, + "loss": 1.0436, + "mean_token_accuracy": 0.6932083368301392, + "num_tokens": 52220792.0, + "step": 2028 + }, + { + "epoch": 0.22282011860311882, + "ewc_loss": 4.202127456665039e-06, + "grad_norm": 82.20884704589844, + "learning_rate": 7.423133235724743e-07, + "loss": 1.0473, + "mean_token_accuracy": 0.7046859264373779, + "num_tokens": 52248973.0, + "step": 2029 + }, + { + "epoch": 0.2229299363057325, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.7210466861724854, + "learning_rate": 7.426793557833088e-07, + "loss": 1.117, + "mean_token_accuracy": 0.6757158041000366, + "num_tokens": 52271488.0, + "step": 2030 + }, + { + "epoch": 0.22303975400834614, + "ewc_loss": 3.814697265625e-06, + "grad_norm": 4.310333251953125, + "learning_rate": 7.430453879941434e-07, + "loss": 1.1573, + "mean_token_accuracy": 0.6622790098190308, + "num_tokens": 52299754.0, + "step": 2031 + }, + { + "epoch": 0.2231495717109598, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.621473550796509, + "learning_rate": 7.43411420204978e-07, + "loss": 1.119, + "mean_token_accuracy": 0.6768057346343994, + "num_tokens": 52324906.0, + "step": 2032 + }, + { + "epoch": 0.22325938941357346, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.6257429122924805, + "learning_rate": 7.437774524158125e-07, + "loss": 1.0065, + "mean_token_accuracy": 0.7008036375045776, + "num_tokens": 52348354.0, + "step": 2033 + }, + { + "epoch": 0.22336920711618713, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 3.129403591156006, + "learning_rate": 7.441434846266471e-07, + "loss": 1.0539, + "mean_token_accuracy": 0.6878081560134888, + "num_tokens": 52369396.0, + "step": 2034 + }, + { + "epoch": 0.2234790248188008, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.708683967590332, + "learning_rate": 7.445095168374816e-07, + "loss": 0.9038, + "mean_token_accuracy": 0.7186470031738281, + "num_tokens": 52391080.0, + "step": 2035 + }, + { + "epoch": 0.22358884252141445, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.8942391872406006, + "learning_rate": 7.448755490483162e-07, + "loss": 0.988, + "mean_token_accuracy": 0.705530047416687, + "num_tokens": 52410968.0, + "step": 2036 + }, + { + "epoch": 0.22369866022402812, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.301386833190918, + "learning_rate": 7.452415812591508e-07, + "loss": 1.1248, + "mean_token_accuracy": 0.6677141189575195, + "num_tokens": 52441502.0, + "step": 2037 + }, + { + "epoch": 0.22380847792664177, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.6968860626220703, + "learning_rate": 7.456076134699853e-07, + "loss": 1.0636, + "mean_token_accuracy": 0.6872447729110718, + "num_tokens": 52464356.0, + "step": 2038 + }, + { + "epoch": 0.22391829562925544, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.4068336486816406, + "learning_rate": 7.459736456808199e-07, + "loss": 1.053, + "mean_token_accuracy": 0.6907786726951599, + "num_tokens": 52491196.0, + "step": 2039 + }, + { + "epoch": 0.2240281133318691, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 3.0843818187713623, + "learning_rate": 7.463396778916544e-07, + "loss": 1.0503, + "mean_token_accuracy": 0.6909305453300476, + "num_tokens": 52510079.0, + "step": 2040 + }, + { + "epoch": 0.22413793103448276, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.5221056938171387, + "learning_rate": 7.467057101024889e-07, + "loss": 1.132, + "mean_token_accuracy": 0.6767834424972534, + "num_tokens": 52536899.0, + "step": 2041 + }, + { + "epoch": 0.2242477487370964, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.4730451107025146, + "learning_rate": 7.470717423133236e-07, + "loss": 1.1387, + "mean_token_accuracy": 0.6724358797073364, + "num_tokens": 52565125.0, + "step": 2042 + }, + { + "epoch": 0.22435756643971008, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.4774763584136963, + "learning_rate": 7.474377745241581e-07, + "loss": 1.0221, + "mean_token_accuracy": 0.6988948583602905, + "num_tokens": 52589724.0, + "step": 2043 + }, + { + "epoch": 0.22446738414232376, + "ewc_loss": 3.769993782043457e-06, + "grad_norm": 2.374798536300659, + "learning_rate": 7.478038067349926e-07, + "loss": 1.1539, + "mean_token_accuracy": 0.6647859811782837, + "num_tokens": 52620905.0, + "step": 2044 + }, + { + "epoch": 0.2245772018449374, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.684352159500122, + "learning_rate": 7.481698389458272e-07, + "loss": 1.0161, + "mean_token_accuracy": 0.7024258375167847, + "num_tokens": 52647382.0, + "step": 2045 + }, + { + "epoch": 0.22468701954755108, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.6850528717041016, + "learning_rate": 7.485358711566617e-07, + "loss": 1.0595, + "mean_token_accuracy": 0.6855864524841309, + "num_tokens": 52670035.0, + "step": 2046 + }, + { + "epoch": 0.22479683725016472, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.1827328205108643, + "learning_rate": 7.489019033674962e-07, + "loss": 1.035, + "mean_token_accuracy": 0.6912655234336853, + "num_tokens": 52702091.0, + "step": 2047 + }, + { + "epoch": 0.2249066549527784, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.657688617706299, + "learning_rate": 7.492679355783309e-07, + "loss": 1.0042, + "mean_token_accuracy": 0.699860155582428, + "num_tokens": 52723476.0, + "step": 2048 + }, + { + "epoch": 0.22501647265539204, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.3651068210601807, + "learning_rate": 7.496339677891654e-07, + "loss": 0.9557, + "mean_token_accuracy": 0.7130042314529419, + "num_tokens": 52750250.0, + "step": 2049 + }, + { + "epoch": 0.22512629035800572, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.520693302154541, + "learning_rate": 7.5e-07, + "loss": 1.0353, + "mean_token_accuracy": 0.6932119131088257, + "num_tokens": 52774913.0, + "step": 2050 + }, + { + "epoch": 0.22523610806061936, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.642148494720459, + "learning_rate": 7.503660322108345e-07, + "loss": 1.1025, + "mean_token_accuracy": 0.6803027391433716, + "num_tokens": 52799061.0, + "step": 2051 + }, + { + "epoch": 0.22534592576323303, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.4130425453186035, + "learning_rate": 7.50732064421669e-07, + "loss": 1.0584, + "mean_token_accuracy": 0.6874498128890991, + "num_tokens": 52827682.0, + "step": 2052 + }, + { + "epoch": 0.22545574346584668, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.5864920616149902, + "learning_rate": 7.510980966325037e-07, + "loss": 1.1042, + "mean_token_accuracy": 0.6744880080223083, + "num_tokens": 52854004.0, + "step": 2053 + }, + { + "epoch": 0.22556556116846035, + "ewc_loss": 3.7848949432373047e-06, + "grad_norm": 2.829897880554199, + "learning_rate": 7.514641288433382e-07, + "loss": 1.0759, + "mean_token_accuracy": 0.6945507526397705, + "num_tokens": 52875334.0, + "step": 2054 + }, + { + "epoch": 0.22567537887107403, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.543952703475952, + "learning_rate": 7.518301610541727e-07, + "loss": 1.0649, + "mean_token_accuracy": 0.6920729875564575, + "num_tokens": 52901072.0, + "step": 2055 + }, + { + "epoch": 0.22578519657368767, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.4124794006347656, + "learning_rate": 7.521961932650073e-07, + "loss": 1.0656, + "mean_token_accuracy": 0.6911738514900208, + "num_tokens": 52929140.0, + "step": 2056 + }, + { + "epoch": 0.22589501427630135, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.6192352771759033, + "learning_rate": 7.525622254758418e-07, + "loss": 1.0043, + "mean_token_accuracy": 0.7014753818511963, + "num_tokens": 52953279.0, + "step": 2057 + }, + { + "epoch": 0.226004831978915, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.3424909114837646, + "learning_rate": 7.529282576866763e-07, + "loss": 1.0983, + "mean_token_accuracy": 0.6753689646720886, + "num_tokens": 52985489.0, + "step": 2058 + }, + { + "epoch": 0.22611464968152867, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.7919485569000244, + "learning_rate": 7.53294289897511e-07, + "loss": 1.021, + "mean_token_accuracy": 0.694530725479126, + "num_tokens": 53008432.0, + "step": 2059 + }, + { + "epoch": 0.2262244673841423, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.723968982696533, + "learning_rate": 7.536603221083455e-07, + "loss": 1.0312, + "mean_token_accuracy": 0.7055860757827759, + "num_tokens": 53030196.0, + "step": 2060 + }, + { + "epoch": 0.226334285086756, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.4389662742614746, + "learning_rate": 7.5402635431918e-07, + "loss": 1.03, + "mean_token_accuracy": 0.7023429274559021, + "num_tokens": 53059442.0, + "step": 2061 + }, + { + "epoch": 0.22644410278936963, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.760772466659546, + "learning_rate": 7.543923865300146e-07, + "loss": 1.0165, + "mean_token_accuracy": 0.7008119225502014, + "num_tokens": 53081527.0, + "step": 2062 + }, + { + "epoch": 0.2265539204919833, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.3627288341522217, + "learning_rate": 7.547584187408491e-07, + "loss": 1.0709, + "mean_token_accuracy": 0.6888355016708374, + "num_tokens": 53109281.0, + "step": 2063 + }, + { + "epoch": 0.22666373819459698, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.5402722358703613, + "learning_rate": 7.551244509516837e-07, + "loss": 1.0259, + "mean_token_accuracy": 0.7041587829589844, + "num_tokens": 53134575.0, + "step": 2064 + }, + { + "epoch": 0.22677355589721063, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.5388238430023193, + "learning_rate": 7.554904831625183e-07, + "loss": 1.028, + "mean_token_accuracy": 0.696446418762207, + "num_tokens": 53163685.0, + "step": 2065 + }, + { + "epoch": 0.2268833735998243, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.4106738567352295, + "learning_rate": 7.558565153733528e-07, + "loss": 1.0597, + "mean_token_accuracy": 0.6916667222976685, + "num_tokens": 53193254.0, + "step": 2066 + }, + { + "epoch": 0.22699319130243795, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.7002477645874023, + "learning_rate": 7.562225475841874e-07, + "loss": 1.0521, + "mean_token_accuracy": 0.6904926300048828, + "num_tokens": 53216268.0, + "step": 2067 + }, + { + "epoch": 0.22710300900505162, + "ewc_loss": 3.7997961044311523e-06, + "grad_norm": 2.545008897781372, + "learning_rate": 7.565885797950219e-07, + "loss": 0.954, + "mean_token_accuracy": 0.713193416595459, + "num_tokens": 53239331.0, + "step": 2068 + }, + { + "epoch": 0.22721282670766527, + "ewc_loss": 3.814697265625e-06, + "grad_norm": 2.4006333351135254, + "learning_rate": 7.569546120058565e-07, + "loss": 1.0881, + "mean_token_accuracy": 0.689787745475769, + "num_tokens": 53269366.0, + "step": 2069 + }, + { + "epoch": 0.22732264441027894, + "ewc_loss": 3.814697265625e-06, + "grad_norm": 2.672851324081421, + "learning_rate": 7.573206442166911e-07, + "loss": 1.0467, + "mean_token_accuracy": 0.6927837133407593, + "num_tokens": 53293551.0, + "step": 2070 + }, + { + "epoch": 0.22743246211289259, + "ewc_loss": 3.814697265625e-06, + "grad_norm": 2.658640146255493, + "learning_rate": 7.576866764275256e-07, + "loss": 0.9942, + "mean_token_accuracy": 0.6991073489189148, + "num_tokens": 53317407.0, + "step": 2071 + }, + { + "epoch": 0.22754227981550626, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.4152722358703613, + "learning_rate": 7.580527086383601e-07, + "loss": 1.1313, + "mean_token_accuracy": 0.68117356300354, + "num_tokens": 53349129.0, + "step": 2072 + }, + { + "epoch": 0.22765209751811993, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.7161295413970947, + "learning_rate": 7.584187408491947e-07, + "loss": 1.0208, + "mean_token_accuracy": 0.6973593235015869, + "num_tokens": 53373807.0, + "step": 2073 + }, + { + "epoch": 0.22776191522073358, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.5784196853637695, + "learning_rate": 7.587847730600292e-07, + "loss": 1.1305, + "mean_token_accuracy": 0.6720976829528809, + "num_tokens": 53398993.0, + "step": 2074 + }, + { + "epoch": 0.22787173292334725, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.338343620300293, + "learning_rate": 7.591508052708638e-07, + "loss": 1.1404, + "mean_token_accuracy": 0.6718421578407288, + "num_tokens": 53430064.0, + "step": 2075 + }, + { + "epoch": 0.2279815506259609, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.3026819229125977, + "learning_rate": 7.595168374816984e-07, + "loss": 1.0153, + "mean_token_accuracy": 0.6915358901023865, + "num_tokens": 53459080.0, + "step": 2076 + }, + { + "epoch": 0.22809136832857457, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.3798398971557617, + "learning_rate": 7.598828696925329e-07, + "loss": 1.1357, + "mean_token_accuracy": 0.6719564199447632, + "num_tokens": 53488249.0, + "step": 2077 + }, + { + "epoch": 0.22820118603118822, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.5071210861206055, + "learning_rate": 7.602489019033674e-07, + "loss": 0.9191, + "mean_token_accuracy": 0.7244435548782349, + "num_tokens": 53512224.0, + "step": 2078 + }, + { + "epoch": 0.2283110037338019, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.5358245372772217, + "learning_rate": 7.60614934114202e-07, + "loss": 1.0705, + "mean_token_accuracy": 0.6842559576034546, + "num_tokens": 53539216.0, + "step": 2079 + }, + { + "epoch": 0.22842082143641554, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.947173595428467, + "learning_rate": 7.609809663250366e-07, + "loss": 0.9845, + "mean_token_accuracy": 0.7056114077568054, + "num_tokens": 53558834.0, + "step": 2080 + }, + { + "epoch": 0.2285306391390292, + "ewc_loss": 3.844499588012695e-06, + "grad_norm": 2.424934148788452, + "learning_rate": 7.613469985358711e-07, + "loss": 1.0553, + "mean_token_accuracy": 0.6890606880187988, + "num_tokens": 53587249.0, + "step": 2081 + }, + { + "epoch": 0.2286404568416429, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.669804096221924, + "learning_rate": 7.617130307467057e-07, + "loss": 1.0165, + "mean_token_accuracy": 0.699536144733429, + "num_tokens": 53610048.0, + "step": 2082 + }, + { + "epoch": 0.22875027454425653, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.328026056289673, + "learning_rate": 7.620790629575402e-07, + "loss": 1.0367, + "mean_token_accuracy": 0.6970611810684204, + "num_tokens": 53639322.0, + "step": 2083 + }, + { + "epoch": 0.2288600922468702, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.9218027591705322, + "learning_rate": 7.624450951683748e-07, + "loss": 0.9868, + "mean_token_accuracy": 0.7075603008270264, + "num_tokens": 53661912.0, + "step": 2084 + }, + { + "epoch": 0.22896990994948385, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.52447772026062, + "learning_rate": 7.628111273792094e-07, + "loss": 1.0284, + "mean_token_accuracy": 0.7007512450218201, + "num_tokens": 53692100.0, + "step": 2085 + }, + { + "epoch": 0.22907972765209753, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.403886318206787, + "learning_rate": 7.631771595900439e-07, + "loss": 1.0898, + "mean_token_accuracy": 0.6814317107200623, + "num_tokens": 53721317.0, + "step": 2086 + }, + { + "epoch": 0.22918954535471117, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.928422451019287, + "learning_rate": 7.635431918008785e-07, + "loss": 1.0963, + "mean_token_accuracy": 0.687997579574585, + "num_tokens": 53745152.0, + "step": 2087 + }, + { + "epoch": 0.22929936305732485, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.8917882442474365, + "learning_rate": 7.63909224011713e-07, + "loss": 0.9982, + "mean_token_accuracy": 0.708737313747406, + "num_tokens": 53766720.0, + "step": 2088 + }, + { + "epoch": 0.2294091807599385, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.466581106185913, + "learning_rate": 7.642752562225475e-07, + "loss": 1.0192, + "mean_token_accuracy": 0.6946593523025513, + "num_tokens": 53794504.0, + "step": 2089 + }, + { + "epoch": 0.22951899846255217, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.56551456451416, + "learning_rate": 7.646412884333821e-07, + "loss": 1.0519, + "mean_token_accuracy": 0.6953766942024231, + "num_tokens": 53820018.0, + "step": 2090 + }, + { + "epoch": 0.2296288161651658, + "ewc_loss": 3.874301910400391e-06, + "grad_norm": 2.2766196727752686, + "learning_rate": 7.650073206442167e-07, + "loss": 1.0796, + "mean_token_accuracy": 0.6862699389457703, + "num_tokens": 53851435.0, + "step": 2091 + }, + { + "epoch": 0.22973863386777948, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.7405638694763184, + "learning_rate": 7.653733528550512e-07, + "loss": 1.0088, + "mean_token_accuracy": 0.6997009515762329, + "num_tokens": 53873241.0, + "step": 2092 + }, + { + "epoch": 0.22984845157039316, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.297576665878296, + "learning_rate": 7.657393850658858e-07, + "loss": 1.0813, + "mean_token_accuracy": 0.68709397315979, + "num_tokens": 53903030.0, + "step": 2093 + }, + { + "epoch": 0.2299582692730068, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.8545432090759277, + "learning_rate": 7.661054172767203e-07, + "loss": 1.0476, + "mean_token_accuracy": 0.6931898593902588, + "num_tokens": 53925218.0, + "step": 2094 + }, + { + "epoch": 0.23006808697562048, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.709419012069702, + "learning_rate": 7.664714494875548e-07, + "loss": 1.1061, + "mean_token_accuracy": 0.6806537508964539, + "num_tokens": 53948843.0, + "step": 2095 + }, + { + "epoch": 0.23017790467823412, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.4033913612365723, + "learning_rate": 7.668374816983895e-07, + "loss": 1.0022, + "mean_token_accuracy": 0.7010000944137573, + "num_tokens": 53977112.0, + "step": 2096 + }, + { + "epoch": 0.2302877223808478, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.5196611881256104, + "learning_rate": 7.67203513909224e-07, + "loss": 1.0332, + "mean_token_accuracy": 0.7015990018844604, + "num_tokens": 54002810.0, + "step": 2097 + }, + { + "epoch": 0.23039754008346144, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.5408120155334473, + "learning_rate": 7.675695461200585e-07, + "loss": 1.0294, + "mean_token_accuracy": 0.7034483551979065, + "num_tokens": 54029780.0, + "step": 2098 + }, + { + "epoch": 0.23050735778607512, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.6439931392669678, + "learning_rate": 7.679355783308931e-07, + "loss": 1.1042, + "mean_token_accuracy": 0.6846734285354614, + "num_tokens": 54054899.0, + "step": 2099 + }, + { + "epoch": 0.23061717548868876, + "ewc_loss": 3.904104232788086e-06, + "grad_norm": 2.335097312927246, + "learning_rate": 7.683016105417276e-07, + "loss": 1.1266, + "mean_token_accuracy": 0.680724024772644, + "num_tokens": 54084137.0, + "step": 2100 + }, + { + "epoch": 0.23072699319130244, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.8290536403656006, + "learning_rate": 7.686676427525623e-07, + "loss": 0.9881, + "mean_token_accuracy": 0.7132527828216553, + "num_tokens": 54108823.0, + "step": 2101 + }, + { + "epoch": 0.2308368108939161, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.733304500579834, + "learning_rate": 7.690336749633968e-07, + "loss": 1.0084, + "mean_token_accuracy": 0.6990115642547607, + "num_tokens": 54131262.0, + "step": 2102 + }, + { + "epoch": 0.23094662859652976, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 3.038031816482544, + "learning_rate": 7.693997071742313e-07, + "loss": 1.0406, + "mean_token_accuracy": 0.690645694732666, + "num_tokens": 54154640.0, + "step": 2103 + }, + { + "epoch": 0.23105644629914343, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.500718355178833, + "learning_rate": 7.697657393850659e-07, + "loss": 1.0212, + "mean_token_accuracy": 0.6979070901870728, + "num_tokens": 54185330.0, + "step": 2104 + }, + { + "epoch": 0.23116626400175708, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.823601722717285, + "learning_rate": 7.701317715959004e-07, + "loss": 0.9919, + "mean_token_accuracy": 0.7069255113601685, + "num_tokens": 54206266.0, + "step": 2105 + }, + { + "epoch": 0.23127608170437075, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.474346160888672, + "learning_rate": 7.704978038067349e-07, + "loss": 1.1023, + "mean_token_accuracy": 0.6746277809143066, + "num_tokens": 54234830.0, + "step": 2106 + }, + { + "epoch": 0.2313858994069844, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.6164276599884033, + "learning_rate": 7.708638360175696e-07, + "loss": 0.9145, + "mean_token_accuracy": 0.7228670120239258, + "num_tokens": 54258518.0, + "step": 2107 + }, + { + "epoch": 0.23149571710959807, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.4848923683166504, + "learning_rate": 7.712298682284041e-07, + "loss": 1.1067, + "mean_token_accuracy": 0.677658200263977, + "num_tokens": 54285139.0, + "step": 2108 + }, + { + "epoch": 0.23160553481221172, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.7324142456054688, + "learning_rate": 7.715959004392386e-07, + "loss": 1.0112, + "mean_token_accuracy": 0.701431930065155, + "num_tokens": 54308341.0, + "step": 2109 + }, + { + "epoch": 0.2317153525148254, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.5708866119384766, + "learning_rate": 7.719619326500732e-07, + "loss": 1.1618, + "mean_token_accuracy": 0.6635240316390991, + "num_tokens": 54338009.0, + "step": 2110 + }, + { + "epoch": 0.23182517021743906, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.7550408840179443, + "learning_rate": 7.723279648609077e-07, + "loss": 1.1107, + "mean_token_accuracy": 0.6780598163604736, + "num_tokens": 54361156.0, + "step": 2111 + }, + { + "epoch": 0.2319349879200527, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.2313196659088135, + "learning_rate": 7.726939970717423e-07, + "loss": 0.9194, + "mean_token_accuracy": 0.7258898019790649, + "num_tokens": 54390645.0, + "step": 2112 + }, + { + "epoch": 0.23204480562266638, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.6689281463623047, + "learning_rate": 7.730600292825769e-07, + "loss": 1.0209, + "mean_token_accuracy": 0.6931312084197998, + "num_tokens": 54412841.0, + "step": 2113 + }, + { + "epoch": 0.23215462332528003, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.4300549030303955, + "learning_rate": 7.734260614934114e-07, + "loss": 0.9974, + "mean_token_accuracy": 0.7036845088005066, + "num_tokens": 54440731.0, + "step": 2114 + }, + { + "epoch": 0.2322644410278937, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.7774147987365723, + "learning_rate": 7.737920937042459e-07, + "loss": 1.0389, + "mean_token_accuracy": 0.6925199627876282, + "num_tokens": 54460803.0, + "step": 2115 + }, + { + "epoch": 0.23237425873050735, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.689892530441284, + "learning_rate": 7.741581259150805e-07, + "loss": 1.0133, + "mean_token_accuracy": 0.7054815292358398, + "num_tokens": 54483265.0, + "step": 2116 + }, + { + "epoch": 0.23248407643312102, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.5373122692108154, + "learning_rate": 7.745241581259151e-07, + "loss": 1.0129, + "mean_token_accuracy": 0.7005741596221924, + "num_tokens": 54509995.0, + "step": 2117 + }, + { + "epoch": 0.23259389413573467, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.787745237350464, + "learning_rate": 7.748901903367497e-07, + "loss": 1.0107, + "mean_token_accuracy": 0.6984561085700989, + "num_tokens": 54533911.0, + "step": 2118 + }, + { + "epoch": 0.23270371183834834, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.8495590686798096, + "learning_rate": 7.752562225475842e-07, + "loss": 1.0173, + "mean_token_accuracy": 0.698468804359436, + "num_tokens": 54555628.0, + "step": 2119 + }, + { + "epoch": 0.23281352954096202, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.432809352874756, + "learning_rate": 7.756222547584187e-07, + "loss": 1.0716, + "mean_token_accuracy": 0.689744234085083, + "num_tokens": 54584663.0, + "step": 2120 + }, + { + "epoch": 0.23292334724357566, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.906867027282715, + "learning_rate": 7.759882869692533e-07, + "loss": 1.0461, + "mean_token_accuracy": 0.6987154483795166, + "num_tokens": 54606332.0, + "step": 2121 + }, + { + "epoch": 0.23303316494618934, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.352980375289917, + "learning_rate": 7.763543191800878e-07, + "loss": 1.129, + "mean_token_accuracy": 0.675078272819519, + "num_tokens": 54639008.0, + "step": 2122 + }, + { + "epoch": 0.23314298264880298, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.3687262535095215, + "learning_rate": 7.767203513909224e-07, + "loss": 1.069, + "mean_token_accuracy": 0.6852549314498901, + "num_tokens": 54668941.0, + "step": 2123 + }, + { + "epoch": 0.23325280035141666, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.7505600452423096, + "learning_rate": 7.77086383601757e-07, + "loss": 1.0187, + "mean_token_accuracy": 0.6973575353622437, + "num_tokens": 54691348.0, + "step": 2124 + }, + { + "epoch": 0.2333626180540303, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.493795394897461, + "learning_rate": 7.774524158125915e-07, + "loss": 1.01, + "mean_token_accuracy": 0.7036486864089966, + "num_tokens": 54715840.0, + "step": 2125 + }, + { + "epoch": 0.23347243575664398, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.505089044570923, + "learning_rate": 7.77818448023426e-07, + "loss": 1.1367, + "mean_token_accuracy": 0.6661624908447266, + "num_tokens": 54742261.0, + "step": 2126 + }, + { + "epoch": 0.23358225345925762, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.319537878036499, + "learning_rate": 7.781844802342606e-07, + "loss": 1.1194, + "mean_token_accuracy": 0.6725835204124451, + "num_tokens": 54775001.0, + "step": 2127 + }, + { + "epoch": 0.2336920711618713, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.930530071258545, + "learning_rate": 7.785505124450952e-07, + "loss": 1.0947, + "mean_token_accuracy": 0.6809040904045105, + "num_tokens": 54800287.0, + "step": 2128 + }, + { + "epoch": 0.23380188886448494, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.480271816253662, + "learning_rate": 7.789165446559297e-07, + "loss": 1.1158, + "mean_token_accuracy": 0.6785691976547241, + "num_tokens": 54827212.0, + "step": 2129 + }, + { + "epoch": 0.23391170656709862, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.5597782135009766, + "learning_rate": 7.792825768667643e-07, + "loss": 1.1886, + "mean_token_accuracy": 0.6564847230911255, + "num_tokens": 54855265.0, + "step": 2130 + }, + { + "epoch": 0.2340215242697123, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.660048007965088, + "learning_rate": 7.796486090775988e-07, + "loss": 1.0995, + "mean_token_accuracy": 0.6750619411468506, + "num_tokens": 54880558.0, + "step": 2131 + }, + { + "epoch": 0.23413134197232593, + "ewc_loss": 3.933906555175781e-06, + "grad_norm": 2.500839948654175, + "learning_rate": 7.800146412884333e-07, + "loss": 1.1312, + "mean_token_accuracy": 0.6737644672393799, + "num_tokens": 54908760.0, + "step": 2132 + }, + { + "epoch": 0.2342411596749396, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.717454433441162, + "learning_rate": 7.80380673499268e-07, + "loss": 0.9478, + "mean_token_accuracy": 0.7131830453872681, + "num_tokens": 54927989.0, + "step": 2133 + }, + { + "epoch": 0.23435097737755325, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.4042863845825195, + "learning_rate": 7.807467057101025e-07, + "loss": 1.076, + "mean_token_accuracy": 0.6855273246765137, + "num_tokens": 54955289.0, + "step": 2134 + }, + { + "epoch": 0.23446079508016693, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.9027225971221924, + "learning_rate": 7.811127379209371e-07, + "loss": 1.0048, + "mean_token_accuracy": 0.6980301141738892, + "num_tokens": 54981124.0, + "step": 2135 + }, + { + "epoch": 0.23457061278278057, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.4377505779266357, + "learning_rate": 7.814787701317716e-07, + "loss": 1.096, + "mean_token_accuracy": 0.6804940700531006, + "num_tokens": 55009205.0, + "step": 2136 + }, + { + "epoch": 0.23468043048539425, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.5162060260772705, + "learning_rate": 7.818448023426061e-07, + "loss": 1.1501, + "mean_token_accuracy": 0.6666420698165894, + "num_tokens": 55036118.0, + "step": 2137 + }, + { + "epoch": 0.2347902481880079, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.549703359603882, + "learning_rate": 7.822108345534407e-07, + "loss": 0.984, + "mean_token_accuracy": 0.7038772106170654, + "num_tokens": 55060467.0, + "step": 2138 + }, + { + "epoch": 0.23490006589062157, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.5346155166625977, + "learning_rate": 7.825768667642753e-07, + "loss": 1.1248, + "mean_token_accuracy": 0.6713274717330933, + "num_tokens": 55087375.0, + "step": 2139 + }, + { + "epoch": 0.23500988359323524, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.548546075820923, + "learning_rate": 7.829428989751098e-07, + "loss": 1.0571, + "mean_token_accuracy": 0.6943579912185669, + "num_tokens": 55112810.0, + "step": 2140 + }, + { + "epoch": 0.2351197012958489, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 3.2640984058380127, + "learning_rate": 7.833089311859444e-07, + "loss": 1.049, + "mean_token_accuracy": 0.6973517537117004, + "num_tokens": 55134221.0, + "step": 2141 + }, + { + "epoch": 0.23522951899846256, + "ewc_loss": 3.9637088775634766e-06, + "grad_norm": 2.484246015548706, + "learning_rate": 7.836749633967789e-07, + "loss": 1.0579, + "mean_token_accuracy": 0.6943751573562622, + "num_tokens": 55161282.0, + "step": 2142 + }, + { + "epoch": 0.2353393367010762, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.5602970123291016, + "learning_rate": 7.840409956076133e-07, + "loss": 1.0793, + "mean_token_accuracy": 0.687018632888794, + "num_tokens": 55185586.0, + "step": 2143 + }, + { + "epoch": 0.23544915440368988, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.4436240196228027, + "learning_rate": 7.844070278184481e-07, + "loss": 1.182, + "mean_token_accuracy": 0.6624596118927002, + "num_tokens": 55214146.0, + "step": 2144 + }, + { + "epoch": 0.23555897210630353, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.8248705863952637, + "learning_rate": 7.847730600292826e-07, + "loss": 1.0297, + "mean_token_accuracy": 0.6914834976196289, + "num_tokens": 55234676.0, + "step": 2145 + }, + { + "epoch": 0.2356687898089172, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.625760793685913, + "learning_rate": 7.85139092240117e-07, + "loss": 0.9641, + "mean_token_accuracy": 0.7098789215087891, + "num_tokens": 55257852.0, + "step": 2146 + }, + { + "epoch": 0.23577860751153085, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 3.01554012298584, + "learning_rate": 7.855051244509517e-07, + "loss": 1.0463, + "mean_token_accuracy": 0.6926167607307434, + "num_tokens": 55280103.0, + "step": 2147 + }, + { + "epoch": 0.23588842521414452, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.450650215148926, + "learning_rate": 7.858711566617861e-07, + "loss": 1.0803, + "mean_token_accuracy": 0.6786541938781738, + "num_tokens": 55306689.0, + "step": 2148 + }, + { + "epoch": 0.2359982429167582, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.4812068939208984, + "learning_rate": 7.862371888726209e-07, + "loss": 1.1174, + "mean_token_accuracy": 0.6751710176467896, + "num_tokens": 55335557.0, + "step": 2149 + }, + { + "epoch": 0.23610806061937184, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.5487568378448486, + "learning_rate": 7.866032210834554e-07, + "loss": 1.067, + "mean_token_accuracy": 0.6861927509307861, + "num_tokens": 55361347.0, + "step": 2150 + }, + { + "epoch": 0.2362178783219855, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.6093146800994873, + "learning_rate": 7.869692532942898e-07, + "loss": 1.0578, + "mean_token_accuracy": 0.6868267059326172, + "num_tokens": 55386252.0, + "step": 2151 + }, + { + "epoch": 0.23632769602459916, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.604315996170044, + "learning_rate": 7.873352855051244e-07, + "loss": 1.0853, + "mean_token_accuracy": 0.6912798881530762, + "num_tokens": 55407899.0, + "step": 2152 + }, + { + "epoch": 0.23643751372721283, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.552558660507202, + "learning_rate": 7.877013177159589e-07, + "loss": 0.8965, + "mean_token_accuracy": 0.7287512421607971, + "num_tokens": 55430617.0, + "step": 2153 + }, + { + "epoch": 0.23654733142982648, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 3.2049455642700195, + "learning_rate": 7.880673499267934e-07, + "loss": 0.8851, + "mean_token_accuracy": 0.7308313250541687, + "num_tokens": 55449753.0, + "step": 2154 + }, + { + "epoch": 0.23665714913244015, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.555739164352417, + "learning_rate": 7.884333821376281e-07, + "loss": 1.0697, + "mean_token_accuracy": 0.6871763467788696, + "num_tokens": 55477552.0, + "step": 2155 + }, + { + "epoch": 0.2367669668350538, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.658311128616333, + "learning_rate": 7.887994143484626e-07, + "loss": 1.0196, + "mean_token_accuracy": 0.6997444033622742, + "num_tokens": 55500986.0, + "step": 2156 + }, + { + "epoch": 0.23687678453766747, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.399174928665161, + "learning_rate": 7.891654465592971e-07, + "loss": 1.0838, + "mean_token_accuracy": 0.6853177547454834, + "num_tokens": 55533557.0, + "step": 2157 + }, + { + "epoch": 0.23698660224028115, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.5446815490722656, + "learning_rate": 7.895314787701317e-07, + "loss": 1.0792, + "mean_token_accuracy": 0.6801092624664307, + "num_tokens": 55562766.0, + "step": 2158 + }, + { + "epoch": 0.2370964199428948, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.4435086250305176, + "learning_rate": 7.898975109809662e-07, + "loss": 1.133, + "mean_token_accuracy": 0.6690609455108643, + "num_tokens": 55590781.0, + "step": 2159 + }, + { + "epoch": 0.23720623764550847, + "ewc_loss": 3.993511199951172e-06, + "grad_norm": 2.6157286167144775, + "learning_rate": 7.902635431918008e-07, + "loss": 1.0272, + "mean_token_accuracy": 0.6972590088844299, + "num_tokens": 55615755.0, + "step": 2160 + }, + { + "epoch": 0.2373160553481221, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.608306646347046, + "learning_rate": 7.906295754026354e-07, + "loss": 1.0187, + "mean_token_accuracy": 0.6997998952865601, + "num_tokens": 55639958.0, + "step": 2161 + }, + { + "epoch": 0.2374258730507358, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.8055529594421387, + "learning_rate": 7.909956076134699e-07, + "loss": 1.0937, + "mean_token_accuracy": 0.6800352334976196, + "num_tokens": 55661257.0, + "step": 2162 + }, + { + "epoch": 0.23753569075334943, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.429582118988037, + "learning_rate": 7.913616398243044e-07, + "loss": 0.9708, + "mean_token_accuracy": 0.719382107257843, + "num_tokens": 55687147.0, + "step": 2163 + }, + { + "epoch": 0.2376455084559631, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.390573501586914, + "learning_rate": 7.91727672035139e-07, + "loss": 1.083, + "mean_token_accuracy": 0.6822113990783691, + "num_tokens": 55718756.0, + "step": 2164 + }, + { + "epoch": 0.23775532615857675, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.8450260162353516, + "learning_rate": 7.920937042459736e-07, + "loss": 1.1302, + "mean_token_accuracy": 0.672884464263916, + "num_tokens": 55745692.0, + "step": 2165 + }, + { + "epoch": 0.23786514386119043, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.4790375232696533, + "learning_rate": 7.924597364568082e-07, + "loss": 1.0231, + "mean_token_accuracy": 0.7040637731552124, + "num_tokens": 55771595.0, + "step": 2166 + }, + { + "epoch": 0.23797496156380407, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.5145010948181152, + "learning_rate": 7.928257686676427e-07, + "loss": 1.0557, + "mean_token_accuracy": 0.6881241798400879, + "num_tokens": 55795574.0, + "step": 2167 + }, + { + "epoch": 0.23808477926641775, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.5027995109558105, + "learning_rate": 7.931918008784772e-07, + "loss": 1.0915, + "mean_token_accuracy": 0.6832192540168762, + "num_tokens": 55823116.0, + "step": 2168 + }, + { + "epoch": 0.23819459696903142, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.5238354206085205, + "learning_rate": 7.935578330893118e-07, + "loss": 1.0613, + "mean_token_accuracy": 0.6860300302505493, + "num_tokens": 55849908.0, + "step": 2169 + }, + { + "epoch": 0.23830441467164507, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.342397451400757, + "learning_rate": 7.939238653001463e-07, + "loss": 1.0869, + "mean_token_accuracy": 0.6815965175628662, + "num_tokens": 55879307.0, + "step": 2170 + }, + { + "epoch": 0.23841423237425874, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.555544376373291, + "learning_rate": 7.942898975109809e-07, + "loss": 1.1263, + "mean_token_accuracy": 0.6740983724594116, + "num_tokens": 55907879.0, + "step": 2171 + }, + { + "epoch": 0.23852405007687238, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.8274166584014893, + "learning_rate": 7.946559297218155e-07, + "loss": 1.1383, + "mean_token_accuracy": 0.6759110689163208, + "num_tokens": 55930065.0, + "step": 2172 + }, + { + "epoch": 0.23863386777948606, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.6906232833862305, + "learning_rate": 7.9502196193265e-07, + "loss": 1.0377, + "mean_token_accuracy": 0.6943973302841187, + "num_tokens": 55954500.0, + "step": 2173 + }, + { + "epoch": 0.2387436854820997, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 20.533218383789062, + "learning_rate": 7.953879941434845e-07, + "loss": 0.9648, + "mean_token_accuracy": 0.7123969197273254, + "num_tokens": 55980294.0, + "step": 2174 + }, + { + "epoch": 0.23885350318471338, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.456688404083252, + "learning_rate": 7.957540263543191e-07, + "loss": 1.0506, + "mean_token_accuracy": 0.6941609382629395, + "num_tokens": 56012046.0, + "step": 2175 + }, + { + "epoch": 0.23896332088732702, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.5397017002105713, + "learning_rate": 7.961200585651537e-07, + "loss": 1.0377, + "mean_token_accuracy": 0.6947847008705139, + "num_tokens": 56039172.0, + "step": 2176 + }, + { + "epoch": 0.2390731385899407, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.760214328765869, + "learning_rate": 7.964860907759882e-07, + "loss": 1.0314, + "mean_token_accuracy": 0.7001148462295532, + "num_tokens": 56061746.0, + "step": 2177 + }, + { + "epoch": 0.23918295629255437, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 3.044323205947876, + "learning_rate": 7.968521229868228e-07, + "loss": 1.0726, + "mean_token_accuracy": 0.6812504529953003, + "num_tokens": 56086553.0, + "step": 2178 + }, + { + "epoch": 0.23929277399516802, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.3806302547454834, + "learning_rate": 7.972181551976573e-07, + "loss": 1.1288, + "mean_token_accuracy": 0.6730474233627319, + "num_tokens": 56118829.0, + "step": 2179 + }, + { + "epoch": 0.2394025916977817, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.5140726566314697, + "learning_rate": 7.975841874084918e-07, + "loss": 1.0752, + "mean_token_accuracy": 0.6843611598014832, + "num_tokens": 56145714.0, + "step": 2180 + }, + { + "epoch": 0.23951240940039534, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.6218454837799072, + "learning_rate": 7.979502196193265e-07, + "loss": 1.0741, + "mean_token_accuracy": 0.6819871664047241, + "num_tokens": 56170868.0, + "step": 2181 + }, + { + "epoch": 0.239622227103009, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.588491439819336, + "learning_rate": 7.98316251830161e-07, + "loss": 1.0483, + "mean_token_accuracy": 0.6930230259895325, + "num_tokens": 56195851.0, + "step": 2182 + }, + { + "epoch": 0.23973204480562266, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.404386520385742, + "learning_rate": 7.986822840409956e-07, + "loss": 1.1066, + "mean_token_accuracy": 0.6861020922660828, + "num_tokens": 56227161.0, + "step": 2183 + }, + { + "epoch": 0.23984186250823633, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.821218490600586, + "learning_rate": 7.990483162518301e-07, + "loss": 1.1163, + "mean_token_accuracy": 0.6769621968269348, + "num_tokens": 56249842.0, + "step": 2184 + }, + { + "epoch": 0.23995168021084998, + "ewc_loss": 4.023313522338867e-06, + "grad_norm": 2.603367328643799, + "learning_rate": 7.994143484626646e-07, + "loss": 1.0718, + "mean_token_accuracy": 0.6869655847549438, + "num_tokens": 56276810.0, + "step": 2185 + }, + { + "epoch": 0.24006149791346365, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.5443782806396484, + "learning_rate": 7.997803806734992e-07, + "loss": 1.021, + "mean_token_accuracy": 0.6966805458068848, + "num_tokens": 56301375.0, + "step": 2186 + }, + { + "epoch": 0.24017131561607732, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.4500343799591064, + "learning_rate": 8.001464128843338e-07, + "loss": 1.0794, + "mean_token_accuracy": 0.6827617883682251, + "num_tokens": 56327061.0, + "step": 2187 + }, + { + "epoch": 0.24028113331869097, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.4147090911865234, + "learning_rate": 8.005124450951683e-07, + "loss": 1.0485, + "mean_token_accuracy": 0.6919220089912415, + "num_tokens": 56357185.0, + "step": 2188 + }, + { + "epoch": 0.24039095102130464, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 16.71766471862793, + "learning_rate": 8.008784773060029e-07, + "loss": 1.0727, + "mean_token_accuracy": 0.6844469308853149, + "num_tokens": 56380411.0, + "step": 2189 + }, + { + "epoch": 0.2405007687239183, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.938382148742676, + "learning_rate": 8.012445095168374e-07, + "loss": 1.0303, + "mean_token_accuracy": 0.702454686164856, + "num_tokens": 56403539.0, + "step": 2190 + }, + { + "epoch": 0.24061058642653196, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.4575324058532715, + "learning_rate": 8.016105417276719e-07, + "loss": 1.0364, + "mean_token_accuracy": 0.7004891037940979, + "num_tokens": 56430207.0, + "step": 2191 + }, + { + "epoch": 0.2407204041291456, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.3157265186309814, + "learning_rate": 8.019765739385066e-07, + "loss": 1.1125, + "mean_token_accuracy": 0.6722331047058105, + "num_tokens": 56460386.0, + "step": 2192 + }, + { + "epoch": 0.24083022183175928, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.528055429458618, + "learning_rate": 8.023426061493411e-07, + "loss": 0.9616, + "mean_token_accuracy": 0.7158318758010864, + "num_tokens": 56487140.0, + "step": 2193 + }, + { + "epoch": 0.24094003953437293, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.7452681064605713, + "learning_rate": 8.027086383601756e-07, + "loss": 1.1336, + "mean_token_accuracy": 0.6688160300254822, + "num_tokens": 56508781.0, + "step": 2194 + }, + { + "epoch": 0.2410498572369866, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.6280760765075684, + "learning_rate": 8.030746705710102e-07, + "loss": 1.0686, + "mean_token_accuracy": 0.6843709349632263, + "num_tokens": 56532472.0, + "step": 2195 + }, + { + "epoch": 0.24115967493960028, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.9046056270599365, + "learning_rate": 8.034407027818447e-07, + "loss": 1.1033, + "mean_token_accuracy": 0.681634247303009, + "num_tokens": 56567185.0, + "step": 2196 + }, + { + "epoch": 0.24126949264221392, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.3027591705322266, + "learning_rate": 8.038067349926793e-07, + "loss": 1.0117, + "mean_token_accuracy": 0.7057594060897827, + "num_tokens": 56596186.0, + "step": 2197 + }, + { + "epoch": 0.2413793103448276, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.4872045516967773, + "learning_rate": 8.041727672035139e-07, + "loss": 0.996, + "mean_token_accuracy": 0.7003580927848816, + "num_tokens": 56621026.0, + "step": 2198 + }, + { + "epoch": 0.24148912804744124, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 3.5095579624176025, + "learning_rate": 8.045387994143484e-07, + "loss": 1.0232, + "mean_token_accuracy": 0.6979854702949524, + "num_tokens": 56641149.0, + "step": 2199 + }, + { + "epoch": 0.24159894575005492, + "ewc_loss": 4.082918167114258e-06, + "grad_norm": 2.6549766063690186, + "learning_rate": 8.04904831625183e-07, + "loss": 1.0117, + "mean_token_accuracy": 0.7007013559341431, + "num_tokens": 56663743.0, + "step": 2200 + }, + { + "epoch": 0.24170876345266856, + "ewc_loss": 4.082918167114258e-06, + "grad_norm": 2.518038034439087, + "learning_rate": 8.052708638360175e-07, + "loss": 0.9939, + "mean_token_accuracy": 0.7147936820983887, + "num_tokens": 56690487.0, + "step": 2201 + }, + { + "epoch": 0.24181858115528224, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.571585178375244, + "learning_rate": 8.05636896046852e-07, + "loss": 1.0143, + "mean_token_accuracy": 0.69435054063797, + "num_tokens": 56714174.0, + "step": 2202 + }, + { + "epoch": 0.24192839885789588, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.4670634269714355, + "learning_rate": 8.060029282576867e-07, + "loss": 1.1186, + "mean_token_accuracy": 0.6737057566642761, + "num_tokens": 56743225.0, + "step": 2203 + }, + { + "epoch": 0.24203821656050956, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.760713815689087, + "learning_rate": 8.063689604685212e-07, + "loss": 1.0806, + "mean_token_accuracy": 0.6796917915344238, + "num_tokens": 56769588.0, + "step": 2204 + }, + { + "epoch": 0.2421480342631232, + "ewc_loss": 4.0531158447265625e-06, + "grad_norm": 2.5421302318573, + "learning_rate": 8.067349926793557e-07, + "loss": 1.0444, + "mean_token_accuracy": 0.6936311721801758, + "num_tokens": 56793483.0, + "step": 2205 + }, + { + "epoch": 0.24225785196573688, + "ewc_loss": 4.082918167114258e-06, + "grad_norm": 2.3393807411193848, + "learning_rate": 8.071010248901903e-07, + "loss": 1.0934, + "mean_token_accuracy": 0.688024640083313, + "num_tokens": 56821789.0, + "step": 2206 + }, + { + "epoch": 0.24236766966835055, + "ewc_loss": 4.082918167114258e-06, + "grad_norm": 2.9979236125946045, + "learning_rate": 8.074670571010248e-07, + "loss": 0.9857, + "mean_token_accuracy": 0.7051205635070801, + "num_tokens": 56847788.0, + "step": 2207 + }, + { + "epoch": 0.2424774873709642, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.512741804122925, + "learning_rate": 8.078330893118594e-07, + "loss": 0.9963, + "mean_token_accuracy": 0.7072113156318665, + "num_tokens": 56874038.0, + "step": 2208 + }, + { + "epoch": 0.24258730507357787, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.7833175659179688, + "learning_rate": 8.08199121522694e-07, + "loss": 1.0691, + "mean_token_accuracy": 0.687990128993988, + "num_tokens": 56895943.0, + "step": 2209 + }, + { + "epoch": 0.24269712277619152, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.6296424865722656, + "learning_rate": 8.085651537335285e-07, + "loss": 1.0116, + "mean_token_accuracy": 0.6986459493637085, + "num_tokens": 56920459.0, + "step": 2210 + }, + { + "epoch": 0.2428069404788052, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.608971118927002, + "learning_rate": 8.08931185944363e-07, + "loss": 1.1411, + "mean_token_accuracy": 0.6705354452133179, + "num_tokens": 56947307.0, + "step": 2211 + }, + { + "epoch": 0.24291675818141883, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.461988687515259, + "learning_rate": 8.092972181551976e-07, + "loss": 1.0677, + "mean_token_accuracy": 0.6878577470779419, + "num_tokens": 56974344.0, + "step": 2212 + }, + { + "epoch": 0.2430265758840325, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.389188051223755, + "learning_rate": 8.096632503660322e-07, + "loss": 1.1896, + "mean_token_accuracy": 0.6652182340621948, + "num_tokens": 57005194.0, + "step": 2213 + }, + { + "epoch": 0.24313639358664615, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.680166482925415, + "learning_rate": 8.100292825768668e-07, + "loss": 0.9859, + "mean_token_accuracy": 0.7009116411209106, + "num_tokens": 57027359.0, + "step": 2214 + }, + { + "epoch": 0.24324621128925983, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.71753191947937, + "learning_rate": 8.103953147877013e-07, + "loss": 1.0214, + "mean_token_accuracy": 0.6981129050254822, + "num_tokens": 57050898.0, + "step": 2215 + }, + { + "epoch": 0.2433560289918735, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.6806082725524902, + "learning_rate": 8.107613469985358e-07, + "loss": 0.9933, + "mean_token_accuracy": 0.7000100612640381, + "num_tokens": 57075058.0, + "step": 2216 + }, + { + "epoch": 0.24346584669448715, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.558060884475708, + "learning_rate": 8.111273792093704e-07, + "loss": 0.9877, + "mean_token_accuracy": 0.7076413631439209, + "num_tokens": 57100386.0, + "step": 2217 + }, + { + "epoch": 0.24357566439710082, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.4697694778442383, + "learning_rate": 8.114934114202049e-07, + "loss": 1.0271, + "mean_token_accuracy": 0.6953341960906982, + "num_tokens": 57126772.0, + "step": 2218 + }, + { + "epoch": 0.24368548209971447, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.5035204887390137, + "learning_rate": 8.118594436310395e-07, + "loss": 1.0737, + "mean_token_accuracy": 0.6880649328231812, + "num_tokens": 57152044.0, + "step": 2219 + }, + { + "epoch": 0.24379529980232814, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.562560796737671, + "learning_rate": 8.122254758418741e-07, + "loss": 1.0752, + "mean_token_accuracy": 0.6894849538803101, + "num_tokens": 57176202.0, + "step": 2220 + }, + { + "epoch": 0.2439051175049418, + "ewc_loss": 4.112720489501953e-06, + "grad_norm": 2.673097848892212, + "learning_rate": 8.125915080527086e-07, + "loss": 1.005, + "mean_token_accuracy": 0.701345682144165, + "num_tokens": 57198532.0, + "step": 2221 + }, + { + "epoch": 0.24401493520755546, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.4830055236816406, + "learning_rate": 8.129575402635431e-07, + "loss": 1.059, + "mean_token_accuracy": 0.6891387104988098, + "num_tokens": 57228410.0, + "step": 2222 + }, + { + "epoch": 0.2441247529101691, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.4391088485717773, + "learning_rate": 8.133235724743777e-07, + "loss": 1.0085, + "mean_token_accuracy": 0.7006601691246033, + "num_tokens": 57257091.0, + "step": 2223 + }, + { + "epoch": 0.24423457061278278, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.51345157623291, + "learning_rate": 8.136896046852123e-07, + "loss": 1.0189, + "mean_token_accuracy": 0.6958520412445068, + "num_tokens": 57282667.0, + "step": 2224 + }, + { + "epoch": 0.24434438831539645, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 16.871469497680664, + "learning_rate": 8.140556368960468e-07, + "loss": 0.969, + "mean_token_accuracy": 0.7099272012710571, + "num_tokens": 57301885.0, + "step": 2225 + }, + { + "epoch": 0.2444542060180101, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.9584689140319824, + "learning_rate": 8.144216691068814e-07, + "loss": 1.0568, + "mean_token_accuracy": 0.7000499963760376, + "num_tokens": 57322637.0, + "step": 2226 + }, + { + "epoch": 0.24456402372062377, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.823904514312744, + "learning_rate": 8.147877013177159e-07, + "loss": 1.0957, + "mean_token_accuracy": 0.685822606086731, + "num_tokens": 57346366.0, + "step": 2227 + }, + { + "epoch": 0.24467384142323742, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.704507350921631, + "learning_rate": 8.151537335285504e-07, + "loss": 1.0779, + "mean_token_accuracy": 0.6929766535758972, + "num_tokens": 57373168.0, + "step": 2228 + }, + { + "epoch": 0.2447836591258511, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.5312297344207764, + "learning_rate": 8.155197657393851e-07, + "loss": 1.0239, + "mean_token_accuracy": 0.694189190864563, + "num_tokens": 57402621.0, + "step": 2229 + }, + { + "epoch": 0.24489347682846474, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.7636220455169678, + "learning_rate": 8.158857979502196e-07, + "loss": 1.0761, + "mean_token_accuracy": 0.6948471069335938, + "num_tokens": 57426230.0, + "step": 2230 + }, + { + "epoch": 0.2450032945310784, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 3.5020711421966553, + "learning_rate": 8.162518301610542e-07, + "loss": 1.0211, + "mean_token_accuracy": 0.7011415958404541, + "num_tokens": 57453725.0, + "step": 2231 + }, + { + "epoch": 0.24511311223369206, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.7167553901672363, + "learning_rate": 8.166178623718887e-07, + "loss": 1.0845, + "mean_token_accuracy": 0.6822607517242432, + "num_tokens": 57484306.0, + "step": 2232 + }, + { + "epoch": 0.24522292993630573, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.635434627532959, + "learning_rate": 8.169838945827232e-07, + "loss": 1.0693, + "mean_token_accuracy": 0.68916916847229, + "num_tokens": 57510966.0, + "step": 2233 + }, + { + "epoch": 0.2453327476389194, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.398151397705078, + "learning_rate": 8.173499267935578e-07, + "loss": 1.051, + "mean_token_accuracy": 0.6889032125473022, + "num_tokens": 57538533.0, + "step": 2234 + }, + { + "epoch": 0.24544256534153305, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.5964465141296387, + "learning_rate": 8.177159590043924e-07, + "loss": 1.0421, + "mean_token_accuracy": 0.6900222897529602, + "num_tokens": 57564275.0, + "step": 2235 + }, + { + "epoch": 0.24555238304414673, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.3666815757751465, + "learning_rate": 8.180819912152269e-07, + "loss": 1.032, + "mean_token_accuracy": 0.6984153985977173, + "num_tokens": 57594837.0, + "step": 2236 + }, + { + "epoch": 0.24566220074676037, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.915759325027466, + "learning_rate": 8.184480234260615e-07, + "loss": 1.0056, + "mean_token_accuracy": 0.7120077610015869, + "num_tokens": 57620478.0, + "step": 2237 + }, + { + "epoch": 0.24577201844937405, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.551933765411377, + "learning_rate": 8.18814055636896e-07, + "loss": 1.1056, + "mean_token_accuracy": 0.681877613067627, + "num_tokens": 57647752.0, + "step": 2238 + }, + { + "epoch": 0.2458818361519877, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.6732001304626465, + "learning_rate": 8.191800878477305e-07, + "loss": 1.025, + "mean_token_accuracy": 0.6988062858581543, + "num_tokens": 57672591.0, + "step": 2239 + }, + { + "epoch": 0.24599165385460137, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.6198270320892334, + "learning_rate": 8.195461200585652e-07, + "loss": 1.0134, + "mean_token_accuracy": 0.6991239190101624, + "num_tokens": 57696420.0, + "step": 2240 + }, + { + "epoch": 0.246101471557215, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.694305658340454, + "learning_rate": 8.199121522693997e-07, + "loss": 1.0435, + "mean_token_accuracy": 0.6897556781768799, + "num_tokens": 57720386.0, + "step": 2241 + }, + { + "epoch": 0.2462112892598287, + "ewc_loss": 4.1425228118896484e-06, + "grad_norm": 2.7678170204162598, + "learning_rate": 8.202781844802342e-07, + "loss": 1.0611, + "mean_token_accuracy": 0.6925514936447144, + "num_tokens": 57746173.0, + "step": 2242 + }, + { + "epoch": 0.24632110696244233, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.6709280014038086, + "learning_rate": 8.206442166910688e-07, + "loss": 1.0587, + "mean_token_accuracy": 0.6889448165893555, + "num_tokens": 57770085.0, + "step": 2243 + }, + { + "epoch": 0.246430924665056, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.2470645904541016, + "learning_rate": 8.210102489019033e-07, + "loss": 0.9699, + "mean_token_accuracy": 0.7086175680160522, + "num_tokens": 57799026.0, + "step": 2244 + }, + { + "epoch": 0.24654074236766968, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.5172934532165527, + "learning_rate": 8.213762811127379e-07, + "loss": 1.0464, + "mean_token_accuracy": 0.6930407285690308, + "num_tokens": 57824626.0, + "step": 2245 + }, + { + "epoch": 0.24665056007028333, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.767514705657959, + "learning_rate": 8.217423133235725e-07, + "loss": 1.0656, + "mean_token_accuracy": 0.6805975437164307, + "num_tokens": 57847648.0, + "step": 2246 + }, + { + "epoch": 0.246760377772897, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.6139447689056396, + "learning_rate": 8.22108345534407e-07, + "loss": 1.0618, + "mean_token_accuracy": 0.6855865716934204, + "num_tokens": 57870239.0, + "step": 2247 + }, + { + "epoch": 0.24687019547551065, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.4275288581848145, + "learning_rate": 8.224743777452416e-07, + "loss": 1.0371, + "mean_token_accuracy": 0.6926683187484741, + "num_tokens": 57899702.0, + "step": 2248 + }, + { + "epoch": 0.24698001317812432, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 3.006992816925049, + "learning_rate": 8.228404099560761e-07, + "loss": 1.0154, + "mean_token_accuracy": 0.7026956081390381, + "num_tokens": 57918400.0, + "step": 2249 + }, + { + "epoch": 0.24708983088073797, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.6427643299102783, + "learning_rate": 8.232064421669106e-07, + "loss": 1.0439, + "mean_token_accuracy": 0.6922581195831299, + "num_tokens": 57940924.0, + "step": 2250 + }, + { + "epoch": 0.24719964858335164, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.9365787506103516, + "learning_rate": 8.235724743777453e-07, + "loss": 0.9678, + "mean_token_accuracy": 0.7174720764160156, + "num_tokens": 57959917.0, + "step": 2251 + }, + { + "epoch": 0.24730946628596528, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.349916696548462, + "learning_rate": 8.239385065885798e-07, + "loss": 1.0918, + "mean_token_accuracy": 0.6884022951126099, + "num_tokens": 57989935.0, + "step": 2252 + }, + { + "epoch": 0.24741928398857896, + "ewc_loss": 4.202127456665039e-06, + "grad_norm": 3.3641679286956787, + "learning_rate": 8.243045387994143e-07, + "loss": 1.0008, + "mean_token_accuracy": 0.7118756175041199, + "num_tokens": 58009541.0, + "step": 2253 + }, + { + "epoch": 0.24752910169119263, + "ewc_loss": 4.202127456665039e-06, + "grad_norm": 2.420602560043335, + "learning_rate": 8.246705710102489e-07, + "loss": 1.0394, + "mean_token_accuracy": 0.6900689601898193, + "num_tokens": 58042087.0, + "step": 2254 + }, + { + "epoch": 0.24763891939380628, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.4920899868011475, + "learning_rate": 8.250366032210834e-07, + "loss": 0.9903, + "mean_token_accuracy": 0.712607741355896, + "num_tokens": 58070764.0, + "step": 2255 + }, + { + "epoch": 0.24774873709641995, + "ewc_loss": 4.172325134277344e-06, + "grad_norm": 2.8987345695495605, + "learning_rate": 8.25402635431918e-07, + "loss": 1.007, + "mean_token_accuracy": 0.7029192447662354, + "num_tokens": 58090387.0, + "step": 2256 + }, + { + "epoch": 0.2478585547990336, + "ewc_loss": 4.202127456665039e-06, + "grad_norm": 2.6692636013031006, + "learning_rate": 8.257686676427526e-07, + "loss": 1.0946, + "mean_token_accuracy": 0.6849846839904785, + "num_tokens": 58113755.0, + "step": 2257 + }, + { + "epoch": 0.24796837250164727, + "ewc_loss": 4.202127456665039e-06, + "grad_norm": 2.7233378887176514, + "learning_rate": 8.261346998535871e-07, + "loss": 0.9723, + "mean_token_accuracy": 0.7150555849075317, + "num_tokens": 58135562.0, + "step": 2258 + }, + { + "epoch": 0.24807819020426092, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.529207229614258, + "learning_rate": 8.265007320644216e-07, + "loss": 1.0828, + "mean_token_accuracy": 0.6795433759689331, + "num_tokens": 58161079.0, + "step": 2259 + }, + { + "epoch": 0.2481880079068746, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.531738042831421, + "learning_rate": 8.268667642752562e-07, + "loss": 0.9758, + "mean_token_accuracy": 0.7158408761024475, + "num_tokens": 58186104.0, + "step": 2260 + }, + { + "epoch": 0.24829782560948824, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.3006625175476074, + "learning_rate": 8.272327964860908e-07, + "loss": 1.114, + "mean_token_accuracy": 0.6764702200889587, + "num_tokens": 58221477.0, + "step": 2261 + }, + { + "epoch": 0.2484076433121019, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.8540842533111572, + "learning_rate": 8.275988286969253e-07, + "loss": 1.0983, + "mean_token_accuracy": 0.6813507676124573, + "num_tokens": 58247976.0, + "step": 2262 + }, + { + "epoch": 0.24851746101471558, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.860126495361328, + "learning_rate": 8.279648609077599e-07, + "loss": 0.9842, + "mean_token_accuracy": 0.7106120586395264, + "num_tokens": 58267331.0, + "step": 2263 + }, + { + "epoch": 0.24862727871732923, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.660446882247925, + "learning_rate": 8.283308931185944e-07, + "loss": 1.0748, + "mean_token_accuracy": 0.6817488670349121, + "num_tokens": 58292582.0, + "step": 2264 + }, + { + "epoch": 0.2487370964199429, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.5737152099609375, + "learning_rate": 8.28696925329429e-07, + "loss": 1.1625, + "mean_token_accuracy": 0.6585350632667542, + "num_tokens": 58318884.0, + "step": 2265 + }, + { + "epoch": 0.24884691412255655, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.6169309616088867, + "learning_rate": 8.290629575402635e-07, + "loss": 1.0483, + "mean_token_accuracy": 0.6901689767837524, + "num_tokens": 58344826.0, + "step": 2266 + }, + { + "epoch": 0.24895673182517022, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.6619062423706055, + "learning_rate": 8.294289897510981e-07, + "loss": 1.054, + "mean_token_accuracy": 0.6888686418533325, + "num_tokens": 58368859.0, + "step": 2267 + }, + { + "epoch": 0.24906654952778387, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.4532470703125, + "learning_rate": 8.297950219619327e-07, + "loss": 1.0237, + "mean_token_accuracy": 0.7012497186660767, + "num_tokens": 58397377.0, + "step": 2268 + }, + { + "epoch": 0.24917636723039754, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.7573676109313965, + "learning_rate": 8.301610541727672e-07, + "loss": 0.9172, + "mean_token_accuracy": 0.7286214828491211, + "num_tokens": 58419411.0, + "step": 2269 + }, + { + "epoch": 0.2492861849330112, + "ewc_loss": 4.231929779052734e-06, + "grad_norm": 2.516880989074707, + "learning_rate": 8.305270863836017e-07, + "loss": 1.0037, + "mean_token_accuracy": 0.6991444826126099, + "num_tokens": 58447935.0, + "step": 2270 + }, + { + "epoch": 0.24939600263562486, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.951261043548584, + "learning_rate": 8.308931185944363e-07, + "loss": 1.0255, + "mean_token_accuracy": 0.6981399059295654, + "num_tokens": 58474769.0, + "step": 2271 + }, + { + "epoch": 0.24950582033823854, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.696732759475708, + "learning_rate": 8.312591508052709e-07, + "loss": 1.1677, + "mean_token_accuracy": 0.659904420375824, + "num_tokens": 58504789.0, + "step": 2272 + }, + { + "epoch": 0.24961563804085218, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.9922330379486084, + "learning_rate": 8.316251830161054e-07, + "loss": 1.0081, + "mean_token_accuracy": 0.6981110572814941, + "num_tokens": 58526700.0, + "step": 2273 + }, + { + "epoch": 0.24972545574346586, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.458139419555664, + "learning_rate": 8.3199121522694e-07, + "loss": 1.059, + "mean_token_accuracy": 0.6827600002288818, + "num_tokens": 58555555.0, + "step": 2274 + }, + { + "epoch": 0.2498352734460795, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.6198697090148926, + "learning_rate": 8.323572474377745e-07, + "loss": 0.8954, + "mean_token_accuracy": 0.7330234050750732, + "num_tokens": 58577667.0, + "step": 2275 + }, + { + "epoch": 0.24994509114869318, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.490997314453125, + "learning_rate": 8.32723279648609e-07, + "loss": 1.0685, + "mean_token_accuracy": 0.6919194459915161, + "num_tokens": 58606031.0, + "step": 2276 + }, + { + "epoch": 0.2500549088513068, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.6001241207122803, + "learning_rate": 8.330893118594437e-07, + "loss": 1.0043, + "mean_token_accuracy": 0.7044879794120789, + "num_tokens": 58631967.0, + "step": 2277 + }, + { + "epoch": 0.25016472655392047, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.555943250656128, + "learning_rate": 8.334553440702782e-07, + "loss": 1.074, + "mean_token_accuracy": 0.6833335161209106, + "num_tokens": 58662429.0, + "step": 2278 + }, + { + "epoch": 0.25027454425653417, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.5045018196105957, + "learning_rate": 8.338213762811127e-07, + "loss": 1.1024, + "mean_token_accuracy": 0.6717073917388916, + "num_tokens": 58689786.0, + "step": 2279 + }, + { + "epoch": 0.2503843619591478, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.700852394104004, + "learning_rate": 8.341874084919473e-07, + "loss": 1.0066, + "mean_token_accuracy": 0.7011431455612183, + "num_tokens": 58711364.0, + "step": 2280 + }, + { + "epoch": 0.25049417966176146, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.6003427505493164, + "learning_rate": 8.345534407027818e-07, + "loss": 1.0736, + "mean_token_accuracy": 0.6853751540184021, + "num_tokens": 58737840.0, + "step": 2281 + }, + { + "epoch": 0.25060399736437516, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.5648679733276367, + "learning_rate": 8.349194729136164e-07, + "loss": 1.1269, + "mean_token_accuracy": 0.6764023303985596, + "num_tokens": 58766062.0, + "step": 2282 + }, + { + "epoch": 0.2507138150669888, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.6480560302734375, + "learning_rate": 8.35285505124451e-07, + "loss": 1.0649, + "mean_token_accuracy": 0.6848391890525818, + "num_tokens": 58792625.0, + "step": 2283 + }, + { + "epoch": 0.25082363276960246, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.8288683891296387, + "learning_rate": 8.356515373352855e-07, + "loss": 1.1861, + "mean_token_accuracy": 0.6606481075286865, + "num_tokens": 58817547.0, + "step": 2284 + }, + { + "epoch": 0.2509334504722161, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.6011970043182373, + "learning_rate": 8.360175695461201e-07, + "loss": 1.0632, + "mean_token_accuracy": 0.6876702904701233, + "num_tokens": 58845967.0, + "step": 2285 + }, + { + "epoch": 0.2510432681748298, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.380491018295288, + "learning_rate": 8.363836017569546e-07, + "loss": 1.0519, + "mean_token_accuracy": 0.6944705247879028, + "num_tokens": 58876861.0, + "step": 2286 + }, + { + "epoch": 0.25115308587744345, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.677917957305908, + "learning_rate": 8.36749633967789e-07, + "loss": 1.0738, + "mean_token_accuracy": 0.6878318786621094, + "num_tokens": 58900848.0, + "step": 2287 + }, + { + "epoch": 0.2512629035800571, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.7010881900787354, + "learning_rate": 8.371156661786238e-07, + "loss": 1.0088, + "mean_token_accuracy": 0.7022932767868042, + "num_tokens": 58922927.0, + "step": 2288 + }, + { + "epoch": 0.25137272128267074, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.848773956298828, + "learning_rate": 8.374816983894583e-07, + "loss": 1.0492, + "mean_token_accuracy": 0.6871328353881836, + "num_tokens": 58951123.0, + "step": 2289 + }, + { + "epoch": 0.25148253898528444, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.8133132457733154, + "learning_rate": 8.378477306002928e-07, + "loss": 1.0797, + "mean_token_accuracy": 0.6849638223648071, + "num_tokens": 58970496.0, + "step": 2290 + }, + { + "epoch": 0.2515923566878981, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.5018975734710693, + "learning_rate": 8.382137628111274e-07, + "loss": 1.0734, + "mean_token_accuracy": 0.6878300309181213, + "num_tokens": 58994989.0, + "step": 2291 + }, + { + "epoch": 0.25170217439051173, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.809682607650757, + "learning_rate": 8.385797950219619e-07, + "loss": 0.9962, + "mean_token_accuracy": 0.7033964395523071, + "num_tokens": 59015235.0, + "step": 2292 + }, + { + "epoch": 0.25181199209312544, + "ewc_loss": 4.26173210144043e-06, + "grad_norm": 2.7871594429016113, + "learning_rate": 8.389458272327965e-07, + "loss": 1.0046, + "mean_token_accuracy": 0.6982036828994751, + "num_tokens": 59038421.0, + "step": 2293 + }, + { + "epoch": 0.2519218097957391, + "ewc_loss": 4.32133674621582e-06, + "grad_norm": 2.6871700286865234, + "learning_rate": 8.393118594436311e-07, + "loss": 1.0969, + "mean_token_accuracy": 0.6731409430503845, + "num_tokens": 59064410.0, + "step": 2294 + }, + { + "epoch": 0.25203162749835273, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 3.063930034637451, + "learning_rate": 8.396778916544656e-07, + "loss": 1.0683, + "mean_token_accuracy": 0.6853833198547363, + "num_tokens": 59087245.0, + "step": 2295 + }, + { + "epoch": 0.2521414452009664, + "ewc_loss": 4.32133674621582e-06, + "grad_norm": 2.5939323902130127, + "learning_rate": 8.400439238653e-07, + "loss": 1.0558, + "mean_token_accuracy": 0.6877503395080566, + "num_tokens": 59111988.0, + "step": 2296 + }, + { + "epoch": 0.2522512629035801, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.7967095375061035, + "learning_rate": 8.404099560761346e-07, + "loss": 1.0557, + "mean_token_accuracy": 0.6968759298324585, + "num_tokens": 59137371.0, + "step": 2297 + }, + { + "epoch": 0.2523610806061937, + "ewc_loss": 4.32133674621582e-06, + "grad_norm": 2.5875706672668457, + "learning_rate": 8.407759882869691e-07, + "loss": 1.0489, + "mean_token_accuracy": 0.6936418414115906, + "num_tokens": 59163108.0, + "step": 2298 + }, + { + "epoch": 0.25247089830880737, + "ewc_loss": 4.32133674621582e-06, + "grad_norm": 2.5273523330688477, + "learning_rate": 8.411420204978039e-07, + "loss": 1.106, + "mean_token_accuracy": 0.6766666769981384, + "num_tokens": 59191057.0, + "step": 2299 + }, + { + "epoch": 0.252580716011421, + "ewc_loss": 4.32133674621582e-06, + "grad_norm": 2.5379884243011475, + "learning_rate": 8.415080527086383e-07, + "loss": 1.1144, + "mean_token_accuracy": 0.6753814220428467, + "num_tokens": 59220690.0, + "step": 2300 + }, + { + "epoch": 0.2526905337140347, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.4928483963012695, + "learning_rate": 8.418740849194728e-07, + "loss": 1.0166, + "mean_token_accuracy": 0.702964186668396, + "num_tokens": 59247310.0, + "step": 2301 + }, + { + "epoch": 0.25280035141664836, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.8762574195861816, + "learning_rate": 8.422401171303074e-07, + "loss": 1.034, + "mean_token_accuracy": 0.7000671029090881, + "num_tokens": 59271797.0, + "step": 2302 + }, + { + "epoch": 0.252910169119262, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.568849563598633, + "learning_rate": 8.426061493411419e-07, + "loss": 1.0242, + "mean_token_accuracy": 0.6989122033119202, + "num_tokens": 59299996.0, + "step": 2303 + }, + { + "epoch": 0.2530199868218757, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6627023220062256, + "learning_rate": 8.429721815519765e-07, + "loss": 1.1049, + "mean_token_accuracy": 0.6786330938339233, + "num_tokens": 59326639.0, + "step": 2304 + }, + { + "epoch": 0.25312980452448935, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6010353565216064, + "learning_rate": 8.433382137628111e-07, + "loss": 1.0378, + "mean_token_accuracy": 0.7015960216522217, + "num_tokens": 59353176.0, + "step": 2305 + }, + { + "epoch": 0.253239622227103, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.360691547393799, + "learning_rate": 8.437042459736456e-07, + "loss": 1.0437, + "mean_token_accuracy": 0.697099506855011, + "num_tokens": 59381590.0, + "step": 2306 + }, + { + "epoch": 0.25334943992971665, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.3181381225585938, + "learning_rate": 8.440702781844801e-07, + "loss": 1.0066, + "mean_token_accuracy": 0.7035209536552429, + "num_tokens": 59411696.0, + "step": 2307 + }, + { + "epoch": 0.25345925763233035, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6970293521881104, + "learning_rate": 8.444363103953147e-07, + "loss": 1.0792, + "mean_token_accuracy": 0.6779310703277588, + "num_tokens": 59438691.0, + "step": 2308 + }, + { + "epoch": 0.253569075334944, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.3372790813446045, + "learning_rate": 8.448023426061493e-07, + "loss": 1.0364, + "mean_token_accuracy": 0.6943007707595825, + "num_tokens": 59467636.0, + "step": 2309 + }, + { + "epoch": 0.25367889303755764, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.389991521835327, + "learning_rate": 8.451683748169838e-07, + "loss": 1.0719, + "mean_token_accuracy": 0.6888909339904785, + "num_tokens": 59496548.0, + "step": 2310 + }, + { + "epoch": 0.25378871074017134, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6792125701904297, + "learning_rate": 8.455344070278184e-07, + "loss": 1.0696, + "mean_token_accuracy": 0.6876989006996155, + "num_tokens": 59518194.0, + "step": 2311 + }, + { + "epoch": 0.253898528442785, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6976144313812256, + "learning_rate": 8.459004392386529e-07, + "loss": 0.9396, + "mean_token_accuracy": 0.7209866642951965, + "num_tokens": 59540317.0, + "step": 2312 + }, + { + "epoch": 0.25400834614539863, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.5235166549682617, + "learning_rate": 8.462664714494874e-07, + "loss": 1.1041, + "mean_token_accuracy": 0.6722278594970703, + "num_tokens": 59569101.0, + "step": 2313 + }, + { + "epoch": 0.2541181638480123, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6719934940338135, + "learning_rate": 8.46632503660322e-07, + "loss": 1.0954, + "mean_token_accuracy": 0.6779186129570007, + "num_tokens": 59597709.0, + "step": 2314 + }, + { + "epoch": 0.254227981550626, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.5675108432769775, + "learning_rate": 8.469985358711566e-07, + "loss": 1.0473, + "mean_token_accuracy": 0.6908532977104187, + "num_tokens": 59626571.0, + "step": 2315 + }, + { + "epoch": 0.2543377992532396, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.7182021141052246, + "learning_rate": 8.473645680819912e-07, + "loss": 1.0487, + "mean_token_accuracy": 0.6841617822647095, + "num_tokens": 59653507.0, + "step": 2316 + }, + { + "epoch": 0.2544476169558533, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.509284019470215, + "learning_rate": 8.477306002928257e-07, + "loss": 1.0135, + "mean_token_accuracy": 0.7089526057243347, + "num_tokens": 59682791.0, + "step": 2317 + }, + { + "epoch": 0.2545574346584669, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.650089979171753, + "learning_rate": 8.480966325036602e-07, + "loss": 1.1462, + "mean_token_accuracy": 0.6680509448051453, + "num_tokens": 59708535.0, + "step": 2318 + }, + { + "epoch": 0.2546672523610806, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 3.0750296115875244, + "learning_rate": 8.484626647144948e-07, + "loss": 1.0155, + "mean_token_accuracy": 0.7017975449562073, + "num_tokens": 59728949.0, + "step": 2319 + }, + { + "epoch": 0.25477707006369427, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.8082942962646484, + "learning_rate": 8.488286969253294e-07, + "loss": 1.0075, + "mean_token_accuracy": 0.7067909240722656, + "num_tokens": 59750730.0, + "step": 2320 + }, + { + "epoch": 0.2548868877663079, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6039092540740967, + "learning_rate": 8.491947291361639e-07, + "loss": 1.0383, + "mean_token_accuracy": 0.6954730153083801, + "num_tokens": 59773449.0, + "step": 2321 + }, + { + "epoch": 0.2549967054689216, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.6018800735473633, + "learning_rate": 8.495607613469985e-07, + "loss": 1.0804, + "mean_token_accuracy": 0.6817448735237122, + "num_tokens": 59797987.0, + "step": 2322 + }, + { + "epoch": 0.25510652317153526, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.8698890209198, + "learning_rate": 8.49926793557833e-07, + "loss": 0.9987, + "mean_token_accuracy": 0.6990377902984619, + "num_tokens": 59819138.0, + "step": 2323 + }, + { + "epoch": 0.2552163408741489, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.7202749252319336, + "learning_rate": 8.502928257686675e-07, + "loss": 1.095, + "mean_token_accuracy": 0.6877470016479492, + "num_tokens": 59842604.0, + "step": 2324 + }, + { + "epoch": 0.25532615857676255, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.4814891815185547, + "learning_rate": 8.506588579795022e-07, + "loss": 1.0601, + "mean_token_accuracy": 0.6898808479309082, + "num_tokens": 59869413.0, + "step": 2325 + }, + { + "epoch": 0.25543597627937625, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.5998685359954834, + "learning_rate": 8.510248901903367e-07, + "loss": 1.0726, + "mean_token_accuracy": 0.6836833357810974, + "num_tokens": 59896911.0, + "step": 2326 + }, + { + "epoch": 0.2555457939819899, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.7986767292022705, + "learning_rate": 8.513909224011712e-07, + "loss": 1.0013, + "mean_token_accuracy": 0.7029064893722534, + "num_tokens": 59920209.0, + "step": 2327 + }, + { + "epoch": 0.25565561168460355, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.7899281978607178, + "learning_rate": 8.517569546120058e-07, + "loss": 1.0912, + "mean_token_accuracy": 0.680747926235199, + "num_tokens": 59946626.0, + "step": 2328 + }, + { + "epoch": 0.25576542938721725, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.596907615661621, + "learning_rate": 8.521229868228403e-07, + "loss": 1.0859, + "mean_token_accuracy": 0.687433123588562, + "num_tokens": 59975887.0, + "step": 2329 + }, + { + "epoch": 0.2558752470898309, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.529597520828247, + "learning_rate": 8.524890190336748e-07, + "loss": 0.972, + "mean_token_accuracy": 0.7120037078857422, + "num_tokens": 60000691.0, + "step": 2330 + }, + { + "epoch": 0.25598506479244454, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.5631625652313232, + "learning_rate": 8.528550512445095e-07, + "loss": 1.0481, + "mean_token_accuracy": 0.6941435933113098, + "num_tokens": 60025315.0, + "step": 2331 + }, + { + "epoch": 0.2560948824950582, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.5515966415405273, + "learning_rate": 8.53221083455344e-07, + "loss": 1.1197, + "mean_token_accuracy": 0.6857434511184692, + "num_tokens": 60057120.0, + "step": 2332 + }, + { + "epoch": 0.2562047001976719, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.490780830383301, + "learning_rate": 8.535871156661786e-07, + "loss": 1.0542, + "mean_token_accuracy": 0.6975852847099304, + "num_tokens": 60084156.0, + "step": 2333 + }, + { + "epoch": 0.25631451790028553, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.7040984630584717, + "learning_rate": 8.539531478770131e-07, + "loss": 1.1083, + "mean_token_accuracy": 0.67948317527771, + "num_tokens": 60109948.0, + "step": 2334 + }, + { + "epoch": 0.2564243356028992, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.996530532836914, + "learning_rate": 8.543191800878476e-07, + "loss": 0.9678, + "mean_token_accuracy": 0.7071768045425415, + "num_tokens": 60129571.0, + "step": 2335 + }, + { + "epoch": 0.2565341533055128, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.530726909637451, + "learning_rate": 8.546852122986823e-07, + "loss": 1.0864, + "mean_token_accuracy": 0.6822535991668701, + "num_tokens": 60157568.0, + "step": 2336 + }, + { + "epoch": 0.2566439710081265, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.552344799041748, + "learning_rate": 8.550512445095168e-07, + "loss": 1.0311, + "mean_token_accuracy": 0.695458173751831, + "num_tokens": 60185034.0, + "step": 2337 + }, + { + "epoch": 0.25675378871074017, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.208643674850464, + "learning_rate": 8.554172767203513e-07, + "loss": 1.0597, + "mean_token_accuracy": 0.6908515691757202, + "num_tokens": 60218705.0, + "step": 2338 + }, + { + "epoch": 0.2568636064133538, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.4220938682556152, + "learning_rate": 8.557833089311859e-07, + "loss": 0.9687, + "mean_token_accuracy": 0.7153484225273132, + "num_tokens": 60243522.0, + "step": 2339 + }, + { + "epoch": 0.2569734241159675, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.4976022243499756, + "learning_rate": 8.561493411420204e-07, + "loss": 1.0388, + "mean_token_accuracy": 0.6982727646827698, + "num_tokens": 60271606.0, + "step": 2340 + }, + { + "epoch": 0.25708324181858117, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.640367269515991, + "learning_rate": 8.56515373352855e-07, + "loss": 1.0659, + "mean_token_accuracy": 0.6915429830551147, + "num_tokens": 60300332.0, + "step": 2341 + }, + { + "epoch": 0.2571930595211948, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.5110111236572266, + "learning_rate": 8.568814055636896e-07, + "loss": 0.9277, + "mean_token_accuracy": 0.7256565093994141, + "num_tokens": 60328449.0, + "step": 2342 + }, + { + "epoch": 0.25730287722380846, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.483893632888794, + "learning_rate": 8.572474377745241e-07, + "loss": 1.0433, + "mean_token_accuracy": 0.6983076930046082, + "num_tokens": 60354910.0, + "step": 2343 + }, + { + "epoch": 0.25741269492642216, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.634507656097412, + "learning_rate": 8.576134699853586e-07, + "loss": 1.0233, + "mean_token_accuracy": 0.702444314956665, + "num_tokens": 60380063.0, + "step": 2344 + }, + { + "epoch": 0.2575225126290358, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.788355588912964, + "learning_rate": 8.579795021961932e-07, + "loss": 1.0782, + "mean_token_accuracy": 0.6862573623657227, + "num_tokens": 60404706.0, + "step": 2345 + }, + { + "epoch": 0.25763233033164945, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.4455690383911133, + "learning_rate": 8.583455344070277e-07, + "loss": 1.0272, + "mean_token_accuracy": 0.7009724378585815, + "num_tokens": 60433576.0, + "step": 2346 + }, + { + "epoch": 0.2577421480342631, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.49923038482666, + "learning_rate": 8.587115666178624e-07, + "loss": 1.0324, + "mean_token_accuracy": 0.6943727731704712, + "num_tokens": 60458975.0, + "step": 2347 + }, + { + "epoch": 0.2578519657368768, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.2969634532928467, + "learning_rate": 8.590775988286969e-07, + "loss": 1.0469, + "mean_token_accuracy": 0.7032598257064819, + "num_tokens": 60490043.0, + "step": 2348 + }, + { + "epoch": 0.25796178343949044, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 3.5031044483184814, + "learning_rate": 8.594436310395314e-07, + "loss": 1.0664, + "mean_token_accuracy": 0.6861348152160645, + "num_tokens": 60512267.0, + "step": 2349 + }, + { + "epoch": 0.2580716011421041, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.7420499324798584, + "learning_rate": 8.59809663250366e-07, + "loss": 1.0771, + "mean_token_accuracy": 0.687653124332428, + "num_tokens": 60537869.0, + "step": 2350 + }, + { + "epoch": 0.2581814188447178, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.6354970932006836, + "learning_rate": 8.601756954612005e-07, + "loss": 1.0376, + "mean_token_accuracy": 0.6971409320831299, + "num_tokens": 60563603.0, + "step": 2351 + }, + { + "epoch": 0.25829123654733144, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.6789379119873047, + "learning_rate": 8.605417276720351e-07, + "loss": 1.0433, + "mean_token_accuracy": 0.6907778978347778, + "num_tokens": 60587192.0, + "step": 2352 + }, + { + "epoch": 0.2584010542499451, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.6141207218170166, + "learning_rate": 8.609077598828697e-07, + "loss": 0.9937, + "mean_token_accuracy": 0.7019480466842651, + "num_tokens": 60610638.0, + "step": 2353 + }, + { + "epoch": 0.25851087195255873, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.5909311771392822, + "learning_rate": 8.612737920937042e-07, + "loss": 1.1, + "mean_token_accuracy": 0.6862166523933411, + "num_tokens": 60636598.0, + "step": 2354 + }, + { + "epoch": 0.25862068965517243, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.3220536708831787, + "learning_rate": 8.616398243045387e-07, + "loss": 1.0912, + "mean_token_accuracy": 0.6840651631355286, + "num_tokens": 60666496.0, + "step": 2355 + }, + { + "epoch": 0.2587305073577861, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.2623610496520996, + "learning_rate": 8.620058565153733e-07, + "loss": 1.0973, + "mean_token_accuracy": 0.6793379783630371, + "num_tokens": 60696995.0, + "step": 2356 + }, + { + "epoch": 0.2588403250603997, + "ewc_loss": 4.351139068603516e-06, + "grad_norm": 2.994971990585327, + "learning_rate": 8.623718887262079e-07, + "loss": 0.9909, + "mean_token_accuracy": 0.7112579345703125, + "num_tokens": 60717345.0, + "step": 2357 + }, + { + "epoch": 0.2589501427630134, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.5599749088287354, + "learning_rate": 8.627379209370424e-07, + "loss": 1.0874, + "mean_token_accuracy": 0.6916272044181824, + "num_tokens": 60744639.0, + "step": 2358 + }, + { + "epoch": 0.25905996046562707, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.3755452632904053, + "learning_rate": 8.63103953147877e-07, + "loss": 1.1179, + "mean_token_accuracy": 0.6790914535522461, + "num_tokens": 60773675.0, + "step": 2359 + }, + { + "epoch": 0.2591697781682407, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.44952654838562, + "learning_rate": 8.634699853587115e-07, + "loss": 1.0417, + "mean_token_accuracy": 0.6952307224273682, + "num_tokens": 60799408.0, + "step": 2360 + }, + { + "epoch": 0.25927959587085436, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.653761386871338, + "learning_rate": 8.63836017569546e-07, + "loss": 1.0946, + "mean_token_accuracy": 0.682200014591217, + "num_tokens": 60823031.0, + "step": 2361 + }, + { + "epoch": 0.25938941357346806, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.3827168941497803, + "learning_rate": 8.642020497803806e-07, + "loss": 1.0691, + "mean_token_accuracy": 0.6913335919380188, + "num_tokens": 60851807.0, + "step": 2362 + }, + { + "epoch": 0.2594992312760817, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 3.2433271408081055, + "learning_rate": 8.645680819912152e-07, + "loss": 1.0938, + "mean_token_accuracy": 0.6830532550811768, + "num_tokens": 60871805.0, + "step": 2363 + }, + { + "epoch": 0.25960904897869536, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.44512939453125, + "learning_rate": 8.649341142020498e-07, + "loss": 1.1918, + "mean_token_accuracy": 0.6554970741271973, + "num_tokens": 60909599.0, + "step": 2364 + }, + { + "epoch": 0.259718866681309, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.4498140811920166, + "learning_rate": 8.653001464128843e-07, + "loss": 1.1076, + "mean_token_accuracy": 0.6828181743621826, + "num_tokens": 60936870.0, + "step": 2365 + }, + { + "epoch": 0.2598286843839227, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 3.1151294708251953, + "learning_rate": 8.656661786237188e-07, + "loss": 1.04, + "mean_token_accuracy": 0.6964946389198303, + "num_tokens": 60958742.0, + "step": 2366 + }, + { + "epoch": 0.25993850208653635, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.6011343002319336, + "learning_rate": 8.660322108345534e-07, + "loss": 1.0613, + "mean_token_accuracy": 0.6937909722328186, + "num_tokens": 60987581.0, + "step": 2367 + }, + { + "epoch": 0.26004831978915, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 3.6789846420288086, + "learning_rate": 8.66398243045388e-07, + "loss": 1.1008, + "mean_token_accuracy": 0.6795464754104614, + "num_tokens": 61015388.0, + "step": 2368 + }, + { + "epoch": 0.2601581374917637, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.377121925354004, + "learning_rate": 8.667642752562225e-07, + "loss": 1.0895, + "mean_token_accuracy": 0.684398889541626, + "num_tokens": 61046490.0, + "step": 2369 + }, + { + "epoch": 0.26026795519437734, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.4846792221069336, + "learning_rate": 8.671303074670571e-07, + "loss": 1.1388, + "mean_token_accuracy": 0.6706165075302124, + "num_tokens": 61076806.0, + "step": 2370 + }, + { + "epoch": 0.260377772896991, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.301617383956909, + "learning_rate": 8.674963396778916e-07, + "loss": 1.1418, + "mean_token_accuracy": 0.6758430600166321, + "num_tokens": 61109820.0, + "step": 2371 + }, + { + "epoch": 0.26048759059960463, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.6385884284973145, + "learning_rate": 8.678623718887261e-07, + "loss": 1.0234, + "mean_token_accuracy": 0.6966225504875183, + "num_tokens": 61131765.0, + "step": 2372 + }, + { + "epoch": 0.26059740830221834, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.527031898498535, + "learning_rate": 8.682284040995608e-07, + "loss": 1.0701, + "mean_token_accuracy": 0.6977831125259399, + "num_tokens": 61160748.0, + "step": 2373 + }, + { + "epoch": 0.260707226004832, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.399729013442993, + "learning_rate": 8.685944363103953e-07, + "loss": 1.0285, + "mean_token_accuracy": 0.7018247842788696, + "num_tokens": 61189525.0, + "step": 2374 + }, + { + "epoch": 0.26081704370744563, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.457503080368042, + "learning_rate": 8.689604685212298e-07, + "loss": 1.0673, + "mean_token_accuracy": 0.6920165419578552, + "num_tokens": 61217544.0, + "step": 2375 + }, + { + "epoch": 0.2609268614100593, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.4844446182250977, + "learning_rate": 8.693265007320644e-07, + "loss": 1.0451, + "mean_token_accuracy": 0.6917779445648193, + "num_tokens": 61244972.0, + "step": 2376 + }, + { + "epoch": 0.261036679112673, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.6524269580841064, + "learning_rate": 8.696925329428989e-07, + "loss": 1.1143, + "mean_token_accuracy": 0.6697016358375549, + "num_tokens": 61272586.0, + "step": 2377 + }, + { + "epoch": 0.2611464968152866, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.828005790710449, + "learning_rate": 8.700585651537334e-07, + "loss": 0.9228, + "mean_token_accuracy": 0.722931981086731, + "num_tokens": 61294173.0, + "step": 2378 + }, + { + "epoch": 0.26125631451790027, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.5931878089904785, + "learning_rate": 8.704245973645681e-07, + "loss": 1.0732, + "mean_token_accuracy": 0.6869845986366272, + "num_tokens": 61321813.0, + "step": 2379 + }, + { + "epoch": 0.26136613222051397, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.5096991062164307, + "learning_rate": 8.707906295754026e-07, + "loss": 1.1906, + "mean_token_accuracy": 0.656647264957428, + "num_tokens": 61351003.0, + "step": 2380 + }, + { + "epoch": 0.2614759499231276, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.619718551635742, + "learning_rate": 8.711566617862372e-07, + "loss": 1.1119, + "mean_token_accuracy": 0.6808627247810364, + "num_tokens": 61376842.0, + "step": 2381 + }, + { + "epoch": 0.26158576762574126, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 3.360616683959961, + "learning_rate": 8.715226939970717e-07, + "loss": 1.0506, + "mean_token_accuracy": 0.6933543086051941, + "num_tokens": 61409980.0, + "step": 2382 + }, + { + "epoch": 0.2616955853283549, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.2901861667633057, + "learning_rate": 8.718887262079062e-07, + "loss": 1.0932, + "mean_token_accuracy": 0.6865062117576599, + "num_tokens": 61442720.0, + "step": 2383 + }, + { + "epoch": 0.2618054030309686, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.6213266849517822, + "learning_rate": 8.722547584187409e-07, + "loss": 1.026, + "mean_token_accuracy": 0.7004711627960205, + "num_tokens": 61468322.0, + "step": 2384 + }, + { + "epoch": 0.26191522073358225, + "ewc_loss": 4.380941390991211e-06, + "grad_norm": 2.708266496658325, + "learning_rate": 8.726207906295754e-07, + "loss": 1.056, + "mean_token_accuracy": 0.6837921142578125, + "num_tokens": 61490210.0, + "step": 2385 + }, + { + "epoch": 0.2620250384361959, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.7523882389068604, + "learning_rate": 8.729868228404099e-07, + "loss": 1.0547, + "mean_token_accuracy": 0.6882104873657227, + "num_tokens": 61513905.0, + "step": 2386 + }, + { + "epoch": 0.2621348561388096, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.878279447555542, + "learning_rate": 8.733528550512445e-07, + "loss": 1.0437, + "mean_token_accuracy": 0.6901856660842896, + "num_tokens": 61533984.0, + "step": 2387 + }, + { + "epoch": 0.26224467384142325, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.560152053833008, + "learning_rate": 8.73718887262079e-07, + "loss": 1.0098, + "mean_token_accuracy": 0.7065852880477905, + "num_tokens": 61559133.0, + "step": 2388 + }, + { + "epoch": 0.2623544915440369, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.498225212097168, + "learning_rate": 8.740849194729136e-07, + "loss": 1.0285, + "mean_token_accuracy": 0.6998670697212219, + "num_tokens": 61583791.0, + "step": 2389 + }, + { + "epoch": 0.26246430924665054, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.848940849304199, + "learning_rate": 8.744509516837482e-07, + "loss": 1.0078, + "mean_token_accuracy": 0.7046577334403992, + "num_tokens": 61604249.0, + "step": 2390 + }, + { + "epoch": 0.26257412694926424, + "ewc_loss": 4.410743713378906e-06, + "grad_norm": 2.793630361557007, + "learning_rate": 8.748169838945827e-07, + "loss": 1.0161, + "mean_token_accuracy": 0.7034224271774292, + "num_tokens": 61627052.0, + "step": 2391 + }, + { + "epoch": 0.2626839446518779, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.42909836769104, + "learning_rate": 8.751830161054172e-07, + "loss": 1.1169, + "mean_token_accuracy": 0.679540753364563, + "num_tokens": 61655979.0, + "step": 2392 + }, + { + "epoch": 0.26279376235449153, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.486208200454712, + "learning_rate": 8.755490483162518e-07, + "loss": 1.1358, + "mean_token_accuracy": 0.6664661169052124, + "num_tokens": 61684326.0, + "step": 2393 + }, + { + "epoch": 0.2629035800571052, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.454923391342163, + "learning_rate": 8.759150805270863e-07, + "loss": 1.1099, + "mean_token_accuracy": 0.6759946942329407, + "num_tokens": 61712239.0, + "step": 2394 + }, + { + "epoch": 0.2630133977597189, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.4623851776123047, + "learning_rate": 8.76281112737921e-07, + "loss": 1.073, + "mean_token_accuracy": 0.687056303024292, + "num_tokens": 61740619.0, + "step": 2395 + }, + { + "epoch": 0.2631232154623325, + "ewc_loss": 4.4405460357666016e-06, + "grad_norm": 2.5369820594787598, + "learning_rate": 8.766471449487555e-07, + "loss": 1.0927, + "mean_token_accuracy": 0.6790661811828613, + "num_tokens": 61766457.0, + "step": 2396 + }, + { + "epoch": 0.2632330331649462, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.5521974563598633, + "learning_rate": 8.7701317715959e-07, + "loss": 1.0624, + "mean_token_accuracy": 0.6937017440795898, + "num_tokens": 61792109.0, + "step": 2397 + }, + { + "epoch": 0.2633428508675599, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.411578893661499, + "learning_rate": 8.773792093704246e-07, + "loss": 1.0372, + "mean_token_accuracy": 0.6946797370910645, + "num_tokens": 61821867.0, + "step": 2398 + }, + { + "epoch": 0.2634526685701735, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.4173760414123535, + "learning_rate": 8.777452415812591e-07, + "loss": 1.0123, + "mean_token_accuracy": 0.7018159031867981, + "num_tokens": 61851076.0, + "step": 2399 + }, + { + "epoch": 0.26356248627278717, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.4075448513031006, + "learning_rate": 8.781112737920937e-07, + "loss": 1.0121, + "mean_token_accuracy": 0.6978650093078613, + "num_tokens": 61883955.0, + "step": 2400 + }, + { + "epoch": 0.2636723039754008, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.4912986755371094, + "learning_rate": 8.784773060029283e-07, + "loss": 0.9558, + "mean_token_accuracy": 0.7135488986968994, + "num_tokens": 61909733.0, + "step": 2401 + }, + { + "epoch": 0.2637821216780145, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.5372509956359863, + "learning_rate": 8.788433382137628e-07, + "loss": 1.0599, + "mean_token_accuracy": 0.6881915330886841, + "num_tokens": 61937771.0, + "step": 2402 + }, + { + "epoch": 0.26389193938062816, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 3.111485481262207, + "learning_rate": 8.792093704245973e-07, + "loss": 1.031, + "mean_token_accuracy": 0.6969989538192749, + "num_tokens": 61958207.0, + "step": 2403 + }, + { + "epoch": 0.2640017570832418, + "ewc_loss": 4.470348358154297e-06, + "grad_norm": 2.6376168727874756, + "learning_rate": 8.795754026354319e-07, + "loss": 1.0819, + "mean_token_accuracy": 0.6782694458961487, + "num_tokens": 61987806.0, + "step": 2404 + }, + { + "epoch": 0.2641115747858555, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.30295991897583, + "learning_rate": 8.799414348462665e-07, + "loss": 1.1444, + "mean_token_accuracy": 0.6684647798538208, + "num_tokens": 62019461.0, + "step": 2405 + }, + { + "epoch": 0.26422139248846915, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.404524326324463, + "learning_rate": 8.80307467057101e-07, + "loss": 1.1411, + "mean_token_accuracy": 0.6720380783081055, + "num_tokens": 62052648.0, + "step": 2406 + }, + { + "epoch": 0.2643312101910828, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.6222829818725586, + "learning_rate": 8.806734992679356e-07, + "loss": 1.052, + "mean_token_accuracy": 0.6991958022117615, + "num_tokens": 62077698.0, + "step": 2407 + }, + { + "epoch": 0.26444102789369645, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.6765859127044678, + "learning_rate": 8.810395314787701e-07, + "loss": 1.0817, + "mean_token_accuracy": 0.694880485534668, + "num_tokens": 62105774.0, + "step": 2408 + }, + { + "epoch": 0.26455084559631015, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.7843403816223145, + "learning_rate": 8.814055636896046e-07, + "loss": 1.0586, + "mean_token_accuracy": 0.6895279884338379, + "num_tokens": 62128899.0, + "step": 2409 + }, + { + "epoch": 0.2646606632989238, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.7869815826416016, + "learning_rate": 8.817715959004392e-07, + "loss": 1.1011, + "mean_token_accuracy": 0.6763942837715149, + "num_tokens": 62152471.0, + "step": 2410 + }, + { + "epoch": 0.26477048100153744, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.8982481956481934, + "learning_rate": 8.821376281112738e-07, + "loss": 1.0723, + "mean_token_accuracy": 0.6904187202453613, + "num_tokens": 62174310.0, + "step": 2411 + }, + { + "epoch": 0.2648802987041511, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.9797043800354004, + "learning_rate": 8.825036603221084e-07, + "loss": 1.0534, + "mean_token_accuracy": 0.704858124256134, + "num_tokens": 62193956.0, + "step": 2412 + }, + { + "epoch": 0.2649901164067648, + "ewc_loss": 4.5299530029296875e-06, + "grad_norm": 3.2652390003204346, + "learning_rate": 8.828696925329429e-07, + "loss": 1.0851, + "mean_token_accuracy": 0.6777913570404053, + "num_tokens": 62220723.0, + "step": 2413 + }, + { + "epoch": 0.26509993410937843, + "ewc_loss": 4.5299530029296875e-06, + "grad_norm": 2.570237398147583, + "learning_rate": 8.832357247437774e-07, + "loss": 1.0781, + "mean_token_accuracy": 0.686992883682251, + "num_tokens": 62246768.0, + "step": 2414 + }, + { + "epoch": 0.2652097518119921, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.451295852661133, + "learning_rate": 8.83601756954612e-07, + "loss": 1.0095, + "mean_token_accuracy": 0.7058476805686951, + "num_tokens": 62274476.0, + "step": 2415 + }, + { + "epoch": 0.2653195695146058, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.6285996437072754, + "learning_rate": 8.839677891654466e-07, + "loss": 1.0041, + "mean_token_accuracy": 0.7069274187088013, + "num_tokens": 62300250.0, + "step": 2416 + }, + { + "epoch": 0.2654293872172194, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.549304962158203, + "learning_rate": 8.843338213762811e-07, + "loss": 1.0077, + "mean_token_accuracy": 0.7094584107398987, + "num_tokens": 62325482.0, + "step": 2417 + }, + { + "epoch": 0.26553920491983307, + "ewc_loss": 4.500150680541992e-06, + "grad_norm": 2.7362139225006104, + "learning_rate": 8.846998535871157e-07, + "loss": 1.0654, + "mean_token_accuracy": 0.6936020851135254, + "num_tokens": 62348385.0, + "step": 2418 + }, + { + "epoch": 0.2656490226224467, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.67624568939209, + "learning_rate": 8.850658857979502e-07, + "loss": 1.0269, + "mean_token_accuracy": 0.6976032853126526, + "num_tokens": 62371942.0, + "step": 2419 + }, + { + "epoch": 0.2657588403250604, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.6219959259033203, + "learning_rate": 8.854319180087847e-07, + "loss": 1.0181, + "mean_token_accuracy": 0.70337975025177, + "num_tokens": 62398163.0, + "step": 2420 + }, + { + "epoch": 0.26586865802767407, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.839409351348877, + "learning_rate": 8.857979502196194e-07, + "loss": 1.0612, + "mean_token_accuracy": 0.682410717010498, + "num_tokens": 62419047.0, + "step": 2421 + }, + { + "epoch": 0.2659784757302877, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.4840025901794434, + "learning_rate": 8.861639824304539e-07, + "loss": 1.1468, + "mean_token_accuracy": 0.679845929145813, + "num_tokens": 62446956.0, + "step": 2422 + }, + { + "epoch": 0.26608829343290136, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.817736864089966, + "learning_rate": 8.865300146412884e-07, + "loss": 1.0806, + "mean_token_accuracy": 0.6872454881668091, + "num_tokens": 62468806.0, + "step": 2423 + }, + { + "epoch": 0.26619811113551506, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.620365858078003, + "learning_rate": 8.86896046852123e-07, + "loss": 1.087, + "mean_token_accuracy": 0.6850847601890564, + "num_tokens": 62493696.0, + "step": 2424 + }, + { + "epoch": 0.2663079288381287, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 3.123250961303711, + "learning_rate": 8.872620790629575e-07, + "loss": 0.957, + "mean_token_accuracy": 0.7130956649780273, + "num_tokens": 62510545.0, + "step": 2425 + }, + { + "epoch": 0.26641774654074235, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.429008960723877, + "learning_rate": 8.87628111273792e-07, + "loss": 1.0436, + "mean_token_accuracy": 0.7005792856216431, + "num_tokens": 62538602.0, + "step": 2426 + }, + { + "epoch": 0.26652756424335605, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 3.0887508392333984, + "learning_rate": 8.879941434846267e-07, + "loss": 0.9554, + "mean_token_accuracy": 0.7134869694709778, + "num_tokens": 62556858.0, + "step": 2427 + }, + { + "epoch": 0.2666373819459697, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.6915950775146484, + "learning_rate": 8.883601756954612e-07, + "loss": 1.0367, + "mean_token_accuracy": 0.6988938450813293, + "num_tokens": 62580368.0, + "step": 2428 + }, + { + "epoch": 0.26674719964858334, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.7513644695281982, + "learning_rate": 8.887262079062958e-07, + "loss": 1.0461, + "mean_token_accuracy": 0.6952831149101257, + "num_tokens": 62605952.0, + "step": 2429 + }, + { + "epoch": 0.266857017351197, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.3782083988189697, + "learning_rate": 8.890922401171303e-07, + "loss": 1.0671, + "mean_token_accuracy": 0.6851272583007812, + "num_tokens": 62635903.0, + "step": 2430 + }, + { + "epoch": 0.2669668350538107, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.3187196254730225, + "learning_rate": 8.894582723279648e-07, + "loss": 1.1405, + "mean_token_accuracy": 0.6666579246520996, + "num_tokens": 62675543.0, + "step": 2431 + }, + { + "epoch": 0.26707665275642434, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.4146485328674316, + "learning_rate": 8.898243045387995e-07, + "loss": 1.1121, + "mean_token_accuracy": 0.6739463806152344, + "num_tokens": 62705144.0, + "step": 2432 + }, + { + "epoch": 0.267186470459038, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 7.914210796356201, + "learning_rate": 8.90190336749634e-07, + "loss": 1.0377, + "mean_token_accuracy": 0.6928262114524841, + "num_tokens": 62730740.0, + "step": 2433 + }, + { + "epoch": 0.2672962881616517, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.856482982635498, + "learning_rate": 8.905563689604685e-07, + "loss": 1.0953, + "mean_token_accuracy": 0.6825999617576599, + "num_tokens": 62755879.0, + "step": 2434 + }, + { + "epoch": 0.26740610586426533, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.3368287086486816, + "learning_rate": 8.909224011713031e-07, + "loss": 1.0273, + "mean_token_accuracy": 0.6994415521621704, + "num_tokens": 62786905.0, + "step": 2435 + }, + { + "epoch": 0.267515923566879, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.5652754306793213, + "learning_rate": 8.912884333821376e-07, + "loss": 1.0727, + "mean_token_accuracy": 0.6899054050445557, + "num_tokens": 62815251.0, + "step": 2436 + }, + { + "epoch": 0.2676257412694926, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.6434173583984375, + "learning_rate": 8.916544655929722e-07, + "loss": 1.0121, + "mean_token_accuracy": 0.7096555233001709, + "num_tokens": 62840071.0, + "step": 2437 + }, + { + "epoch": 0.2677355589721063, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.618504762649536, + "learning_rate": 8.920204978038068e-07, + "loss": 0.9751, + "mean_token_accuracy": 0.7100191712379456, + "num_tokens": 62863599.0, + "step": 2438 + }, + { + "epoch": 0.26784537667471997, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.6833770275115967, + "learning_rate": 8.923865300146413e-07, + "loss": 1.0192, + "mean_token_accuracy": 0.6990859508514404, + "num_tokens": 62886063.0, + "step": 2439 + }, + { + "epoch": 0.2679551943773336, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.604475259780884, + "learning_rate": 8.927525622254757e-07, + "loss": 1.0448, + "mean_token_accuracy": 0.7017289996147156, + "num_tokens": 62911701.0, + "step": 2440 + }, + { + "epoch": 0.26806501207994726, + "ewc_loss": 4.559755325317383e-06, + "grad_norm": 2.8172829151153564, + "learning_rate": 8.931185944363104e-07, + "loss": 0.9944, + "mean_token_accuracy": 0.7054427266120911, + "num_tokens": 62931389.0, + "step": 2441 + }, + { + "epoch": 0.26817482978256096, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.648512840270996, + "learning_rate": 8.934846266471448e-07, + "loss": 1.1247, + "mean_token_accuracy": 0.6720985174179077, + "num_tokens": 62959298.0, + "step": 2442 + }, + { + "epoch": 0.2682846474851746, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.450502395629883, + "learning_rate": 8.938506588579794e-07, + "loss": 1.0248, + "mean_token_accuracy": 0.7056952714920044, + "num_tokens": 62987228.0, + "step": 2443 + }, + { + "epoch": 0.26839446518778826, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.4098997116088867, + "learning_rate": 8.94216691068814e-07, + "loss": 1.0314, + "mean_token_accuracy": 0.6925404071807861, + "num_tokens": 63016848.0, + "step": 2444 + }, + { + "epoch": 0.26850428289040196, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.4822988510131836, + "learning_rate": 8.945827232796485e-07, + "loss": 1.1343, + "mean_token_accuracy": 0.6738220453262329, + "num_tokens": 63047314.0, + "step": 2445 + }, + { + "epoch": 0.2686141005930156, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 7.877511501312256, + "learning_rate": 8.949487554904832e-07, + "loss": 1.138, + "mean_token_accuracy": 0.670467734336853, + "num_tokens": 63081054.0, + "step": 2446 + }, + { + "epoch": 0.26872391829562925, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.847480535507202, + "learning_rate": 8.953147877013176e-07, + "loss": 1.0403, + "mean_token_accuracy": 0.6952928900718689, + "num_tokens": 63102763.0, + "step": 2447 + }, + { + "epoch": 0.2688337359982429, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.859973192214966, + "learning_rate": 8.956808199121522e-07, + "loss": 1.0481, + "mean_token_accuracy": 0.6917316317558289, + "num_tokens": 63124068.0, + "step": 2448 + }, + { + "epoch": 0.2689435537008566, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.6164255142211914, + "learning_rate": 8.960468521229869e-07, + "loss": 0.9979, + "mean_token_accuracy": 0.7059139609336853, + "num_tokens": 63145050.0, + "step": 2449 + }, + { + "epoch": 0.26905337140347024, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.6736159324645996, + "learning_rate": 8.964128843338213e-07, + "loss": 1.0077, + "mean_token_accuracy": 0.710344135761261, + "num_tokens": 63167810.0, + "step": 2450 + }, + { + "epoch": 0.2691631891060839, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 2.5345733165740967, + "learning_rate": 8.967789165446558e-07, + "loss": 1.0189, + "mean_token_accuracy": 0.7029804587364197, + "num_tokens": 63197274.0, + "step": 2451 + }, + { + "epoch": 0.26927300680869753, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.6900246143341064, + "learning_rate": 8.971449487554904e-07, + "loss": 1.0709, + "mean_token_accuracy": 0.6856694221496582, + "num_tokens": 63221700.0, + "step": 2452 + }, + { + "epoch": 0.26938282451131124, + "ewc_loss": 4.589557647705078e-06, + "grad_norm": 3.0043561458587646, + "learning_rate": 8.97510980966325e-07, + "loss": 1.0929, + "mean_token_accuracy": 0.6780841946601868, + "num_tokens": 63248117.0, + "step": 2453 + }, + { + "epoch": 0.2694926422139249, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.619666814804077, + "learning_rate": 8.978770131771595e-07, + "loss": 0.9571, + "mean_token_accuracy": 0.7152932286262512, + "num_tokens": 63271257.0, + "step": 2454 + }, + { + "epoch": 0.26960245991653853, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.8069841861724854, + "learning_rate": 8.982430453879941e-07, + "loss": 1.054, + "mean_token_accuracy": 0.6871421337127686, + "num_tokens": 63293129.0, + "step": 2455 + }, + { + "epoch": 0.26971227761915223, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.807743549346924, + "learning_rate": 8.986090775988286e-07, + "loss": 1.0383, + "mean_token_accuracy": 0.695462167263031, + "num_tokens": 63315767.0, + "step": 2456 + }, + { + "epoch": 0.2698220953217659, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.4755308628082275, + "learning_rate": 8.989751098096631e-07, + "loss": 1.0606, + "mean_token_accuracy": 0.694550633430481, + "num_tokens": 63345045.0, + "step": 2457 + }, + { + "epoch": 0.2699319130243795, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.5434932708740234, + "learning_rate": 8.993411420204977e-07, + "loss": 1.028, + "mean_token_accuracy": 0.6951562166213989, + "num_tokens": 63370935.0, + "step": 2458 + }, + { + "epoch": 0.27004173072699317, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.628283739089966, + "learning_rate": 8.997071742313323e-07, + "loss": 1.0533, + "mean_token_accuracy": 0.6854817867279053, + "num_tokens": 63395023.0, + "step": 2459 + }, + { + "epoch": 0.27015154842960687, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.2846829891204834, + "learning_rate": 9.000732064421668e-07, + "loss": 1.1325, + "mean_token_accuracy": 0.6722767353057861, + "num_tokens": 63429774.0, + "step": 2460 + }, + { + "epoch": 0.2702613661322205, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.5740575790405273, + "learning_rate": 9.004392386530014e-07, + "loss": 0.9895, + "mean_token_accuracy": 0.705103874206543, + "num_tokens": 63454711.0, + "step": 2461 + }, + { + "epoch": 0.27037118383483416, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.599170207977295, + "learning_rate": 9.008052708638359e-07, + "loss": 1.019, + "mean_token_accuracy": 0.6961575746536255, + "num_tokens": 63481256.0, + "step": 2462 + }, + { + "epoch": 0.27048100153744786, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.800779342651367, + "learning_rate": 9.011713030746705e-07, + "loss": 1.0894, + "mean_token_accuracy": 0.6811692118644714, + "num_tokens": 63507253.0, + "step": 2463 + }, + { + "epoch": 0.2705908192400615, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.5840935707092285, + "learning_rate": 9.015373352855051e-07, + "loss": 1.0353, + "mean_token_accuracy": 0.6850954294204712, + "num_tokens": 63532003.0, + "step": 2464 + }, + { + "epoch": 0.27070063694267515, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.952382802963257, + "learning_rate": 9.019033674963396e-07, + "loss": 1.0233, + "mean_token_accuracy": 0.6966198682785034, + "num_tokens": 63554003.0, + "step": 2465 + }, + { + "epoch": 0.2708104546452888, + "ewc_loss": 4.6193599700927734e-06, + "grad_norm": 2.8965439796447754, + "learning_rate": 9.022693997071742e-07, + "loss": 0.9177, + "mean_token_accuracy": 0.7252035737037659, + "num_tokens": 63574058.0, + "step": 2466 + }, + { + "epoch": 0.2709202723479025, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.93721342086792, + "learning_rate": 9.026354319180087e-07, + "loss": 0.9968, + "mean_token_accuracy": 0.7022666931152344, + "num_tokens": 63596234.0, + "step": 2467 + }, + { + "epoch": 0.27103009005051615, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.509594678878784, + "learning_rate": 9.030014641288432e-07, + "loss": 1.0901, + "mean_token_accuracy": 0.6841424703598022, + "num_tokens": 63627837.0, + "step": 2468 + }, + { + "epoch": 0.2711399077531298, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.3728930950164795, + "learning_rate": 9.033674963396779e-07, + "loss": 1.1565, + "mean_token_accuracy": 0.6659811735153198, + "num_tokens": 63661751.0, + "step": 2469 + }, + { + "epoch": 0.27124972545574344, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.5911691188812256, + "learning_rate": 9.037335285505124e-07, + "loss": 1.0751, + "mean_token_accuracy": 0.6884036064147949, + "num_tokens": 63686230.0, + "step": 2470 + }, + { + "epoch": 0.27135954315835714, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.7469935417175293, + "learning_rate": 9.040995607613469e-07, + "loss": 1.1271, + "mean_token_accuracy": 0.6707720756530762, + "num_tokens": 63710974.0, + "step": 2471 + }, + { + "epoch": 0.2714693608609708, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.481248378753662, + "learning_rate": 9.044655929721815e-07, + "loss": 1.118, + "mean_token_accuracy": 0.6742540597915649, + "num_tokens": 63741057.0, + "step": 2472 + }, + { + "epoch": 0.27157917856358443, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.560588836669922, + "learning_rate": 9.04831625183016e-07, + "loss": 1.1201, + "mean_token_accuracy": 0.6744987964630127, + "num_tokens": 63766625.0, + "step": 2473 + }, + { + "epoch": 0.27168899626619814, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.4806063175201416, + "learning_rate": 9.051976573938505e-07, + "loss": 1.1089, + "mean_token_accuracy": 0.6784241199493408, + "num_tokens": 63792679.0, + "step": 2474 + }, + { + "epoch": 0.2717988139688118, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.5051887035369873, + "learning_rate": 9.055636896046852e-07, + "loss": 1.0689, + "mean_token_accuracy": 0.6879732608795166, + "num_tokens": 63818259.0, + "step": 2475 + }, + { + "epoch": 0.2719086316714254, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.9572908878326416, + "learning_rate": 9.059297218155197e-07, + "loss": 1.0046, + "mean_token_accuracy": 0.7006509304046631, + "num_tokens": 63839630.0, + "step": 2476 + }, + { + "epoch": 0.2720184493740391, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.437575578689575, + "learning_rate": 9.062957540263542e-07, + "loss": 1.014, + "mean_token_accuracy": 0.6997363567352295, + "num_tokens": 63868019.0, + "step": 2477 + }, + { + "epoch": 0.2721282670766528, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.557609796524048, + "learning_rate": 9.066617862371888e-07, + "loss": 1.049, + "mean_token_accuracy": 0.7003025412559509, + "num_tokens": 63896304.0, + "step": 2478 + }, + { + "epoch": 0.2722380847792664, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.5899672508239746, + "learning_rate": 9.070278184480233e-07, + "loss": 0.9805, + "mean_token_accuracy": 0.7066357135772705, + "num_tokens": 63925436.0, + "step": 2479 + }, + { + "epoch": 0.27234790248188007, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.5683140754699707, + "learning_rate": 9.07393850658858e-07, + "loss": 1.031, + "mean_token_accuracy": 0.6961545348167419, + "num_tokens": 63949487.0, + "step": 2480 + }, + { + "epoch": 0.27245772018449377, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 2.6214938163757324, + "learning_rate": 9.077598828696925e-07, + "loss": 0.9601, + "mean_token_accuracy": 0.721745491027832, + "num_tokens": 63972779.0, + "step": 2481 + }, + { + "epoch": 0.2725675378871074, + "ewc_loss": 4.649162292480469e-06, + "grad_norm": 3.124258279800415, + "learning_rate": 9.08125915080527e-07, + "loss": 1.0253, + "mean_token_accuracy": 0.6982495784759521, + "num_tokens": 63990229.0, + "step": 2482 + }, + { + "epoch": 0.27267735558972106, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.7123446464538574, + "learning_rate": 9.084919472913616e-07, + "loss": 1.1206, + "mean_token_accuracy": 0.6708346605300903, + "num_tokens": 64014964.0, + "step": 2483 + }, + { + "epoch": 0.2727871732923347, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.5728275775909424, + "learning_rate": 9.088579795021961e-07, + "loss": 1.026, + "mean_token_accuracy": 0.6973230838775635, + "num_tokens": 64041383.0, + "step": 2484 + }, + { + "epoch": 0.2728969909949484, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.6813902854919434, + "learning_rate": 9.092240117130307e-07, + "loss": 1.0911, + "mean_token_accuracy": 0.678705096244812, + "num_tokens": 64070705.0, + "step": 2485 + }, + { + "epoch": 0.27300680869756205, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.3490753173828125, + "learning_rate": 9.095900439238653e-07, + "loss": 0.9538, + "mean_token_accuracy": 0.7115997076034546, + "num_tokens": 64098527.0, + "step": 2486 + }, + { + "epoch": 0.2731166264001757, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.2668418884277344, + "learning_rate": 9.099560761346998e-07, + "loss": 1.0067, + "mean_token_accuracy": 0.6992180347442627, + "num_tokens": 64129915.0, + "step": 2487 + }, + { + "epoch": 0.27322644410278935, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.716686487197876, + "learning_rate": 9.103221083455343e-07, + "loss": 1.0644, + "mean_token_accuracy": 0.6878433227539062, + "num_tokens": 64154983.0, + "step": 2488 + }, + { + "epoch": 0.27333626180540305, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.5637564659118652, + "learning_rate": 9.106881405563689e-07, + "loss": 1.1744, + "mean_token_accuracy": 0.6716464757919312, + "num_tokens": 64186278.0, + "step": 2489 + }, + { + "epoch": 0.2734460795080167, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.6504883766174316, + "learning_rate": 9.110541727672034e-07, + "loss": 1.0731, + "mean_token_accuracy": 0.6864583492279053, + "num_tokens": 64211406.0, + "step": 2490 + }, + { + "epoch": 0.27355589721063034, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.6327884197235107, + "learning_rate": 9.11420204978038e-07, + "loss": 0.9848, + "mean_token_accuracy": 0.7103030681610107, + "num_tokens": 64236208.0, + "step": 2491 + }, + { + "epoch": 0.27366571491324404, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.3532676696777344, + "learning_rate": 9.117862371888726e-07, + "loss": 1.1532, + "mean_token_accuracy": 0.6656771302223206, + "num_tokens": 64269029.0, + "step": 2492 + }, + { + "epoch": 0.2737755326158577, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.7497754096984863, + "learning_rate": 9.121522693997071e-07, + "loss": 1.1166, + "mean_token_accuracy": 0.6684539318084717, + "num_tokens": 64295333.0, + "step": 2493 + }, + { + "epoch": 0.27388535031847133, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.3922739028930664, + "learning_rate": 9.125183016105416e-07, + "loss": 1.0751, + "mean_token_accuracy": 0.6825655698776245, + "num_tokens": 64325395.0, + "step": 2494 + }, + { + "epoch": 0.273995168021085, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.4046175479888916, + "learning_rate": 9.128843338213762e-07, + "loss": 1.0641, + "mean_token_accuracy": 0.6847623586654663, + "num_tokens": 64353858.0, + "step": 2495 + }, + { + "epoch": 0.2741049857236987, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.6187362670898438, + "learning_rate": 9.132503660322108e-07, + "loss": 1.0882, + "mean_token_accuracy": 0.6797979474067688, + "num_tokens": 64378320.0, + "step": 2496 + }, + { + "epoch": 0.2742148034263123, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.646937847137451, + "learning_rate": 9.136163982430454e-07, + "loss": 0.9902, + "mean_token_accuracy": 0.7082725763320923, + "num_tokens": 64402520.0, + "step": 2497 + }, + { + "epoch": 0.27432462112892597, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.670508861541748, + "learning_rate": 9.139824304538799e-07, + "loss": 0.9419, + "mean_token_accuracy": 0.7210433483123779, + "num_tokens": 64424227.0, + "step": 2498 + }, + { + "epoch": 0.2744344388315396, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 3.170335054397583, + "learning_rate": 9.143484626647144e-07, + "loss": 1.1473, + "mean_token_accuracy": 0.6724306344985962, + "num_tokens": 64449016.0, + "step": 2499 + }, + { + "epoch": 0.2745442565341533, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.6649723052978516, + "learning_rate": 9.14714494875549e-07, + "loss": 0.8919, + "mean_token_accuracy": 0.7285615801811218, + "num_tokens": 64471600.0, + "step": 2500 + }, + { + "epoch": 0.27465407423676697, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.529963731765747, + "learning_rate": 9.150805270863836e-07, + "loss": 1.1265, + "mean_token_accuracy": 0.677562952041626, + "num_tokens": 64501456.0, + "step": 2501 + }, + { + "epoch": 0.2747638919393806, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.73770809173584, + "learning_rate": 9.154465592972181e-07, + "loss": 1.0436, + "mean_token_accuracy": 0.6971580982208252, + "num_tokens": 64525896.0, + "step": 2502 + }, + { + "epoch": 0.2748737096419943, + "ewc_loss": 4.678964614868164e-06, + "grad_norm": 2.4150335788726807, + "learning_rate": 9.158125915080527e-07, + "loss": 1.0965, + "mean_token_accuracy": 0.6841912269592285, + "num_tokens": 64556903.0, + "step": 2503 + }, + { + "epoch": 0.27498352734460796, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.5233335494995117, + "learning_rate": 9.161786237188872e-07, + "loss": 1.0906, + "mean_token_accuracy": 0.679123044013977, + "num_tokens": 64586040.0, + "step": 2504 + }, + { + "epoch": 0.2750933450472216, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.602991819381714, + "learning_rate": 9.165446559297217e-07, + "loss": 1.1111, + "mean_token_accuracy": 0.6799520254135132, + "num_tokens": 64611675.0, + "step": 2505 + }, + { + "epoch": 0.27520316274983525, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.4836246967315674, + "learning_rate": 9.169106881405563e-07, + "loss": 0.9773, + "mean_token_accuracy": 0.7145770788192749, + "num_tokens": 64637574.0, + "step": 2506 + }, + { + "epoch": 0.27531298045244895, + "ewc_loss": 4.738569259643555e-06, + "grad_norm": 4.0810723304748535, + "learning_rate": 9.172767203513909e-07, + "loss": 0.9196, + "mean_token_accuracy": 0.7233456373214722, + "num_tokens": 64659812.0, + "step": 2507 + }, + { + "epoch": 0.2754227981550626, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.8589205741882324, + "learning_rate": 9.176427525622254e-07, + "loss": 1.1233, + "mean_token_accuracy": 0.6761519908905029, + "num_tokens": 64683019.0, + "step": 2508 + }, + { + "epoch": 0.27553261585767624, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.770695686340332, + "learning_rate": 9.1800878477306e-07, + "loss": 1.0645, + "mean_token_accuracy": 0.6938708424568176, + "num_tokens": 64708468.0, + "step": 2509 + }, + { + "epoch": 0.27564243356028995, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.2905309200286865, + "learning_rate": 9.183748169838945e-07, + "loss": 1.0825, + "mean_token_accuracy": 0.6779452562332153, + "num_tokens": 64740161.0, + "step": 2510 + }, + { + "epoch": 0.2757522512629036, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.845520257949829, + "learning_rate": 9.18740849194729e-07, + "loss": 0.9997, + "mean_token_accuracy": 0.7040570974349976, + "num_tokens": 64759273.0, + "step": 2511 + }, + { + "epoch": 0.27586206896551724, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.382179021835327, + "learning_rate": 9.191068814055637e-07, + "loss": 1.026, + "mean_token_accuracy": 0.697716474533081, + "num_tokens": 64788926.0, + "step": 2512 + }, + { + "epoch": 0.2759718866681309, + "ewc_loss": 4.76837158203125e-06, + "grad_norm": 3.041628360748291, + "learning_rate": 9.194729136163982e-07, + "loss": 1.0827, + "mean_token_accuracy": 0.686798095703125, + "num_tokens": 64817940.0, + "step": 2513 + }, + { + "epoch": 0.2760817043707446, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.2671632766723633, + "learning_rate": 9.198389458272328e-07, + "loss": 1.1658, + "mean_token_accuracy": 0.6776609420776367, + "num_tokens": 64855373.0, + "step": 2514 + }, + { + "epoch": 0.27619152207335823, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.6077260971069336, + "learning_rate": 9.202049780380673e-07, + "loss": 1.039, + "mean_token_accuracy": 0.6911690831184387, + "num_tokens": 64881957.0, + "step": 2515 + }, + { + "epoch": 0.2763013397759719, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.485456705093384, + "learning_rate": 9.205710102489018e-07, + "loss": 0.987, + "mean_token_accuracy": 0.7067626118659973, + "num_tokens": 64908585.0, + "step": 2516 + }, + { + "epoch": 0.2764111574785855, + "ewc_loss": 4.738569259643555e-06, + "grad_norm": 2.918294668197632, + "learning_rate": 9.209370424597365e-07, + "loss": 0.9981, + "mean_token_accuracy": 0.7033885717391968, + "num_tokens": 64930242.0, + "step": 2517 + }, + { + "epoch": 0.2765209751811992, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.3706610202789307, + "learning_rate": 9.21303074670571e-07, + "loss": 1.0135, + "mean_token_accuracy": 0.6996246576309204, + "num_tokens": 64959196.0, + "step": 2518 + }, + { + "epoch": 0.27663079288381287, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.400463104248047, + "learning_rate": 9.216691068814055e-07, + "loss": 1.0447, + "mean_token_accuracy": 0.6898550987243652, + "num_tokens": 64987422.0, + "step": 2519 + }, + { + "epoch": 0.2767406105864265, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.419595241546631, + "learning_rate": 9.220351390922401e-07, + "loss": 1.1228, + "mean_token_accuracy": 0.6798293590545654, + "num_tokens": 65015218.0, + "step": 2520 + }, + { + "epoch": 0.2768504282890402, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.6608288288116455, + "learning_rate": 9.224011713030746e-07, + "loss": 1.0476, + "mean_token_accuracy": 0.6955654621124268, + "num_tokens": 65042379.0, + "step": 2521 + }, + { + "epoch": 0.27696024599165386, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.5920569896698, + "learning_rate": 9.227672035139091e-07, + "loss": 1.0745, + "mean_token_accuracy": 0.6840567588806152, + "num_tokens": 65069290.0, + "step": 2522 + }, + { + "epoch": 0.2770700636942675, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.576998472213745, + "learning_rate": 9.231332357247438e-07, + "loss": 1.0122, + "mean_token_accuracy": 0.6962178945541382, + "num_tokens": 65094108.0, + "step": 2523 + }, + { + "epoch": 0.27717988139688116, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.5043904781341553, + "learning_rate": 9.234992679355783e-07, + "loss": 1.0048, + "mean_token_accuracy": 0.7050111889839172, + "num_tokens": 65122212.0, + "step": 2524 + }, + { + "epoch": 0.27728969909949486, + "ewc_loss": 4.738569259643555e-06, + "grad_norm": 2.541691541671753, + "learning_rate": 9.238653001464128e-07, + "loss": 1.1018, + "mean_token_accuracy": 0.6818089485168457, + "num_tokens": 65153331.0, + "step": 2525 + }, + { + "epoch": 0.2773995168021085, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.655834913253784, + "learning_rate": 9.242313323572474e-07, + "loss": 1.0161, + "mean_token_accuracy": 0.7026268243789673, + "num_tokens": 65178441.0, + "step": 2526 + }, + { + "epoch": 0.27750933450472215, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.422938108444214, + "learning_rate": 9.245973645680819e-07, + "loss": 0.9816, + "mean_token_accuracy": 0.7072689533233643, + "num_tokens": 65207757.0, + "step": 2527 + }, + { + "epoch": 0.2776191522073358, + "ewc_loss": 4.738569259643555e-06, + "grad_norm": 2.6670103073120117, + "learning_rate": 9.249633967789166e-07, + "loss": 1.0967, + "mean_token_accuracy": 0.6860682964324951, + "num_tokens": 65236719.0, + "step": 2528 + }, + { + "epoch": 0.2777289699099495, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.6037356853485107, + "learning_rate": 9.253294289897511e-07, + "loss": 0.9643, + "mean_token_accuracy": 0.7138371467590332, + "num_tokens": 65260911.0, + "step": 2529 + }, + { + "epoch": 0.27783878761256314, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.863950490951538, + "learning_rate": 9.256954612005856e-07, + "loss": 0.9835, + "mean_token_accuracy": 0.7080260515213013, + "num_tokens": 65281523.0, + "step": 2530 + }, + { + "epoch": 0.2779486053151768, + "ewc_loss": 4.708766937255859e-06, + "grad_norm": 2.827763795852661, + "learning_rate": 9.260614934114202e-07, + "loss": 1.0535, + "mean_token_accuracy": 0.6863628625869751, + "num_tokens": 65304936.0, + "step": 2531 + }, + { + "epoch": 0.2780584230177905, + "ewc_loss": 4.738569259643555e-06, + "grad_norm": 2.481062889099121, + "learning_rate": 9.264275256222547e-07, + "loss": 0.925, + "mean_token_accuracy": 0.7210436463356018, + "num_tokens": 65329771.0, + "step": 2532 + }, + { + "epoch": 0.27816824072040414, + "ewc_loss": 4.738569259643555e-06, + "grad_norm": 2.808281660079956, + "learning_rate": 9.267935578330893e-07, + "loss": 1.137, + "mean_token_accuracy": 0.6737910509109497, + "num_tokens": 65352739.0, + "step": 2533 + }, + { + "epoch": 0.2782780584230178, + "ewc_loss": 4.76837158203125e-06, + "grad_norm": 2.554011344909668, + "learning_rate": 9.271595900439239e-07, + "loss": 1.0826, + "mean_token_accuracy": 0.6958728432655334, + "num_tokens": 65381469.0, + "step": 2534 + }, + { + "epoch": 0.27838787612563143, + "ewc_loss": 4.76837158203125e-06, + "grad_norm": 2.901075601577759, + "learning_rate": 9.275256222547584e-07, + "loss": 0.9846, + "mean_token_accuracy": 0.708708643913269, + "num_tokens": 65402894.0, + "step": 2535 + }, + { + "epoch": 0.27849769382824513, + "ewc_loss": 4.76837158203125e-06, + "grad_norm": 2.4282686710357666, + "learning_rate": 9.278916544655929e-07, + "loss": 1.1241, + "mean_token_accuracy": 0.6821788549423218, + "num_tokens": 65433634.0, + "step": 2536 + }, + { + "epoch": 0.2786075115308588, + "ewc_loss": 4.798173904418945e-06, + "grad_norm": 2.957411766052246, + "learning_rate": 9.282576866764275e-07, + "loss": 0.9232, + "mean_token_accuracy": 0.7211246490478516, + "num_tokens": 65452164.0, + "step": 2537 + }, + { + "epoch": 0.2787173292334724, + "ewc_loss": 4.76837158203125e-06, + "grad_norm": 2.5579042434692383, + "learning_rate": 9.28623718887262e-07, + "loss": 1.1062, + "mean_token_accuracy": 0.685935914516449, + "num_tokens": 65478967.0, + "step": 2538 + }, + { + "epoch": 0.2788271469360861, + "ewc_loss": 4.76837158203125e-06, + "grad_norm": 2.8689157962799072, + "learning_rate": 9.289897510980966e-07, + "loss": 0.9804, + "mean_token_accuracy": 0.709541916847229, + "num_tokens": 65499590.0, + "step": 2539 + }, + { + "epoch": 0.27893696463869977, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 3.003847599029541, + "learning_rate": 9.293557833089312e-07, + "loss": 0.995, + "mean_token_accuracy": 0.7049936056137085, + "num_tokens": 65525619.0, + "step": 2540 + }, + { + "epoch": 0.2790467823413134, + "ewc_loss": 4.798173904418945e-06, + "grad_norm": 2.8625738620758057, + "learning_rate": 9.297218155197657e-07, + "loss": 1.006, + "mean_token_accuracy": 0.7033507823944092, + "num_tokens": 65547328.0, + "step": 2541 + }, + { + "epoch": 0.27915660004392706, + "ewc_loss": 4.798173904418945e-06, + "grad_norm": 2.7513535022735596, + "learning_rate": 9.300878477306002e-07, + "loss": 1.1148, + "mean_token_accuracy": 0.6748993992805481, + "num_tokens": 65574613.0, + "step": 2542 + }, + { + "epoch": 0.27926641774654076, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.4071438312530518, + "learning_rate": 9.304538799414348e-07, + "loss": 0.9939, + "mean_token_accuracy": 0.7012161016464233, + "num_tokens": 65603136.0, + "step": 2543 + }, + { + "epoch": 0.2793762354491544, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.6816904544830322, + "learning_rate": 9.308199121522694e-07, + "loss": 1.0344, + "mean_token_accuracy": 0.6950516700744629, + "num_tokens": 65628228.0, + "step": 2544 + }, + { + "epoch": 0.27948605315176805, + "ewc_loss": 4.798173904418945e-06, + "grad_norm": 2.687631368637085, + "learning_rate": 9.31185944363104e-07, + "loss": 1.0645, + "mean_token_accuracy": 0.6869293451309204, + "num_tokens": 65655181.0, + "step": 2545 + }, + { + "epoch": 0.2795958708543817, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 3.24198842048645, + "learning_rate": 9.315519765739385e-07, + "loss": 1.1486, + "mean_token_accuracy": 0.6706545352935791, + "num_tokens": 65684935.0, + "step": 2546 + }, + { + "epoch": 0.2797056885569954, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 3.6474945545196533, + "learning_rate": 9.31918008784773e-07, + "loss": 1.0348, + "mean_token_accuracy": 0.693332850933075, + "num_tokens": 65708943.0, + "step": 2547 + }, + { + "epoch": 0.27981550625960905, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.4299848079681396, + "learning_rate": 9.322840409956076e-07, + "loss": 1.0193, + "mean_token_accuracy": 0.7065441608428955, + "num_tokens": 65737123.0, + "step": 2548 + }, + { + "epoch": 0.2799253239622227, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.635169744491577, + "learning_rate": 9.326500732064422e-07, + "loss": 1.0679, + "mean_token_accuracy": 0.6945580840110779, + "num_tokens": 65760927.0, + "step": 2549 + }, + { + "epoch": 0.2800351416648364, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.4589693546295166, + "learning_rate": 9.330161054172767e-07, + "loss": 1.0634, + "mean_token_accuracy": 0.6864904165267944, + "num_tokens": 65788313.0, + "step": 2550 + }, + { + "epoch": 0.28014495936745004, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.5073511600494385, + "learning_rate": 9.333821376281113e-07, + "loss": 1.0513, + "mean_token_accuracy": 0.6896622776985168, + "num_tokens": 65812857.0, + "step": 2551 + }, + { + "epoch": 0.2802547770700637, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.633845806121826, + "learning_rate": 9.337481698389458e-07, + "loss": 1.01, + "mean_token_accuracy": 0.7070047855377197, + "num_tokens": 65836955.0, + "step": 2552 + }, + { + "epoch": 0.28036459477267733, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.6539130210876465, + "learning_rate": 9.341142020497803e-07, + "loss": 1.0409, + "mean_token_accuracy": 0.6960406303405762, + "num_tokens": 65863603.0, + "step": 2553 + }, + { + "epoch": 0.28047441247529104, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.903167963027954, + "learning_rate": 9.344802342606149e-07, + "loss": 1.0715, + "mean_token_accuracy": 0.6914840936660767, + "num_tokens": 65890120.0, + "step": 2554 + }, + { + "epoch": 0.2805842301779047, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.3298308849334717, + "learning_rate": 9.348462664714495e-07, + "loss": 1.067, + "mean_token_accuracy": 0.6941474676132202, + "num_tokens": 65926236.0, + "step": 2555 + }, + { + "epoch": 0.2806940478805183, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.802232265472412, + "learning_rate": 9.35212298682284e-07, + "loss": 0.9863, + "mean_token_accuracy": 0.7083454132080078, + "num_tokens": 65947643.0, + "step": 2556 + }, + { + "epoch": 0.28080386558313203, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.809793710708618, + "learning_rate": 9.355783308931186e-07, + "loss": 1.01, + "mean_token_accuracy": 0.7083069086074829, + "num_tokens": 65970667.0, + "step": 2557 + }, + { + "epoch": 0.2809136832857457, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.603386163711548, + "learning_rate": 9.359443631039531e-07, + "loss": 1.0868, + "mean_token_accuracy": 0.6856590509414673, + "num_tokens": 65996991.0, + "step": 2558 + }, + { + "epoch": 0.2810235009883593, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.393782615661621, + "learning_rate": 9.363103953147876e-07, + "loss": 1.05, + "mean_token_accuracy": 0.6916482448577881, + "num_tokens": 66027766.0, + "step": 2559 + }, + { + "epoch": 0.28113331869097297, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 3.0479540824890137, + "learning_rate": 9.366764275256223e-07, + "loss": 0.9544, + "mean_token_accuracy": 0.7130408883094788, + "num_tokens": 66046417.0, + "step": 2560 + }, + { + "epoch": 0.28124313639358667, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.790631055831909, + "learning_rate": 9.370424597364568e-07, + "loss": 0.9979, + "mean_token_accuracy": 0.7060693502426147, + "num_tokens": 66073856.0, + "step": 2561 + }, + { + "epoch": 0.2813529540962003, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.810547113418579, + "learning_rate": 9.374084919472914e-07, + "loss": 1.039, + "mean_token_accuracy": 0.6956982612609863, + "num_tokens": 66101429.0, + "step": 2562 + }, + { + "epoch": 0.28146277179881396, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.4922351837158203, + "learning_rate": 9.377745241581259e-07, + "loss": 1.1325, + "mean_token_accuracy": 0.6700186729431152, + "num_tokens": 66129613.0, + "step": 2563 + }, + { + "epoch": 0.2815725895014276, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.4742119312286377, + "learning_rate": 9.381405563689604e-07, + "loss": 1.0577, + "mean_token_accuracy": 0.6874447464942932, + "num_tokens": 66159666.0, + "step": 2564 + }, + { + "epoch": 0.2816824072040413, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.831101417541504, + "learning_rate": 9.385065885797951e-07, + "loss": 1.0302, + "mean_token_accuracy": 0.6989683508872986, + "num_tokens": 66181171.0, + "step": 2565 + }, + { + "epoch": 0.28179222490665495, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.706727981567383, + "learning_rate": 9.388726207906296e-07, + "loss": 0.9656, + "mean_token_accuracy": 0.7189221978187561, + "num_tokens": 66205013.0, + "step": 2566 + }, + { + "epoch": 0.2819020426092686, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.6280317306518555, + "learning_rate": 9.392386530014641e-07, + "loss": 1.036, + "mean_token_accuracy": 0.6994551420211792, + "num_tokens": 66230664.0, + "step": 2567 + }, + { + "epoch": 0.2820118603118823, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.394341230392456, + "learning_rate": 9.396046852122987e-07, + "loss": 1.0336, + "mean_token_accuracy": 0.6975067853927612, + "num_tokens": 66260812.0, + "step": 2568 + }, + { + "epoch": 0.28212167801449595, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.669647455215454, + "learning_rate": 9.399707174231332e-07, + "loss": 1.0618, + "mean_token_accuracy": 0.6842178702354431, + "num_tokens": 66282638.0, + "step": 2569 + }, + { + "epoch": 0.2822314957171096, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.3096776008605957, + "learning_rate": 9.403367496339677e-07, + "loss": 1.089, + "mean_token_accuracy": 0.6846877336502075, + "num_tokens": 66315689.0, + "step": 2570 + }, + { + "epoch": 0.28234131341972324, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.4689133167266846, + "learning_rate": 9.407027818448024e-07, + "loss": 1.0, + "mean_token_accuracy": 0.7047750949859619, + "num_tokens": 66343772.0, + "step": 2571 + }, + { + "epoch": 0.28245113112233694, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.732281446456909, + "learning_rate": 9.410688140556369e-07, + "loss": 0.9338, + "mean_token_accuracy": 0.7195010781288147, + "num_tokens": 66366792.0, + "step": 2572 + }, + { + "epoch": 0.2825609488249506, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.787184238433838, + "learning_rate": 9.414348462664714e-07, + "loss": 1.0034, + "mean_token_accuracy": 0.7008455991744995, + "num_tokens": 66389908.0, + "step": 2573 + }, + { + "epoch": 0.28267076652756423, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.419356107711792, + "learning_rate": 9.41800878477306e-07, + "loss": 1.0869, + "mean_token_accuracy": 0.6860706806182861, + "num_tokens": 66423245.0, + "step": 2574 + }, + { + "epoch": 0.2827805842301779, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.5597751140594482, + "learning_rate": 9.421669106881405e-07, + "loss": 1.1157, + "mean_token_accuracy": 0.6713166832923889, + "num_tokens": 66453024.0, + "step": 2575 + }, + { + "epoch": 0.2828904019327916, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.604501247406006, + "learning_rate": 9.425329428989752e-07, + "loss": 1.0509, + "mean_token_accuracy": 0.6933062076568604, + "num_tokens": 66479736.0, + "step": 2576 + }, + { + "epoch": 0.2830002196354052, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.3984642028808594, + "learning_rate": 9.428989751098097e-07, + "loss": 1.0926, + "mean_token_accuracy": 0.6771796345710754, + "num_tokens": 66509498.0, + "step": 2577 + }, + { + "epoch": 0.28311003733801887, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.644822597503662, + "learning_rate": 9.432650073206442e-07, + "loss": 1.0462, + "mean_token_accuracy": 0.6945559978485107, + "num_tokens": 66534943.0, + "step": 2578 + }, + { + "epoch": 0.2832198550406326, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.573430061340332, + "learning_rate": 9.436310395314788e-07, + "loss": 1.0324, + "mean_token_accuracy": 0.6954767107963562, + "num_tokens": 66561088.0, + "step": 2579 + }, + { + "epoch": 0.2833296727432462, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.6331980228424072, + "learning_rate": 9.439970717423133e-07, + "loss": 0.9155, + "mean_token_accuracy": 0.7278672456741333, + "num_tokens": 66584705.0, + "step": 2580 + }, + { + "epoch": 0.28343949044585987, + "ewc_loss": 4.827976226806641e-06, + "grad_norm": 2.3154244422912598, + "learning_rate": 9.443631039531479e-07, + "loss": 1.0952, + "mean_token_accuracy": 0.6799588799476624, + "num_tokens": 66617156.0, + "step": 2581 + }, + { + "epoch": 0.2835493081484735, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.568204641342163, + "learning_rate": 9.447291361639825e-07, + "loss": 1.1152, + "mean_token_accuracy": 0.6706490516662598, + "num_tokens": 66644407.0, + "step": 2582 + }, + { + "epoch": 0.2836591258510872, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.390292167663574, + "learning_rate": 9.45095168374817e-07, + "loss": 1.1437, + "mean_token_accuracy": 0.6745580434799194, + "num_tokens": 66680089.0, + "step": 2583 + }, + { + "epoch": 0.28376894355370086, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.2302396297454834, + "learning_rate": 9.454612005856515e-07, + "loss": 1.0567, + "mean_token_accuracy": 0.6899533271789551, + "num_tokens": 66716109.0, + "step": 2584 + }, + { + "epoch": 0.2838787612563145, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.6883716583251953, + "learning_rate": 9.458272327964861e-07, + "loss": 1.0804, + "mean_token_accuracy": 0.6905311346054077, + "num_tokens": 66740383.0, + "step": 2585 + }, + { + "epoch": 0.2839885789589282, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.3544740676879883, + "learning_rate": 9.461932650073206e-07, + "loss": 1.0787, + "mean_token_accuracy": 0.6914777755737305, + "num_tokens": 66774475.0, + "step": 2586 + }, + { + "epoch": 0.28409839666154185, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.4138972759246826, + "learning_rate": 9.465592972181552e-07, + "loss": 0.9685, + "mean_token_accuracy": 0.7150387763977051, + "num_tokens": 66802774.0, + "step": 2587 + }, + { + "epoch": 0.2842082143641555, + "ewc_loss": 4.857778549194336e-06, + "grad_norm": 2.863680124282837, + "learning_rate": 9.469253294289898e-07, + "loss": 1.0137, + "mean_token_accuracy": 0.6965728998184204, + "num_tokens": 66822965.0, + "step": 2588 + }, + { + "epoch": 0.28431803206676914, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.63836932182312, + "learning_rate": 9.472913616398243e-07, + "loss": 1.0799, + "mean_token_accuracy": 0.6885419487953186, + "num_tokens": 66852734.0, + "step": 2589 + }, + { + "epoch": 0.28442784976938285, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.7363927364349365, + "learning_rate": 9.476573938506587e-07, + "loss": 0.9654, + "mean_token_accuracy": 0.7190787196159363, + "num_tokens": 66877006.0, + "step": 2590 + }, + { + "epoch": 0.2845376674719965, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.609618663787842, + "learning_rate": 9.480234260614933e-07, + "loss": 1.0908, + "mean_token_accuracy": 0.6866106390953064, + "num_tokens": 66902300.0, + "step": 2591 + }, + { + "epoch": 0.28464748517461014, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.623418092727661, + "learning_rate": 9.48389458272328e-07, + "loss": 0.9613, + "mean_token_accuracy": 0.718974769115448, + "num_tokens": 66930991.0, + "step": 2592 + }, + { + "epoch": 0.2847573028772238, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.65986704826355, + "learning_rate": 9.487554904831626e-07, + "loss": 1.0555, + "mean_token_accuracy": 0.6946996450424194, + "num_tokens": 66957004.0, + "step": 2593 + }, + { + "epoch": 0.2848671205798375, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.885326862335205, + "learning_rate": 9.49121522693997e-07, + "loss": 1.0613, + "mean_token_accuracy": 0.6995856761932373, + "num_tokens": 66980931.0, + "step": 2594 + }, + { + "epoch": 0.28497693828245113, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.73132061958313, + "learning_rate": 9.494875549048315e-07, + "loss": 1.0182, + "mean_token_accuracy": 0.7056807279586792, + "num_tokens": 67005148.0, + "step": 2595 + }, + { + "epoch": 0.2850867559850648, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.981828212738037, + "learning_rate": 9.498535871156661e-07, + "loss": 1.0482, + "mean_token_accuracy": 0.6805493235588074, + "num_tokens": 67024780.0, + "step": 2596 + }, + { + "epoch": 0.2851965736876785, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 2.6566755771636963, + "learning_rate": 9.502196193265007e-07, + "loss": 1.0685, + "mean_token_accuracy": 0.6913908123970032, + "num_tokens": 67051088.0, + "step": 2597 + }, + { + "epoch": 0.2853063913902921, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.921665906906128, + "learning_rate": 9.505856515373352e-07, + "loss": 1.0358, + "mean_token_accuracy": 0.6987087726593018, + "num_tokens": 67073148.0, + "step": 2598 + }, + { + "epoch": 0.28541620909290577, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.507297992706299, + "learning_rate": 9.509516837481698e-07, + "loss": 1.1108, + "mean_token_accuracy": 0.6835572719573975, + "num_tokens": 67103289.0, + "step": 2599 + }, + { + "epoch": 0.2855260267955194, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.416438102722168, + "learning_rate": 9.513177159590043e-07, + "loss": 1.1641, + "mean_token_accuracy": 0.6614166498184204, + "num_tokens": 67134940.0, + "step": 2600 + }, + { + "epoch": 0.2856358444981331, + "ewc_loss": 4.887580871582031e-06, + "grad_norm": 3.5462350845336914, + "learning_rate": 9.516837481698388e-07, + "loss": 1.0711, + "mean_token_accuracy": 0.6883537769317627, + "num_tokens": 67162861.0, + "step": 2601 + }, + { + "epoch": 0.28574566220074676, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.8628509044647217, + "learning_rate": 9.520497803806734e-07, + "loss": 1.0179, + "mean_token_accuracy": 0.6977555751800537, + "num_tokens": 67183291.0, + "step": 2602 + }, + { + "epoch": 0.2858554799033604, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.6558752059936523, + "learning_rate": 9.52415812591508e-07, + "loss": 1.0509, + "mean_token_accuracy": 0.6932531595230103, + "num_tokens": 67208977.0, + "step": 2603 + }, + { + "epoch": 0.28596529760597406, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.7020792961120605, + "learning_rate": 9.527818448023425e-07, + "loss": 1.0251, + "mean_token_accuracy": 0.699738621711731, + "num_tokens": 67233401.0, + "step": 2604 + }, + { + "epoch": 0.28607511530858776, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.5330538749694824, + "learning_rate": 9.531478770131771e-07, + "loss": 1.0712, + "mean_token_accuracy": 0.6959665417671204, + "num_tokens": 67258898.0, + "step": 2605 + }, + { + "epoch": 0.2861849330112014, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.580514907836914, + "learning_rate": 9.535139092240116e-07, + "loss": 1.0274, + "mean_token_accuracy": 0.6995062828063965, + "num_tokens": 67288247.0, + "step": 2606 + }, + { + "epoch": 0.28629475071381505, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.516535997390747, + "learning_rate": 9.538799414348462e-07, + "loss": 1.0422, + "mean_token_accuracy": 0.6962271332740784, + "num_tokens": 67312209.0, + "step": 2607 + }, + { + "epoch": 0.28640456841642875, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.686091423034668, + "learning_rate": 9.542459736456808e-07, + "loss": 1.042, + "mean_token_accuracy": 0.6991592049598694, + "num_tokens": 67336959.0, + "step": 2608 + }, + { + "epoch": 0.2865143861190424, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.7539122104644775, + "learning_rate": 9.546120058565154e-07, + "loss": 1.034, + "mean_token_accuracy": 0.6933692693710327, + "num_tokens": 67365192.0, + "step": 2609 + }, + { + "epoch": 0.28662420382165604, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.567200183868408, + "learning_rate": 9.549780380673498e-07, + "loss": 1.0363, + "mean_token_accuracy": 0.6929412484169006, + "num_tokens": 67392262.0, + "step": 2610 + }, + { + "epoch": 0.2867340215242697, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.459355354309082, + "learning_rate": 9.553440702781844e-07, + "loss": 1.0399, + "mean_token_accuracy": 0.6994613409042358, + "num_tokens": 67419572.0, + "step": 2611 + }, + { + "epoch": 0.2868438392268834, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.755584716796875, + "learning_rate": 9.55710102489019e-07, + "loss": 0.9817, + "mean_token_accuracy": 0.7162162065505981, + "num_tokens": 67440259.0, + "step": 2612 + }, + { + "epoch": 0.28695365692949704, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.7534725666046143, + "learning_rate": 9.560761346998536e-07, + "loss": 0.9815, + "mean_token_accuracy": 0.7070250511169434, + "num_tokens": 67464103.0, + "step": 2613 + }, + { + "epoch": 0.2870634746321107, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.5710175037384033, + "learning_rate": 9.56442166910688e-07, + "loss": 1.1312, + "mean_token_accuracy": 0.6711863875389099, + "num_tokens": 67498304.0, + "step": 2614 + }, + { + "epoch": 0.2871732923347244, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.9411630630493164, + "learning_rate": 9.568081991215226e-07, + "loss": 1.1128, + "mean_token_accuracy": 0.6749492883682251, + "num_tokens": 67521821.0, + "step": 2615 + }, + { + "epoch": 0.28728311003733803, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.876370668411255, + "learning_rate": 9.571742313323572e-07, + "loss": 0.9988, + "mean_token_accuracy": 0.7071506977081299, + "num_tokens": 67541624.0, + "step": 2616 + }, + { + "epoch": 0.2873929277399517, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.752110719680786, + "learning_rate": 9.575402635431916e-07, + "loss": 1.0218, + "mean_token_accuracy": 0.7069950699806213, + "num_tokens": 67563863.0, + "step": 2617 + }, + { + "epoch": 0.2875027454425653, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.7138078212738037, + "learning_rate": 9.579062957540262e-07, + "loss": 1.1088, + "mean_token_accuracy": 0.6743017435073853, + "num_tokens": 67588335.0, + "step": 2618 + }, + { + "epoch": 0.287612563145179, + "ewc_loss": 4.9173831939697266e-06, + "grad_norm": 2.9501149654388428, + "learning_rate": 9.582723279648608e-07, + "loss": 0.9705, + "mean_token_accuracy": 0.708426833152771, + "num_tokens": 67610513.0, + "step": 2619 + }, + { + "epoch": 0.28772238084779267, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.3445210456848145, + "learning_rate": 9.586383601756954e-07, + "loss": 1.1677, + "mean_token_accuracy": 0.6629142761230469, + "num_tokens": 67643857.0, + "step": 2620 + }, + { + "epoch": 0.2878321985504063, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.6717567443847656, + "learning_rate": 9.5900439238653e-07, + "loss": 1.0157, + "mean_token_accuracy": 0.6965851783752441, + "num_tokens": 67665589.0, + "step": 2621 + }, + { + "epoch": 0.28794201625301996, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.40105938911438, + "learning_rate": 9.593704245973644e-07, + "loss": 1.0526, + "mean_token_accuracy": 0.6938616633415222, + "num_tokens": 67695418.0, + "step": 2622 + }, + { + "epoch": 0.28805183395563366, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.7907979488372803, + "learning_rate": 9.59736456808199e-07, + "loss": 1.0197, + "mean_token_accuracy": 0.6951720714569092, + "num_tokens": 67716503.0, + "step": 2623 + }, + { + "epoch": 0.2881616516582473, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.4955687522888184, + "learning_rate": 9.601024890190336e-07, + "loss": 1.0277, + "mean_token_accuracy": 0.7022870182991028, + "num_tokens": 67743036.0, + "step": 2624 + }, + { + "epoch": 0.28827146936086095, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.4804365634918213, + "learning_rate": 9.604685212298682e-07, + "loss": 1.1096, + "mean_token_accuracy": 0.676706075668335, + "num_tokens": 67775477.0, + "step": 2625 + }, + { + "epoch": 0.28838128706347466, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.9714787006378174, + "learning_rate": 9.608345534407028e-07, + "loss": 1.0611, + "mean_token_accuracy": 0.6923242211341858, + "num_tokens": 67796246.0, + "step": 2626 + }, + { + "epoch": 0.2884911047660883, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 3.0513718128204346, + "learning_rate": 9.612005856515372e-07, + "loss": 1.0172, + "mean_token_accuracy": 0.7070760130882263, + "num_tokens": 67820758.0, + "step": 2627 + }, + { + "epoch": 0.28860092246870195, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.6071572303771973, + "learning_rate": 9.615666178623718e-07, + "loss": 1.0428, + "mean_token_accuracy": 0.6948078870773315, + "num_tokens": 67844105.0, + "step": 2628 + }, + { + "epoch": 0.2887107401713156, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.418055295944214, + "learning_rate": 9.619326500732064e-07, + "loss": 1.1177, + "mean_token_accuracy": 0.6705290675163269, + "num_tokens": 67876536.0, + "step": 2629 + }, + { + "epoch": 0.2888205578739293, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.6963069438934326, + "learning_rate": 9.62298682284041e-07, + "loss": 1.0123, + "mean_token_accuracy": 0.706440806388855, + "num_tokens": 67906726.0, + "step": 2630 + }, + { + "epoch": 0.28893037557654294, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.493476390838623, + "learning_rate": 9.626647144948754e-07, + "loss": 1.0621, + "mean_token_accuracy": 0.68027263879776, + "num_tokens": 67934853.0, + "step": 2631 + }, + { + "epoch": 0.2890401932791566, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.603715181350708, + "learning_rate": 9.6303074670571e-07, + "loss": 1.0519, + "mean_token_accuracy": 0.7008368372917175, + "num_tokens": 67960487.0, + "step": 2632 + }, + { + "epoch": 0.2891500109817703, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.6416659355163574, + "learning_rate": 9.633967789165446e-07, + "loss": 1.0328, + "mean_token_accuracy": 0.6940655708312988, + "num_tokens": 67985099.0, + "step": 2633 + }, + { + "epoch": 0.28925982868438394, + "ewc_loss": 4.947185516357422e-06, + "grad_norm": 2.9990522861480713, + "learning_rate": 9.63762811127379e-07, + "loss": 1.1351, + "mean_token_accuracy": 0.6730419397354126, + "num_tokens": 68005907.0, + "step": 2634 + }, + { + "epoch": 0.2893696463869976, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.9216952323913574, + "learning_rate": 9.641288433382138e-07, + "loss": 1.0017, + "mean_token_accuracy": 0.7070378065109253, + "num_tokens": 68025604.0, + "step": 2635 + }, + { + "epoch": 0.2894794640896112, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.4744648933410645, + "learning_rate": 9.644948755490482e-07, + "loss": 0.9276, + "mean_token_accuracy": 0.7235783934593201, + "num_tokens": 68053834.0, + "step": 2636 + }, + { + "epoch": 0.28958928179222493, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 3.0535471439361572, + "learning_rate": 9.648609077598828e-07, + "loss": 1.0565, + "mean_token_accuracy": 0.6881935000419617, + "num_tokens": 68075654.0, + "step": 2637 + }, + { + "epoch": 0.2896990994948386, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.745163679122925, + "learning_rate": 9.652269399707174e-07, + "loss": 1.0333, + "mean_token_accuracy": 0.7002219557762146, + "num_tokens": 68100679.0, + "step": 2638 + }, + { + "epoch": 0.2898089171974522, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.7268006801605225, + "learning_rate": 9.655929721815518e-07, + "loss": 1.0777, + "mean_token_accuracy": 0.6894052028656006, + "num_tokens": 68126152.0, + "step": 2639 + }, + { + "epoch": 0.28991873490006587, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.2444093227386475, + "learning_rate": 9.659590043923866e-07, + "loss": 1.0646, + "mean_token_accuracy": 0.6909871101379395, + "num_tokens": 68160904.0, + "step": 2640 + }, + { + "epoch": 0.29002855260267957, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.5314321517944336, + "learning_rate": 9.66325036603221e-07, + "loss": 1.1431, + "mean_token_accuracy": 0.6687562465667725, + "num_tokens": 68194352.0, + "step": 2641 + }, + { + "epoch": 0.2901383703052932, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.634032964706421, + "learning_rate": 9.666910688140556e-07, + "loss": 1.0633, + "mean_token_accuracy": 0.6930225491523743, + "num_tokens": 68220095.0, + "step": 2642 + }, + { + "epoch": 0.29024818800790686, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 3.0178024768829346, + "learning_rate": 9.670571010248902e-07, + "loss": 1.0227, + "mean_token_accuracy": 0.7013494968414307, + "num_tokens": 68245461.0, + "step": 2643 + }, + { + "epoch": 0.29035800571052056, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.6319403648376465, + "learning_rate": 9.674231332357246e-07, + "loss": 1.0333, + "mean_token_accuracy": 0.6900444626808167, + "num_tokens": 68270879.0, + "step": 2644 + }, + { + "epoch": 0.2904678234131342, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.7891743183135986, + "learning_rate": 9.677891654465592e-07, + "loss": 0.9634, + "mean_token_accuracy": 0.7082006335258484, + "num_tokens": 68292893.0, + "step": 2645 + }, + { + "epoch": 0.29057764111574785, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.5301594734191895, + "learning_rate": 9.681551976573938e-07, + "loss": 1.0734, + "mean_token_accuracy": 0.6819683313369751, + "num_tokens": 68319075.0, + "step": 2646 + }, + { + "epoch": 0.2906874588183615, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.7324130535125732, + "learning_rate": 9.685212298682284e-07, + "loss": 0.981, + "mean_token_accuracy": 0.708739161491394, + "num_tokens": 68342098.0, + "step": 2647 + }, + { + "epoch": 0.2907972765209752, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.657069206237793, + "learning_rate": 9.688872620790628e-07, + "loss": 1.1232, + "mean_token_accuracy": 0.6695528626441956, + "num_tokens": 68371225.0, + "step": 2648 + }, + { + "epoch": 0.29090709422358885, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.74035906791687, + "learning_rate": 9.692532942898974e-07, + "loss": 1.0804, + "mean_token_accuracy": 0.6855120658874512, + "num_tokens": 68398188.0, + "step": 2649 + }, + { + "epoch": 0.2910169119262025, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.6368014812469482, + "learning_rate": 9.69619326500732e-07, + "loss": 0.9972, + "mean_token_accuracy": 0.7082095146179199, + "num_tokens": 68425520.0, + "step": 2650 + }, + { + "epoch": 0.29112672962881614, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.2803738117218018, + "learning_rate": 9.699853587115666e-07, + "loss": 1.1116, + "mean_token_accuracy": 0.676292896270752, + "num_tokens": 68462456.0, + "step": 2651 + }, + { + "epoch": 0.29123654733142984, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.6482396125793457, + "learning_rate": 9.703513909224012e-07, + "loss": 1.1241, + "mean_token_accuracy": 0.6757659912109375, + "num_tokens": 68489331.0, + "step": 2652 + }, + { + "epoch": 0.2913463650340435, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.5110819339752197, + "learning_rate": 9.707174231332356e-07, + "loss": 1.0685, + "mean_token_accuracy": 0.6870224475860596, + "num_tokens": 68516051.0, + "step": 2653 + }, + { + "epoch": 0.29145618273665713, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.660747766494751, + "learning_rate": 9.710834553440702e-07, + "loss": 1.0715, + "mean_token_accuracy": 0.6927706003189087, + "num_tokens": 68539805.0, + "step": 2654 + }, + { + "epoch": 0.29156600043927083, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.539722204208374, + "learning_rate": 9.714494875549048e-07, + "loss": 1.0351, + "mean_token_accuracy": 0.6970239877700806, + "num_tokens": 68565221.0, + "step": 2655 + }, + { + "epoch": 0.2916758181418845, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.4642069339752197, + "learning_rate": 9.718155197657394e-07, + "loss": 1.0247, + "mean_token_accuracy": 0.6999880075454712, + "num_tokens": 68593703.0, + "step": 2656 + }, + { + "epoch": 0.2917856358444981, + "ewc_loss": 4.976987838745117e-06, + "grad_norm": 2.682854652404785, + "learning_rate": 9.72181551976574e-07, + "loss": 1.0229, + "mean_token_accuracy": 0.6974052786827087, + "num_tokens": 68615507.0, + "step": 2657 + }, + { + "epoch": 0.29189545354711177, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.4371540546417236, + "learning_rate": 9.725475841874084e-07, + "loss": 1.0965, + "mean_token_accuracy": 0.6958386898040771, + "num_tokens": 68644423.0, + "step": 2658 + }, + { + "epoch": 0.2920052712497255, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.550438404083252, + "learning_rate": 9.72913616398243e-07, + "loss": 1.0607, + "mean_token_accuracy": 0.6916356086730957, + "num_tokens": 68670329.0, + "step": 2659 + }, + { + "epoch": 0.2921150889523391, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.9057679176330566, + "learning_rate": 9.732796486090776e-07, + "loss": 1.0707, + "mean_token_accuracy": 0.695080578327179, + "num_tokens": 68691542.0, + "step": 2660 + }, + { + "epoch": 0.29222490665495277, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.7927350997924805, + "learning_rate": 9.736456808199122e-07, + "loss": 1.0195, + "mean_token_accuracy": 0.699195384979248, + "num_tokens": 68713333.0, + "step": 2661 + }, + { + "epoch": 0.29233472435756647, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.8065595626831055, + "learning_rate": 9.740117130307466e-07, + "loss": 1.0232, + "mean_token_accuracy": 0.704075038433075, + "num_tokens": 68735818.0, + "step": 2662 + }, + { + "epoch": 0.2924445420601801, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 3.094172239303589, + "learning_rate": 9.743777452415812e-07, + "loss": 1.0508, + "mean_token_accuracy": 0.6940575838088989, + "num_tokens": 68762899.0, + "step": 2663 + }, + { + "epoch": 0.29255435976279376, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.758479118347168, + "learning_rate": 9.747437774524158e-07, + "loss": 1.0665, + "mean_token_accuracy": 0.6969051361083984, + "num_tokens": 68786750.0, + "step": 2664 + }, + { + "epoch": 0.2926641774654074, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.758315086364746, + "learning_rate": 9.751098096632502e-07, + "loss": 1.016, + "mean_token_accuracy": 0.7013729810714722, + "num_tokens": 68810131.0, + "step": 2665 + }, + { + "epoch": 0.2927739951680211, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.580413579940796, + "learning_rate": 9.754758418740848e-07, + "loss": 1.1253, + "mean_token_accuracy": 0.6751608848571777, + "num_tokens": 68837302.0, + "step": 2666 + }, + { + "epoch": 0.29288381287063475, + "ewc_loss": 5.036592483520508e-06, + "grad_norm": 2.431051731109619, + "learning_rate": 9.758418740849194e-07, + "loss": 1.0475, + "mean_token_accuracy": 0.698937714099884, + "num_tokens": 68866848.0, + "step": 2667 + }, + { + "epoch": 0.2929936305732484, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.6791861057281494, + "learning_rate": 9.76207906295754e-07, + "loss": 1.1043, + "mean_token_accuracy": 0.681839108467102, + "num_tokens": 68892221.0, + "step": 2668 + }, + { + "epoch": 0.29310344827586204, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.4318952560424805, + "learning_rate": 9.765739385065886e-07, + "loss": 1.0138, + "mean_token_accuracy": 0.7015700340270996, + "num_tokens": 68920022.0, + "step": 2669 + }, + { + "epoch": 0.29321326597847575, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.875666618347168, + "learning_rate": 9.76939970717423e-07, + "loss": 1.0137, + "mean_token_accuracy": 0.7080872058868408, + "num_tokens": 68941852.0, + "step": 2670 + }, + { + "epoch": 0.2933230836810894, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.5272862911224365, + "learning_rate": 9.773060029282576e-07, + "loss": 1.0965, + "mean_token_accuracy": 0.6799914836883545, + "num_tokens": 68973292.0, + "step": 2671 + }, + { + "epoch": 0.29343290138370304, + "ewc_loss": 5.0067901611328125e-06, + "grad_norm": 2.878734588623047, + "learning_rate": 9.776720351390922e-07, + "loss": 1.0849, + "mean_token_accuracy": 0.687480092048645, + "num_tokens": 68995369.0, + "step": 2672 + }, + { + "epoch": 0.29354271908631674, + "ewc_loss": 5.036592483520508e-06, + "grad_norm": 2.616992712020874, + "learning_rate": 9.780380673499268e-07, + "loss": 1.0553, + "mean_token_accuracy": 0.6832789182662964, + "num_tokens": 69018419.0, + "step": 2673 + }, + { + "epoch": 0.2936525367889304, + "ewc_loss": 5.036592483520508e-06, + "grad_norm": 2.9384000301361084, + "learning_rate": 9.784040995607614e-07, + "loss": 1.0066, + "mean_token_accuracy": 0.7089236378669739, + "num_tokens": 69039570.0, + "step": 2674 + }, + { + "epoch": 0.29376235449154403, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.613862991333008, + "learning_rate": 9.787701317715958e-07, + "loss": 1.006, + "mean_token_accuracy": 0.7080096006393433, + "num_tokens": 69065295.0, + "step": 2675 + }, + { + "epoch": 0.2938721721941577, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.7602744102478027, + "learning_rate": 9.791361639824304e-07, + "loss": 1.0255, + "mean_token_accuracy": 0.7059082388877869, + "num_tokens": 69088292.0, + "step": 2676 + }, + { + "epoch": 0.2939819898967714, + "ewc_loss": 5.036592483520508e-06, + "grad_norm": 3.287862777709961, + "learning_rate": 9.79502196193265e-07, + "loss": 1.0103, + "mean_token_accuracy": 0.6987050175666809, + "num_tokens": 69109009.0, + "step": 2677 + }, + { + "epoch": 0.294091807599385, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.5174403190612793, + "learning_rate": 9.798682284040996e-07, + "loss": 1.0042, + "mean_token_accuracy": 0.7107946872711182, + "num_tokens": 69135296.0, + "step": 2678 + }, + { + "epoch": 0.29420162530199867, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.4592764377593994, + "learning_rate": 9.80234260614934e-07, + "loss": 1.1003, + "mean_token_accuracy": 0.6759325265884399, + "num_tokens": 69167474.0, + "step": 2679 + }, + { + "epoch": 0.2943114430046123, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.326784133911133, + "learning_rate": 9.806002928257686e-07, + "loss": 1.0792, + "mean_token_accuracy": 0.6892108917236328, + "num_tokens": 69199833.0, + "step": 2680 + }, + { + "epoch": 0.294421260707226, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.489962100982666, + "learning_rate": 9.809663250366032e-07, + "loss": 1.0819, + "mean_token_accuracy": 0.6847513914108276, + "num_tokens": 69228030.0, + "step": 2681 + }, + { + "epoch": 0.29453107840983966, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.580197811126709, + "learning_rate": 9.813323572474376e-07, + "loss": 1.0765, + "mean_token_accuracy": 0.6791373491287231, + "num_tokens": 69253176.0, + "step": 2682 + }, + { + "epoch": 0.2946408961124533, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.528961658477783, + "learning_rate": 9.816983894582724e-07, + "loss": 0.9974, + "mean_token_accuracy": 0.707474410533905, + "num_tokens": 69279049.0, + "step": 2683 + }, + { + "epoch": 0.294750713815067, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.2879481315612793, + "learning_rate": 9.820644216691068e-07, + "loss": 1.0845, + "mean_token_accuracy": 0.6856510639190674, + "num_tokens": 69311963.0, + "step": 2684 + }, + { + "epoch": 0.29486053151768066, + "ewc_loss": 5.066394805908203e-06, + "grad_norm": 2.6829090118408203, + "learning_rate": 9.824304538799414e-07, + "loss": 0.9853, + "mean_token_accuracy": 0.7116718888282776, + "num_tokens": 69338490.0, + "step": 2685 + }, + { + "epoch": 0.2949703492202943, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.784580945968628, + "learning_rate": 9.82796486090776e-07, + "loss": 1.0156, + "mean_token_accuracy": 0.6980859041213989, + "num_tokens": 69360211.0, + "step": 2686 + }, + { + "epoch": 0.29508016692290795, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.573455810546875, + "learning_rate": 9.831625183016104e-07, + "loss": 1.1812, + "mean_token_accuracy": 0.6710090041160583, + "num_tokens": 69389732.0, + "step": 2687 + }, + { + "epoch": 0.29518998462552165, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.700512647628784, + "learning_rate": 9.835285505124452e-07, + "loss": 0.9504, + "mean_token_accuracy": 0.7143370509147644, + "num_tokens": 69412227.0, + "step": 2688 + }, + { + "epoch": 0.2952998023281353, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.7370948791503906, + "learning_rate": 9.838945827232796e-07, + "loss": 1.071, + "mean_token_accuracy": 0.6887654066085815, + "num_tokens": 69435342.0, + "step": 2689 + }, + { + "epoch": 0.29540962003074894, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.335191249847412, + "learning_rate": 9.842606149341142e-07, + "loss": 1.0738, + "mean_token_accuracy": 0.6825157999992371, + "num_tokens": 69466875.0, + "step": 2690 + }, + { + "epoch": 0.29551943773336264, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.6177690029144287, + "learning_rate": 9.846266471449488e-07, + "loss": 1.0707, + "mean_token_accuracy": 0.6979342103004456, + "num_tokens": 69493073.0, + "step": 2691 + }, + { + "epoch": 0.2956292554359763, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 5.321713924407959, + "learning_rate": 9.849926793557832e-07, + "loss": 0.9992, + "mean_token_accuracy": 0.7022485733032227, + "num_tokens": 69520144.0, + "step": 2692 + }, + { + "epoch": 0.29573907313858994, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.6430606842041016, + "learning_rate": 9.853587115666178e-07, + "loss": 1.0278, + "mean_token_accuracy": 0.6968560218811035, + "num_tokens": 69545286.0, + "step": 2693 + }, + { + "epoch": 0.2958488908412036, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.7118754386901855, + "learning_rate": 9.857247437774524e-07, + "loss": 1.1607, + "mean_token_accuracy": 0.671230137348175, + "num_tokens": 69575666.0, + "step": 2694 + }, + { + "epoch": 0.2959587085438173, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.740034341812134, + "learning_rate": 9.86090775988287e-07, + "loss": 1.0221, + "mean_token_accuracy": 0.6972548961639404, + "num_tokens": 69597333.0, + "step": 2695 + }, + { + "epoch": 0.29606852624643093, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 3.029672384262085, + "learning_rate": 9.864568081991214e-07, + "loss": 0.9944, + "mean_token_accuracy": 0.7049082517623901, + "num_tokens": 69615995.0, + "step": 2696 + }, + { + "epoch": 0.2961783439490446, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.6356019973754883, + "learning_rate": 9.86822840409956e-07, + "loss": 1.1363, + "mean_token_accuracy": 0.6738467216491699, + "num_tokens": 69640263.0, + "step": 2697 + }, + { + "epoch": 0.2962881616516582, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.6772940158843994, + "learning_rate": 9.871888726207906e-07, + "loss": 1.0433, + "mean_token_accuracy": 0.7018763422966003, + "num_tokens": 69662277.0, + "step": 2698 + }, + { + "epoch": 0.2963979793542719, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.4119455814361572, + "learning_rate": 9.875549048316252e-07, + "loss": 1.1206, + "mean_token_accuracy": 0.6766518354415894, + "num_tokens": 69694745.0, + "step": 2699 + }, + { + "epoch": 0.29650779705688557, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.648787021636963, + "learning_rate": 9.879209370424598e-07, + "loss": 1.0492, + "mean_token_accuracy": 0.6926373243331909, + "num_tokens": 69720316.0, + "step": 2700 + }, + { + "epoch": 0.2966176147594992, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.5067248344421387, + "learning_rate": 9.882869692532942e-07, + "loss": 1.0713, + "mean_token_accuracy": 0.6861767768859863, + "num_tokens": 69750100.0, + "step": 2701 + }, + { + "epoch": 0.2967274324621129, + "ewc_loss": 5.0961971282958984e-06, + "grad_norm": 2.5114896297454834, + "learning_rate": 9.886530014641288e-07, + "loss": 1.0823, + "mean_token_accuracy": 0.6866015195846558, + "num_tokens": 69777792.0, + "step": 2702 + }, + { + "epoch": 0.29683725016472656, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.5291922092437744, + "learning_rate": 9.890190336749634e-07, + "loss": 1.1022, + "mean_token_accuracy": 0.6710503101348877, + "num_tokens": 69804220.0, + "step": 2703 + }, + { + "epoch": 0.2969470678673402, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.409245729446411, + "learning_rate": 9.89385065885798e-07, + "loss": 1.1045, + "mean_token_accuracy": 0.6727712154388428, + "num_tokens": 69837893.0, + "step": 2704 + }, + { + "epoch": 0.29705688556995385, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.5973353385925293, + "learning_rate": 9.897510980966326e-07, + "loss": 1.0193, + "mean_token_accuracy": 0.7074886560440063, + "num_tokens": 69862269.0, + "step": 2705 + }, + { + "epoch": 0.29716670327256756, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.5477235317230225, + "learning_rate": 9.90117130307467e-07, + "loss": 1.117, + "mean_token_accuracy": 0.6729037761688232, + "num_tokens": 69893396.0, + "step": 2706 + }, + { + "epoch": 0.2972765209751812, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.548902750015259, + "learning_rate": 9.904831625183016e-07, + "loss": 1.0263, + "mean_token_accuracy": 0.6941673755645752, + "num_tokens": 69920648.0, + "step": 2707 + }, + { + "epoch": 0.29738633867779485, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.4109435081481934, + "learning_rate": 9.908491947291362e-07, + "loss": 1.0328, + "mean_token_accuracy": 0.6948115825653076, + "num_tokens": 69947077.0, + "step": 2708 + }, + { + "epoch": 0.29749615638040855, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.4829494953155518, + "learning_rate": 9.912152269399708e-07, + "loss": 1.1128, + "mean_token_accuracy": 0.681320309638977, + "num_tokens": 69977260.0, + "step": 2709 + }, + { + "epoch": 0.2976059740830222, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.5404903888702393, + "learning_rate": 9.915812591508052e-07, + "loss": 1.0347, + "mean_token_accuracy": 0.700497031211853, + "num_tokens": 70006314.0, + "step": 2710 + }, + { + "epoch": 0.29771579178563584, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.7568535804748535, + "learning_rate": 9.919472913616398e-07, + "loss": 1.0269, + "mean_token_accuracy": 0.6986655592918396, + "num_tokens": 70033098.0, + "step": 2711 + }, + { + "epoch": 0.2978256094882495, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.4577760696411133, + "learning_rate": 9.923133235724744e-07, + "loss": 1.044, + "mean_token_accuracy": 0.7013403177261353, + "num_tokens": 70062200.0, + "step": 2712 + }, + { + "epoch": 0.2979354271908632, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.4145267009735107, + "learning_rate": 9.926793557833088e-07, + "loss": 1.1041, + "mean_token_accuracy": 0.6934283971786499, + "num_tokens": 70089711.0, + "step": 2713 + }, + { + "epoch": 0.29804524489347684, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.436330556869507, + "learning_rate": 9.930453879941434e-07, + "loss": 0.9594, + "mean_token_accuracy": 0.7167000770568848, + "num_tokens": 70113978.0, + "step": 2714 + }, + { + "epoch": 0.2981550625960905, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.5852601528167725, + "learning_rate": 9.93411420204978e-07, + "loss": 1.0201, + "mean_token_accuracy": 0.6976138949394226, + "num_tokens": 70138238.0, + "step": 2715 + }, + { + "epoch": 0.2982648802987041, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.7189314365386963, + "learning_rate": 9.937774524158126e-07, + "loss": 1.0003, + "mean_token_accuracy": 0.7089048624038696, + "num_tokens": 70161906.0, + "step": 2716 + }, + { + "epoch": 0.29837469800131783, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.507108449935913, + "learning_rate": 9.941434846266472e-07, + "loss": 0.9961, + "mean_token_accuracy": 0.7044294476509094, + "num_tokens": 70192723.0, + "step": 2717 + }, + { + "epoch": 0.2984845157039315, + "ewc_loss": 5.125999450683594e-06, + "grad_norm": 2.63385009765625, + "learning_rate": 9.945095168374816e-07, + "loss": 1.0265, + "mean_token_accuracy": 0.6982603669166565, + "num_tokens": 70221730.0, + "step": 2718 + }, + { + "epoch": 0.2985943334065451, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.439483642578125, + "learning_rate": 9.948755490483162e-07, + "loss": 1.006, + "mean_token_accuracy": 0.7061803340911865, + "num_tokens": 70249072.0, + "step": 2719 + }, + { + "epoch": 0.2987041511091588, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.9760406017303467, + "learning_rate": 9.952415812591508e-07, + "loss": 1.0041, + "mean_token_accuracy": 0.7020602226257324, + "num_tokens": 70270451.0, + "step": 2720 + }, + { + "epoch": 0.29881396881177247, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.901176929473877, + "learning_rate": 9.956076134699854e-07, + "loss": 0.9526, + "mean_token_accuracy": 0.7198678851127625, + "num_tokens": 70290888.0, + "step": 2721 + }, + { + "epoch": 0.2989237865143861, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.849487543106079, + "learning_rate": 9.9597364568082e-07, + "loss": 1.1235, + "mean_token_accuracy": 0.6739348769187927, + "num_tokens": 70314777.0, + "step": 2722 + }, + { + "epoch": 0.29903360421699976, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.485194683074951, + "learning_rate": 9.963396778916544e-07, + "loss": 1.0135, + "mean_token_accuracy": 0.7034711837768555, + "num_tokens": 70343692.0, + "step": 2723 + }, + { + "epoch": 0.29914342191961346, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.55098032951355, + "learning_rate": 9.96705710102489e-07, + "loss": 1.03, + "mean_token_accuracy": 0.6931015849113464, + "num_tokens": 70367561.0, + "step": 2724 + }, + { + "epoch": 0.2992532396222271, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 16.73497200012207, + "learning_rate": 9.970717423133236e-07, + "loss": 0.9629, + "mean_token_accuracy": 0.7081543207168579, + "num_tokens": 70389900.0, + "step": 2725 + }, + { + "epoch": 0.29936305732484075, + "ewc_loss": 5.155801773071289e-06, + "grad_norm": 2.8910000324249268, + "learning_rate": 9.974377745241582e-07, + "loss": 1.003, + "mean_token_accuracy": 0.7062951326370239, + "num_tokens": 70415989.0, + "step": 2726 + }, + { + "epoch": 0.2994728750274544, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.4278697967529297, + "learning_rate": 9.978038067349926e-07, + "loss": 1.0961, + "mean_token_accuracy": 0.6803566217422485, + "num_tokens": 70448872.0, + "step": 2727 + }, + { + "epoch": 0.2995826927300681, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.505617618560791, + "learning_rate": 9.981698389458272e-07, + "loss": 1.0655, + "mean_token_accuracy": 0.6907676458358765, + "num_tokens": 70479507.0, + "step": 2728 + }, + { + "epoch": 0.29969251043268175, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.462087869644165, + "learning_rate": 9.985358711566618e-07, + "loss": 1.0108, + "mean_token_accuracy": 0.6993687152862549, + "num_tokens": 70510493.0, + "step": 2729 + }, + { + "epoch": 0.2998023281352954, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.8757710456848145, + "learning_rate": 9.989019033674961e-07, + "loss": 0.9727, + "mean_token_accuracy": 0.7077076435089111, + "num_tokens": 70532115.0, + "step": 2730 + }, + { + "epoch": 0.2999121458379091, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.2940053939819336, + "learning_rate": 9.99267935578331e-07, + "loss": 0.9434, + "mean_token_accuracy": 0.709068775177002, + "num_tokens": 70563634.0, + "step": 2731 + }, + { + "epoch": 0.30002196354052274, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.468924045562744, + "learning_rate": 9.996339677891654e-07, + "loss": 1.0323, + "mean_token_accuracy": 0.7044469118118286, + "num_tokens": 70591530.0, + "step": 2732 + }, + { + "epoch": 0.3001317812431364, + "ewc_loss": 5.185604095458984e-06, + "grad_norm": 2.933274030685425, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6904266476631165, + "num_tokens": 70611898.0, + "step": 2733 + }, + { + "epoch": 0.30024159894575003, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.9350638389587402, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.719449520111084, + "num_tokens": 70631779.0, + "step": 2734 + }, + { + "epoch": 0.30035141664836373, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.516130208969116, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.6940640211105347, + "num_tokens": 70660246.0, + "step": 2735 + }, + { + "epoch": 0.3004612343509774, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.576777696609497, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.6887282133102417, + "num_tokens": 70691353.0, + "step": 2736 + }, + { + "epoch": 0.300571052053591, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.796156167984009, + "learning_rate": 1e-06, + "loss": 1.1272, + "mean_token_accuracy": 0.6720760464668274, + "num_tokens": 70714143.0, + "step": 2737 + }, + { + "epoch": 0.3006808697562047, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.837113857269287, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6860107183456421, + "num_tokens": 70738976.0, + "step": 2738 + }, + { + "epoch": 0.3007906874588184, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.8032495975494385, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.7118974924087524, + "num_tokens": 70760448.0, + "step": 2739 + }, + { + "epoch": 0.300900505161432, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.4978573322296143, + "learning_rate": 1e-06, + "loss": 1.1283, + "mean_token_accuracy": 0.6673420667648315, + "num_tokens": 70786483.0, + "step": 2740 + }, + { + "epoch": 0.30101032286404567, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.596740484237671, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7084255218505859, + "num_tokens": 70809208.0, + "step": 2741 + }, + { + "epoch": 0.30112014056665937, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.7269155979156494, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7059305906295776, + "num_tokens": 70834608.0, + "step": 2742 + }, + { + "epoch": 0.301229958269273, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.429164171218872, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7101683616638184, + "num_tokens": 70861261.0, + "step": 2743 + }, + { + "epoch": 0.30133977597188666, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.533841371536255, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6853523254394531, + "num_tokens": 70889551.0, + "step": 2744 + }, + { + "epoch": 0.3014495936745003, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.5970711708068848, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6906023025512695, + "num_tokens": 70916162.0, + "step": 2745 + }, + { + "epoch": 0.301559411377114, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.577275276184082, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.706688404083252, + "num_tokens": 70944163.0, + "step": 2746 + }, + { + "epoch": 0.30166922907972765, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.635080337524414, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7033056020736694, + "num_tokens": 70970646.0, + "step": 2747 + }, + { + "epoch": 0.3017790467823413, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.2702600955963135, + "learning_rate": 1e-06, + "loss": 1.1123, + "mean_token_accuracy": 0.6756154298782349, + "num_tokens": 71003040.0, + "step": 2748 + }, + { + "epoch": 0.301888864484955, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 8.799352645874023, + "learning_rate": 1e-06, + "loss": 1.1046, + "mean_token_accuracy": 0.6794520616531372, + "num_tokens": 71034419.0, + "step": 2749 + }, + { + "epoch": 0.30199868218756865, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.790011405944824, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6928912401199341, + "num_tokens": 71060266.0, + "step": 2750 + }, + { + "epoch": 0.3021084998901823, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 16.718299865722656, + "learning_rate": 1e-06, + "loss": 1.1484, + "mean_token_accuracy": 0.6642529964447021, + "num_tokens": 71085939.0, + "step": 2751 + }, + { + "epoch": 0.30221831759279594, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.8961973190307617, + "learning_rate": 1e-06, + "loss": 1.0523, + "mean_token_accuracy": 0.6900638341903687, + "num_tokens": 71110518.0, + "step": 2752 + }, + { + "epoch": 0.30232813529540964, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.8116581439971924, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.7060725092887878, + "num_tokens": 71132356.0, + "step": 2753 + }, + { + "epoch": 0.3024379529980233, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.723808765411377, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.690561830997467, + "num_tokens": 71155990.0, + "step": 2754 + }, + { + "epoch": 0.30254777070063693, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.856123208999634, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7155237197875977, + "num_tokens": 71176187.0, + "step": 2755 + }, + { + "epoch": 0.3026575884032506, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.5592663288116455, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.697028636932373, + "num_tokens": 71200764.0, + "step": 2756 + }, + { + "epoch": 0.3027674061058643, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.64915132522583, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6923973560333252, + "num_tokens": 71228226.0, + "step": 2757 + }, + { + "epoch": 0.3028772238084779, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.6151599884033203, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6888425350189209, + "num_tokens": 71252338.0, + "step": 2758 + }, + { + "epoch": 0.30298704151109157, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.7466466426849365, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6791922450065613, + "num_tokens": 71276029.0, + "step": 2759 + }, + { + "epoch": 0.30309685921370527, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.5595195293426514, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7148339152336121, + "num_tokens": 71300798.0, + "step": 2760 + }, + { + "epoch": 0.3032066769163189, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.357083559036255, + "learning_rate": 1e-06, + "loss": 1.132, + "mean_token_accuracy": 0.6697695255279541, + "num_tokens": 71335051.0, + "step": 2761 + }, + { + "epoch": 0.30331649461893256, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 2.660358190536499, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6864631772041321, + "num_tokens": 71361661.0, + "step": 2762 + }, + { + "epoch": 0.3034263123215462, + "ewc_loss": 5.21540641784668e-06, + "grad_norm": 3.269141912460327, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.6971182823181152, + "num_tokens": 71382572.0, + "step": 2763 + }, + { + "epoch": 0.3035361300241599, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.9097678661346436, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7124857902526855, + "num_tokens": 71403425.0, + "step": 2764 + }, + { + "epoch": 0.30364594772677356, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6825215816497803, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6926168203353882, + "num_tokens": 71428908.0, + "step": 2765 + }, + { + "epoch": 0.3037557654293872, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.560243606567383, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7142676711082458, + "num_tokens": 71454772.0, + "step": 2766 + }, + { + "epoch": 0.3038655831320009, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6794509887695312, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6823383569717407, + "num_tokens": 71482693.0, + "step": 2767 + }, + { + "epoch": 0.30397540083461455, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.7076337337493896, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7138175368309021, + "num_tokens": 71504932.0, + "step": 2768 + }, + { + "epoch": 0.3040852185372282, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 3.092952013015747, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7065443992614746, + "num_tokens": 71524792.0, + "step": 2769 + }, + { + "epoch": 0.30419503623984184, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.5028467178344727, + "learning_rate": 1e-06, + "loss": 1.1653, + "mean_token_accuracy": 0.6621488332748413, + "num_tokens": 71555938.0, + "step": 2770 + }, + { + "epoch": 0.30430485394245554, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.8536155223846436, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6831897497177124, + "num_tokens": 71582278.0, + "step": 2771 + }, + { + "epoch": 0.3044146716450692, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6263253688812256, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6930465698242188, + "num_tokens": 71611249.0, + "step": 2772 + }, + { + "epoch": 0.30452448934768284, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4233782291412354, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6930370330810547, + "num_tokens": 71640985.0, + "step": 2773 + }, + { + "epoch": 0.3046343070502965, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.7699007987976074, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7019639015197754, + "num_tokens": 71663987.0, + "step": 2774 + }, + { + "epoch": 0.3047441247529102, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.995734691619873, + "learning_rate": 1e-06, + "loss": 1.1146, + "mean_token_accuracy": 0.6786407232284546, + "num_tokens": 71688019.0, + "step": 2775 + }, + { + "epoch": 0.30485394245552383, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.643099308013916, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7131969928741455, + "num_tokens": 71715223.0, + "step": 2776 + }, + { + "epoch": 0.3049637601581375, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6454286575317383, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6759301424026489, + "num_tokens": 71742935.0, + "step": 2777 + }, + { + "epoch": 0.3050735778607512, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6238343715667725, + "learning_rate": 1e-06, + "loss": 1.1648, + "mean_token_accuracy": 0.6654417514801025, + "num_tokens": 71774364.0, + "step": 2778 + }, + { + "epoch": 0.3051833955633648, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.395746946334839, + "learning_rate": 1e-06, + "loss": 1.1661, + "mean_token_accuracy": 0.6668973565101624, + "num_tokens": 71808027.0, + "step": 2779 + }, + { + "epoch": 0.30529321326597847, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.7433080673217773, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7067239284515381, + "num_tokens": 71830722.0, + "step": 2780 + }, + { + "epoch": 0.3054030309685921, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.7793164253234863, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.69773268699646, + "num_tokens": 71852928.0, + "step": 2781 + }, + { + "epoch": 0.3055128486712058, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4482381343841553, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6900689005851746, + "num_tokens": 71882433.0, + "step": 2782 + }, + { + "epoch": 0.30562266637381946, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4421913623809814, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6848160624504089, + "num_tokens": 71914594.0, + "step": 2783 + }, + { + "epoch": 0.3057324840764331, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6842076778411865, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6792610883712769, + "num_tokens": 71938171.0, + "step": 2784 + }, + { + "epoch": 0.30584230177904675, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.475001335144043, + "learning_rate": 1e-06, + "loss": 1.1244, + "mean_token_accuracy": 0.6883859634399414, + "num_tokens": 71966619.0, + "step": 2785 + }, + { + "epoch": 0.30595211948166046, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.5684335231781006, + "learning_rate": 1e-06, + "loss": 1.1459, + "mean_token_accuracy": 0.6648547053337097, + "num_tokens": 71996236.0, + "step": 2786 + }, + { + "epoch": 0.3060619371842741, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6590521335601807, + "learning_rate": 1e-06, + "loss": 1.1247, + "mean_token_accuracy": 0.6739175915718079, + "num_tokens": 72020186.0, + "step": 2787 + }, + { + "epoch": 0.30617175488688775, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4796299934387207, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.703526496887207, + "num_tokens": 72045722.0, + "step": 2788 + }, + { + "epoch": 0.30628157258950145, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.7395684719085693, + "learning_rate": 1e-06, + "loss": 1.1415, + "mean_token_accuracy": 0.6715550422668457, + "num_tokens": 72072531.0, + "step": 2789 + }, + { + "epoch": 0.3063913902921151, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.7476115226745605, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6950649619102478, + "num_tokens": 72098928.0, + "step": 2790 + }, + { + "epoch": 0.30650120799472874, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.605599880218506, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6849485635757446, + "num_tokens": 72123697.0, + "step": 2791 + }, + { + "epoch": 0.3066110256973424, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 3.1352555751800537, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7025445699691772, + "num_tokens": 72144223.0, + "step": 2792 + }, + { + "epoch": 0.3067208433999561, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 3.6249887943267822, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7055926322937012, + "num_tokens": 72167845.0, + "step": 2793 + }, + { + "epoch": 0.30683066110256974, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.689741373062134, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6953438520431519, + "num_tokens": 72190607.0, + "step": 2794 + }, + { + "epoch": 0.3069404788051834, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6881563663482666, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.687892735004425, + "num_tokens": 72216993.0, + "step": 2795 + }, + { + "epoch": 0.3070502965077971, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.680356502532959, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.704883873462677, + "num_tokens": 72240619.0, + "step": 2796 + }, + { + "epoch": 0.30716011421041073, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.879760503768921, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6859328746795654, + "num_tokens": 72261970.0, + "step": 2797 + }, + { + "epoch": 0.3072699319130244, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.807443857192993, + "learning_rate": 1e-06, + "loss": 1.1205, + "mean_token_accuracy": 0.6783934831619263, + "num_tokens": 72286977.0, + "step": 2798 + }, + { + "epoch": 0.307379749615638, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6354832649230957, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6880558729171753, + "num_tokens": 72312612.0, + "step": 2799 + }, + { + "epoch": 0.3074895673182517, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 3.1635353565216064, + "learning_rate": 1e-06, + "loss": 1.1077, + "mean_token_accuracy": 0.6828405857086182, + "num_tokens": 72342397.0, + "step": 2800 + }, + { + "epoch": 0.30759938502086537, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4471869468688965, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6833078861236572, + "num_tokens": 72373431.0, + "step": 2801 + }, + { + "epoch": 0.307709202723479, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.737555503845215, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6918208599090576, + "num_tokens": 72399898.0, + "step": 2802 + }, + { + "epoch": 0.30781902042609266, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6240956783294678, + "learning_rate": 1e-06, + "loss": 0.8419, + "mean_token_accuracy": 0.749521017074585, + "num_tokens": 72421490.0, + "step": 2803 + }, + { + "epoch": 0.30792883812870636, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.8395023345947266, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.714465320110321, + "num_tokens": 72441399.0, + "step": 2804 + }, + { + "epoch": 0.30803865583132, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.9358930587768555, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7190612554550171, + "num_tokens": 72465359.0, + "step": 2805 + }, + { + "epoch": 0.30814847353393365, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.5264480113983154, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7050498127937317, + "num_tokens": 72492832.0, + "step": 2806 + }, + { + "epoch": 0.30825829123654735, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6873157024383545, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6888387799263, + "num_tokens": 72517740.0, + "step": 2807 + }, + { + "epoch": 0.308368108939161, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4490506649017334, + "learning_rate": 1e-06, + "loss": 0.9465, + "mean_token_accuracy": 0.716074526309967, + "num_tokens": 72546634.0, + "step": 2808 + }, + { + "epoch": 0.30847792664177465, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.441460609436035, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.6940792798995972, + "num_tokens": 72574884.0, + "step": 2809 + }, + { + "epoch": 0.3085877443443883, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.594169855117798, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7101012468338013, + "num_tokens": 72599568.0, + "step": 2810 + }, + { + "epoch": 0.308697562047002, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.692450761795044, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7034585475921631, + "num_tokens": 72623881.0, + "step": 2811 + }, + { + "epoch": 0.30880737974961564, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.3624284267425537, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6957091093063354, + "num_tokens": 72656182.0, + "step": 2812 + }, + { + "epoch": 0.3089171974522293, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5398755073547363, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7122248411178589, + "num_tokens": 72681467.0, + "step": 2813 + }, + { + "epoch": 0.309027015154843, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.793994665145874, + "learning_rate": 1e-06, + "loss": 1.1259, + "mean_token_accuracy": 0.6812294125556946, + "num_tokens": 72704525.0, + "step": 2814 + }, + { + "epoch": 0.30913683285745663, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.6683237552642822, + "learning_rate": 1e-06, + "loss": 1.0779, + "mean_token_accuracy": 0.6843199133872986, + "num_tokens": 72729940.0, + "step": 2815 + }, + { + "epoch": 0.3092466505600703, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.6906867027282715, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7073836326599121, + "num_tokens": 72754018.0, + "step": 2816 + }, + { + "epoch": 0.3093564682626839, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6591804027557373, + "learning_rate": 1e-06, + "loss": 1.1114, + "mean_token_accuracy": 0.6817785501480103, + "num_tokens": 72780723.0, + "step": 2817 + }, + { + "epoch": 0.3094662859652976, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.6590425968170166, + "learning_rate": 1e-06, + "loss": 1.0989, + "mean_token_accuracy": 0.6783696413040161, + "num_tokens": 72809617.0, + "step": 2818 + }, + { + "epoch": 0.3095761036679113, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.926297903060913, + "learning_rate": 1e-06, + "loss": 0.8847, + "mean_token_accuracy": 0.7326509356498718, + "num_tokens": 72829097.0, + "step": 2819 + }, + { + "epoch": 0.3096859213705249, + "ewc_loss": 5.245208740234375e-06, + "grad_norm": 2.4982895851135254, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7046534419059753, + "num_tokens": 72856357.0, + "step": 2820 + }, + { + "epoch": 0.30979573907313857, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.6685242652893066, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7068101167678833, + "num_tokens": 72882233.0, + "step": 2821 + }, + { + "epoch": 0.30990555677575227, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.7312591075897217, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7071499824523926, + "num_tokens": 72905153.0, + "step": 2822 + }, + { + "epoch": 0.3100153744783659, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.9860734939575195, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.716645359992981, + "num_tokens": 72926296.0, + "step": 2823 + }, + { + "epoch": 0.31012519218097956, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5082719326019287, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6897889375686646, + "num_tokens": 72955950.0, + "step": 2824 + }, + { + "epoch": 0.31023500988359326, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.6507482528686523, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6789966821670532, + "num_tokens": 72982794.0, + "step": 2825 + }, + { + "epoch": 0.3103448275862069, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.524142265319824, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.6970819234848022, + "num_tokens": 73007331.0, + "step": 2826 + }, + { + "epoch": 0.31045464528882055, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.930906295776367, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.693468451499939, + "num_tokens": 73027103.0, + "step": 2827 + }, + { + "epoch": 0.3105644629914342, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.676459789276123, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6934208273887634, + "num_tokens": 73049063.0, + "step": 2828 + }, + { + "epoch": 0.3106742806940479, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.435579538345337, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.6941827535629272, + "num_tokens": 73078869.0, + "step": 2829 + }, + { + "epoch": 0.31078409839666155, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.6404614448547363, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6964962482452393, + "num_tokens": 73105191.0, + "step": 2830 + }, + { + "epoch": 0.3108939160992752, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5925042629241943, + "learning_rate": 1e-06, + "loss": 1.0984, + "mean_token_accuracy": 0.6866328716278076, + "num_tokens": 73131768.0, + "step": 2831 + }, + { + "epoch": 0.31100373380188884, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5922515392303467, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6956195831298828, + "num_tokens": 73158912.0, + "step": 2832 + }, + { + "epoch": 0.31111355150450254, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.7908565998077393, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.6987905502319336, + "num_tokens": 73183322.0, + "step": 2833 + }, + { + "epoch": 0.3112233692071162, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5094401836395264, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6964693665504456, + "num_tokens": 73212338.0, + "step": 2834 + }, + { + "epoch": 0.31133318690972983, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.579771041870117, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6882607936859131, + "num_tokens": 73241951.0, + "step": 2835 + }, + { + "epoch": 0.31144300461234353, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5178253650665283, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7159672379493713, + "num_tokens": 73269042.0, + "step": 2836 + }, + { + "epoch": 0.3115528223149572, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.826657772064209, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.6981613039970398, + "num_tokens": 73295104.0, + "step": 2837 + }, + { + "epoch": 0.3116626400175708, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.5799829959869385, + "learning_rate": 1e-06, + "loss": 1.1035, + "mean_token_accuracy": 0.6824434995651245, + "num_tokens": 73322207.0, + "step": 2838 + }, + { + "epoch": 0.31177245772018447, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.90230393409729, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.689008355140686, + "num_tokens": 73344542.0, + "step": 2839 + }, + { + "epoch": 0.31188227542279817, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.7585930824279785, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.70269775390625, + "num_tokens": 73366839.0, + "step": 2840 + }, + { + "epoch": 0.3119920931254118, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.614034414291382, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.6964725255966187, + "num_tokens": 73390568.0, + "step": 2841 + }, + { + "epoch": 0.31210191082802546, + "ewc_loss": 5.27501106262207e-06, + "grad_norm": 2.692077398300171, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7103999853134155, + "num_tokens": 73411547.0, + "step": 2842 + }, + { + "epoch": 0.31221172853063917, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.739210605621338, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6929970979690552, + "num_tokens": 73436034.0, + "step": 2843 + }, + { + "epoch": 0.3123215462332528, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.4908158779144287, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6927679181098938, + "num_tokens": 73461319.0, + "step": 2844 + }, + { + "epoch": 0.31243136393586646, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.8563029766082764, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.690990686416626, + "num_tokens": 73483607.0, + "step": 2845 + }, + { + "epoch": 0.3125411816384801, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.550739049911499, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6802005767822266, + "num_tokens": 73510148.0, + "step": 2846 + }, + { + "epoch": 0.3126509993410938, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 3.22902774810791, + "learning_rate": 1e-06, + "loss": 0.8734, + "mean_token_accuracy": 0.7338007092475891, + "num_tokens": 73527606.0, + "step": 2847 + }, + { + "epoch": 0.31276081704370745, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.949418067932129, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.6992303729057312, + "num_tokens": 73548554.0, + "step": 2848 + }, + { + "epoch": 0.3128706347463211, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.421678066253662, + "learning_rate": 1e-06, + "loss": 0.9233, + "mean_token_accuracy": 0.7232682704925537, + "num_tokens": 73575796.0, + "step": 2849 + }, + { + "epoch": 0.31298045244893474, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.45884108543396, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6971120238304138, + "num_tokens": 73605313.0, + "step": 2850 + }, + { + "epoch": 0.31309027015154844, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.6479949951171875, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6874600648880005, + "num_tokens": 73630136.0, + "step": 2851 + }, + { + "epoch": 0.3132000878541621, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.7566299438476562, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7110249996185303, + "num_tokens": 73650701.0, + "step": 2852 + }, + { + "epoch": 0.31330990555677574, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.6078195571899414, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7167139649391174, + "num_tokens": 73675451.0, + "step": 2853 + }, + { + "epoch": 0.31341972325938944, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.747133255004883, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.689655065536499, + "num_tokens": 73699631.0, + "step": 2854 + }, + { + "epoch": 0.3135295409620031, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.421956777572632, + "learning_rate": 1e-06, + "loss": 1.1821, + "mean_token_accuracy": 0.6668689846992493, + "num_tokens": 73734841.0, + "step": 2855 + }, + { + "epoch": 0.31363935866461673, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.496131658554077, + "learning_rate": 1e-06, + "loss": 1.0995, + "mean_token_accuracy": 0.6758323907852173, + "num_tokens": 73766927.0, + "step": 2856 + }, + { + "epoch": 0.3137491763672304, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.4638075828552246, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6960417032241821, + "num_tokens": 73794957.0, + "step": 2857 + }, + { + "epoch": 0.3138589940698441, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.8882341384887695, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.6989628076553345, + "num_tokens": 73814952.0, + "step": 2858 + }, + { + "epoch": 0.3139688117724577, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.5591626167297363, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6902790665626526, + "num_tokens": 73840972.0, + "step": 2859 + }, + { + "epoch": 0.31407862947507137, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.4668796062469482, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6812117695808411, + "num_tokens": 73871707.0, + "step": 2860 + }, + { + "epoch": 0.314188447177685, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.702655792236328, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7045074105262756, + "num_tokens": 73896564.0, + "step": 2861 + }, + { + "epoch": 0.3142982648802987, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.611280918121338, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7017477750778198, + "num_tokens": 73923668.0, + "step": 2862 + }, + { + "epoch": 0.31440808258291236, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.501915216445923, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6860900521278381, + "num_tokens": 73949813.0, + "step": 2863 + }, + { + "epoch": 0.314517900285526, + "ewc_loss": 5.304813385009766e-06, + "grad_norm": 2.8000974655151367, + "learning_rate": 1e-06, + "loss": 1.1009, + "mean_token_accuracy": 0.6864693760871887, + "num_tokens": 73972394.0, + "step": 2864 + }, + { + "epoch": 0.3146277179881397, + "ewc_loss": 5.334615707397461e-06, + "grad_norm": 2.6697683334350586, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.6940702199935913, + "num_tokens": 73996907.0, + "step": 2865 + }, + { + "epoch": 0.31473753569075336, + "ewc_loss": 5.334615707397461e-06, + "grad_norm": 2.602569341659546, + "learning_rate": 1e-06, + "loss": 0.9273, + "mean_token_accuracy": 0.7234959006309509, + "num_tokens": 74023279.0, + "step": 2866 + }, + { + "epoch": 0.314847353393367, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.394803524017334, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6961524486541748, + "num_tokens": 74055139.0, + "step": 2867 + }, + { + "epoch": 0.31495717109598065, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.79976224899292, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7019789218902588, + "num_tokens": 74077381.0, + "step": 2868 + }, + { + "epoch": 0.31506698879859435, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.7125439643859863, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6956050395965576, + "num_tokens": 74102507.0, + "step": 2869 + }, + { + "epoch": 0.315176806501208, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.5996744632720947, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6811217665672302, + "num_tokens": 74131402.0, + "step": 2870 + }, + { + "epoch": 0.31528662420382164, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.9986565113067627, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7120493650436401, + "num_tokens": 74150490.0, + "step": 2871 + }, + { + "epoch": 0.31539644190643534, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.5835256576538086, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6790914535522461, + "num_tokens": 74177232.0, + "step": 2872 + }, + { + "epoch": 0.315506259609049, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.685460090637207, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.7143339514732361, + "num_tokens": 74201663.0, + "step": 2873 + }, + { + "epoch": 0.31561607731166264, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.5788965225219727, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.699978232383728, + "num_tokens": 74225721.0, + "step": 2874 + }, + { + "epoch": 0.3157258950142763, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.3764400482177734, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6990532875061035, + "num_tokens": 74254833.0, + "step": 2875 + }, + { + "epoch": 0.31583571271689, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.5777804851531982, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.6954109072685242, + "num_tokens": 74280086.0, + "step": 2876 + }, + { + "epoch": 0.31594553041950363, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.529968500137329, + "learning_rate": 1e-06, + "loss": 1.1622, + "mean_token_accuracy": 0.6784085035324097, + "num_tokens": 74308202.0, + "step": 2877 + }, + { + "epoch": 0.3160553481221173, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.6888587474823, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.6944443583488464, + "num_tokens": 74340124.0, + "step": 2878 + }, + { + "epoch": 0.3161651658247309, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.897219181060791, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7014894485473633, + "num_tokens": 74363775.0, + "step": 2879 + }, + { + "epoch": 0.3162749835273446, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.7713704109191895, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7120339274406433, + "num_tokens": 74385059.0, + "step": 2880 + }, + { + "epoch": 0.31638480122995827, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.350865125656128, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6887146234512329, + "num_tokens": 74415274.0, + "step": 2881 + }, + { + "epoch": 0.3164946189325719, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 3.092069387435913, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6938601732254028, + "num_tokens": 74433509.0, + "step": 2882 + }, + { + "epoch": 0.3166044366351856, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.57500958442688, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.7075452208518982, + "num_tokens": 74456649.0, + "step": 2883 + }, + { + "epoch": 0.31671425433779926, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.7994837760925293, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.6960873603820801, + "num_tokens": 74478535.0, + "step": 2884 + }, + { + "epoch": 0.3168240720404129, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.6588361263275146, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.708114743232727, + "num_tokens": 74501778.0, + "step": 2885 + }, + { + "epoch": 0.31693388974302655, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.504124879837036, + "learning_rate": 1e-06, + "loss": 1.1287, + "mean_token_accuracy": 0.6732875108718872, + "num_tokens": 74530204.0, + "step": 2886 + }, + { + "epoch": 0.31704370744564025, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.7866320610046387, + "learning_rate": 1e-06, + "loss": 0.869, + "mean_token_accuracy": 0.7350136041641235, + "num_tokens": 74552726.0, + "step": 2887 + }, + { + "epoch": 0.3171535251482539, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.730982542037964, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7025039196014404, + "num_tokens": 74574343.0, + "step": 2888 + }, + { + "epoch": 0.31726334285086755, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.831271171569824, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7031186819076538, + "num_tokens": 74597489.0, + "step": 2889 + }, + { + "epoch": 0.31737316055348125, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.38266921043396, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6961847543716431, + "num_tokens": 74630000.0, + "step": 2890 + }, + { + "epoch": 0.3174829782560949, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.5545237064361572, + "learning_rate": 1e-06, + "loss": 1.1361, + "mean_token_accuracy": 0.6759889721870422, + "num_tokens": 74656838.0, + "step": 2891 + }, + { + "epoch": 0.31759279595870854, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.565382480621338, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6888538002967834, + "num_tokens": 74680459.0, + "step": 2892 + }, + { + "epoch": 0.3177026136613222, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.359236240386963, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7056766748428345, + "num_tokens": 74709266.0, + "step": 2893 + }, + { + "epoch": 0.3178124313639359, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.4240968227386475, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7136403918266296, + "num_tokens": 74737699.0, + "step": 2894 + }, + { + "epoch": 0.31792224906654953, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.7004904747009277, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7051906585693359, + "num_tokens": 74761079.0, + "step": 2895 + }, + { + "epoch": 0.3180320667691632, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.758204698562622, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6999235153198242, + "num_tokens": 74790938.0, + "step": 2896 + }, + { + "epoch": 0.3181418844717768, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.3867595195770264, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7055854797363281, + "num_tokens": 74819771.0, + "step": 2897 + }, + { + "epoch": 0.3182517021743905, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.6500585079193115, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6968219876289368, + "num_tokens": 74845767.0, + "step": 2898 + }, + { + "epoch": 0.3183615198770042, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.471114158630371, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7047138214111328, + "num_tokens": 74873917.0, + "step": 2899 + }, + { + "epoch": 0.3184713375796178, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.6253089904785156, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7144560813903809, + "num_tokens": 74898683.0, + "step": 2900 + }, + { + "epoch": 0.3185811552822315, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.5766382217407227, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6889957189559937, + "num_tokens": 74923942.0, + "step": 2901 + }, + { + "epoch": 0.31869097298484517, + "ewc_loss": 5.364418029785156e-06, + "grad_norm": 2.5264668464660645, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7073134183883667, + "num_tokens": 74950771.0, + "step": 2902 + }, + { + "epoch": 0.3188007906874588, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.8877687454223633, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6930537819862366, + "num_tokens": 74973253.0, + "step": 2903 + }, + { + "epoch": 0.31891060839007246, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.6434853076934814, + "learning_rate": 1e-06, + "loss": 1.1178, + "mean_token_accuracy": 0.678510308265686, + "num_tokens": 74999470.0, + "step": 2904 + }, + { + "epoch": 0.31902042609268616, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.9470229148864746, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6851704120635986, + "num_tokens": 75019191.0, + "step": 2905 + }, + { + "epoch": 0.3191302437952998, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.548671245574951, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6841878890991211, + "num_tokens": 75047634.0, + "step": 2906 + }, + { + "epoch": 0.31924006149791345, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.480318069458008, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7030782699584961, + "num_tokens": 75077206.0, + "step": 2907 + }, + { + "epoch": 0.3193498792005271, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.7111620903015137, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6978355646133423, + "num_tokens": 75103531.0, + "step": 2908 + }, + { + "epoch": 0.3194596969031408, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.284599781036377, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7113766670227051, + "num_tokens": 75135218.0, + "step": 2909 + }, + { + "epoch": 0.31956951460575445, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.6376616954803467, + "learning_rate": 1e-06, + "loss": 1.1116, + "mean_token_accuracy": 0.6700279712677002, + "num_tokens": 75162189.0, + "step": 2910 + }, + { + "epoch": 0.3196793323083681, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.777684450149536, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.72407466173172, + "num_tokens": 75183522.0, + "step": 2911 + }, + { + "epoch": 0.3197891500109818, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.523418664932251, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7027181386947632, + "num_tokens": 75211860.0, + "step": 2912 + }, + { + "epoch": 0.31989896771359544, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.2723233699798584, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6825605630874634, + "num_tokens": 75250506.0, + "step": 2913 + }, + { + "epoch": 0.3200087854162091, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.2926788330078125, + "learning_rate": 1e-06, + "loss": 1.1476, + "mean_token_accuracy": 0.6723284721374512, + "num_tokens": 75285783.0, + "step": 2914 + }, + { + "epoch": 0.32011860311882273, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.468907117843628, + "learning_rate": 1e-06, + "loss": 1.1799, + "mean_token_accuracy": 0.6544479727745056, + "num_tokens": 75318059.0, + "step": 2915 + }, + { + "epoch": 0.32022842082143643, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.4914841651916504, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6931533813476562, + "num_tokens": 75345039.0, + "step": 2916 + }, + { + "epoch": 0.3203382385240501, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.338000535964966, + "learning_rate": 1e-06, + "loss": 1.1279, + "mean_token_accuracy": 0.6809494495391846, + "num_tokens": 75374543.0, + "step": 2917 + }, + { + "epoch": 0.3204480562266637, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.738084316253662, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6988140344619751, + "num_tokens": 75397039.0, + "step": 2918 + }, + { + "epoch": 0.3205578739292774, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.554776191711426, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6887246370315552, + "num_tokens": 75422381.0, + "step": 2919 + }, + { + "epoch": 0.32066769163189107, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.9415578842163086, + "learning_rate": 1e-06, + "loss": 0.9332, + "mean_token_accuracy": 0.717664897441864, + "num_tokens": 75446255.0, + "step": 2920 + }, + { + "epoch": 0.3207775093345047, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.739316463470459, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7188667058944702, + "num_tokens": 75468639.0, + "step": 2921 + }, + { + "epoch": 0.32088732703711836, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.530405044555664, + "learning_rate": 1e-06, + "loss": 0.9376, + "mean_token_accuracy": 0.7181963920593262, + "num_tokens": 75492994.0, + "step": 2922 + }, + { + "epoch": 0.32099714473973207, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.4341351985931396, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6893236637115479, + "num_tokens": 75520922.0, + "step": 2923 + }, + { + "epoch": 0.3211069624423457, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.463870048522949, + "learning_rate": 1e-06, + "loss": 1.1067, + "mean_token_accuracy": 0.6866979002952576, + "num_tokens": 75548894.0, + "step": 2924 + }, + { + "epoch": 0.32121678014495936, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.819854259490967, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6899926662445068, + "num_tokens": 75569737.0, + "step": 2925 + }, + { + "epoch": 0.321326597847573, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.5807790756225586, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.6985512375831604, + "num_tokens": 75594651.0, + "step": 2926 + }, + { + "epoch": 0.3214364155501867, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.495934247970581, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6957700252532959, + "num_tokens": 75620646.0, + "step": 2927 + }, + { + "epoch": 0.32154623325280035, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.732422351837158, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.712134599685669, + "num_tokens": 75642082.0, + "step": 2928 + }, + { + "epoch": 0.321656050955414, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.485128164291382, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6857584714889526, + "num_tokens": 75671471.0, + "step": 2929 + }, + { + "epoch": 0.3217658686580277, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.5321085453033447, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6837632060050964, + "num_tokens": 75698969.0, + "step": 2930 + }, + { + "epoch": 0.32187568636064134, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.6388964653015137, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7250633835792542, + "num_tokens": 75726030.0, + "step": 2931 + }, + { + "epoch": 0.321985504063255, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.7067580223083496, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7082127928733826, + "num_tokens": 75749806.0, + "step": 2932 + }, + { + "epoch": 0.32209532176586864, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 2.8837270736694336, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6992062926292419, + "num_tokens": 75774356.0, + "step": 2933 + }, + { + "epoch": 0.32220513946848234, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.7960665225982666, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6839390993118286, + "num_tokens": 75801477.0, + "step": 2934 + }, + { + "epoch": 0.322314957171096, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.5261130332946777, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7089777588844299, + "num_tokens": 75826278.0, + "step": 2935 + }, + { + "epoch": 0.32242477487370963, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 80.53695678710938, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6978000402450562, + "num_tokens": 75853384.0, + "step": 2936 + }, + { + "epoch": 0.3225345925763233, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.9256515502929688, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.7144752740859985, + "num_tokens": 75875598.0, + "step": 2937 + }, + { + "epoch": 0.322644410278937, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.730199098587036, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.697263777256012, + "num_tokens": 75901014.0, + "step": 2938 + }, + { + "epoch": 0.3227542279815506, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 16.69749641418457, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7048759460449219, + "num_tokens": 75927081.0, + "step": 2939 + }, + { + "epoch": 0.32286404568416427, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.829906463623047, + "learning_rate": 1e-06, + "loss": 0.8938, + "mean_token_accuracy": 0.7338824272155762, + "num_tokens": 75947874.0, + "step": 2940 + }, + { + "epoch": 0.32297386338677797, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.5922861099243164, + "learning_rate": 1e-06, + "loss": 1.1257, + "mean_token_accuracy": 0.6759582757949829, + "num_tokens": 75974473.0, + "step": 2941 + }, + { + "epoch": 0.3230836810893916, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.8927032947540283, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7025012373924255, + "num_tokens": 75997171.0, + "step": 2942 + }, + { + "epoch": 0.32319349879200526, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7489113807678223, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.7021099328994751, + "num_tokens": 76020257.0, + "step": 2943 + }, + { + "epoch": 0.3233033164946189, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.7799031734466553, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7023555636405945, + "num_tokens": 76042908.0, + "step": 2944 + }, + { + "epoch": 0.3234131341972326, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.712552547454834, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6961485743522644, + "num_tokens": 76071996.0, + "step": 2945 + }, + { + "epoch": 0.32352295189984626, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.556138515472412, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7213406562805176, + "num_tokens": 76095474.0, + "step": 2946 + }, + { + "epoch": 0.3236327696024599, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.731301784515381, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7094883918762207, + "num_tokens": 76119182.0, + "step": 2947 + }, + { + "epoch": 0.3237425873050736, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.624314069747925, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7104094624519348, + "num_tokens": 76141518.0, + "step": 2948 + }, + { + "epoch": 0.32385240500768725, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.5030548572540283, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.6953144073486328, + "num_tokens": 76166749.0, + "step": 2949 + }, + { + "epoch": 0.3239622227103009, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.483731985092163, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.7183504104614258, + "num_tokens": 76194031.0, + "step": 2950 + }, + { + "epoch": 0.32407204041291454, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.5694825649261475, + "learning_rate": 1e-06, + "loss": 1.1674, + "mean_token_accuracy": 0.6702003479003906, + "num_tokens": 76220946.0, + "step": 2951 + }, + { + "epoch": 0.32418185811552824, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.523216962814331, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6967096328735352, + "num_tokens": 76247852.0, + "step": 2952 + }, + { + "epoch": 0.3242916758181419, + "ewc_loss": 5.3942203521728516e-06, + "grad_norm": 3.112985849380493, + "learning_rate": 1e-06, + "loss": 0.8745, + "mean_token_accuracy": 0.7423304319381714, + "num_tokens": 76265952.0, + "step": 2953 + }, + { + "epoch": 0.32440149352075554, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.510991334915161, + "learning_rate": 1e-06, + "loss": 1.1044, + "mean_token_accuracy": 0.6850488185882568, + "num_tokens": 76295479.0, + "step": 2954 + }, + { + "epoch": 0.3245113112233692, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.687267303466797, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7114598751068115, + "num_tokens": 76319832.0, + "step": 2955 + }, + { + "epoch": 0.3246211289259829, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.857944965362549, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6977607011795044, + "num_tokens": 76346144.0, + "step": 2956 + }, + { + "epoch": 0.32473094662859653, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.8634469509124756, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7116277813911438, + "num_tokens": 76373994.0, + "step": 2957 + }, + { + "epoch": 0.3248407643312102, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.5041308403015137, + "learning_rate": 1e-06, + "loss": 1.1525, + "mean_token_accuracy": 0.6656064987182617, + "num_tokens": 76404158.0, + "step": 2958 + }, + { + "epoch": 0.3249505820338239, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.457312822341919, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6873951554298401, + "num_tokens": 76440125.0, + "step": 2959 + }, + { + "epoch": 0.3250603997364375, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.868006944656372, + "learning_rate": 1e-06, + "loss": 1.1011, + "mean_token_accuracy": 0.6798990964889526, + "num_tokens": 76464590.0, + "step": 2960 + }, + { + "epoch": 0.32517021743905117, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.685170888900757, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6834303736686707, + "num_tokens": 76488566.0, + "step": 2961 + }, + { + "epoch": 0.3252800351416648, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.4200000762939453, + "learning_rate": 1e-06, + "loss": 1.1446, + "mean_token_accuracy": 0.6798770427703857, + "num_tokens": 76517572.0, + "step": 2962 + }, + { + "epoch": 0.3253898528442785, + "ewc_loss": 5.424022674560547e-06, + "grad_norm": 2.732839345932007, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6872826218605042, + "num_tokens": 76539776.0, + "step": 2963 + }, + { + "epoch": 0.32549967054689216, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.453883171081543, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7169736623764038, + "num_tokens": 76566605.0, + "step": 2964 + }, + { + "epoch": 0.3256094882495058, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.5786592960357666, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7081461548805237, + "num_tokens": 76591486.0, + "step": 2965 + }, + { + "epoch": 0.3257193059521195, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 3.5797765254974365, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.702592134475708, + "num_tokens": 76611866.0, + "step": 2966 + }, + { + "epoch": 0.32582912365473315, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.6882708072662354, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.697138786315918, + "num_tokens": 76637126.0, + "step": 2967 + }, + { + "epoch": 0.3259389413573468, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.755662441253662, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.6976105570793152, + "num_tokens": 76659032.0, + "step": 2968 + }, + { + "epoch": 0.32604875905996045, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.7527167797088623, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7016492486000061, + "num_tokens": 76681382.0, + "step": 2969 + }, + { + "epoch": 0.32615857676257415, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 16.758586883544922, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7007542848587036, + "num_tokens": 76703872.0, + "step": 2970 + }, + { + "epoch": 0.3262683944651878, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7986700534820557, + "learning_rate": 1e-06, + "loss": 1.1539, + "mean_token_accuracy": 0.66608726978302, + "num_tokens": 76732208.0, + "step": 2971 + }, + { + "epoch": 0.32637821216780144, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.52970552444458, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6857852935791016, + "num_tokens": 76761485.0, + "step": 2972 + }, + { + "epoch": 0.3264880298704151, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.684451103210449, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6964151263237, + "num_tokens": 76790799.0, + "step": 2973 + }, + { + "epoch": 0.3265978475730288, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 3.272062301635742, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6907707452774048, + "num_tokens": 76814140.0, + "step": 2974 + }, + { + "epoch": 0.32670766527564243, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7191901206970215, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6903048753738403, + "num_tokens": 76836439.0, + "step": 2975 + }, + { + "epoch": 0.3268174829782561, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.739623546600342, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7039029598236084, + "num_tokens": 76857722.0, + "step": 2976 + }, + { + "epoch": 0.3269273006808698, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.9498376846313477, + "learning_rate": 1e-06, + "loss": 1.1771, + "mean_token_accuracy": 0.6649751663208008, + "num_tokens": 76883370.0, + "step": 2977 + }, + { + "epoch": 0.3270371183834834, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.686758279800415, + "learning_rate": 1e-06, + "loss": 1.0932, + "mean_token_accuracy": 0.6759109497070312, + "num_tokens": 76907397.0, + "step": 2978 + }, + { + "epoch": 0.3271469360860971, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.6130290031433105, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.6932886838912964, + "num_tokens": 76931225.0, + "step": 2979 + }, + { + "epoch": 0.3272567537887107, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.6491246223449707, + "learning_rate": 1e-06, + "loss": 1.125, + "mean_token_accuracy": 0.6679288148880005, + "num_tokens": 76956296.0, + "step": 2980 + }, + { + "epoch": 0.3273665714913244, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.5407445430755615, + "learning_rate": 1e-06, + "loss": 0.9554, + "mean_token_accuracy": 0.7137235403060913, + "num_tokens": 76979668.0, + "step": 2981 + }, + { + "epoch": 0.32747638919393807, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.8795583248138428, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.6983282566070557, + "num_tokens": 76999757.0, + "step": 2982 + }, + { + "epoch": 0.3275862068965517, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.3426437377929688, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7082411050796509, + "num_tokens": 77031664.0, + "step": 2983 + }, + { + "epoch": 0.32769602459916536, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.9581892490386963, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.6943744421005249, + "num_tokens": 77050012.0, + "step": 2984 + }, + { + "epoch": 0.32780584230177906, + "ewc_loss": 5.453824996948242e-06, + "grad_norm": 2.848080635070801, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.6937429308891296, + "num_tokens": 77071442.0, + "step": 2985 + }, + { + "epoch": 0.3279156600043927, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.713630437850952, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6869083046913147, + "num_tokens": 77098835.0, + "step": 2986 + }, + { + "epoch": 0.32802547770700635, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.438883066177368, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7113344073295593, + "num_tokens": 77125012.0, + "step": 2987 + }, + { + "epoch": 0.32813529540962005, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.732163429260254, + "learning_rate": 1e-06, + "loss": 0.9278, + "mean_token_accuracy": 0.7243415117263794, + "num_tokens": 77148158.0, + "step": 2988 + }, + { + "epoch": 0.3282451131122337, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.586514949798584, + "learning_rate": 1e-06, + "loss": 1.1037, + "mean_token_accuracy": 0.6788538694381714, + "num_tokens": 77175883.0, + "step": 2989 + }, + { + "epoch": 0.32835493081484735, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7998759746551514, + "learning_rate": 1e-06, + "loss": 1.1475, + "mean_token_accuracy": 0.6673452854156494, + "num_tokens": 77200164.0, + "step": 2990 + }, + { + "epoch": 0.328464748517461, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.597806692123413, + "learning_rate": 1e-06, + "loss": 1.1329, + "mean_token_accuracy": 0.6766621470451355, + "num_tokens": 77226914.0, + "step": 2991 + }, + { + "epoch": 0.3285745662200747, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.575913429260254, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.694622814655304, + "num_tokens": 77253262.0, + "step": 2992 + }, + { + "epoch": 0.32868438392268834, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.607362747192383, + "learning_rate": 1e-06, + "loss": 0.9166, + "mean_token_accuracy": 0.727433443069458, + "num_tokens": 77280602.0, + "step": 2993 + }, + { + "epoch": 0.328794201625302, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.452496290206909, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.699059247970581, + "num_tokens": 77308632.0, + "step": 2994 + }, + { + "epoch": 0.3289040193279157, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7040493488311768, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6929243803024292, + "num_tokens": 77331211.0, + "step": 2995 + }, + { + "epoch": 0.32901383703052933, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.471482992172241, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7136315107345581, + "num_tokens": 77358266.0, + "step": 2996 + }, + { + "epoch": 0.329123654733143, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.85471510887146, + "learning_rate": 1e-06, + "loss": 1.1271, + "mean_token_accuracy": 0.6861962080001831, + "num_tokens": 77382026.0, + "step": 2997 + }, + { + "epoch": 0.3292334724357566, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.4066827297210693, + "learning_rate": 1e-06, + "loss": 1.1628, + "mean_token_accuracy": 0.670923113822937, + "num_tokens": 77420022.0, + "step": 2998 + }, + { + "epoch": 0.3293432901383703, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 3.1422460079193115, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6883885860443115, + "num_tokens": 77451856.0, + "step": 2999 + }, + { + "epoch": 0.32945310784098397, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.5289411544799805, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6805214881896973, + "num_tokens": 77481883.0, + "step": 3000 + }, + { + "epoch": 0.3295629255435976, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.9160735607147217, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7158057689666748, + "num_tokens": 77501260.0, + "step": 3001 + }, + { + "epoch": 0.32967274324621126, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7698090076446533, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7037065625190735, + "num_tokens": 77522736.0, + "step": 3002 + }, + { + "epoch": 0.32978256094882497, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.4562606811523438, + "learning_rate": 1e-06, + "loss": 1.1675, + "mean_token_accuracy": 0.6583571434020996, + "num_tokens": 77552652.0, + "step": 3003 + }, + { + "epoch": 0.3298923786514386, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.550048351287842, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.723253607749939, + "num_tokens": 77577360.0, + "step": 3004 + }, + { + "epoch": 0.33000219635405226, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.4225735664367676, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7070478200912476, + "num_tokens": 77606752.0, + "step": 3005 + }, + { + "epoch": 0.33011201405666596, + "ewc_loss": 5.4836273193359375e-06, + "grad_norm": 2.7607362270355225, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7067536115646362, + "num_tokens": 77628709.0, + "step": 3006 + }, + { + "epoch": 0.3302218317592796, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.6350135803222656, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6790813207626343, + "num_tokens": 77656252.0, + "step": 3007 + }, + { + "epoch": 0.33033164946189325, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5150210857391357, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6982762813568115, + "num_tokens": 77681961.0, + "step": 3008 + }, + { + "epoch": 0.3304414671645069, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.586507558822632, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.6990537047386169, + "num_tokens": 77709226.0, + "step": 3009 + }, + { + "epoch": 0.3305512848671206, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.9709062576293945, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7150481343269348, + "num_tokens": 77726517.0, + "step": 3010 + }, + { + "epoch": 0.33066110256973424, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.50819730758667, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6986333131790161, + "num_tokens": 77752521.0, + "step": 3011 + }, + { + "epoch": 0.3307709202723479, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.585209846496582, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.695361852645874, + "num_tokens": 77779017.0, + "step": 3012 + }, + { + "epoch": 0.33088073797496154, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.6689064502716064, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6768147945404053, + "num_tokens": 77803179.0, + "step": 3013 + }, + { + "epoch": 0.33099055567757524, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 16.685028076171875, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7119655609130859, + "num_tokens": 77830529.0, + "step": 3014 + }, + { + "epoch": 0.3311003733801889, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.8520278930664062, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.69632887840271, + "num_tokens": 77857041.0, + "step": 3015 + }, + { + "epoch": 0.33121019108280253, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.8448071479797363, + "learning_rate": 1e-06, + "loss": 1.1221, + "mean_token_accuracy": 0.6789891123771667, + "num_tokens": 77881178.0, + "step": 3016 + }, + { + "epoch": 0.33132000878541623, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.6990280151367188, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6815468072891235, + "num_tokens": 77910440.0, + "step": 3017 + }, + { + "epoch": 0.3314298264880299, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.340526580810547, + "learning_rate": 1e-06, + "loss": 1.1251, + "mean_token_accuracy": 0.6742355823516846, + "num_tokens": 77941588.0, + "step": 3018 + }, + { + "epoch": 0.3315396441906435, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.703197479248047, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7125213146209717, + "num_tokens": 77962746.0, + "step": 3019 + }, + { + "epoch": 0.33164946189325717, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.743617534637451, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6837882995605469, + "num_tokens": 77985228.0, + "step": 3020 + }, + { + "epoch": 0.33175927959587087, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.4980716705322266, + "learning_rate": 1e-06, + "loss": 1.119, + "mean_token_accuracy": 0.6709906458854675, + "num_tokens": 78015834.0, + "step": 3021 + }, + { + "epoch": 0.3318690972984845, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.7316949367523193, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6998074054718018, + "num_tokens": 78038463.0, + "step": 3022 + }, + { + "epoch": 0.33197891500109816, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5306177139282227, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6933717727661133, + "num_tokens": 78068067.0, + "step": 3023 + }, + { + "epoch": 0.33208873270371186, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 3.0085256099700928, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6957474946975708, + "num_tokens": 78099716.0, + "step": 3024 + }, + { + "epoch": 0.3321985504063255, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.7892847061157227, + "learning_rate": 1e-06, + "loss": 0.9139, + "mean_token_accuracy": 0.726272463798523, + "num_tokens": 78123803.0, + "step": 3025 + }, + { + "epoch": 0.33230836810893916, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.598210573196411, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7058883905410767, + "num_tokens": 78148900.0, + "step": 3026 + }, + { + "epoch": 0.3324181858115528, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 5.258559703826904, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7013287544250488, + "num_tokens": 78172082.0, + "step": 3027 + }, + { + "epoch": 0.3325280035141665, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.6877377033233643, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7128103971481323, + "num_tokens": 78197140.0, + "step": 3028 + }, + { + "epoch": 0.33263782121678015, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.3777475357055664, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.674619734287262, + "num_tokens": 78229344.0, + "step": 3029 + }, + { + "epoch": 0.3327476389193938, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5340116024017334, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6946622133255005, + "num_tokens": 78257350.0, + "step": 3030 + }, + { + "epoch": 0.33285745662200744, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.4417526721954346, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7075644731521606, + "num_tokens": 78286663.0, + "step": 3031 + }, + { + "epoch": 0.33296727432462114, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 3.3649508953094482, + "learning_rate": 1e-06, + "loss": 1.0845, + "mean_token_accuracy": 0.6861521005630493, + "num_tokens": 78318532.0, + "step": 3032 + }, + { + "epoch": 0.3330770920272348, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.594550848007202, + "learning_rate": 1e-06, + "loss": 1.1331, + "mean_token_accuracy": 0.6695882081985474, + "num_tokens": 78342676.0, + "step": 3033 + }, + { + "epoch": 0.33318690972984844, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.9838123321533203, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7021044492721558, + "num_tokens": 78364760.0, + "step": 3034 + }, + { + "epoch": 0.33329672743246214, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.836677074432373, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6940866708755493, + "num_tokens": 78388601.0, + "step": 3035 + }, + { + "epoch": 0.3334065451350758, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5813426971435547, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6934657096862793, + "num_tokens": 78415549.0, + "step": 3036 + }, + { + "epoch": 0.33351636283768943, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5424180030822754, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7030164003372192, + "num_tokens": 78442685.0, + "step": 3037 + }, + { + "epoch": 0.3336261805403031, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5831141471862793, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6949372887611389, + "num_tokens": 78469472.0, + "step": 3038 + }, + { + "epoch": 0.3337359982429168, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.4744679927825928, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7099723815917969, + "num_tokens": 78501465.0, + "step": 3039 + }, + { + "epoch": 0.3338458159455304, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5791513919830322, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.713676929473877, + "num_tokens": 78525782.0, + "step": 3040 + }, + { + "epoch": 0.33395563364814407, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 3.150723695755005, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6977160573005676, + "num_tokens": 78550247.0, + "step": 3041 + }, + { + "epoch": 0.33406545135075777, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5226244926452637, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7031179666519165, + "num_tokens": 78576229.0, + "step": 3042 + }, + { + "epoch": 0.3341752690533714, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.395068883895874, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6938873529434204, + "num_tokens": 78603794.0, + "step": 3043 + }, + { + "epoch": 0.33428508675598506, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 3.089137315750122, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7004244327545166, + "num_tokens": 78624317.0, + "step": 3044 + }, + { + "epoch": 0.3343949044585987, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.5925910472869873, + "learning_rate": 1e-06, + "loss": 1.0946, + "mean_token_accuracy": 0.6800893545150757, + "num_tokens": 78649693.0, + "step": 3045 + }, + { + "epoch": 0.3345047221612124, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.4708807468414307, + "learning_rate": 1e-06, + "loss": 0.9364, + "mean_token_accuracy": 0.7253265976905823, + "num_tokens": 78676792.0, + "step": 3046 + }, + { + "epoch": 0.33461453986382605, + "ewc_loss": 5.513429641723633e-06, + "grad_norm": 2.6823530197143555, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7107914686203003, + "num_tokens": 78698812.0, + "step": 3047 + }, + { + "epoch": 0.3347243575664397, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.7186527252197266, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.6964028477668762, + "num_tokens": 78723823.0, + "step": 3048 + }, + { + "epoch": 0.33483417526905335, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.449545383453369, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6869969964027405, + "num_tokens": 78751359.0, + "step": 3049 + }, + { + "epoch": 0.33494399297166705, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.4561350345611572, + "learning_rate": 1e-06, + "loss": 1.1413, + "mean_token_accuracy": 0.6675990223884583, + "num_tokens": 78778325.0, + "step": 3050 + }, + { + "epoch": 0.3350538106742807, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.729686975479126, + "learning_rate": 1e-06, + "loss": 0.9429, + "mean_token_accuracy": 0.7195333242416382, + "num_tokens": 78800096.0, + "step": 3051 + }, + { + "epoch": 0.33516362837689434, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.888745069503784, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.6992971897125244, + "num_tokens": 78820036.0, + "step": 3052 + }, + { + "epoch": 0.33527344607950804, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5604970455169678, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6944385170936584, + "num_tokens": 78845369.0, + "step": 3053 + }, + { + "epoch": 0.3353832637821217, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.4835383892059326, + "learning_rate": 1e-06, + "loss": 1.1352, + "mean_token_accuracy": 0.6734727025032043, + "num_tokens": 78878373.0, + "step": 3054 + }, + { + "epoch": 0.33549308148473533, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.6871888637542725, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6943555474281311, + "num_tokens": 78900627.0, + "step": 3055 + }, + { + "epoch": 0.335602899187349, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.38515567779541, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6896028518676758, + "num_tokens": 78932358.0, + "step": 3056 + }, + { + "epoch": 0.3357127168899627, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.7190520763397217, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.6969286203384399, + "num_tokens": 78954929.0, + "step": 3057 + }, + { + "epoch": 0.3358225345925763, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.3277103900909424, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.712714672088623, + "num_tokens": 78984236.0, + "step": 3058 + }, + { + "epoch": 0.33593235229519, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.321666955947876, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6924847364425659, + "num_tokens": 79013737.0, + "step": 3059 + }, + { + "epoch": 0.3360421699978036, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.783889055252075, + "learning_rate": 1e-06, + "loss": 1.1108, + "mean_token_accuracy": 0.677791953086853, + "num_tokens": 79047783.0, + "step": 3060 + }, + { + "epoch": 0.3361519877004173, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.844667673110962, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.6967271566390991, + "num_tokens": 79069512.0, + "step": 3061 + }, + { + "epoch": 0.33626180540303097, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.4951725006103516, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7204346060752869, + "num_tokens": 79098817.0, + "step": 3062 + }, + { + "epoch": 0.3363716231056446, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.814774990081787, + "learning_rate": 1e-06, + "loss": 1.0754, + "mean_token_accuracy": 0.6835675239562988, + "num_tokens": 79119523.0, + "step": 3063 + }, + { + "epoch": 0.3364814408082583, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.490600347518921, + "learning_rate": 1e-06, + "loss": 1.1329, + "mean_token_accuracy": 0.6738593578338623, + "num_tokens": 79148617.0, + "step": 3064 + }, + { + "epoch": 0.33659125851087196, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5746662616729736, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7004324197769165, + "num_tokens": 79174100.0, + "step": 3065 + }, + { + "epoch": 0.3367010762134856, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5178275108337402, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.7069503664970398, + "num_tokens": 79200804.0, + "step": 3066 + }, + { + "epoch": 0.33681089391609925, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.712451696395874, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7018297910690308, + "num_tokens": 79223988.0, + "step": 3067 + }, + { + "epoch": 0.33692071161871295, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 3.210588216781616, + "learning_rate": 1e-06, + "loss": 1.123, + "mean_token_accuracy": 0.67616206407547, + "num_tokens": 79248404.0, + "step": 3068 + }, + { + "epoch": 0.3370305293213266, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.441189765930176, + "learning_rate": 1e-06, + "loss": 1.1219, + "mean_token_accuracy": 0.6830304265022278, + "num_tokens": 79277072.0, + "step": 3069 + }, + { + "epoch": 0.33714034702394025, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.836710214614868, + "learning_rate": 1e-06, + "loss": 1.1391, + "mean_token_accuracy": 0.6702258586883545, + "num_tokens": 79304757.0, + "step": 3070 + }, + { + "epoch": 0.33725016472655395, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.2413330078125, + "learning_rate": 1e-06, + "loss": 1.0736, + "mean_token_accuracy": 0.6863430738449097, + "num_tokens": 79339350.0, + "step": 3071 + }, + { + "epoch": 0.3373599824291676, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.36324143409729, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6806010603904724, + "num_tokens": 79369818.0, + "step": 3072 + }, + { + "epoch": 0.33746980013178124, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.3267619609832764, + "learning_rate": 1e-06, + "loss": 1.1248, + "mean_token_accuracy": 0.6799865961074829, + "num_tokens": 79401881.0, + "step": 3073 + }, + { + "epoch": 0.3375796178343949, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 3.038303852081299, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.7008332014083862, + "num_tokens": 79420457.0, + "step": 3074 + }, + { + "epoch": 0.3376894355370086, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.4282991886138916, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6960183382034302, + "num_tokens": 79450330.0, + "step": 3075 + }, + { + "epoch": 0.33779925323962223, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5737528800964355, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.694786787033081, + "num_tokens": 79477626.0, + "step": 3076 + }, + { + "epoch": 0.3379090709422359, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.7715890407562256, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6919234991073608, + "num_tokens": 79500703.0, + "step": 3077 + }, + { + "epoch": 0.3380188886448495, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.5993330478668213, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6827800869941711, + "num_tokens": 79526708.0, + "step": 3078 + }, + { + "epoch": 0.3381287063474632, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.2909624576568604, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.6991041302680969, + "num_tokens": 79561806.0, + "step": 3079 + }, + { + "epoch": 0.33823852405007687, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.5206563472747803, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7041476964950562, + "num_tokens": 79587764.0, + "step": 3080 + }, + { + "epoch": 0.3383483417526905, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.4301257133483887, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7097177505493164, + "num_tokens": 79616144.0, + "step": 3081 + }, + { + "epoch": 0.3384581594553042, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.6752564907073975, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7057653069496155, + "num_tokens": 79638796.0, + "step": 3082 + }, + { + "epoch": 0.33856797715791787, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.587470293045044, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7149749994277954, + "num_tokens": 79662566.0, + "step": 3083 + }, + { + "epoch": 0.3386777948605315, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.565143585205078, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7040063738822937, + "num_tokens": 79691734.0, + "step": 3084 + }, + { + "epoch": 0.33878761256314516, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.443964719772339, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6977781653404236, + "num_tokens": 79720234.0, + "step": 3085 + }, + { + "epoch": 0.33889743026575886, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.5292608737945557, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6887283325195312, + "num_tokens": 79744274.0, + "step": 3086 + }, + { + "epoch": 0.3390072479683725, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.5249712467193604, + "learning_rate": 1e-06, + "loss": 1.0466, + "mean_token_accuracy": 0.6931440830230713, + "num_tokens": 79771274.0, + "step": 3087 + }, + { + "epoch": 0.33911706567098615, + "ewc_loss": 5.543231964111328e-06, + "grad_norm": 2.605064630508423, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6812881231307983, + "num_tokens": 79797808.0, + "step": 3088 + }, + { + "epoch": 0.3392268833735998, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 3.00411319732666, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7030667066574097, + "num_tokens": 79818490.0, + "step": 3089 + }, + { + "epoch": 0.3393367010762135, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 2.57053804397583, + "learning_rate": 1e-06, + "loss": 1.1044, + "mean_token_accuracy": 0.6768724918365479, + "num_tokens": 79846912.0, + "step": 3090 + }, + { + "epoch": 0.33944651877882714, + "ewc_loss": 5.5730342864990234e-06, + "grad_norm": 3.234274387359619, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7162560820579529, + "num_tokens": 79862416.0, + "step": 3091 + }, + { + "epoch": 0.3395563364814408, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.559687852859497, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6836686730384827, + "num_tokens": 79888186.0, + "step": 3092 + }, + { + "epoch": 0.3396661541840545, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.5355372428894043, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6890808343887329, + "num_tokens": 79914312.0, + "step": 3093 + }, + { + "epoch": 0.33977597188666814, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.697425603866577, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7057496309280396, + "num_tokens": 79938231.0, + "step": 3094 + }, + { + "epoch": 0.3398857895892818, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.491617441177368, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7138100862503052, + "num_tokens": 79965742.0, + "step": 3095 + }, + { + "epoch": 0.33999560729189543, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.6309797763824463, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6974865198135376, + "num_tokens": 80000024.0, + "step": 3096 + }, + { + "epoch": 0.34010542499450913, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.4813857078552246, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6966379880905151, + "num_tokens": 80032041.0, + "step": 3097 + }, + { + "epoch": 0.3402152426971228, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.511599063873291, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6839218139648438, + "num_tokens": 80057633.0, + "step": 3098 + }, + { + "epoch": 0.3403250603997364, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.5676252841949463, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.694496750831604, + "num_tokens": 80083175.0, + "step": 3099 + }, + { + "epoch": 0.3404348781023501, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.630333423614502, + "learning_rate": 1e-06, + "loss": 1.1269, + "mean_token_accuracy": 0.6784474849700928, + "num_tokens": 80112222.0, + "step": 3100 + }, + { + "epoch": 0.34054469580496377, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 3.135476589202881, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7033700942993164, + "num_tokens": 80129502.0, + "step": 3101 + }, + { + "epoch": 0.3406545135075774, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.641080856323242, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.710969090461731, + "num_tokens": 80152719.0, + "step": 3102 + }, + { + "epoch": 0.34076433121019106, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 3.0611326694488525, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7111087441444397, + "num_tokens": 80177076.0, + "step": 3103 + }, + { + "epoch": 0.34087414891280476, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.5315499305725098, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6939138770103455, + "num_tokens": 80204022.0, + "step": 3104 + }, + { + "epoch": 0.3409839666154184, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.7468652725219727, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.708479642868042, + "num_tokens": 80229727.0, + "step": 3105 + }, + { + "epoch": 0.34109378431803206, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.7290165424346924, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7159119844436646, + "num_tokens": 80251886.0, + "step": 3106 + }, + { + "epoch": 0.3412036020206457, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.386699914932251, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6786395907402039, + "num_tokens": 80283712.0, + "step": 3107 + }, + { + "epoch": 0.3413134197232594, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.6043524742126465, + "learning_rate": 1e-06, + "loss": 1.123, + "mean_token_accuracy": 0.6811718940734863, + "num_tokens": 80311758.0, + "step": 3108 + }, + { + "epoch": 0.34142323742587305, + "ewc_loss": 5.602836608886719e-06, + "grad_norm": 2.681752920150757, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.6938295364379883, + "num_tokens": 80333423.0, + "step": 3109 + }, + { + "epoch": 0.3415330551284867, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.6934821605682373, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.6811926960945129, + "num_tokens": 80358027.0, + "step": 3110 + }, + { + "epoch": 0.3416428728311004, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.801616907119751, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6941691040992737, + "num_tokens": 80385582.0, + "step": 3111 + }, + { + "epoch": 0.34175269053371404, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.8826732635498047, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.713031530380249, + "num_tokens": 80415533.0, + "step": 3112 + }, + { + "epoch": 0.3418625082363277, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.6014745235443115, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7135345339775085, + "num_tokens": 80439223.0, + "step": 3113 + }, + { + "epoch": 0.34197232593894134, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.8245394229888916, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.6994662284851074, + "num_tokens": 80462023.0, + "step": 3114 + }, + { + "epoch": 0.34208214364155504, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.7767181396484375, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6945980191230774, + "num_tokens": 80484919.0, + "step": 3115 + }, + { + "epoch": 0.3421919613441687, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.576561212539673, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7211301326751709, + "num_tokens": 80508170.0, + "step": 3116 + }, + { + "epoch": 0.34230177904678233, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.5323452949523926, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6842815279960632, + "num_tokens": 80539222.0, + "step": 3117 + }, + { + "epoch": 0.34241159674939603, + "ewc_loss": 5.632638931274414e-06, + "grad_norm": 2.707674026489258, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7097752690315247, + "num_tokens": 80562836.0, + "step": 3118 + }, + { + "epoch": 0.3425214144520097, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.633833169937134, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.698874831199646, + "num_tokens": 80588438.0, + "step": 3119 + }, + { + "epoch": 0.3426312321546233, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.667928457260132, + "learning_rate": 1e-06, + "loss": 1.1215, + "mean_token_accuracy": 0.6776556968688965, + "num_tokens": 80615939.0, + "step": 3120 + }, + { + "epoch": 0.34274104985723697, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.471308469772339, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6869901418685913, + "num_tokens": 80642335.0, + "step": 3121 + }, + { + "epoch": 0.34285086755985067, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.6287789344787598, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6886144876480103, + "num_tokens": 80665781.0, + "step": 3122 + }, + { + "epoch": 0.3429606852624643, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.342426061630249, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6956440806388855, + "num_tokens": 80696047.0, + "step": 3123 + }, + { + "epoch": 0.34307050296507796, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.765054702758789, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7051233053207397, + "num_tokens": 80719122.0, + "step": 3124 + }, + { + "epoch": 0.3431803206676916, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.526378631591797, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6965669989585876, + "num_tokens": 80747066.0, + "step": 3125 + }, + { + "epoch": 0.3432901383703053, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.5456197261810303, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6991252899169922, + "num_tokens": 80775529.0, + "step": 3126 + }, + { + "epoch": 0.34339995607291895, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.9326703548431396, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.6863821744918823, + "num_tokens": 80806422.0, + "step": 3127 + }, + { + "epoch": 0.3435097737755326, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.5214967727661133, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.6986647844314575, + "num_tokens": 80829206.0, + "step": 3128 + }, + { + "epoch": 0.3436195914781463, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.380208730697632, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6885030269622803, + "num_tokens": 80859251.0, + "step": 3129 + }, + { + "epoch": 0.34372940918075995, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.8640687465667725, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7043031454086304, + "num_tokens": 80879585.0, + "step": 3130 + }, + { + "epoch": 0.3438392268833736, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 3.131011486053467, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7134410738945007, + "num_tokens": 80896459.0, + "step": 3131 + }, + { + "epoch": 0.34394904458598724, + "ewc_loss": 5.662441253662109e-06, + "grad_norm": 2.920919895172119, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7061619758605957, + "num_tokens": 80916349.0, + "step": 3132 + }, + { + "epoch": 0.34405886228860094, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.686357259750366, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7115360498428345, + "num_tokens": 80940274.0, + "step": 3133 + }, + { + "epoch": 0.3441686799912146, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.579859495162964, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6925492286682129, + "num_tokens": 80967458.0, + "step": 3134 + }, + { + "epoch": 0.34427849769382823, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.805424451828003, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.709579586982727, + "num_tokens": 80987311.0, + "step": 3135 + }, + { + "epoch": 0.3443883153964419, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.9791929721832275, + "learning_rate": 1e-06, + "loss": 1.167, + "mean_token_accuracy": 0.6632063984870911, + "num_tokens": 81009634.0, + "step": 3136 + }, + { + "epoch": 0.3444981330990556, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.9767873287200928, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7012923359870911, + "num_tokens": 81029324.0, + "step": 3137 + }, + { + "epoch": 0.3446079508016692, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.7822837829589844, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6862340569496155, + "num_tokens": 81060117.0, + "step": 3138 + }, + { + "epoch": 0.3447177685042829, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.7757203578948975, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7028002142906189, + "num_tokens": 81084163.0, + "step": 3139 + }, + { + "epoch": 0.3448275862068966, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.4810492992401123, + "learning_rate": 1e-06, + "loss": 1.1067, + "mean_token_accuracy": 0.681496262550354, + "num_tokens": 81115321.0, + "step": 3140 + }, + { + "epoch": 0.3449374039095102, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.870401382446289, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7027768492698669, + "num_tokens": 81136307.0, + "step": 3141 + }, + { + "epoch": 0.34504722161212387, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.569124460220337, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6842936873435974, + "num_tokens": 81163479.0, + "step": 3142 + }, + { + "epoch": 0.3451570393147375, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.7921674251556396, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6844373345375061, + "num_tokens": 81188105.0, + "step": 3143 + }, + { + "epoch": 0.3452668570173512, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.5571160316467285, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6975600123405457, + "num_tokens": 81217529.0, + "step": 3144 + }, + { + "epoch": 0.34537667471996486, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.337984323501587, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7041788101196289, + "num_tokens": 81245442.0, + "step": 3145 + }, + { + "epoch": 0.3454864924225785, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.4536848068237305, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7097277641296387, + "num_tokens": 81273374.0, + "step": 3146 + }, + { + "epoch": 0.3455963101251922, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.587360382080078, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7032692432403564, + "num_tokens": 81299150.0, + "step": 3147 + }, + { + "epoch": 0.34570612782780585, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.593020439147949, + "learning_rate": 1e-06, + "loss": 1.1033, + "mean_token_accuracy": 0.6816689372062683, + "num_tokens": 81323114.0, + "step": 3148 + }, + { + "epoch": 0.3458159455304195, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.9068078994750977, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7198905944824219, + "num_tokens": 81343642.0, + "step": 3149 + }, + { + "epoch": 0.34592576323303315, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.583311080932617, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7127209901809692, + "num_tokens": 81371186.0, + "step": 3150 + }, + { + "epoch": 0.34603558093564685, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.496035099029541, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.6990969181060791, + "num_tokens": 81398073.0, + "step": 3151 + }, + { + "epoch": 0.3461453986382605, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.401154041290283, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7040805220603943, + "num_tokens": 81426255.0, + "step": 3152 + }, + { + "epoch": 0.34625521634087414, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.626333713531494, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6902797222137451, + "num_tokens": 81452329.0, + "step": 3153 + }, + { + "epoch": 0.3463650340434878, + "ewc_loss": 5.692243576049805e-06, + "grad_norm": 2.5696678161621094, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7243392467498779, + "num_tokens": 81475750.0, + "step": 3154 + }, + { + "epoch": 0.3464748517461015, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.9498326778411865, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6966946721076965, + "num_tokens": 81499098.0, + "step": 3155 + }, + { + "epoch": 0.34658466944871513, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5352706909179688, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.6827295422554016, + "num_tokens": 81525567.0, + "step": 3156 + }, + { + "epoch": 0.3466944871513288, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.772688627243042, + "learning_rate": 1e-06, + "loss": 1.1201, + "mean_token_accuracy": 0.6710298657417297, + "num_tokens": 81552757.0, + "step": 3157 + }, + { + "epoch": 0.3468043048539425, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.511253595352173, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6877574920654297, + "num_tokens": 81579060.0, + "step": 3158 + }, + { + "epoch": 0.3469141225565561, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.539167642593384, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6816025972366333, + "num_tokens": 81608542.0, + "step": 3159 + }, + { + "epoch": 0.34702394025916977, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.30292010307312, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.700685977935791, + "num_tokens": 81639879.0, + "step": 3160 + }, + { + "epoch": 0.3471337579617834, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.4331789016723633, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.711773693561554, + "num_tokens": 81664867.0, + "step": 3161 + }, + { + "epoch": 0.3472435756643971, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.7102248668670654, + "learning_rate": 1e-06, + "loss": 1.1193, + "mean_token_accuracy": 0.6785982251167297, + "num_tokens": 81691172.0, + "step": 3162 + }, + { + "epoch": 0.34735339336701077, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.286951780319214, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7042419910430908, + "num_tokens": 81722600.0, + "step": 3163 + }, + { + "epoch": 0.3474632110696244, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 3.4808642864227295, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7195621728897095, + "num_tokens": 81737797.0, + "step": 3164 + }, + { + "epoch": 0.34757302877223806, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.9173734188079834, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7091565132141113, + "num_tokens": 81757397.0, + "step": 3165 + }, + { + "epoch": 0.34768284647485176, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.815382957458496, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7144314050674438, + "num_tokens": 81778433.0, + "step": 3166 + }, + { + "epoch": 0.3477926641774654, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.369415283203125, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6906015276908875, + "num_tokens": 81808813.0, + "step": 3167 + }, + { + "epoch": 0.34790248188007905, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5253655910491943, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.679314136505127, + "num_tokens": 81835433.0, + "step": 3168 + }, + { + "epoch": 0.34801229958269275, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.905320644378662, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7056149840354919, + "num_tokens": 81855762.0, + "step": 3169 + }, + { + "epoch": 0.3481221172853064, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.672325611114502, + "learning_rate": 1e-06, + "loss": 1.1281, + "mean_token_accuracy": 0.6690599918365479, + "num_tokens": 81880430.0, + "step": 3170 + }, + { + "epoch": 0.34823193498792004, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.7315385341644287, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.7126444578170776, + "num_tokens": 81903500.0, + "step": 3171 + }, + { + "epoch": 0.3483417526905337, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.37910532951355, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7092412710189819, + "num_tokens": 81935057.0, + "step": 3172 + }, + { + "epoch": 0.3484515703931474, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.751023292541504, + "learning_rate": 1e-06, + "loss": 1.1111, + "mean_token_accuracy": 0.6746385097503662, + "num_tokens": 81959076.0, + "step": 3173 + }, + { + "epoch": 0.34856138809576104, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.7664899826049805, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7061761617660522, + "num_tokens": 81981974.0, + "step": 3174 + }, + { + "epoch": 0.3486712057983747, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.2380528450012207, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.68320232629776, + "num_tokens": 82015685.0, + "step": 3175 + }, + { + "epoch": 0.3487810235009884, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.90069580078125, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7040136456489563, + "num_tokens": 82037034.0, + "step": 3176 + }, + { + "epoch": 0.34889084120360203, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.696666955947876, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7071282863616943, + "num_tokens": 82060036.0, + "step": 3177 + }, + { + "epoch": 0.3490006589062157, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5736083984375, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6953324675559998, + "num_tokens": 82086924.0, + "step": 3178 + }, + { + "epoch": 0.3491104766088293, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.69624400138855, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7016974091529846, + "num_tokens": 82110692.0, + "step": 3179 + }, + { + "epoch": 0.349220294311443, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 3.021556854248047, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.6865631937980652, + "num_tokens": 82133124.0, + "step": 3180 + }, + { + "epoch": 0.34933011201405667, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.530648946762085, + "learning_rate": 1e-06, + "loss": 0.8974, + "mean_token_accuracy": 0.7311481237411499, + "num_tokens": 82160041.0, + "step": 3181 + }, + { + "epoch": 0.3494399297166703, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.963670015335083, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7024785280227661, + "num_tokens": 82183128.0, + "step": 3182 + }, + { + "epoch": 0.34954974741928396, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.727365255355835, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.7010256052017212, + "num_tokens": 82206128.0, + "step": 3183 + }, + { + "epoch": 0.34965956512189766, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.8957817554473877, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7046405076980591, + "num_tokens": 82227187.0, + "step": 3184 + }, + { + "epoch": 0.3497693828245113, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5942628383636475, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7084528803825378, + "num_tokens": 82250941.0, + "step": 3185 + }, + { + "epoch": 0.34987920052712496, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.437403440475464, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6906946301460266, + "num_tokens": 82279220.0, + "step": 3186 + }, + { + "epoch": 0.34998901822973866, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.562225580215454, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6858892440795898, + "num_tokens": 82308794.0, + "step": 3187 + }, + { + "epoch": 0.3500988359323523, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.699167013168335, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.7023141980171204, + "num_tokens": 82333167.0, + "step": 3188 + }, + { + "epoch": 0.35020865363496595, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.799980401992798, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.7138554453849792, + "num_tokens": 82356761.0, + "step": 3189 + }, + { + "epoch": 0.3503184713375796, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.7224671840667725, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.7026571035385132, + "num_tokens": 82377746.0, + "step": 3190 + }, + { + "epoch": 0.3504282890401933, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.587646245956421, + "learning_rate": 1e-06, + "loss": 1.161, + "mean_token_accuracy": 0.6677265167236328, + "num_tokens": 82404722.0, + "step": 3191 + }, + { + "epoch": 0.35053810674280694, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.521228313446045, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7024684548377991, + "num_tokens": 82431008.0, + "step": 3192 + }, + { + "epoch": 0.3506479244454206, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.4669158458709717, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6817807555198669, + "num_tokens": 82459447.0, + "step": 3193 + }, + { + "epoch": 0.3507577421480343, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.567476987838745, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6870821714401245, + "num_tokens": 82485581.0, + "step": 3194 + }, + { + "epoch": 0.35086755985064794, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.4154579639434814, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7011882066726685, + "num_tokens": 82512694.0, + "step": 3195 + }, + { + "epoch": 0.3509773775532616, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.4704878330230713, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7146527767181396, + "num_tokens": 82537247.0, + "step": 3196 + }, + { + "epoch": 0.35108719525587523, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 3.072100877761841, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7154147028923035, + "num_tokens": 82558570.0, + "step": 3197 + }, + { + "epoch": 0.35119701295848893, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5118253231048584, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6902703046798706, + "num_tokens": 82587299.0, + "step": 3198 + }, + { + "epoch": 0.3513068306611026, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.2578091621398926, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6896538138389587, + "num_tokens": 82619644.0, + "step": 3199 + }, + { + "epoch": 0.3514166483637162, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.655944347381592, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7042475938796997, + "num_tokens": 82644335.0, + "step": 3200 + }, + { + "epoch": 0.35152646606632987, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.3447413444519043, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6909977197647095, + "num_tokens": 82674506.0, + "step": 3201 + }, + { + "epoch": 0.35163628376894357, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5218443870544434, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.68653404712677, + "num_tokens": 82702472.0, + "step": 3202 + }, + { + "epoch": 0.3517461014715572, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.458307981491089, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6768718361854553, + "num_tokens": 82730656.0, + "step": 3203 + }, + { + "epoch": 0.35185591917417086, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5095374584198, + "learning_rate": 1e-06, + "loss": 1.1315, + "mean_token_accuracy": 0.6709751486778259, + "num_tokens": 82761816.0, + "step": 3204 + }, + { + "epoch": 0.35196573687678456, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.683053493499756, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7108264565467834, + "num_tokens": 82785818.0, + "step": 3205 + }, + { + "epoch": 0.3520755545793982, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.524235248565674, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.699670672416687, + "num_tokens": 82813666.0, + "step": 3206 + }, + { + "epoch": 0.35218537228201185, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.5307602882385254, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.7054463028907776, + "num_tokens": 82840296.0, + "step": 3207 + }, + { + "epoch": 0.3522951899846255, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 3.0950472354888916, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7009918689727783, + "num_tokens": 82857720.0, + "step": 3208 + }, + { + "epoch": 0.3524050076872392, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.7475733757019043, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6811926960945129, + "num_tokens": 82882989.0, + "step": 3209 + }, + { + "epoch": 0.35251482538985285, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.653156280517578, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7245622873306274, + "num_tokens": 82908373.0, + "step": 3210 + }, + { + "epoch": 0.3526246430924665, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.383192777633667, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7020065784454346, + "num_tokens": 82939675.0, + "step": 3211 + }, + { + "epoch": 0.35273446079508014, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.8101649284362793, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.7033078670501709, + "num_tokens": 82964508.0, + "step": 3212 + }, + { + "epoch": 0.35284427849769384, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.703441619873047, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7204889059066772, + "num_tokens": 82986542.0, + "step": 3213 + }, + { + "epoch": 0.3529540962003075, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.6360881328582764, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7012695074081421, + "num_tokens": 83010692.0, + "step": 3214 + }, + { + "epoch": 0.35306391390292113, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.4570107460021973, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7168398499488831, + "num_tokens": 83037047.0, + "step": 3215 + }, + { + "epoch": 0.35317373160553484, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.539469003677368, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6962624788284302, + "num_tokens": 83063122.0, + "step": 3216 + }, + { + "epoch": 0.3532835493081485, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.509960412979126, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6954120397567749, + "num_tokens": 83091684.0, + "step": 3217 + }, + { + "epoch": 0.3533933670107621, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.5270705223083496, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6945574283599854, + "num_tokens": 83119244.0, + "step": 3218 + }, + { + "epoch": 0.3535031847133758, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.8106844425201416, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.7049771547317505, + "num_tokens": 83141800.0, + "step": 3219 + }, + { + "epoch": 0.3536130024159895, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.6646666526794434, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.6971653699874878, + "num_tokens": 83170081.0, + "step": 3220 + }, + { + "epoch": 0.3537228201186031, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.912198781967163, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.6997178792953491, + "num_tokens": 83190689.0, + "step": 3221 + }, + { + "epoch": 0.35383263782121677, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.505571126937866, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6752899885177612, + "num_tokens": 83223907.0, + "step": 3222 + }, + { + "epoch": 0.35394245552383047, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.7688279151916504, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6938484907150269, + "num_tokens": 83245462.0, + "step": 3223 + }, + { + "epoch": 0.3540522732264441, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 3.035001277923584, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7038259506225586, + "num_tokens": 83265293.0, + "step": 3224 + }, + { + "epoch": 0.35416209092905776, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.646008253097534, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7101568579673767, + "num_tokens": 83291068.0, + "step": 3225 + }, + { + "epoch": 0.3542719086316714, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.654942512512207, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6895037293434143, + "num_tokens": 83318510.0, + "step": 3226 + }, + { + "epoch": 0.3543817263342851, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.2484352588653564, + "learning_rate": 1e-06, + "loss": 1.0965, + "mean_token_accuracy": 0.6765990257263184, + "num_tokens": 83353576.0, + "step": 3227 + }, + { + "epoch": 0.35449154403689875, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.6273598670959473, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6903051137924194, + "num_tokens": 83376074.0, + "step": 3228 + }, + { + "epoch": 0.3546013617395124, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.422793388366699, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.69748854637146, + "num_tokens": 83401544.0, + "step": 3229 + }, + { + "epoch": 0.35471117944212605, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.236560821533203, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6944036483764648, + "num_tokens": 83433777.0, + "step": 3230 + }, + { + "epoch": 0.35482099714473975, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.745703935623169, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6877443194389343, + "num_tokens": 83459923.0, + "step": 3231 + }, + { + "epoch": 0.3549308148473534, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.554687261581421, + "learning_rate": 1e-06, + "loss": 1.1012, + "mean_token_accuracy": 0.675767719745636, + "num_tokens": 83487301.0, + "step": 3232 + }, + { + "epoch": 0.35504063254996704, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.7975776195526123, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7090439796447754, + "num_tokens": 83507683.0, + "step": 3233 + }, + { + "epoch": 0.35515045025258074, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.7129173278808594, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6805853843688965, + "num_tokens": 83533092.0, + "step": 3234 + }, + { + "epoch": 0.3552602679551944, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.617487668991089, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6854275465011597, + "num_tokens": 83562221.0, + "step": 3235 + }, + { + "epoch": 0.35537008565780803, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.9603800773620605, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.70135498046875, + "num_tokens": 83587927.0, + "step": 3236 + }, + { + "epoch": 0.3554799033604217, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.32327938079834, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6939758658409119, + "num_tokens": 83619553.0, + "step": 3237 + }, + { + "epoch": 0.3555897210630354, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.288710594177246, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.722243070602417, + "num_tokens": 83652859.0, + "step": 3238 + }, + { + "epoch": 0.355699538765649, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.5291171073913574, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6894475817680359, + "num_tokens": 83679314.0, + "step": 3239 + }, + { + "epoch": 0.35580935646826267, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.4468071460723877, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6930667161941528, + "num_tokens": 83704588.0, + "step": 3240 + }, + { + "epoch": 0.3559191741708763, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.353062152862549, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7000732421875, + "num_tokens": 83732073.0, + "step": 3241 + }, + { + "epoch": 0.35602899187349, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.400574207305908, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6793627738952637, + "num_tokens": 83761233.0, + "step": 3242 + }, + { + "epoch": 0.35613880957610367, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.611809730529785, + "learning_rate": 1e-06, + "loss": 1.1067, + "mean_token_accuracy": 0.6722958087921143, + "num_tokens": 83789309.0, + "step": 3243 + }, + { + "epoch": 0.3562486272787173, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.384413242340088, + "learning_rate": 1e-06, + "loss": 1.1186, + "mean_token_accuracy": 0.6823117733001709, + "num_tokens": 83822253.0, + "step": 3244 + }, + { + "epoch": 0.356358444981331, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.397286891937256, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7037417888641357, + "num_tokens": 83850718.0, + "step": 3245 + }, + { + "epoch": 0.35646826268394466, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.587209939956665, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6952676773071289, + "num_tokens": 83874658.0, + "step": 3246 + }, + { + "epoch": 0.3565780803865583, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.6427154541015625, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7186747789382935, + "num_tokens": 83899291.0, + "step": 3247 + }, + { + "epoch": 0.35668789808917195, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.539194107055664, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7187972664833069, + "num_tokens": 83922590.0, + "step": 3248 + }, + { + "epoch": 0.35679771579178565, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.4591667652130127, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6971125602722168, + "num_tokens": 83950687.0, + "step": 3249 + }, + { + "epoch": 0.3569075334943993, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.4701151847839355, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.694294810295105, + "num_tokens": 83977564.0, + "step": 3250 + }, + { + "epoch": 0.35701735119701294, + "ewc_loss": 5.7220458984375e-06, + "grad_norm": 2.784031867980957, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6788907051086426, + "num_tokens": 84003184.0, + "step": 3251 + }, + { + "epoch": 0.35712716889962665, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.3896484375, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6977368593215942, + "num_tokens": 84031308.0, + "step": 3252 + }, + { + "epoch": 0.3572369866022403, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.5695137977600098, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.673729658126831, + "num_tokens": 84057005.0, + "step": 3253 + }, + { + "epoch": 0.35734680430485394, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.625513792037964, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6864234209060669, + "num_tokens": 84084931.0, + "step": 3254 + }, + { + "epoch": 0.3574566220074676, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.770439624786377, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6813492774963379, + "num_tokens": 84108380.0, + "step": 3255 + }, + { + "epoch": 0.3575664397100813, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.794046401977539, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7181777954101562, + "num_tokens": 84130087.0, + "step": 3256 + }, + { + "epoch": 0.35767625741269493, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.595189094543457, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7033393979072571, + "num_tokens": 84154892.0, + "step": 3257 + }, + { + "epoch": 0.3577860751153086, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.7645578384399414, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7066473364830017, + "num_tokens": 84176496.0, + "step": 3258 + }, + { + "epoch": 0.3578958928179222, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.814446449279785, + "learning_rate": 1e-06, + "loss": 1.083, + "mean_token_accuracy": 0.6845828294754028, + "num_tokens": 84200797.0, + "step": 3259 + }, + { + "epoch": 0.3580057105205359, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.632488489151001, + "learning_rate": 1e-06, + "loss": 1.131, + "mean_token_accuracy": 0.6842373609542847, + "num_tokens": 84226155.0, + "step": 3260 + }, + { + "epoch": 0.35811552822314957, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.677311420440674, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.715744137763977, + "num_tokens": 84249673.0, + "step": 3261 + }, + { + "epoch": 0.3582253459257632, + "ewc_loss": 5.751848220825195e-06, + "grad_norm": 2.718899965286255, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6976057291030884, + "num_tokens": 84273724.0, + "step": 3262 + }, + { + "epoch": 0.3583351636283769, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.7992513179779053, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6897468566894531, + "num_tokens": 84299628.0, + "step": 3263 + }, + { + "epoch": 0.35844498133099056, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.798631191253662, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7088899612426758, + "num_tokens": 84327527.0, + "step": 3264 + }, + { + "epoch": 0.3585547990336042, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.602181911468506, + "learning_rate": 1e-06, + "loss": 1.0935, + "mean_token_accuracy": 0.6871308088302612, + "num_tokens": 84354889.0, + "step": 3265 + }, + { + "epoch": 0.35866461673621786, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.7895936965942383, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6946630477905273, + "num_tokens": 84380522.0, + "step": 3266 + }, + { + "epoch": 0.35877443443883156, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.613813877105713, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7091073393821716, + "num_tokens": 84405266.0, + "step": 3267 + }, + { + "epoch": 0.3588842521414452, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.8417415618896484, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7161600589752197, + "num_tokens": 84428101.0, + "step": 3268 + }, + { + "epoch": 0.35899406984405885, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.7520813941955566, + "learning_rate": 1e-06, + "loss": 1.1249, + "mean_token_accuracy": 0.6754797697067261, + "num_tokens": 84454829.0, + "step": 3269 + }, + { + "epoch": 0.35910388754667255, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.6516900062561035, + "learning_rate": 1e-06, + "loss": 0.9303, + "mean_token_accuracy": 0.7229932546615601, + "num_tokens": 84475796.0, + "step": 3270 + }, + { + "epoch": 0.3592137052492862, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.4782180786132812, + "learning_rate": 1e-06, + "loss": 1.1155, + "mean_token_accuracy": 0.672696590423584, + "num_tokens": 84506712.0, + "step": 3271 + }, + { + "epoch": 0.35932352295189984, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.854517698287964, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7051514983177185, + "num_tokens": 84527613.0, + "step": 3272 + }, + { + "epoch": 0.3594333406545135, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.746999740600586, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.6936733722686768, + "num_tokens": 84552379.0, + "step": 3273 + }, + { + "epoch": 0.3595431583571272, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.3967432975769043, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6778444051742554, + "num_tokens": 84583957.0, + "step": 3274 + }, + { + "epoch": 0.35965297605974084, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.355003595352173, + "learning_rate": 1e-06, + "loss": 1.124, + "mean_token_accuracy": 0.6704739332199097, + "num_tokens": 84616061.0, + "step": 3275 + }, + { + "epoch": 0.3597627937623545, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.723776340484619, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.692245602607727, + "num_tokens": 84639966.0, + "step": 3276 + }, + { + "epoch": 0.35987261146496813, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.4273064136505127, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6975909471511841, + "num_tokens": 84670805.0, + "step": 3277 + }, + { + "epoch": 0.35998242916758183, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.577141761779785, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6929799318313599, + "num_tokens": 84695492.0, + "step": 3278 + }, + { + "epoch": 0.3600922468701955, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.547342300415039, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7223255634307861, + "num_tokens": 84721170.0, + "step": 3279 + }, + { + "epoch": 0.3602020645728091, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 3.0334174633026123, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6900395154953003, + "num_tokens": 84746589.0, + "step": 3280 + }, + { + "epoch": 0.3603118822754228, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.6448142528533936, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6932341456413269, + "num_tokens": 84770710.0, + "step": 3281 + }, + { + "epoch": 0.36042169997803647, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.3073971271514893, + "learning_rate": 1e-06, + "loss": 1.1698, + "mean_token_accuracy": 0.6634035110473633, + "num_tokens": 84806101.0, + "step": 3282 + }, + { + "epoch": 0.3605315176806501, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.428530216217041, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7084286212921143, + "num_tokens": 84835710.0, + "step": 3283 + }, + { + "epoch": 0.36064133538326376, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.590639114379883, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.691123366355896, + "num_tokens": 84863843.0, + "step": 3284 + }, + { + "epoch": 0.36075115308587746, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.6920928955078125, + "learning_rate": 1e-06, + "loss": 1.0765, + "mean_token_accuracy": 0.6853451728820801, + "num_tokens": 84886682.0, + "step": 3285 + }, + { + "epoch": 0.3608609707884911, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.899095296859741, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6930177807807922, + "num_tokens": 84907655.0, + "step": 3286 + }, + { + "epoch": 0.36097078849110475, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.487926483154297, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.6911777853965759, + "num_tokens": 84936194.0, + "step": 3287 + }, + { + "epoch": 0.3610806061937184, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.49721097946167, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6833666563034058, + "num_tokens": 84964512.0, + "step": 3288 + }, + { + "epoch": 0.3611904238963321, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.6255533695220947, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6963951587677002, + "num_tokens": 84991502.0, + "step": 3289 + }, + { + "epoch": 0.36130024159894575, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.7047436237335205, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6859991550445557, + "num_tokens": 85013129.0, + "step": 3290 + }, + { + "epoch": 0.3614100593015594, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.713458299636841, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7147481441497803, + "num_tokens": 85036620.0, + "step": 3291 + }, + { + "epoch": 0.3615198770041731, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.8125011920928955, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6914551258087158, + "num_tokens": 85060171.0, + "step": 3292 + }, + { + "epoch": 0.36162969470678674, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.5278639793395996, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6833814382553101, + "num_tokens": 85086553.0, + "step": 3293 + }, + { + "epoch": 0.3617395124094004, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.3967764377593994, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.6900588870048523, + "num_tokens": 85118680.0, + "step": 3294 + }, + { + "epoch": 0.36184933011201403, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.6171066761016846, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7104734778404236, + "num_tokens": 85142298.0, + "step": 3295 + }, + { + "epoch": 0.36195914781462774, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.949885368347168, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7083368301391602, + "num_tokens": 85170484.0, + "step": 3296 + }, + { + "epoch": 0.3620689655172414, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.5638821125030518, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.71540367603302, + "num_tokens": 85196017.0, + "step": 3297 + }, + { + "epoch": 0.362178783219855, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.8093340396881104, + "learning_rate": 1e-06, + "loss": 0.9294, + "mean_token_accuracy": 0.7221471071243286, + "num_tokens": 85216845.0, + "step": 3298 + }, + { + "epoch": 0.36228860092246873, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.605668067932129, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7096033692359924, + "num_tokens": 85243824.0, + "step": 3299 + }, + { + "epoch": 0.3623984186250824, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.473285675048828, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.7019424438476562, + "num_tokens": 85274198.0, + "step": 3300 + }, + { + "epoch": 0.362508236327696, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.5928773880004883, + "learning_rate": 1e-06, + "loss": 1.0623, + "mean_token_accuracy": 0.6956966519355774, + "num_tokens": 85303239.0, + "step": 3301 + }, + { + "epoch": 0.36261805403030967, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.507960557937622, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7040790319442749, + "num_tokens": 85330757.0, + "step": 3302 + }, + { + "epoch": 0.36272787173292337, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.6075010299682617, + "learning_rate": 1e-06, + "loss": 0.9002, + "mean_token_accuracy": 0.7291606664657593, + "num_tokens": 85354065.0, + "step": 3303 + }, + { + "epoch": 0.362837689435537, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 16.701251983642578, + "learning_rate": 1e-06, + "loss": 0.8811, + "mean_token_accuracy": 0.7370626926422119, + "num_tokens": 85377501.0, + "step": 3304 + }, + { + "epoch": 0.36294750713815066, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.6591103076934814, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6909551024436951, + "num_tokens": 85405009.0, + "step": 3305 + }, + { + "epoch": 0.3630573248407643, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.8570125102996826, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7228339910507202, + "num_tokens": 85425905.0, + "step": 3306 + }, + { + "epoch": 0.363167142543378, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.622549057006836, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7047672271728516, + "num_tokens": 85452029.0, + "step": 3307 + }, + { + "epoch": 0.36327696024599165, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.564406633377075, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6998482346534729, + "num_tokens": 85478947.0, + "step": 3308 + }, + { + "epoch": 0.3633867779486053, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.215937376022339, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6955122947692871, + "num_tokens": 85510887.0, + "step": 3309 + }, + { + "epoch": 0.363496595651219, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.4638633728027344, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6917674541473389, + "num_tokens": 85539105.0, + "step": 3310 + }, + { + "epoch": 0.36360641335383265, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.870534896850586, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7141705751419067, + "num_tokens": 85561827.0, + "step": 3311 + }, + { + "epoch": 0.3637162310564463, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.672739267349243, + "learning_rate": 1e-06, + "loss": 1.1349, + "mean_token_accuracy": 0.6692079305648804, + "num_tokens": 85594193.0, + "step": 3312 + }, + { + "epoch": 0.36382604875905994, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 3.1466195583343506, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6914807558059692, + "num_tokens": 85619805.0, + "step": 3313 + }, + { + "epoch": 0.36393586646167364, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 3.022625207901001, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.700652539730072, + "num_tokens": 85640275.0, + "step": 3314 + }, + { + "epoch": 0.3640456841642873, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.6218760013580322, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6867858171463013, + "num_tokens": 85665126.0, + "step": 3315 + }, + { + "epoch": 0.36415550186690093, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 3.0311684608459473, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.7002121210098267, + "num_tokens": 85689650.0, + "step": 3316 + }, + { + "epoch": 0.3642653195695146, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.8147478103637695, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6915767192840576, + "num_tokens": 85720567.0, + "step": 3317 + }, + { + "epoch": 0.3643751372721283, + "ewc_loss": 5.781650543212891e-06, + "grad_norm": 2.6284170150756836, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.691871702671051, + "num_tokens": 85744555.0, + "step": 3318 + }, + { + "epoch": 0.3644849549747419, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.4430177211761475, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.6981557011604309, + "num_tokens": 85769197.0, + "step": 3319 + }, + { + "epoch": 0.36459477267735557, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.601250410079956, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7017290592193604, + "num_tokens": 85794927.0, + "step": 3320 + }, + { + "epoch": 0.3647045903799693, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.480236768722534, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7022719383239746, + "num_tokens": 85820362.0, + "step": 3321 + }, + { + "epoch": 0.3648144080825829, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.7377655506134033, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7203505039215088, + "num_tokens": 85841839.0, + "step": 3322 + }, + { + "epoch": 0.36492422578519657, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.675879955291748, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6867406368255615, + "num_tokens": 85867377.0, + "step": 3323 + }, + { + "epoch": 0.3650340434878102, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.537325859069824, + "learning_rate": 1e-06, + "loss": 1.094, + "mean_token_accuracy": 0.6802972555160522, + "num_tokens": 85894728.0, + "step": 3324 + }, + { + "epoch": 0.3651438611904239, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.5444462299346924, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7114945650100708, + "num_tokens": 85920460.0, + "step": 3325 + }, + { + "epoch": 0.36525367889303756, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.598299741744995, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6871049404144287, + "num_tokens": 85947417.0, + "step": 3326 + }, + { + "epoch": 0.3653634965956512, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.3368871212005615, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6829768419265747, + "num_tokens": 85978475.0, + "step": 3327 + }, + { + "epoch": 0.3654733142982649, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.590372323989868, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6926605105400085, + "num_tokens": 86005268.0, + "step": 3328 + }, + { + "epoch": 0.36558313200087855, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.3949637413024902, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7098218202590942, + "num_tokens": 86032447.0, + "step": 3329 + }, + { + "epoch": 0.3656929497034922, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.749741554260254, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6815719604492188, + "num_tokens": 86055965.0, + "step": 3330 + }, + { + "epoch": 0.36580276740610584, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.546908378601074, + "learning_rate": 1e-06, + "loss": 1.1011, + "mean_token_accuracy": 0.6739706993103027, + "num_tokens": 86083407.0, + "step": 3331 + }, + { + "epoch": 0.36591258510871955, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.576511859893799, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7048264741897583, + "num_tokens": 86107077.0, + "step": 3332 + }, + { + "epoch": 0.3660224028113332, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.7152252197265625, + "learning_rate": 1e-06, + "loss": 1.0974, + "mean_token_accuracy": 0.6877135634422302, + "num_tokens": 86131381.0, + "step": 3333 + }, + { + "epoch": 0.36613222051394684, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.8933355808258057, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6883531808853149, + "num_tokens": 86154122.0, + "step": 3334 + }, + { + "epoch": 0.3662420382165605, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.7962567806243896, + "learning_rate": 1e-06, + "loss": 1.0832, + "mean_token_accuracy": 0.6912553310394287, + "num_tokens": 86180267.0, + "step": 3335 + }, + { + "epoch": 0.3663518559191742, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.2359142303466797, + "learning_rate": 1e-06, + "loss": 1.1191, + "mean_token_accuracy": 0.6762961149215698, + "num_tokens": 86215594.0, + "step": 3336 + }, + { + "epoch": 0.36646167362178783, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.4721388816833496, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6901159286499023, + "num_tokens": 86246071.0, + "step": 3337 + }, + { + "epoch": 0.3665714913244015, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.3632779121398926, + "learning_rate": 1e-06, + "loss": 1.1427, + "mean_token_accuracy": 0.6681175231933594, + "num_tokens": 86277975.0, + "step": 3338 + }, + { + "epoch": 0.3666813090270152, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.466552734375, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6931948661804199, + "num_tokens": 86304784.0, + "step": 3339 + }, + { + "epoch": 0.3667911267296288, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.501328706741333, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7098544836044312, + "num_tokens": 86334727.0, + "step": 3340 + }, + { + "epoch": 0.36690094443224247, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.5106539726257324, + "learning_rate": 1e-06, + "loss": 1.1268, + "mean_token_accuracy": 0.6907024383544922, + "num_tokens": 86360301.0, + "step": 3341 + }, + { + "epoch": 0.3670107621348561, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.584935426712036, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7082974314689636, + "num_tokens": 86384913.0, + "step": 3342 + }, + { + "epoch": 0.3671205798374698, + "ewc_loss": 5.811452865600586e-06, + "grad_norm": 2.6277804374694824, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7008106708526611, + "num_tokens": 86410825.0, + "step": 3343 + }, + { + "epoch": 0.36723039754008346, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.5389678478240967, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6915473937988281, + "num_tokens": 86436639.0, + "step": 3344 + }, + { + "epoch": 0.3673402152426971, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.489583730697632, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.6968071460723877, + "num_tokens": 86465164.0, + "step": 3345 + }, + { + "epoch": 0.3674500329453108, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.7271173000335693, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6894948482513428, + "num_tokens": 86487983.0, + "step": 3346 + }, + { + "epoch": 0.36755985064792446, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.3602805137634277, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7040400505065918, + "num_tokens": 86515243.0, + "step": 3347 + }, + { + "epoch": 0.3676696683505381, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.6183362007141113, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7080901861190796, + "num_tokens": 86540172.0, + "step": 3348 + }, + { + "epoch": 0.36777948605315175, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.2990479469299316, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6796683073043823, + "num_tokens": 86571138.0, + "step": 3349 + }, + { + "epoch": 0.36788930375576545, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.3100879192352295, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.6825119256973267, + "num_tokens": 86604797.0, + "step": 3350 + }, + { + "epoch": 0.3679991214583791, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.5668671131134033, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7100663781166077, + "num_tokens": 86628636.0, + "step": 3351 + }, + { + "epoch": 0.36810893916099274, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.32942795753479, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6915109157562256, + "num_tokens": 86663120.0, + "step": 3352 + }, + { + "epoch": 0.3682187568636064, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.4912428855895996, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7153496146202087, + "num_tokens": 86689552.0, + "step": 3353 + }, + { + "epoch": 0.3683285745662201, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.789067268371582, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6979593634605408, + "num_tokens": 86712863.0, + "step": 3354 + }, + { + "epoch": 0.36843839226883374, + "ewc_loss": 5.841255187988281e-06, + "grad_norm": 2.638559341430664, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.6897733807563782, + "num_tokens": 86736563.0, + "step": 3355 + }, + { + "epoch": 0.3685482099714474, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.7103829383850098, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7032729387283325, + "num_tokens": 86764256.0, + "step": 3356 + }, + { + "epoch": 0.3686580276740611, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6899874210357666, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7163751125335693, + "num_tokens": 86787797.0, + "step": 3357 + }, + { + "epoch": 0.36876784537667473, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.507874011993408, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.712792158126831, + "num_tokens": 86815390.0, + "step": 3358 + }, + { + "epoch": 0.3688776630792884, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.492039203643799, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7304738759994507, + "num_tokens": 86846028.0, + "step": 3359 + }, + { + "epoch": 0.368987480781902, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6831576824188232, + "learning_rate": 1e-06, + "loss": 1.1083, + "mean_token_accuracy": 0.6802272796630859, + "num_tokens": 86869114.0, + "step": 3360 + }, + { + "epoch": 0.3690972984845157, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6267011165618896, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6934465169906616, + "num_tokens": 86893563.0, + "step": 3361 + }, + { + "epoch": 0.36920711618712937, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6515719890594482, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7010975480079651, + "num_tokens": 86916057.0, + "step": 3362 + }, + { + "epoch": 0.369316933889743, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.5412254333496094, + "learning_rate": 1e-06, + "loss": 1.0753, + "mean_token_accuracy": 0.6852668523788452, + "num_tokens": 86944193.0, + "step": 3363 + }, + { + "epoch": 0.36942675159235666, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 16.726341247558594, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6945791840553284, + "num_tokens": 86973130.0, + "step": 3364 + }, + { + "epoch": 0.36953656929497036, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.8268635272979736, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7158610820770264, + "num_tokens": 86997045.0, + "step": 3365 + }, + { + "epoch": 0.369646386997584, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.7830400466918945, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6876566410064697, + "num_tokens": 87021295.0, + "step": 3366 + }, + { + "epoch": 0.36975620470019765, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.481935739517212, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6984759569168091, + "num_tokens": 87046413.0, + "step": 3367 + }, + { + "epoch": 0.36986602240281136, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.7260499000549316, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.691772997379303, + "num_tokens": 87073882.0, + "step": 3368 + }, + { + "epoch": 0.369975840105425, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.500028371810913, + "learning_rate": 1e-06, + "loss": 1.1196, + "mean_token_accuracy": 0.6831362247467041, + "num_tokens": 87100385.0, + "step": 3369 + }, + { + "epoch": 0.37008565780803865, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.531155586242676, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7118091583251953, + "num_tokens": 87126593.0, + "step": 3370 + }, + { + "epoch": 0.3701954755106523, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.3050124645233154, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6855183243751526, + "num_tokens": 87161071.0, + "step": 3371 + }, + { + "epoch": 0.370305293213266, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.614809513092041, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6955246925354004, + "num_tokens": 87186272.0, + "step": 3372 + }, + { + "epoch": 0.37041511091587964, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.3872456550598145, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7046982049942017, + "num_tokens": 87212120.0, + "step": 3373 + }, + { + "epoch": 0.3705249286184933, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.8190395832061768, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7060421705245972, + "num_tokens": 87231883.0, + "step": 3374 + }, + { + "epoch": 0.370634746321107, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.4161651134490967, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.6833316087722778, + "num_tokens": 87260370.0, + "step": 3375 + }, + { + "epoch": 0.37074456402372064, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.504392385482788, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6926552653312683, + "num_tokens": 87286894.0, + "step": 3376 + }, + { + "epoch": 0.3708543817263343, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.3070764541625977, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6807612776756287, + "num_tokens": 87321398.0, + "step": 3377 + }, + { + "epoch": 0.3709641994289479, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.7582948207855225, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.712308943271637, + "num_tokens": 87345279.0, + "step": 3378 + }, + { + "epoch": 0.37107401713156163, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6689200401306152, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7017963528633118, + "num_tokens": 87369573.0, + "step": 3379 + }, + { + "epoch": 0.3711838348341753, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6791558265686035, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6826487183570862, + "num_tokens": 87396176.0, + "step": 3380 + }, + { + "epoch": 0.3712936525367889, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.627056121826172, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6930351853370667, + "num_tokens": 87426871.0, + "step": 3381 + }, + { + "epoch": 0.37140347023940257, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.7717316150665283, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6999877691268921, + "num_tokens": 87453349.0, + "step": 3382 + }, + { + "epoch": 0.37151328794201627, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.515002489089966, + "learning_rate": 1e-06, + "loss": 1.1455, + "mean_token_accuracy": 0.6943769454956055, + "num_tokens": 87484742.0, + "step": 3383 + }, + { + "epoch": 0.3716231056446299, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.7324376106262207, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7120988368988037, + "num_tokens": 87506704.0, + "step": 3384 + }, + { + "epoch": 0.37173292334724356, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.605482339859009, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6921480894088745, + "num_tokens": 87529839.0, + "step": 3385 + }, + { + "epoch": 0.37184274104985726, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.61011004447937, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6889461874961853, + "num_tokens": 87558485.0, + "step": 3386 + }, + { + "epoch": 0.3719525587524709, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 3.3702869415283203, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7054029703140259, + "num_tokens": 87586352.0, + "step": 3387 + }, + { + "epoch": 0.37206237645508455, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.9679534435272217, + "learning_rate": 1e-06, + "loss": 1.1274, + "mean_token_accuracy": 0.6870761513710022, + "num_tokens": 87614672.0, + "step": 3388 + }, + { + "epoch": 0.3721721941576982, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.410343885421753, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7118836641311646, + "num_tokens": 87643313.0, + "step": 3389 + }, + { + "epoch": 0.3722820118603119, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.4269537925720215, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6951963901519775, + "num_tokens": 87670451.0, + "step": 3390 + }, + { + "epoch": 0.37239182956292555, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.8083410263061523, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.684352457523346, + "num_tokens": 87692044.0, + "step": 3391 + }, + { + "epoch": 0.3725016472655392, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.3708245754241943, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7017014622688293, + "num_tokens": 87721089.0, + "step": 3392 + }, + { + "epoch": 0.37261146496815284, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 3.104275703430176, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6993225812911987, + "num_tokens": 87743663.0, + "step": 3393 + }, + { + "epoch": 0.37272128267076654, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.817039966583252, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.70749831199646, + "num_tokens": 87764986.0, + "step": 3394 + }, + { + "epoch": 0.3728311003733802, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.609612464904785, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7056546211242676, + "num_tokens": 87790380.0, + "step": 3395 + }, + { + "epoch": 0.37294091807599383, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6164698600769043, + "learning_rate": 1e-06, + "loss": 1.0765, + "mean_token_accuracy": 0.6865447759628296, + "num_tokens": 87814012.0, + "step": 3396 + }, + { + "epoch": 0.37305073577860753, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.580397367477417, + "learning_rate": 1e-06, + "loss": 1.0905, + "mean_token_accuracy": 0.6878156661987305, + "num_tokens": 87838307.0, + "step": 3397 + }, + { + "epoch": 0.3731605534812212, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.631387948989868, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6977519392967224, + "num_tokens": 87864332.0, + "step": 3398 + }, + { + "epoch": 0.3732703711838348, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.9100677967071533, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6900543570518494, + "num_tokens": 87884903.0, + "step": 3399 + }, + { + "epoch": 0.37338018888644847, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.604137659072876, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.6952813863754272, + "num_tokens": 87908543.0, + "step": 3400 + }, + { + "epoch": 0.3734900065890622, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.710853338241577, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7055951356887817, + "num_tokens": 87930022.0, + "step": 3401 + }, + { + "epoch": 0.3735998242916758, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.32706356048584, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6892264485359192, + "num_tokens": 87961748.0, + "step": 3402 + }, + { + "epoch": 0.37370964199428947, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.374804973602295, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7056812644004822, + "num_tokens": 87995355.0, + "step": 3403 + }, + { + "epoch": 0.37381945969690317, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 3.04378342628479, + "learning_rate": 1e-06, + "loss": 1.0927, + "mean_token_accuracy": 0.6777762174606323, + "num_tokens": 88026996.0, + "step": 3404 + }, + { + "epoch": 0.3739292773995168, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.510782241821289, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7184262871742249, + "num_tokens": 88052140.0, + "step": 3405 + }, + { + "epoch": 0.37403909510213046, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.8146018981933594, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6876193284988403, + "num_tokens": 88073640.0, + "step": 3406 + }, + { + "epoch": 0.3741489128047441, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6739015579223633, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6960582733154297, + "num_tokens": 88096965.0, + "step": 3407 + }, + { + "epoch": 0.3742587305073578, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.29378080368042, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6888479590415955, + "num_tokens": 88128966.0, + "step": 3408 + }, + { + "epoch": 0.37436854820997145, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.5229103565216064, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.709255039691925, + "num_tokens": 88152934.0, + "step": 3409 + }, + { + "epoch": 0.3744783659125851, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.773959159851074, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6826019287109375, + "num_tokens": 88175382.0, + "step": 3410 + }, + { + "epoch": 0.37458818361519874, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.4233474731445312, + "learning_rate": 1e-06, + "loss": 1.1258, + "mean_token_accuracy": 0.6840710043907166, + "num_tokens": 88204938.0, + "step": 3411 + }, + { + "epoch": 0.37469800131781245, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.752901792526245, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.6979819536209106, + "num_tokens": 88226259.0, + "step": 3412 + }, + { + "epoch": 0.3748078190204261, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.8527021408081055, + "learning_rate": 1e-06, + "loss": 1.1122, + "mean_token_accuracy": 0.6719470024108887, + "num_tokens": 88251893.0, + "step": 3413 + }, + { + "epoch": 0.37491763672303974, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.503216028213501, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7033432126045227, + "num_tokens": 88277786.0, + "step": 3414 + }, + { + "epoch": 0.37502745442565344, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.8534059524536133, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7006555795669556, + "num_tokens": 88300143.0, + "step": 3415 + }, + { + "epoch": 0.3751372721282671, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.799994945526123, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7013038992881775, + "num_tokens": 88323457.0, + "step": 3416 + }, + { + "epoch": 0.37524708983088073, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.82295823097229, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7093772292137146, + "num_tokens": 88343643.0, + "step": 3417 + }, + { + "epoch": 0.3753569075334944, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.3575241565704346, + "learning_rate": 1e-06, + "loss": 1.1255, + "mean_token_accuracy": 0.6698559522628784, + "num_tokens": 88373513.0, + "step": 3418 + }, + { + "epoch": 0.3754667252361081, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.4396095275878906, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6890098452568054, + "num_tokens": 88402245.0, + "step": 3419 + }, + { + "epoch": 0.3755765429387217, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.7857894897460938, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6921026110649109, + "num_tokens": 88430919.0, + "step": 3420 + }, + { + "epoch": 0.37568636064133537, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.522022247314453, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7084053754806519, + "num_tokens": 88460174.0, + "step": 3421 + }, + { + "epoch": 0.37579617834394907, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.522280216217041, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6874221563339233, + "num_tokens": 88488063.0, + "step": 3422 + }, + { + "epoch": 0.3759059960465627, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.3711981773376465, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.691695511341095, + "num_tokens": 88520294.0, + "step": 3423 + }, + { + "epoch": 0.37601581374917636, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.2894387245178223, + "learning_rate": 1e-06, + "loss": 1.11, + "mean_token_accuracy": 0.6761735677719116, + "num_tokens": 88554544.0, + "step": 3424 + }, + { + "epoch": 0.37612563145179, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.575169563293457, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7065541744232178, + "num_tokens": 88578934.0, + "step": 3425 + }, + { + "epoch": 0.3762354491544037, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.7690000534057617, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7028264999389648, + "num_tokens": 88600495.0, + "step": 3426 + }, + { + "epoch": 0.37634526685701736, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.4937844276428223, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7107672691345215, + "num_tokens": 88627034.0, + "step": 3427 + }, + { + "epoch": 0.376455084559631, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.687783718109131, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.6982094645500183, + "num_tokens": 88650799.0, + "step": 3428 + }, + { + "epoch": 0.37656490226224465, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.817441701889038, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7055004835128784, + "num_tokens": 88670895.0, + "step": 3429 + }, + { + "epoch": 0.37667471996485835, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.679645299911499, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.692072868347168, + "num_tokens": 88693919.0, + "step": 3430 + }, + { + "epoch": 0.376784537667472, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.671046495437622, + "learning_rate": 1e-06, + "loss": 1.1152, + "mean_token_accuracy": 0.6730226278305054, + "num_tokens": 88721418.0, + "step": 3431 + }, + { + "epoch": 0.37689435537008564, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.6675753593444824, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7171578407287598, + "num_tokens": 88745403.0, + "step": 3432 + }, + { + "epoch": 0.37700417307269934, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.341871500015259, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6820418238639832, + "num_tokens": 88780492.0, + "step": 3433 + }, + { + "epoch": 0.377113990775313, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.399491786956787, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6865644454956055, + "num_tokens": 88811625.0, + "step": 3434 + }, + { + "epoch": 0.37722380847792664, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.4755873680114746, + "learning_rate": 1e-06, + "loss": 1.1651, + "mean_token_accuracy": 0.6760313510894775, + "num_tokens": 88843377.0, + "step": 3435 + }, + { + "epoch": 0.3773336261805403, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.549638509750366, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7031643390655518, + "num_tokens": 88871222.0, + "step": 3436 + }, + { + "epoch": 0.377443443883154, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.11604642868042, + "learning_rate": 1e-06, + "loss": 1.1228, + "mean_token_accuracy": 0.6710485219955444, + "num_tokens": 88906003.0, + "step": 3437 + }, + { + "epoch": 0.37755326158576763, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.6090385913848877, + "learning_rate": 1e-06, + "loss": 1.1125, + "mean_token_accuracy": 0.6814054846763611, + "num_tokens": 88934175.0, + "step": 3438 + }, + { + "epoch": 0.3776630792883813, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.4217875003814697, + "learning_rate": 1e-06, + "loss": 1.128, + "mean_token_accuracy": 0.6738500595092773, + "num_tokens": 88967305.0, + "step": 3439 + }, + { + "epoch": 0.3777728969909949, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.5553390979766846, + "learning_rate": 1e-06, + "loss": 1.1321, + "mean_token_accuracy": 0.6719250679016113, + "num_tokens": 88995058.0, + "step": 3440 + }, + { + "epoch": 0.3778827146936086, + "ewc_loss": 5.8710575103759766e-06, + "grad_norm": 2.854107141494751, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.685825765132904, + "num_tokens": 89016825.0, + "step": 3441 + }, + { + "epoch": 0.37799253239622227, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.583155632019043, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.6968756318092346, + "num_tokens": 89044034.0, + "step": 3442 + }, + { + "epoch": 0.3781023500988359, + "ewc_loss": 5.900859832763672e-06, + "grad_norm": 2.6726620197296143, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6989080905914307, + "num_tokens": 89068052.0, + "step": 3443 + }, + { + "epoch": 0.3782121678014496, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.5064656734466553, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7103807926177979, + "num_tokens": 89094888.0, + "step": 3444 + }, + { + "epoch": 0.37832198550406326, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.4415557384490967, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.6992026567459106, + "num_tokens": 89120885.0, + "step": 3445 + }, + { + "epoch": 0.3784318032066769, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.4580564498901367, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7132003903388977, + "num_tokens": 89148606.0, + "step": 3446 + }, + { + "epoch": 0.37854162090929055, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.7842302322387695, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7225213050842285, + "num_tokens": 89170649.0, + "step": 3447 + }, + { + "epoch": 0.37865143861190426, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.5646793842315674, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6878378391265869, + "num_tokens": 89196599.0, + "step": 3448 + }, + { + "epoch": 0.3787612563145179, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 3.0269689559936523, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7063471674919128, + "num_tokens": 89218130.0, + "step": 3449 + }, + { + "epoch": 0.37887107401713155, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.795320987701416, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.692419707775116, + "num_tokens": 89242478.0, + "step": 3450 + }, + { + "epoch": 0.37898089171974525, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.682795286178589, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6916239857673645, + "num_tokens": 89267931.0, + "step": 3451 + }, + { + "epoch": 0.3790907094223589, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.5476534366607666, + "learning_rate": 1e-06, + "loss": 0.9785, + "mean_token_accuracy": 0.7189674377441406, + "num_tokens": 89298706.0, + "step": 3452 + }, + { + "epoch": 0.37920052712497254, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.5613162517547607, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.707207202911377, + "num_tokens": 89325375.0, + "step": 3453 + }, + { + "epoch": 0.3793103448275862, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.7060489654541016, + "learning_rate": 1e-06, + "loss": 0.9775, + "mean_token_accuracy": 0.7085207104682922, + "num_tokens": 89346526.0, + "step": 3454 + }, + { + "epoch": 0.3794201625301999, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.544368267059326, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6900359392166138, + "num_tokens": 89372830.0, + "step": 3455 + }, + { + "epoch": 0.37952998023281354, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.5305545330047607, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6974303126335144, + "num_tokens": 89396497.0, + "step": 3456 + }, + { + "epoch": 0.3796397979354272, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.5803377628326416, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7049925327301025, + "num_tokens": 89422444.0, + "step": 3457 + }, + { + "epoch": 0.3797496156380408, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.573986768722534, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6808978319168091, + "num_tokens": 89447722.0, + "step": 3458 + }, + { + "epoch": 0.37985943334065453, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.518519163131714, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7199646234512329, + "num_tokens": 89472695.0, + "step": 3459 + }, + { + "epoch": 0.3799692510432682, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.6232011318206787, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.690087080001831, + "num_tokens": 89504156.0, + "step": 3460 + }, + { + "epoch": 0.3800790687458818, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.3041176795959473, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6915359497070312, + "num_tokens": 89533887.0, + "step": 3461 + }, + { + "epoch": 0.3801888864484955, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.714977502822876, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7027379870414734, + "num_tokens": 89558518.0, + "step": 3462 + }, + { + "epoch": 0.38029870415110917, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.7453126907348633, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6892021298408508, + "num_tokens": 89579890.0, + "step": 3463 + }, + { + "epoch": 0.3804085218537228, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.535055637359619, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7016713619232178, + "num_tokens": 89604732.0, + "step": 3464 + }, + { + "epoch": 0.38051833955633646, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.8259453773498535, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7082880735397339, + "num_tokens": 89625793.0, + "step": 3465 + }, + { + "epoch": 0.38062815725895016, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.705991744995117, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7102546691894531, + "num_tokens": 89648859.0, + "step": 3466 + }, + { + "epoch": 0.3807379749615638, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.413628339767456, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6958940029144287, + "num_tokens": 89679663.0, + "step": 3467 + }, + { + "epoch": 0.38084779266417745, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.503060817718506, + "learning_rate": 1e-06, + "loss": 1.1141, + "mean_token_accuracy": 0.6787940859794617, + "num_tokens": 89705345.0, + "step": 3468 + }, + { + "epoch": 0.3809576103667911, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.689408302307129, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6910281181335449, + "num_tokens": 89730700.0, + "step": 3469 + }, + { + "epoch": 0.3810674280694048, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.323671579360962, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7093394994735718, + "num_tokens": 89763152.0, + "step": 3470 + }, + { + "epoch": 0.38117724577201845, + "ewc_loss": 5.930662155151367e-06, + "grad_norm": 2.744509220123291, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6884607672691345, + "num_tokens": 89786840.0, + "step": 3471 + }, + { + "epoch": 0.3812870634746321, + "ewc_loss": 5.990266799926758e-06, + "grad_norm": 2.652702808380127, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7104862928390503, + "num_tokens": 89811531.0, + "step": 3472 + }, + { + "epoch": 0.3813968811772458, + "ewc_loss": 5.990266799926758e-06, + "grad_norm": 2.618638038635254, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6921305656433105, + "num_tokens": 89837392.0, + "step": 3473 + }, + { + "epoch": 0.38150669887985944, + "ewc_loss": 5.990266799926758e-06, + "grad_norm": 2.6774041652679443, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7110572457313538, + "num_tokens": 89861657.0, + "step": 3474 + }, + { + "epoch": 0.3816165165824731, + "ewc_loss": 5.990266799926758e-06, + "grad_norm": 2.657684087753296, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6980605721473694, + "num_tokens": 89888089.0, + "step": 3475 + }, + { + "epoch": 0.38172633428508673, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 3.104940176010132, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.6968487501144409, + "num_tokens": 89922136.0, + "step": 3476 + }, + { + "epoch": 0.38183615198770043, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.5631165504455566, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7055988311767578, + "num_tokens": 89946651.0, + "step": 3477 + }, + { + "epoch": 0.3819459696903141, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.9301483631134033, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.6991882920265198, + "num_tokens": 89970358.0, + "step": 3478 + }, + { + "epoch": 0.3820557873929277, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.708820104598999, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6866781115531921, + "num_tokens": 89994077.0, + "step": 3479 + }, + { + "epoch": 0.3821656050955414, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.7702393531799316, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6860854029655457, + "num_tokens": 90017250.0, + "step": 3480 + }, + { + "epoch": 0.3822754227981551, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.4774420261383057, + "learning_rate": 1e-06, + "loss": 1.1084, + "mean_token_accuracy": 0.6786786317825317, + "num_tokens": 90046866.0, + "step": 3481 + }, + { + "epoch": 0.3823852405007687, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.862135648727417, + "learning_rate": 1e-06, + "loss": 1.1374, + "mean_token_accuracy": 0.6763011813163757, + "num_tokens": 90070820.0, + "step": 3482 + }, + { + "epoch": 0.38249505820338237, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.4148943424224854, + "learning_rate": 1e-06, + "loss": 1.0932, + "mean_token_accuracy": 0.682287871837616, + "num_tokens": 90103266.0, + "step": 3483 + }, + { + "epoch": 0.38260487590599607, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.7689287662506104, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7039887309074402, + "num_tokens": 90128277.0, + "step": 3484 + }, + { + "epoch": 0.3827146936086097, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.5131876468658447, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6889430284500122, + "num_tokens": 90158627.0, + "step": 3485 + }, + { + "epoch": 0.38282451131122336, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.483161687850952, + "learning_rate": 1e-06, + "loss": 1.112, + "mean_token_accuracy": 0.6759645342826843, + "num_tokens": 90188177.0, + "step": 3486 + }, + { + "epoch": 0.382934329013837, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.277494192123413, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6852846145629883, + "num_tokens": 90219370.0, + "step": 3487 + }, + { + "epoch": 0.3830441467164507, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.8596842288970947, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7250714898109436, + "num_tokens": 90239794.0, + "step": 3488 + }, + { + "epoch": 0.38315396441906435, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.327120542526245, + "learning_rate": 1e-06, + "loss": 1.1112, + "mean_token_accuracy": 0.676780104637146, + "num_tokens": 90270780.0, + "step": 3489 + }, + { + "epoch": 0.383263782121678, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.5893707275390625, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7148138880729675, + "num_tokens": 90299170.0, + "step": 3490 + }, + { + "epoch": 0.3833735998242917, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.492239475250244, + "learning_rate": 1e-06, + "loss": 1.1357, + "mean_token_accuracy": 0.6725925207138062, + "num_tokens": 90330360.0, + "step": 3491 + }, + { + "epoch": 0.38348341752690535, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.8943395614624023, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7052329778671265, + "num_tokens": 90350344.0, + "step": 3492 + }, + { + "epoch": 0.383593235229519, + "ewc_loss": 5.9604644775390625e-06, + "grad_norm": 2.6783807277679443, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7168005704879761, + "num_tokens": 90374020.0, + "step": 3493 + }, + { + "epoch": 0.38370305293213264, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.6994261741638184, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7009636759757996, + "num_tokens": 90398968.0, + "step": 3494 + }, + { + "epoch": 0.38381287063474634, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.6866507530212402, + "learning_rate": 1e-06, + "loss": 0.9041, + "mean_token_accuracy": 0.732031524181366, + "num_tokens": 90424099.0, + "step": 3495 + }, + { + "epoch": 0.38392268833736, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.6997578144073486, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6938135623931885, + "num_tokens": 90447219.0, + "step": 3496 + }, + { + "epoch": 0.38403250603997363, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.52132511138916, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7065427303314209, + "num_tokens": 90475098.0, + "step": 3497 + }, + { + "epoch": 0.38414232374258733, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 3.0653152465820312, + "learning_rate": 1e-06, + "loss": 0.8983, + "mean_token_accuracy": 0.733822762966156, + "num_tokens": 90493385.0, + "step": 3498 + }, + { + "epoch": 0.384252141445201, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.534874439239502, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7017971873283386, + "num_tokens": 90524253.0, + "step": 3499 + }, + { + "epoch": 0.3843619591478146, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.327497959136963, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6899173259735107, + "num_tokens": 90554973.0, + "step": 3500 + }, + { + "epoch": 0.38447177685042827, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.3802478313446045, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.7003397941589355, + "num_tokens": 90584515.0, + "step": 3501 + }, + { + "epoch": 0.38458159455304197, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.7570998668670654, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6969490051269531, + "num_tokens": 90608481.0, + "step": 3502 + }, + { + "epoch": 0.3846914122556556, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.6256446838378906, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6925101280212402, + "num_tokens": 90633367.0, + "step": 3503 + }, + { + "epoch": 0.38480122995826926, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 16.839805603027344, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.6984160542488098, + "num_tokens": 90657309.0, + "step": 3504 + }, + { + "epoch": 0.3849110476608829, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.8523881435394287, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6978512406349182, + "num_tokens": 90681447.0, + "step": 3505 + }, + { + "epoch": 0.3850208653634966, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.781587839126587, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.6855370998382568, + "num_tokens": 90705941.0, + "step": 3506 + }, + { + "epoch": 0.38513068306611026, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.3899319171905518, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7107184529304504, + "num_tokens": 90734216.0, + "step": 3507 + }, + { + "epoch": 0.3852405007687239, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.406719207763672, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7010307312011719, + "num_tokens": 90761487.0, + "step": 3508 + }, + { + "epoch": 0.3853503184713376, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.639899969100952, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6872802972793579, + "num_tokens": 90789622.0, + "step": 3509 + }, + { + "epoch": 0.38546013617395125, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.374972343444824, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.690335750579834, + "num_tokens": 90819427.0, + "step": 3510 + }, + { + "epoch": 0.3855699538765649, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.507310628890991, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6840455532073975, + "num_tokens": 90846329.0, + "step": 3511 + }, + { + "epoch": 0.38567977157917854, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.4295244216918945, + "learning_rate": 1e-06, + "loss": 1.1399, + "mean_token_accuracy": 0.6684302091598511, + "num_tokens": 90875571.0, + "step": 3512 + }, + { + "epoch": 0.38578958928179224, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.8806068897247314, + "learning_rate": 1e-06, + "loss": 0.9072, + "mean_token_accuracy": 0.7293354868888855, + "num_tokens": 90895193.0, + "step": 3513 + }, + { + "epoch": 0.3858994069844059, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.3968112468719482, + "learning_rate": 1e-06, + "loss": 1.1367, + "mean_token_accuracy": 0.67540043592453, + "num_tokens": 90926516.0, + "step": 3514 + }, + { + "epoch": 0.38600922468701954, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.6197314262390137, + "learning_rate": 1e-06, + "loss": 1.1054, + "mean_token_accuracy": 0.6742212176322937, + "num_tokens": 90955106.0, + "step": 3515 + }, + { + "epoch": 0.3861190423896332, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.5227181911468506, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7002938389778137, + "num_tokens": 90980991.0, + "step": 3516 + }, + { + "epoch": 0.3862288600922469, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.7272090911865234, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6953179240226746, + "num_tokens": 91004316.0, + "step": 3517 + }, + { + "epoch": 0.38633867779486053, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.495058059692383, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7223361730575562, + "num_tokens": 91027761.0, + "step": 3518 + }, + { + "epoch": 0.3864484954974742, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.7374789714813232, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.6835724711418152, + "num_tokens": 91050506.0, + "step": 3519 + }, + { + "epoch": 0.3865583132000879, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.3407042026519775, + "learning_rate": 1e-06, + "loss": 1.101, + "mean_token_accuracy": 0.675617516040802, + "num_tokens": 91082101.0, + "step": 3520 + }, + { + "epoch": 0.3866681309027015, + "ewc_loss": 6.020069122314453e-06, + "grad_norm": 2.9072206020355225, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7164764404296875, + "num_tokens": 91104010.0, + "step": 3521 + }, + { + "epoch": 0.38677794860531517, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.5841338634490967, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.7038834095001221, + "num_tokens": 91129990.0, + "step": 3522 + }, + { + "epoch": 0.3868877663079288, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.614794969558716, + "learning_rate": 1e-06, + "loss": 1.1382, + "mean_token_accuracy": 0.6717204451560974, + "num_tokens": 91158919.0, + "step": 3523 + }, + { + "epoch": 0.3869975840105425, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.6662721633911133, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7036525011062622, + "num_tokens": 91180887.0, + "step": 3524 + }, + { + "epoch": 0.38710740171315616, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.3559954166412354, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.68837970495224, + "num_tokens": 91211256.0, + "step": 3525 + }, + { + "epoch": 0.3872172194157698, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.582122564315796, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6971975564956665, + "num_tokens": 91235786.0, + "step": 3526 + }, + { + "epoch": 0.3873270371183835, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.7054085731506348, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.6949029564857483, + "num_tokens": 91259530.0, + "step": 3527 + }, + { + "epoch": 0.38743685482099716, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.734018564224243, + "learning_rate": 1e-06, + "loss": 1.0953, + "mean_token_accuracy": 0.6801902055740356, + "num_tokens": 91284855.0, + "step": 3528 + }, + { + "epoch": 0.3875466725236108, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.5805649757385254, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7207961082458496, + "num_tokens": 91311673.0, + "step": 3529 + }, + { + "epoch": 0.38765649022622445, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.636470079421997, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.6902590990066528, + "num_tokens": 91337540.0, + "step": 3530 + }, + { + "epoch": 0.38776630792883815, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.5396320819854736, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6937183141708374, + "num_tokens": 91363407.0, + "step": 3531 + }, + { + "epoch": 0.3878761256314518, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 3.0090556144714355, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.6971666812896729, + "num_tokens": 91382935.0, + "step": 3532 + }, + { + "epoch": 0.38798594333406544, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.8967483043670654, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.6956124305725098, + "num_tokens": 91407494.0, + "step": 3533 + }, + { + "epoch": 0.3880957610366791, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.5268895626068115, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.716904878616333, + "num_tokens": 91434532.0, + "step": 3534 + }, + { + "epoch": 0.3882055787392928, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 3.0372838973999023, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7273440361022949, + "num_tokens": 91453366.0, + "step": 3535 + }, + { + "epoch": 0.38831539644190644, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.538698673248291, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.6944554448127747, + "num_tokens": 91479978.0, + "step": 3536 + }, + { + "epoch": 0.3884252141445201, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.4157681465148926, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6891448497772217, + "num_tokens": 91512790.0, + "step": 3537 + }, + { + "epoch": 0.3885350318471338, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.4622879028320312, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.697699785232544, + "num_tokens": 91539101.0, + "step": 3538 + }, + { + "epoch": 0.38864484954974743, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.809969663619995, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6922168731689453, + "num_tokens": 91561374.0, + "step": 3539 + }, + { + "epoch": 0.3887546672523611, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.35274600982666, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7036185264587402, + "num_tokens": 91589580.0, + "step": 3540 + }, + { + "epoch": 0.3888644849549747, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.665019989013672, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6905245780944824, + "num_tokens": 91612399.0, + "step": 3541 + }, + { + "epoch": 0.3889743026575884, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.3579530715942383, + "learning_rate": 1e-06, + "loss": 1.1201, + "mean_token_accuracy": 0.6769756078720093, + "num_tokens": 91644091.0, + "step": 3542 + }, + { + "epoch": 0.38908412036020207, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.4208803176879883, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.695767879486084, + "num_tokens": 91670530.0, + "step": 3543 + }, + { + "epoch": 0.3891939380628157, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.7925007343292236, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.7034388184547424, + "num_tokens": 91694669.0, + "step": 3544 + }, + { + "epoch": 0.38930375576542936, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.3273823261260986, + "learning_rate": 1e-06, + "loss": 1.0831, + "mean_token_accuracy": 0.6836446523666382, + "num_tokens": 91728597.0, + "step": 3545 + }, + { + "epoch": 0.38941357346804306, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.5309877395629883, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6949296593666077, + "num_tokens": 91756801.0, + "step": 3546 + }, + { + "epoch": 0.3895233911706567, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.904376983642578, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6827719807624817, + "num_tokens": 91778500.0, + "step": 3547 + }, + { + "epoch": 0.38963320887327035, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.688845634460449, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.694786548614502, + "num_tokens": 91804944.0, + "step": 3548 + }, + { + "epoch": 0.38974302657588406, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.4205424785614014, + "learning_rate": 1e-06, + "loss": 0.9431, + "mean_token_accuracy": 0.724564254283905, + "num_tokens": 91833600.0, + "step": 3549 + }, + { + "epoch": 0.3898528442784977, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.459934711456299, + "learning_rate": 1e-06, + "loss": 0.9154, + "mean_token_accuracy": 0.7250970602035522, + "num_tokens": 91856897.0, + "step": 3550 + }, + { + "epoch": 0.38996266198111135, + "ewc_loss": 6.0498714447021484e-06, + "grad_norm": 2.6534674167633057, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6930578947067261, + "num_tokens": 91878769.0, + "step": 3551 + }, + { + "epoch": 0.390072479683725, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.6233201026916504, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7080588340759277, + "num_tokens": 91901129.0, + "step": 3552 + }, + { + "epoch": 0.3901822973863387, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.73706316947937, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7260928750038147, + "num_tokens": 91921789.0, + "step": 3553 + }, + { + "epoch": 0.39029211508895234, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.5075979232788086, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.6892490386962891, + "num_tokens": 91949252.0, + "step": 3554 + }, + { + "epoch": 0.390401932791566, + "ewc_loss": 6.079673767089844e-06, + "grad_norm": 2.832798480987549, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7136285901069641, + "num_tokens": 91970893.0, + "step": 3555 + }, + { + "epoch": 0.3905117504941797, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.3799450397491455, + "learning_rate": 1e-06, + "loss": 1.1308, + "mean_token_accuracy": 0.6803985834121704, + "num_tokens": 92003459.0, + "step": 3556 + }, + { + "epoch": 0.39062156819679333, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.613823890686035, + "learning_rate": 1e-06, + "loss": 1.1167, + "mean_token_accuracy": 0.6813399791717529, + "num_tokens": 92032949.0, + "step": 3557 + }, + { + "epoch": 0.390731385899407, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.5383596420288086, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.6959161758422852, + "num_tokens": 92058753.0, + "step": 3558 + }, + { + "epoch": 0.3908412036020206, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.6130526065826416, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6707499623298645, + "num_tokens": 92086817.0, + "step": 3559 + }, + { + "epoch": 0.3909510213046343, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.4971442222595215, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6875098347663879, + "num_tokens": 92114863.0, + "step": 3560 + }, + { + "epoch": 0.391060839007248, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.520451545715332, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6896615624427795, + "num_tokens": 92143174.0, + "step": 3561 + }, + { + "epoch": 0.3911706567098616, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.481832981109619, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7040041089057922, + "num_tokens": 92169271.0, + "step": 3562 + }, + { + "epoch": 0.39128047441247527, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.3407881259918213, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6984468698501587, + "num_tokens": 92197530.0, + "step": 3563 + }, + { + "epoch": 0.39139029211508897, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.414149284362793, + "learning_rate": 1e-06, + "loss": 1.1314, + "mean_token_accuracy": 0.666530430316925, + "num_tokens": 92226522.0, + "step": 3564 + }, + { + "epoch": 0.3915001098177026, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.4676108360290527, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6863359212875366, + "num_tokens": 92255728.0, + "step": 3565 + }, + { + "epoch": 0.39160992752031626, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.2856974601745605, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7051973938941956, + "num_tokens": 92288415.0, + "step": 3566 + }, + { + "epoch": 0.39171974522292996, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.651789903640747, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7051274180412292, + "num_tokens": 92310582.0, + "step": 3567 + }, + { + "epoch": 0.3918295629255436, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.5755412578582764, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6882612705230713, + "num_tokens": 92337192.0, + "step": 3568 + }, + { + "epoch": 0.39193938062815725, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.503605842590332, + "learning_rate": 1e-06, + "loss": 1.1233, + "mean_token_accuracy": 0.6743719577789307, + "num_tokens": 92366075.0, + "step": 3569 + }, + { + "epoch": 0.3920491983307709, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 3.0742545127868652, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6938776969909668, + "num_tokens": 92391433.0, + "step": 3570 + }, + { + "epoch": 0.3921590160333846, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.583159923553467, + "learning_rate": 1e-06, + "loss": 1.1441, + "mean_token_accuracy": 0.6702705025672913, + "num_tokens": 92419463.0, + "step": 3571 + }, + { + "epoch": 0.39226883373599825, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.9285385608673096, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7005721926689148, + "num_tokens": 92441456.0, + "step": 3572 + }, + { + "epoch": 0.3923786514386119, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.487521171569824, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7185750603675842, + "num_tokens": 92466494.0, + "step": 3573 + }, + { + "epoch": 0.3924884691412256, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.5351152420043945, + "learning_rate": 1e-06, + "loss": 1.0929, + "mean_token_accuracy": 0.6832704544067383, + "num_tokens": 92491099.0, + "step": 3574 + }, + { + "epoch": 0.39259828684383924, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.718688726425171, + "learning_rate": 1e-06, + "loss": 0.939, + "mean_token_accuracy": 0.7196530103683472, + "num_tokens": 92518006.0, + "step": 3575 + }, + { + "epoch": 0.3927081045464529, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.495401382446289, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7058387398719788, + "num_tokens": 92542700.0, + "step": 3576 + }, + { + "epoch": 0.39281792224906653, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.7420949935913086, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6929008364677429, + "num_tokens": 92566279.0, + "step": 3577 + }, + { + "epoch": 0.39292773995168023, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.4020729064941406, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.6843070983886719, + "num_tokens": 92599121.0, + "step": 3578 + }, + { + "epoch": 0.3930375576542939, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.6072895526885986, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6983070969581604, + "num_tokens": 92624001.0, + "step": 3579 + }, + { + "epoch": 0.3931473753569075, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.463974952697754, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6982924938201904, + "num_tokens": 92652425.0, + "step": 3580 + }, + { + "epoch": 0.39325719305952117, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.648153066635132, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6952548027038574, + "num_tokens": 92674724.0, + "step": 3581 + }, + { + "epoch": 0.39336701076213487, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.5371434688568115, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6932784914970398, + "num_tokens": 92699667.0, + "step": 3582 + }, + { + "epoch": 0.3934768284647485, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.4654836654663086, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6871927380561829, + "num_tokens": 92727517.0, + "step": 3583 + }, + { + "epoch": 0.39358664616736216, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.9128129482269287, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7070097923278809, + "num_tokens": 92747373.0, + "step": 3584 + }, + { + "epoch": 0.39369646386997587, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.3646202087402344, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.7044489979743958, + "num_tokens": 92775832.0, + "step": 3585 + }, + { + "epoch": 0.3938062815725895, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.2999446392059326, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.694949746131897, + "num_tokens": 92807378.0, + "step": 3586 + }, + { + "epoch": 0.39391609927520316, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.5961642265319824, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7022231221199036, + "num_tokens": 92832155.0, + "step": 3587 + }, + { + "epoch": 0.3940259169778168, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.6477367877960205, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7020038366317749, + "num_tokens": 92856290.0, + "step": 3588 + }, + { + "epoch": 0.3941357346804305, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 2.552173376083374, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.6931025981903076, + "num_tokens": 92881250.0, + "step": 3589 + }, + { + "epoch": 0.39424555238304415, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.710070848464966, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6919963359832764, + "num_tokens": 92904024.0, + "step": 3590 + }, + { + "epoch": 0.3943553700856578, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.5329041481018066, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.6991453766822815, + "num_tokens": 92932387.0, + "step": 3591 + }, + { + "epoch": 0.39446518778827144, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.456613063812256, + "learning_rate": 1e-06, + "loss": 1.1011, + "mean_token_accuracy": 0.6798316240310669, + "num_tokens": 92959209.0, + "step": 3592 + }, + { + "epoch": 0.39457500549088514, + "ewc_loss": 6.109476089477539e-06, + "grad_norm": 3.503390312194824, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7157073020935059, + "num_tokens": 92977796.0, + "step": 3593 + }, + { + "epoch": 0.3946848231934988, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.592787027359009, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7134065628051758, + "num_tokens": 93002954.0, + "step": 3594 + }, + { + "epoch": 0.39479464089611244, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.4461724758148193, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.686542272567749, + "num_tokens": 93033062.0, + "step": 3595 + }, + { + "epoch": 0.39490445859872614, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.6796035766601562, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.689817488193512, + "num_tokens": 93059527.0, + "step": 3596 + }, + { + "epoch": 0.3950142763013398, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.638726234436035, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7140241861343384, + "num_tokens": 93085994.0, + "step": 3597 + }, + { + "epoch": 0.39512409400395343, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.6487205028533936, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6896345615386963, + "num_tokens": 93110793.0, + "step": 3598 + }, + { + "epoch": 0.3952339117065671, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.5449936389923096, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7108340263366699, + "num_tokens": 93137040.0, + "step": 3599 + }, + { + "epoch": 0.3953437294091808, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.604318141937256, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7000060081481934, + "num_tokens": 93160651.0, + "step": 3600 + }, + { + "epoch": 0.3954535471117944, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.6975419521331787, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7061958312988281, + "num_tokens": 93189203.0, + "step": 3601 + }, + { + "epoch": 0.39556336481440807, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.595681667327881, + "learning_rate": 1e-06, + "loss": 1.1131, + "mean_token_accuracy": 0.6805546283721924, + "num_tokens": 93215152.0, + "step": 3602 + }, + { + "epoch": 0.39567318251702177, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.53161883354187, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7073974609375, + "num_tokens": 93242481.0, + "step": 3603 + }, + { + "epoch": 0.3957830002196354, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.316448450088501, + "learning_rate": 1e-06, + "loss": 1.1025, + "mean_token_accuracy": 0.678152859210968, + "num_tokens": 93273852.0, + "step": 3604 + }, + { + "epoch": 0.39589281792224906, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.6768271923065186, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.698310375213623, + "num_tokens": 93297600.0, + "step": 3605 + }, + { + "epoch": 0.3960026356248627, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.55218505859375, + "learning_rate": 1e-06, + "loss": 1.1497, + "mean_token_accuracy": 0.6689123511314392, + "num_tokens": 93326205.0, + "step": 3606 + }, + { + "epoch": 0.3961124533274764, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.4006898403167725, + "learning_rate": 1e-06, + "loss": 1.134, + "mean_token_accuracy": 0.6728807687759399, + "num_tokens": 93356719.0, + "step": 3607 + }, + { + "epoch": 0.39622227103009006, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.7700092792510986, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7177213430404663, + "num_tokens": 93379681.0, + "step": 3608 + }, + { + "epoch": 0.3963320887327037, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.412174701690674, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6903274059295654, + "num_tokens": 93407884.0, + "step": 3609 + }, + { + "epoch": 0.39644190643531735, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.5309951305389404, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6955955028533936, + "num_tokens": 93434797.0, + "step": 3610 + }, + { + "epoch": 0.39655172413793105, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.415175676345825, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.698208212852478, + "num_tokens": 93461110.0, + "step": 3611 + }, + { + "epoch": 0.3966615418405447, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.401036262512207, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6912983059883118, + "num_tokens": 93493577.0, + "step": 3612 + }, + { + "epoch": 0.39677135954315834, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.744295358657837, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7065801620483398, + "num_tokens": 93516458.0, + "step": 3613 + }, + { + "epoch": 0.39688117724577204, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.5857760906219482, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7087966799736023, + "num_tokens": 93543958.0, + "step": 3614 + }, + { + "epoch": 0.3969909949483857, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 3.240496873855591, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.70586758852005, + "num_tokens": 93560899.0, + "step": 3615 + }, + { + "epoch": 0.39710081265099934, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.8931386470794678, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.712786853313446, + "num_tokens": 93579658.0, + "step": 3616 + }, + { + "epoch": 0.397210630353613, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.4619393348693848, + "learning_rate": 1e-06, + "loss": 0.8666, + "mean_token_accuracy": 0.7404886484146118, + "num_tokens": 93605222.0, + "step": 3617 + }, + { + "epoch": 0.3973204480562267, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.508112907409668, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7022976875305176, + "num_tokens": 93631550.0, + "step": 3618 + }, + { + "epoch": 0.39743026575884033, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.316103219985962, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.689361572265625, + "num_tokens": 93661110.0, + "step": 3619 + }, + { + "epoch": 0.397540083461454, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.882127046585083, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7005050182342529, + "num_tokens": 93680548.0, + "step": 3620 + }, + { + "epoch": 0.3976499011640676, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.3950819969177246, + "learning_rate": 1e-06, + "loss": 1.1987, + "mean_token_accuracy": 0.6556878685951233, + "num_tokens": 93710057.0, + "step": 3621 + }, + { + "epoch": 0.3977597188666813, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.400852680206299, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.6897757053375244, + "num_tokens": 93742099.0, + "step": 3622 + }, + { + "epoch": 0.39786953656929497, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.3951985836029053, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7202826738357544, + "num_tokens": 93767255.0, + "step": 3623 + }, + { + "epoch": 0.3979793542719086, + "ewc_loss": 6.139278411865234e-06, + "grad_norm": 2.846817970275879, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.703718900680542, + "num_tokens": 93787179.0, + "step": 3624 + }, + { + "epoch": 0.3980891719745223, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.321993827819824, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7160776853561401, + "num_tokens": 93818656.0, + "step": 3625 + }, + { + "epoch": 0.39819898967713596, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.57086443901062, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6957276463508606, + "num_tokens": 93842689.0, + "step": 3626 + }, + { + "epoch": 0.3983088073797496, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6375954151153564, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7007060647010803, + "num_tokens": 93866051.0, + "step": 3627 + }, + { + "epoch": 0.39841862508236325, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.4531993865966797, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7087507247924805, + "num_tokens": 93894339.0, + "step": 3628 + }, + { + "epoch": 0.39852844278497696, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.9444046020507812, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6856156587600708, + "num_tokens": 93917128.0, + "step": 3629 + }, + { + "epoch": 0.3986382604875906, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.7784595489501953, + "learning_rate": 1e-06, + "loss": 0.9536, + "mean_token_accuracy": 0.7161898612976074, + "num_tokens": 93937068.0, + "step": 3630 + }, + { + "epoch": 0.39874807819020425, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6238369941711426, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7050113677978516, + "num_tokens": 93961825.0, + "step": 3631 + }, + { + "epoch": 0.39885789589281795, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.7491345405578613, + "learning_rate": 1e-06, + "loss": 1.1247, + "mean_token_accuracy": 0.6713199615478516, + "num_tokens": 93986820.0, + "step": 3632 + }, + { + "epoch": 0.3989677135954316, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.803274631500244, + "learning_rate": 1e-06, + "loss": 1.1082, + "mean_token_accuracy": 0.6800397038459778, + "num_tokens": 94009183.0, + "step": 3633 + }, + { + "epoch": 0.39907753129804524, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5954067707061768, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.7045807838439941, + "num_tokens": 94035169.0, + "step": 3634 + }, + { + "epoch": 0.3991873490006589, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5092551708221436, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6964079141616821, + "num_tokens": 94058915.0, + "step": 3635 + }, + { + "epoch": 0.3992971667032726, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5543785095214844, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7019639611244202, + "num_tokens": 94087673.0, + "step": 3636 + }, + { + "epoch": 0.39940698440588623, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.364841938018799, + "learning_rate": 1e-06, + "loss": 1.1204, + "mean_token_accuracy": 0.6770498752593994, + "num_tokens": 94117111.0, + "step": 3637 + }, + { + "epoch": 0.3995168021084999, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.723457098007202, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.709955096244812, + "num_tokens": 94142061.0, + "step": 3638 + }, + { + "epoch": 0.3996266198111135, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5646474361419678, + "learning_rate": 1e-06, + "loss": 0.9359, + "mean_token_accuracy": 0.7226620316505432, + "num_tokens": 94166473.0, + "step": 3639 + }, + { + "epoch": 0.3997364375137272, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.32977294921875, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7190917730331421, + "num_tokens": 94195339.0, + "step": 3640 + }, + { + "epoch": 0.3998462552163409, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.470792531967163, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7021580934524536, + "num_tokens": 94220333.0, + "step": 3641 + }, + { + "epoch": 0.3999560729189545, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.4446561336517334, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7008194923400879, + "num_tokens": 94247191.0, + "step": 3642 + }, + { + "epoch": 0.4000658906215682, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5502235889434814, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6775559782981873, + "num_tokens": 94272074.0, + "step": 3643 + }, + { + "epoch": 0.40017570832418187, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.9564523696899414, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7208727598190308, + "num_tokens": 94292812.0, + "step": 3644 + }, + { + "epoch": 0.4002855260267955, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.953199863433838, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6915377378463745, + "num_tokens": 94317623.0, + "step": 3645 + }, + { + "epoch": 0.40039534372940916, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.2991487979888916, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6899703145027161, + "num_tokens": 94348024.0, + "step": 3646 + }, + { + "epoch": 0.40050516143202286, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.632188320159912, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6937342882156372, + "num_tokens": 94370827.0, + "step": 3647 + }, + { + "epoch": 0.4006149791346365, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.645382881164551, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6826669573783875, + "num_tokens": 94395141.0, + "step": 3648 + }, + { + "epoch": 0.40072479683725015, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.3852152824401855, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6852986216545105, + "num_tokens": 94427195.0, + "step": 3649 + }, + { + "epoch": 0.4008346145398638, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.415656089782715, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6811084747314453, + "num_tokens": 94456986.0, + "step": 3650 + }, + { + "epoch": 0.4009444322424775, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.7084221839904785, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.6899020671844482, + "num_tokens": 94486412.0, + "step": 3651 + }, + { + "epoch": 0.40105424994509115, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.5556960105895996, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.6972795128822327, + "num_tokens": 94513341.0, + "step": 3652 + }, + { + "epoch": 0.4011640676477048, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.5805439949035645, + "learning_rate": 1e-06, + "loss": 1.1176, + "mean_token_accuracy": 0.6884400248527527, + "num_tokens": 94539325.0, + "step": 3653 + }, + { + "epoch": 0.4012738853503185, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 3.3715081214904785, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6877404451370239, + "num_tokens": 94569341.0, + "step": 3654 + }, + { + "epoch": 0.40138370305293214, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.712573766708374, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7021421194076538, + "num_tokens": 94590125.0, + "step": 3655 + }, + { + "epoch": 0.4014935207555458, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.4687373638153076, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.684279203414917, + "num_tokens": 94618712.0, + "step": 3656 + }, + { + "epoch": 0.40160333845815943, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.665369987487793, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.718093991279602, + "num_tokens": 94639634.0, + "step": 3657 + }, + { + "epoch": 0.40171315616077313, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.730825185775757, + "learning_rate": 1e-06, + "loss": 1.0816, + "mean_token_accuracy": 0.6796334385871887, + "num_tokens": 94663874.0, + "step": 3658 + }, + { + "epoch": 0.4018229738633868, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6352367401123047, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6832247972488403, + "num_tokens": 94692854.0, + "step": 3659 + }, + { + "epoch": 0.4019327915660004, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.987473964691162, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7054246664047241, + "num_tokens": 94711951.0, + "step": 3660 + }, + { + "epoch": 0.4020426092686141, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5612988471984863, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6956009864807129, + "num_tokens": 94738139.0, + "step": 3661 + }, + { + "epoch": 0.40215242697122777, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 20.52474594116211, + "learning_rate": 1e-06, + "loss": 1.1596, + "mean_token_accuracy": 0.6654316186904907, + "num_tokens": 94769276.0, + "step": 3662 + }, + { + "epoch": 0.4022622446738414, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.802159547805786, + "learning_rate": 1e-06, + "loss": 1.154, + "mean_token_accuracy": 0.6690926551818848, + "num_tokens": 94793987.0, + "step": 3663 + }, + { + "epoch": 0.40237206237645506, + "ewc_loss": 6.258487701416016e-06, + "grad_norm": 2.925814151763916, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7016706466674805, + "num_tokens": 94815453.0, + "step": 3664 + }, + { + "epoch": 0.40248188007906877, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.4612338542938232, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7130304574966431, + "num_tokens": 94842720.0, + "step": 3665 + }, + { + "epoch": 0.4025916977816824, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.584127902984619, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.6855574250221252, + "num_tokens": 94869987.0, + "step": 3666 + }, + { + "epoch": 0.40270151548429606, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.7764549255371094, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.6969814300537109, + "num_tokens": 94891386.0, + "step": 3667 + }, + { + "epoch": 0.4028113331869097, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5727100372314453, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6937804222106934, + "num_tokens": 94915405.0, + "step": 3668 + }, + { + "epoch": 0.4029211508895234, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.705463171005249, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6943435072898865, + "num_tokens": 94940529.0, + "step": 3669 + }, + { + "epoch": 0.40303096859213705, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.7325873374938965, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7008989453315735, + "num_tokens": 94963891.0, + "step": 3670 + }, + { + "epoch": 0.4031407862947507, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.7321345806121826, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.6934640407562256, + "num_tokens": 94986480.0, + "step": 3671 + }, + { + "epoch": 0.4032506039973644, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.535767078399658, + "learning_rate": 1e-06, + "loss": 1.0923, + "mean_token_accuracy": 0.6802390217781067, + "num_tokens": 95012838.0, + "step": 3672 + }, + { + "epoch": 0.40336042169997804, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.6982929706573486, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.6937456727027893, + "num_tokens": 95034266.0, + "step": 3673 + }, + { + "epoch": 0.4034702394025917, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.750460386276245, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6872717142105103, + "num_tokens": 95055943.0, + "step": 3674 + }, + { + "epoch": 0.40358005710520534, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6109626293182373, + "learning_rate": 1e-06, + "loss": 1.138, + "mean_token_accuracy": 0.6784898042678833, + "num_tokens": 95083037.0, + "step": 3675 + }, + { + "epoch": 0.40368987480781904, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6983277797698975, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6870236992835999, + "num_tokens": 95109674.0, + "step": 3676 + }, + { + "epoch": 0.4037996925104327, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 3.057863235473633, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.6974707245826721, + "num_tokens": 95128170.0, + "step": 3677 + }, + { + "epoch": 0.40390951021304633, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.463066577911377, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6890591979026794, + "num_tokens": 95158536.0, + "step": 3678 + }, + { + "epoch": 0.40401932791566003, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.3467187881469727, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.7026056051254272, + "num_tokens": 95188713.0, + "step": 3679 + }, + { + "epoch": 0.4041291456182737, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6222872734069824, + "learning_rate": 1e-06, + "loss": 1.1271, + "mean_token_accuracy": 0.6700888276100159, + "num_tokens": 95214405.0, + "step": 3680 + }, + { + "epoch": 0.4042389633208873, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.4428374767303467, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.7179683446884155, + "num_tokens": 95244165.0, + "step": 3681 + }, + { + "epoch": 0.40434878102350097, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5969486236572266, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.7099428176879883, + "num_tokens": 95267477.0, + "step": 3682 + }, + { + "epoch": 0.40445859872611467, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.495156764984131, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6875487565994263, + "num_tokens": 95295044.0, + "step": 3683 + }, + { + "epoch": 0.4045684164287283, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.6914727687835693, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6956653594970703, + "num_tokens": 95319298.0, + "step": 3684 + }, + { + "epoch": 0.40467823413134196, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.5175864696502686, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7081124782562256, + "num_tokens": 95344447.0, + "step": 3685 + }, + { + "epoch": 0.4047880518339556, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.855774164199829, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.6991549730300903, + "num_tokens": 95364424.0, + "step": 3686 + }, + { + "epoch": 0.4048978695365693, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.7827508449554443, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6906647682189941, + "num_tokens": 95387166.0, + "step": 3687 + }, + { + "epoch": 0.40500768723918296, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5744335651397705, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6884403228759766, + "num_tokens": 95414783.0, + "step": 3688 + }, + { + "epoch": 0.4051175049417966, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.6130332946777344, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.696505606174469, + "num_tokens": 95440817.0, + "step": 3689 + }, + { + "epoch": 0.4052273226444103, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.7943408489227295, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7000906467437744, + "num_tokens": 95463800.0, + "step": 3690 + }, + { + "epoch": 0.40533714034702395, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.2969255447387695, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7063449025154114, + "num_tokens": 95493112.0, + "step": 3691 + }, + { + "epoch": 0.4054469580496376, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.48555850982666, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7044063806533813, + "num_tokens": 95519677.0, + "step": 3692 + }, + { + "epoch": 0.40555677575225124, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.3854660987854004, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6915618777275085, + "num_tokens": 95549291.0, + "step": 3693 + }, + { + "epoch": 0.40566659345486494, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.6418471336364746, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7148478031158447, + "num_tokens": 95575323.0, + "step": 3694 + }, + { + "epoch": 0.4057764111574786, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.6559555530548096, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7068679928779602, + "num_tokens": 95601929.0, + "step": 3695 + }, + { + "epoch": 0.40588622886009224, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.428189754486084, + "learning_rate": 1e-06, + "loss": 1.1182, + "mean_token_accuracy": 0.6818333864212036, + "num_tokens": 95632256.0, + "step": 3696 + }, + { + "epoch": 0.4059960465627059, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.4787402153015137, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7090241312980652, + "num_tokens": 95657228.0, + "step": 3697 + }, + { + "epoch": 0.4061058642653196, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.558269739151001, + "learning_rate": 1e-06, + "loss": 1.1154, + "mean_token_accuracy": 0.6776638031005859, + "num_tokens": 95683329.0, + "step": 3698 + }, + { + "epoch": 0.40621568196793323, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.4556005001068115, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6905841827392578, + "num_tokens": 95712366.0, + "step": 3699 + }, + { + "epoch": 0.4063254996705469, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.315873146057129, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6887468099594116, + "num_tokens": 95741651.0, + "step": 3700 + }, + { + "epoch": 0.4064353173731606, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.380241632461548, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6995366811752319, + "num_tokens": 95773832.0, + "step": 3701 + }, + { + "epoch": 0.4065451350757742, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.565208911895752, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6998920440673828, + "num_tokens": 95801416.0, + "step": 3702 + }, + { + "epoch": 0.40665495277838787, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.5872068405151367, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6983356475830078, + "num_tokens": 95826058.0, + "step": 3703 + }, + { + "epoch": 0.4067647704810015, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.208901882171631, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6882786750793457, + "num_tokens": 95859076.0, + "step": 3704 + }, + { + "epoch": 0.4068745881836152, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.404930830001831, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.7013463973999023, + "num_tokens": 95884058.0, + "step": 3705 + }, + { + "epoch": 0.40698440588622886, + "ewc_loss": 6.16908073425293e-06, + "grad_norm": 2.6226556301116943, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.709366500377655, + "num_tokens": 95906494.0, + "step": 3706 + }, + { + "epoch": 0.4070942235888425, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.476302146911621, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6873412132263184, + "num_tokens": 95933524.0, + "step": 3707 + }, + { + "epoch": 0.4072040412914562, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.6260578632354736, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7070881724357605, + "num_tokens": 95957723.0, + "step": 3708 + }, + { + "epoch": 0.40731385899406986, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.599574327468872, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.695662260055542, + "num_tokens": 95980974.0, + "step": 3709 + }, + { + "epoch": 0.4074236766966835, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.4235072135925293, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6964186429977417, + "num_tokens": 96010658.0, + "step": 3710 + }, + { + "epoch": 0.40753349439929715, + "ewc_loss": 6.198883056640625e-06, + "grad_norm": 2.8551244735717773, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6786290407180786, + "num_tokens": 96036358.0, + "step": 3711 + }, + { + "epoch": 0.40764331210191085, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.557687759399414, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.6928343176841736, + "num_tokens": 96062798.0, + "step": 3712 + }, + { + "epoch": 0.4077531298045245, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5375938415527344, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6877626180648804, + "num_tokens": 96089971.0, + "step": 3713 + }, + { + "epoch": 0.40786294750713814, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5101752281188965, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6862174272537231, + "num_tokens": 96116747.0, + "step": 3714 + }, + { + "epoch": 0.4079727652097518, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.6598331928253174, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6844540238380432, + "num_tokens": 96142612.0, + "step": 3715 + }, + { + "epoch": 0.4080825829123655, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.634232759475708, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.690934419631958, + "num_tokens": 96167683.0, + "step": 3716 + }, + { + "epoch": 0.40819240061497913, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.4706759452819824, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7095454931259155, + "num_tokens": 96195221.0, + "step": 3717 + }, + { + "epoch": 0.4083022183175928, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5406723022460938, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7108705639839172, + "num_tokens": 96221480.0, + "step": 3718 + }, + { + "epoch": 0.4084120360202065, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.8582069873809814, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6961069107055664, + "num_tokens": 96242839.0, + "step": 3719 + }, + { + "epoch": 0.4085218537228201, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.981107234954834, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6883018612861633, + "num_tokens": 96264354.0, + "step": 3720 + }, + { + "epoch": 0.4086316714254338, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.4031429290771484, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7134416699409485, + "num_tokens": 96292974.0, + "step": 3721 + }, + { + "epoch": 0.4087414891280474, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.8733763694763184, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7020399570465088, + "num_tokens": 96317311.0, + "step": 3722 + }, + { + "epoch": 0.4088513068306611, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5756680965423584, + "learning_rate": 1e-06, + "loss": 1.0989, + "mean_token_accuracy": 0.6752489805221558, + "num_tokens": 96344098.0, + "step": 3723 + }, + { + "epoch": 0.40896112453327477, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.616513252258301, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6972033977508545, + "num_tokens": 96368361.0, + "step": 3724 + }, + { + "epoch": 0.4090709422358884, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5100467205047607, + "learning_rate": 1e-06, + "loss": 0.9995, + "mean_token_accuracy": 0.706436812877655, + "num_tokens": 96394469.0, + "step": 3725 + }, + { + "epoch": 0.40918075993850206, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.559513807296753, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6914798617362976, + "num_tokens": 96419691.0, + "step": 3726 + }, + { + "epoch": 0.40929057764111576, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.3138794898986816, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6974263787269592, + "num_tokens": 96453773.0, + "step": 3727 + }, + { + "epoch": 0.4094003953437294, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.6566245555877686, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7164067625999451, + "num_tokens": 96478387.0, + "step": 3728 + }, + { + "epoch": 0.40951021304634305, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.6228623390197754, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6937648057937622, + "num_tokens": 96504213.0, + "step": 3729 + }, + { + "epoch": 0.40962003074895675, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.2551262378692627, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6839309334754944, + "num_tokens": 96536107.0, + "step": 3730 + }, + { + "epoch": 0.4097298484515704, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.761390209197998, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.7071636915206909, + "num_tokens": 96561775.0, + "step": 3731 + }, + { + "epoch": 0.40983966615418405, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.6239173412323, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7148160934448242, + "num_tokens": 96584036.0, + "step": 3732 + }, + { + "epoch": 0.4099494838567977, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.8977789878845215, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7129369974136353, + "num_tokens": 96605831.0, + "step": 3733 + }, + { + "epoch": 0.4100593015594114, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 3.0533483028411865, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7041140794754028, + "num_tokens": 96627335.0, + "step": 3734 + }, + { + "epoch": 0.41016911926202504, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.634887933731079, + "learning_rate": 1e-06, + "loss": 1.1147, + "mean_token_accuracy": 0.6883327960968018, + "num_tokens": 96655776.0, + "step": 3735 + }, + { + "epoch": 0.4102789369646387, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.491126298904419, + "learning_rate": 1e-06, + "loss": 1.1362, + "mean_token_accuracy": 0.6809571981430054, + "num_tokens": 96686329.0, + "step": 3736 + }, + { + "epoch": 0.4103887546672524, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.2685546875, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.70904541015625, + "num_tokens": 96717402.0, + "step": 3737 + }, + { + "epoch": 0.41049857236986603, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5504097938537598, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7108326554298401, + "num_tokens": 96743834.0, + "step": 3738 + }, + { + "epoch": 0.4106083900724797, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5994679927825928, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.6940262317657471, + "num_tokens": 96766567.0, + "step": 3739 + }, + { + "epoch": 0.4107182077750933, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.5973191261291504, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6895634531974792, + "num_tokens": 96793264.0, + "step": 3740 + }, + { + "epoch": 0.410828025477707, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.472118854522705, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6885503530502319, + "num_tokens": 96825333.0, + "step": 3741 + }, + { + "epoch": 0.41093784318032067, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.333080530166626, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7175979614257812, + "num_tokens": 96853811.0, + "step": 3742 + }, + { + "epoch": 0.4110476608829343, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.6105659008026123, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.6978899836540222, + "num_tokens": 96878822.0, + "step": 3743 + }, + { + "epoch": 0.41115747858554796, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.7576987743377686, + "learning_rate": 1e-06, + "loss": 0.9211, + "mean_token_accuracy": 0.7266280651092529, + "num_tokens": 96901200.0, + "step": 3744 + }, + { + "epoch": 0.41126729628816167, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 3.0350358486175537, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6982245445251465, + "num_tokens": 96920484.0, + "step": 3745 + }, + { + "epoch": 0.4113771139907753, + "ewc_loss": 6.22868537902832e-06, + "grad_norm": 2.760338544845581, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6999529600143433, + "num_tokens": 96944095.0, + "step": 3746 + }, + { + "epoch": 0.41148693169338896, + "ewc_loss": 6.288290023803711e-06, + "grad_norm": 2.87420654296875, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7004808187484741, + "num_tokens": 96964442.0, + "step": 3747 + }, + { + "epoch": 0.41159674939600266, + "ewc_loss": 6.288290023803711e-06, + "grad_norm": 2.6252448558807373, + "learning_rate": 1e-06, + "loss": 1.0965, + "mean_token_accuracy": 0.684437096118927, + "num_tokens": 96994310.0, + "step": 3748 + }, + { + "epoch": 0.4117065670986163, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.882385730743408, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7128198742866516, + "num_tokens": 97015086.0, + "step": 3749 + }, + { + "epoch": 0.41181638480122995, + "ewc_loss": 6.288290023803711e-06, + "grad_norm": 2.732330799102783, + "learning_rate": 1e-06, + "loss": 1.1172, + "mean_token_accuracy": 0.6804968118667603, + "num_tokens": 97038406.0, + "step": 3750 + }, + { + "epoch": 0.4119262025038436, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.593235969543457, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6880962252616882, + "num_tokens": 97063504.0, + "step": 3751 + }, + { + "epoch": 0.4120360202064573, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.997941732406616, + "learning_rate": 1e-06, + "loss": 1.0998, + "mean_token_accuracy": 0.6755004525184631, + "num_tokens": 97087001.0, + "step": 3752 + }, + { + "epoch": 0.41214583790907094, + "ewc_loss": 6.3478946685791016e-06, + "grad_norm": 2.466094732284546, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.708284854888916, + "num_tokens": 97114967.0, + "step": 3753 + }, + { + "epoch": 0.4122556556116846, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.6411731243133545, + "learning_rate": 1e-06, + "loss": 0.9526, + "mean_token_accuracy": 0.7163732051849365, + "num_tokens": 97137484.0, + "step": 3754 + }, + { + "epoch": 0.4123654733142983, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.6560070514678955, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.6985814571380615, + "num_tokens": 97164287.0, + "step": 3755 + }, + { + "epoch": 0.41247529101691194, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.8497426509857178, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7193393707275391, + "num_tokens": 97185153.0, + "step": 3756 + }, + { + "epoch": 0.4125851087195256, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.4355998039245605, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.694330096244812, + "num_tokens": 97211905.0, + "step": 3757 + }, + { + "epoch": 0.41269492642213923, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.475653648376465, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6947503089904785, + "num_tokens": 97238603.0, + "step": 3758 + }, + { + "epoch": 0.41280474412475293, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.3039138317108154, + "learning_rate": 1e-06, + "loss": 1.1111, + "mean_token_accuracy": 0.6752815842628479, + "num_tokens": 97271073.0, + "step": 3759 + }, + { + "epoch": 0.4129145618273666, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.4603357315063477, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.6976481676101685, + "num_tokens": 97297350.0, + "step": 3760 + }, + { + "epoch": 0.4130243795299802, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.55469012260437, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7022420763969421, + "num_tokens": 97321488.0, + "step": 3761 + }, + { + "epoch": 0.41313419723259387, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.4597561359405518, + "learning_rate": 1e-06, + "loss": 1.1508, + "mean_token_accuracy": 0.6784597635269165, + "num_tokens": 97349133.0, + "step": 3762 + }, + { + "epoch": 0.41324401493520757, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.5577392578125, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6965765953063965, + "num_tokens": 97373337.0, + "step": 3763 + }, + { + "epoch": 0.4133538326378212, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.8719255924224854, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7282518148422241, + "num_tokens": 97393305.0, + "step": 3764 + }, + { + "epoch": 0.41346365034043486, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.438542127609253, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6966948509216309, + "num_tokens": 97420834.0, + "step": 3765 + }, + { + "epoch": 0.41357346804304856, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.606804847717285, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6925405263900757, + "num_tokens": 97447875.0, + "step": 3766 + }, + { + "epoch": 0.4136832857456622, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.724414587020874, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7147442102432251, + "num_tokens": 97468715.0, + "step": 3767 + }, + { + "epoch": 0.41379310344827586, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.704890727996826, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6960352659225464, + "num_tokens": 97491436.0, + "step": 3768 + }, + { + "epoch": 0.4139029211508895, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.363579034805298, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6898437142372131, + "num_tokens": 97521771.0, + "step": 3769 + }, + { + "epoch": 0.4140127388535032, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.6863210201263428, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6962265372276306, + "num_tokens": 97548123.0, + "step": 3770 + }, + { + "epoch": 0.41412255655611685, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.595433235168457, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7179752588272095, + "num_tokens": 97569764.0, + "step": 3771 + }, + { + "epoch": 0.4142323742587305, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.4310247898101807, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.6955817937850952, + "num_tokens": 97598522.0, + "step": 3772 + }, + { + "epoch": 0.41434219196134414, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.5296525955200195, + "learning_rate": 1e-06, + "loss": 0.9084, + "mean_token_accuracy": 0.7324916124343872, + "num_tokens": 97623658.0, + "step": 3773 + }, + { + "epoch": 0.41445200966395784, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.304225444793701, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.686954140663147, + "num_tokens": 97654071.0, + "step": 3774 + }, + { + "epoch": 0.4145618273665715, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.57875394821167, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6924890279769897, + "num_tokens": 97677831.0, + "step": 3775 + }, + { + "epoch": 0.41467164506918514, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.6601529121398926, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.6839991807937622, + "num_tokens": 97702503.0, + "step": 3776 + }, + { + "epoch": 0.41478146277179884, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.864341974258423, + "learning_rate": 1e-06, + "loss": 1.095, + "mean_token_accuracy": 0.685158908367157, + "num_tokens": 97724802.0, + "step": 3777 + }, + { + "epoch": 0.4148912804744125, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.7336254119873047, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7100058197975159, + "num_tokens": 97749234.0, + "step": 3778 + }, + { + "epoch": 0.41500109817702613, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.6460800170898438, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7080718278884888, + "num_tokens": 97770698.0, + "step": 3779 + }, + { + "epoch": 0.4151109158796398, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.216691493988037, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6971608996391296, + "num_tokens": 97794172.0, + "step": 3780 + }, + { + "epoch": 0.4152207335822535, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.8919477462768555, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.7003651857376099, + "num_tokens": 97815066.0, + "step": 3781 + }, + { + "epoch": 0.4153305512848671, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.5500242710113525, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7246605753898621, + "num_tokens": 97841199.0, + "step": 3782 + }, + { + "epoch": 0.41544036898748077, + "ewc_loss": 6.318092346191406e-06, + "grad_norm": 2.625279188156128, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6898074746131897, + "num_tokens": 97865472.0, + "step": 3783 + }, + { + "epoch": 0.41555018669009447, + "ewc_loss": 6.3478946685791016e-06, + "grad_norm": 2.821624517440796, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7096131443977356, + "num_tokens": 97886053.0, + "step": 3784 + }, + { + "epoch": 0.4156600043927081, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6997179985046387, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7024957537651062, + "num_tokens": 97908195.0, + "step": 3785 + }, + { + "epoch": 0.41576982209532176, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4619362354278564, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7050063610076904, + "num_tokens": 97935081.0, + "step": 3786 + }, + { + "epoch": 0.4158796397979354, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.290421485900879, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6765626072883606, + "num_tokens": 97966325.0, + "step": 3787 + }, + { + "epoch": 0.4159894575005491, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.589918613433838, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.696009635925293, + "num_tokens": 97991457.0, + "step": 3788 + }, + { + "epoch": 0.41609927520316276, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.5374555587768555, + "learning_rate": 1e-06, + "loss": 1.0914, + "mean_token_accuracy": 0.684409499168396, + "num_tokens": 98018341.0, + "step": 3789 + }, + { + "epoch": 0.4162090929057764, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.536405563354492, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7045835852622986, + "num_tokens": 98045098.0, + "step": 3790 + }, + { + "epoch": 0.41631891060839005, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.7092103958129883, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6834361553192139, + "num_tokens": 98069357.0, + "step": 3791 + }, + { + "epoch": 0.41642872831100375, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6434781551361084, + "learning_rate": 1e-06, + "loss": 1.1167, + "mean_token_accuracy": 0.6776641607284546, + "num_tokens": 98094848.0, + "step": 3792 + }, + { + "epoch": 0.4165385460136174, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.71097993850708, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6828001737594604, + "num_tokens": 98118179.0, + "step": 3793 + }, + { + "epoch": 0.41664836371623104, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.760702610015869, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6924536228179932, + "num_tokens": 98142211.0, + "step": 3794 + }, + { + "epoch": 0.41675818141884474, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6358559131622314, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6927413940429688, + "num_tokens": 98168425.0, + "step": 3795 + }, + { + "epoch": 0.4168679991214584, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.1331253051757812, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7086663842201233, + "num_tokens": 98191140.0, + "step": 3796 + }, + { + "epoch": 0.41697781682407203, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.398984909057617, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7085964679718018, + "num_tokens": 98221268.0, + "step": 3797 + }, + { + "epoch": 0.4170876345266857, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.359027624130249, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7108890414237976, + "num_tokens": 98243642.0, + "step": 3798 + }, + { + "epoch": 0.4171974522292994, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.7153842449188232, + "learning_rate": 1e-06, + "loss": 0.9192, + "mean_token_accuracy": 0.726402997970581, + "num_tokens": 98265808.0, + "step": 3799 + }, + { + "epoch": 0.417307269931913, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.354713201522827, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7075321078300476, + "num_tokens": 98293589.0, + "step": 3800 + }, + { + "epoch": 0.4174170876345267, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.398090362548828, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6904416084289551, + "num_tokens": 98321281.0, + "step": 3801 + }, + { + "epoch": 0.4175269053371403, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6010074615478516, + "learning_rate": 1e-06, + "loss": 1.1217, + "mean_token_accuracy": 0.6750731468200684, + "num_tokens": 98347302.0, + "step": 3802 + }, + { + "epoch": 0.417636723039754, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6579947471618652, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.6740204691886902, + "num_tokens": 98375310.0, + "step": 3803 + }, + { + "epoch": 0.41774654074236767, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.3347599506378174, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7127944827079773, + "num_tokens": 98405962.0, + "step": 3804 + }, + { + "epoch": 0.4178563584449813, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.689107894897461, + "learning_rate": 1e-06, + "loss": 0.9189, + "mean_token_accuracy": 0.7311879396438599, + "num_tokens": 98429344.0, + "step": 3805 + }, + { + "epoch": 0.417966176147595, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4884657859802246, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.6991593241691589, + "num_tokens": 98455018.0, + "step": 3806 + }, + { + "epoch": 0.41807599385020866, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.431579351425171, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.7061471939086914, + "num_tokens": 98484221.0, + "step": 3807 + }, + { + "epoch": 0.4181858115528223, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4254181385040283, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6925152540206909, + "num_tokens": 98515513.0, + "step": 3808 + }, + { + "epoch": 0.41829562925543595, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.823185443878174, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7113470435142517, + "num_tokens": 98537078.0, + "step": 3809 + }, + { + "epoch": 0.41840544695804965, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.091449499130249, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7008618712425232, + "num_tokens": 98556703.0, + "step": 3810 + }, + { + "epoch": 0.4185152646606633, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4629833698272705, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.686327338218689, + "num_tokens": 98586006.0, + "step": 3811 + }, + { + "epoch": 0.41862508236327695, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6598308086395264, + "learning_rate": 1e-06, + "loss": 1.1082, + "mean_token_accuracy": 0.6793650388717651, + "num_tokens": 98609590.0, + "step": 3812 + }, + { + "epoch": 0.41873490006589065, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.9609179496765137, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.702782928943634, + "num_tokens": 98632955.0, + "step": 3813 + }, + { + "epoch": 0.4188447177685043, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.76230525970459, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.6969112753868103, + "num_tokens": 98654276.0, + "step": 3814 + }, + { + "epoch": 0.41895453547111794, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.949908971786499, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6971846222877502, + "num_tokens": 98676588.0, + "step": 3815 + }, + { + "epoch": 0.4190643531737316, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.46414852142334, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6888484954833984, + "num_tokens": 98704836.0, + "step": 3816 + }, + { + "epoch": 0.4191741708763453, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4333484172821045, + "learning_rate": 1e-06, + "loss": 1.1485, + "mean_token_accuracy": 0.6676526069641113, + "num_tokens": 98736500.0, + "step": 3817 + }, + { + "epoch": 0.41928398857895893, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.8960559368133545, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6964935660362244, + "num_tokens": 98758217.0, + "step": 3818 + }, + { + "epoch": 0.4193938062815726, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.724586248397827, + "learning_rate": 1e-06, + "loss": 1.0921, + "mean_token_accuracy": 0.6771459579467773, + "num_tokens": 98782642.0, + "step": 3819 + }, + { + "epoch": 0.4195036239841862, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.614978313446045, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.6959428787231445, + "num_tokens": 98809715.0, + "step": 3820 + }, + { + "epoch": 0.4196134416867999, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.681581735610962, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7050029039382935, + "num_tokens": 98832396.0, + "step": 3821 + }, + { + "epoch": 0.41972325938941357, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.770413398742676, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.7040003538131714, + "num_tokens": 98855032.0, + "step": 3822 + }, + { + "epoch": 0.4198330770920272, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.426614761352539, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.678826093673706, + "num_tokens": 98885502.0, + "step": 3823 + }, + { + "epoch": 0.4199428947946409, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.591599941253662, + "learning_rate": 1e-06, + "loss": 1.1235, + "mean_token_accuracy": 0.6790338754653931, + "num_tokens": 98910833.0, + "step": 3824 + }, + { + "epoch": 0.42005271249725457, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.60549259185791, + "learning_rate": 1e-06, + "loss": 1.1228, + "mean_token_accuracy": 0.6745155453681946, + "num_tokens": 98939922.0, + "step": 3825 + }, + { + "epoch": 0.4201625301998682, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.553507089614868, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6945741176605225, + "num_tokens": 98966447.0, + "step": 3826 + }, + { + "epoch": 0.42027234790248186, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.0710244178771973, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.698512852191925, + "num_tokens": 98990069.0, + "step": 3827 + }, + { + "epoch": 0.42038216560509556, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.3786416053771973, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7075057625770569, + "num_tokens": 99018779.0, + "step": 3828 + }, + { + "epoch": 0.4204919833077092, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.489861011505127, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6795778274536133, + "num_tokens": 99046989.0, + "step": 3829 + }, + { + "epoch": 0.42060180101032285, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.502211809158325, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6879181861877441, + "num_tokens": 99075830.0, + "step": 3830 + }, + { + "epoch": 0.42071161871293655, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6837735176086426, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6900804042816162, + "num_tokens": 99104043.0, + "step": 3831 + }, + { + "epoch": 0.4208214364155502, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4063193798065186, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6871310472488403, + "num_tokens": 99134589.0, + "step": 3832 + }, + { + "epoch": 0.42093125411816384, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.590507984161377, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6845007538795471, + "num_tokens": 99159009.0, + "step": 3833 + }, + { + "epoch": 0.4210410718207775, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4526045322418213, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6836546063423157, + "num_tokens": 99188690.0, + "step": 3834 + }, + { + "epoch": 0.4211508895233912, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.7290217876434326, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7169750332832336, + "num_tokens": 99209626.0, + "step": 3835 + }, + { + "epoch": 0.42126070722600484, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.690190076828003, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6894748210906982, + "num_tokens": 99237195.0, + "step": 3836 + }, + { + "epoch": 0.4213705249286185, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6964194774627686, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7137682437896729, + "num_tokens": 99265505.0, + "step": 3837 + }, + { + "epoch": 0.42148034263123213, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.7177960872650146, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.71500164270401, + "num_tokens": 99290547.0, + "step": 3838 + }, + { + "epoch": 0.42159016033384583, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6911227703094482, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6991645097732544, + "num_tokens": 99313701.0, + "step": 3839 + }, + { + "epoch": 0.4216999780364595, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.0508921146392822, + "learning_rate": 1e-06, + "loss": 1.1256, + "mean_token_accuracy": 0.6746827363967896, + "num_tokens": 99344109.0, + "step": 3840 + }, + { + "epoch": 0.4218097957390731, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.509549617767334, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7058706879615784, + "num_tokens": 99370893.0, + "step": 3841 + }, + { + "epoch": 0.4219196134416868, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.806063652038574, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.6998449563980103, + "num_tokens": 99393832.0, + "step": 3842 + }, + { + "epoch": 0.42202943114430047, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.8069450855255127, + "learning_rate": 1e-06, + "loss": 1.1681, + "mean_token_accuracy": 0.6648629307746887, + "num_tokens": 99418060.0, + "step": 3843 + }, + { + "epoch": 0.4221392488469141, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.605942487716675, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7083433866500854, + "num_tokens": 99441684.0, + "step": 3844 + }, + { + "epoch": 0.42224906654952776, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.554788589477539, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7005733847618103, + "num_tokens": 99469380.0, + "step": 3845 + }, + { + "epoch": 0.42235888425214146, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.9821813106536865, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.7253337502479553, + "num_tokens": 99487015.0, + "step": 3846 + }, + { + "epoch": 0.4224687019547551, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.7345266342163086, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6854038238525391, + "num_tokens": 99513084.0, + "step": 3847 + }, + { + "epoch": 0.42257851965736876, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6137759685516357, + "learning_rate": 1e-06, + "loss": 1.0844, + "mean_token_accuracy": 0.6885929107666016, + "num_tokens": 99538790.0, + "step": 3848 + }, + { + "epoch": 0.4226883373599824, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6214077472686768, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7087097764015198, + "num_tokens": 99563439.0, + "step": 3849 + }, + { + "epoch": 0.4227981550625961, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6287715435028076, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7216962575912476, + "num_tokens": 99586655.0, + "step": 3850 + }, + { + "epoch": 0.42290797276520975, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.660102367401123, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6830168962478638, + "num_tokens": 99611237.0, + "step": 3851 + }, + { + "epoch": 0.4230177904678234, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.2230067253112793, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.691925048828125, + "num_tokens": 99645810.0, + "step": 3852 + }, + { + "epoch": 0.4231276081704371, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 3.003911256790161, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6960498094558716, + "num_tokens": 99664956.0, + "step": 3853 + }, + { + "epoch": 0.42323742587305074, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.387397527694702, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7132579684257507, + "num_tokens": 99693325.0, + "step": 3854 + }, + { + "epoch": 0.4233472435756644, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.9683873653411865, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.713998556137085, + "num_tokens": 99715867.0, + "step": 3855 + }, + { + "epoch": 0.42345706127827804, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.4778449535369873, + "learning_rate": 1e-06, + "loss": 1.1494, + "mean_token_accuracy": 0.6700150966644287, + "num_tokens": 99745782.0, + "step": 3856 + }, + { + "epoch": 0.42356687898089174, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.735725164413452, + "learning_rate": 1e-06, + "loss": 1.0991, + "mean_token_accuracy": 0.6806539297103882, + "num_tokens": 99768440.0, + "step": 3857 + }, + { + "epoch": 0.4236766966835054, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.8274338245391846, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6908764839172363, + "num_tokens": 99789971.0, + "step": 3858 + }, + { + "epoch": 0.42378651438611903, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.709984540939331, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.697382926940918, + "num_tokens": 99813570.0, + "step": 3859 + }, + { + "epoch": 0.42389633208873273, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.699333667755127, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6846805214881897, + "num_tokens": 99837657.0, + "step": 3860 + }, + { + "epoch": 0.4240061497913464, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.589203357696533, + "learning_rate": 1e-06, + "loss": 1.1619, + "mean_token_accuracy": 0.6625912189483643, + "num_tokens": 99868728.0, + "step": 3861 + }, + { + "epoch": 0.42411596749396, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.586731433868408, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6796150207519531, + "num_tokens": 99896969.0, + "step": 3862 + }, + { + "epoch": 0.42422578519657367, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.7336196899414062, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7075526714324951, + "num_tokens": 99919296.0, + "step": 3863 + }, + { + "epoch": 0.42433560289918737, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 16.612333297729492, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7070806622505188, + "num_tokens": 99941534.0, + "step": 3864 + }, + { + "epoch": 0.424445420601801, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.489332437515259, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7038276195526123, + "num_tokens": 99968505.0, + "step": 3865 + }, + { + "epoch": 0.42455523830441466, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.511770009994507, + "learning_rate": 1e-06, + "loss": 0.922, + "mean_token_accuracy": 0.7279453277587891, + "num_tokens": 99993455.0, + "step": 3866 + }, + { + "epoch": 0.4246650560070283, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 4.71940279006958, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6971105337142944, + "num_tokens": 100018287.0, + "step": 3867 + }, + { + "epoch": 0.424774873709642, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.803173780441284, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7080267667770386, + "num_tokens": 100040541.0, + "step": 3868 + }, + { + "epoch": 0.42488469141225566, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.4994280338287354, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7151982188224792, + "num_tokens": 100066442.0, + "step": 3869 + }, + { + "epoch": 0.4249945091148693, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.662198305130005, + "learning_rate": 1e-06, + "loss": 0.8338, + "mean_token_accuracy": 0.7504631280899048, + "num_tokens": 100086709.0, + "step": 3870 + }, + { + "epoch": 0.425104326817483, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.636258840560913, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7111960649490356, + "num_tokens": 100110542.0, + "step": 3871 + }, + { + "epoch": 0.42521414452009665, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.370692253112793, + "learning_rate": 1e-06, + "loss": 1.1537, + "mean_token_accuracy": 0.6699944734573364, + "num_tokens": 100141320.0, + "step": 3872 + }, + { + "epoch": 0.4253239622227103, + "ewc_loss": 6.377696990966797e-06, + "grad_norm": 2.6613264083862305, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.683151125907898, + "num_tokens": 100164223.0, + "step": 3873 + }, + { + "epoch": 0.42543377992532394, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6987099647521973, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7085787057876587, + "num_tokens": 100185941.0, + "step": 3874 + }, + { + "epoch": 0.42554359762793764, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.5509111881256104, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7069014310836792, + "num_tokens": 100210732.0, + "step": 3875 + }, + { + "epoch": 0.4256534153305513, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.503291130065918, + "learning_rate": 1e-06, + "loss": 1.1134, + "mean_token_accuracy": 0.6803455352783203, + "num_tokens": 100238137.0, + "step": 3876 + }, + { + "epoch": 0.42576323303316493, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.560911178588867, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6879695653915405, + "num_tokens": 100264186.0, + "step": 3877 + }, + { + "epoch": 0.4258730507357786, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.3319008350372314, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6959336996078491, + "num_tokens": 100294331.0, + "step": 3878 + }, + { + "epoch": 0.4259828684383923, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6035892963409424, + "learning_rate": 1e-06, + "loss": 0.8993, + "mean_token_accuracy": 0.7302674055099487, + "num_tokens": 100316130.0, + "step": 3879 + }, + { + "epoch": 0.4260926861410059, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.688103675842285, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6971518993377686, + "num_tokens": 100340052.0, + "step": 3880 + }, + { + "epoch": 0.4262025038436196, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6539225578308105, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6949852705001831, + "num_tokens": 100365026.0, + "step": 3881 + }, + { + "epoch": 0.4263123215462333, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.699413537979126, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6947908401489258, + "num_tokens": 100386885.0, + "step": 3882 + }, + { + "epoch": 0.4264221392488469, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.5383522510528564, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6799153089523315, + "num_tokens": 100413857.0, + "step": 3883 + }, + { + "epoch": 0.42653195695146057, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.1463778018951416, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.698089063167572, + "num_tokens": 100448443.0, + "step": 3884 + }, + { + "epoch": 0.4266417746540742, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6259994506835938, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.724984884262085, + "num_tokens": 100473036.0, + "step": 3885 + }, + { + "epoch": 0.4267515923566879, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.8764867782592773, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6939647197723389, + "num_tokens": 100492808.0, + "step": 3886 + }, + { + "epoch": 0.42686141005930156, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.7496883869171143, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6838581562042236, + "num_tokens": 100516357.0, + "step": 3887 + }, + { + "epoch": 0.4269712277619152, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6049211025238037, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7053998708724976, + "num_tokens": 100539907.0, + "step": 3888 + }, + { + "epoch": 0.4270810454645289, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.357560634613037, + "learning_rate": 1e-06, + "loss": 0.9775, + "mean_token_accuracy": 0.7122368812561035, + "num_tokens": 100570127.0, + "step": 3889 + }, + { + "epoch": 0.42719086316714255, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 3.1207048892974854, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7207462787628174, + "num_tokens": 100587307.0, + "step": 3890 + }, + { + "epoch": 0.4273006808697562, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.7422685623168945, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6924475431442261, + "num_tokens": 100610619.0, + "step": 3891 + }, + { + "epoch": 0.42741049857236985, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.4739575386047363, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.69169682264328, + "num_tokens": 100639567.0, + "step": 3892 + }, + { + "epoch": 0.42752031627498355, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.873749256134033, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.701867938041687, + "num_tokens": 100659524.0, + "step": 3893 + }, + { + "epoch": 0.4276301339775972, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.686293601989746, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.707596480846405, + "num_tokens": 100681819.0, + "step": 3894 + }, + { + "epoch": 0.42773995168021084, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.345717668533325, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6989980936050415, + "num_tokens": 100715905.0, + "step": 3895 + }, + { + "epoch": 0.4278497693828245, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.358665943145752, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6909571886062622, + "num_tokens": 100751228.0, + "step": 3896 + }, + { + "epoch": 0.4279595870854382, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.788731336593628, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7006592154502869, + "num_tokens": 100773363.0, + "step": 3897 + }, + { + "epoch": 0.42806940478805183, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.4102981090545654, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6862764358520508, + "num_tokens": 100801727.0, + "step": 3898 + }, + { + "epoch": 0.4281792224906655, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.2710330486297607, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7014743089675903, + "num_tokens": 100834173.0, + "step": 3899 + }, + { + "epoch": 0.4282890401932792, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.8136026859283447, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6974288821220398, + "num_tokens": 100855435.0, + "step": 3900 + }, + { + "epoch": 0.4283988578958928, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.5148751735687256, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6873672604560852, + "num_tokens": 100884327.0, + "step": 3901 + }, + { + "epoch": 0.42850867559850647, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6433629989624023, + "learning_rate": 1e-06, + "loss": 1.1076, + "mean_token_accuracy": 0.6748247742652893, + "num_tokens": 100910211.0, + "step": 3902 + }, + { + "epoch": 0.4286184933011201, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.6403043270111084, + "learning_rate": 1e-06, + "loss": 1.1098, + "mean_token_accuracy": 0.6818851232528687, + "num_tokens": 100935401.0, + "step": 3903 + }, + { + "epoch": 0.4287283110037338, + "ewc_loss": 6.407499313354492e-06, + "grad_norm": 2.7184858322143555, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.7097669839859009, + "num_tokens": 100958877.0, + "step": 3904 + }, + { + "epoch": 0.42883812870634747, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.6195266246795654, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7102104425430298, + "num_tokens": 100980312.0, + "step": 3905 + }, + { + "epoch": 0.4289479464089611, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.3266847133636475, + "learning_rate": 1e-06, + "loss": 0.9785, + "mean_token_accuracy": 0.714004397392273, + "num_tokens": 101008596.0, + "step": 3906 + }, + { + "epoch": 0.4290577641115748, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.475642681121826, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7152805328369141, + "num_tokens": 101033996.0, + "step": 3907 + }, + { + "epoch": 0.42916758181418846, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.635448455810547, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7024767398834229, + "num_tokens": 101057963.0, + "step": 3908 + }, + { + "epoch": 0.4292773995168021, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 3.0046133995056152, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6929329633712769, + "num_tokens": 101077804.0, + "step": 3909 + }, + { + "epoch": 0.42938721721941575, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.9135451316833496, + "learning_rate": 1e-06, + "loss": 1.124, + "mean_token_accuracy": 0.6825132966041565, + "num_tokens": 101100214.0, + "step": 3910 + }, + { + "epoch": 0.42949703492202945, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.8504045009613037, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7213733196258545, + "num_tokens": 101124305.0, + "step": 3911 + }, + { + "epoch": 0.4296068526246431, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.5322041511535645, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7182726860046387, + "num_tokens": 101150289.0, + "step": 3912 + }, + { + "epoch": 0.42971667032725674, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.5621583461761475, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6873524188995361, + "num_tokens": 101178339.0, + "step": 3913 + }, + { + "epoch": 0.4298264880298704, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.6510610580444336, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.711434006690979, + "num_tokens": 101202614.0, + "step": 3914 + }, + { + "epoch": 0.4299363057324841, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.618427038192749, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7183324694633484, + "num_tokens": 101224528.0, + "step": 3915 + }, + { + "epoch": 0.43004612343509774, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.391887903213501, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7304983139038086, + "num_tokens": 101249037.0, + "step": 3916 + }, + { + "epoch": 0.4301559411377114, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.282829523086548, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6919044256210327, + "num_tokens": 101282444.0, + "step": 3917 + }, + { + "epoch": 0.4302657588403251, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.5279934406280518, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7053841948509216, + "num_tokens": 101313358.0, + "step": 3918 + }, + { + "epoch": 0.43037557654293873, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.4123289585113525, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6954265236854553, + "num_tokens": 101340581.0, + "step": 3919 + }, + { + "epoch": 0.4304853942455524, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.846118927001953, + "learning_rate": 1e-06, + "loss": 0.9194, + "mean_token_accuracy": 0.7127892971038818, + "num_tokens": 101361533.0, + "step": 3920 + }, + { + "epoch": 0.430595211948166, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.710690498352051, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7077524065971375, + "num_tokens": 101383326.0, + "step": 3921 + }, + { + "epoch": 0.4307050296507797, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.661881446838379, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.7028847336769104, + "num_tokens": 101409081.0, + "step": 3922 + }, + { + "epoch": 0.43081484735339337, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.513122081756592, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6942998170852661, + "num_tokens": 101438699.0, + "step": 3923 + }, + { + "epoch": 0.430924665056007, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.4503743648529053, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7088871598243713, + "num_tokens": 101464864.0, + "step": 3924 + }, + { + "epoch": 0.43103448275862066, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.703038454055786, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.6865729093551636, + "num_tokens": 101488078.0, + "step": 3925 + }, + { + "epoch": 0.43114430046123436, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.6193785667419434, + "learning_rate": 1e-06, + "loss": 1.1344, + "mean_token_accuracy": 0.6786962747573853, + "num_tokens": 101514050.0, + "step": 3926 + }, + { + "epoch": 0.431254118163848, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.7911088466644287, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7198168039321899, + "num_tokens": 101536723.0, + "step": 3927 + }, + { + "epoch": 0.43136393586646166, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.5170674324035645, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6856818795204163, + "num_tokens": 101561970.0, + "step": 3928 + }, + { + "epoch": 0.43147375356907536, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.4713215827941895, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.6999221444129944, + "num_tokens": 101589583.0, + "step": 3929 + }, + { + "epoch": 0.431583571271689, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.680755615234375, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6805227398872375, + "num_tokens": 101621673.0, + "step": 3930 + }, + { + "epoch": 0.43169338897430265, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.681867837905884, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6886958479881287, + "num_tokens": 101644065.0, + "step": 3931 + }, + { + "epoch": 0.4318032066769163, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 3.0513105392456055, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7265055179595947, + "num_tokens": 101662750.0, + "step": 3932 + }, + { + "epoch": 0.43191302437953, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.4051125049591064, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7071974277496338, + "num_tokens": 101688990.0, + "step": 3933 + }, + { + "epoch": 0.43202284208214364, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.693505048751831, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.691642165184021, + "num_tokens": 101710686.0, + "step": 3934 + }, + { + "epoch": 0.4321326597847573, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.568145990371704, + "learning_rate": 1e-06, + "loss": 1.1015, + "mean_token_accuracy": 0.6887067556381226, + "num_tokens": 101737899.0, + "step": 3935 + }, + { + "epoch": 0.432242477487371, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.4274425506591797, + "learning_rate": 1e-06, + "loss": 1.0959, + "mean_token_accuracy": 0.6827366352081299, + "num_tokens": 101766454.0, + "step": 3936 + }, + { + "epoch": 0.43235229518998464, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.5160655975341797, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7119872570037842, + "num_tokens": 101791654.0, + "step": 3937 + }, + { + "epoch": 0.4324621128925983, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.5114030838012695, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6977053880691528, + "num_tokens": 101817589.0, + "step": 3938 + }, + { + "epoch": 0.43257193059521193, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7507996559143066, + "learning_rate": 1e-06, + "loss": 1.0727, + "mean_token_accuracy": 0.6908038854598999, + "num_tokens": 101847859.0, + "step": 3939 + }, + { + "epoch": 0.43268174829782563, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.965963125228882, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.6960165500640869, + "num_tokens": 101871930.0, + "step": 3940 + }, + { + "epoch": 0.4327915660004393, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.57610821723938, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.701336145401001, + "num_tokens": 101898638.0, + "step": 3941 + }, + { + "epoch": 0.4329013837030529, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.547147274017334, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6889834403991699, + "num_tokens": 101926998.0, + "step": 3942 + }, + { + "epoch": 0.43301120140566657, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.877470016479492, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6877257823944092, + "num_tokens": 101949123.0, + "step": 3943 + }, + { + "epoch": 0.43312101910828027, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.660344362258911, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7112245559692383, + "num_tokens": 101971407.0, + "step": 3944 + }, + { + "epoch": 0.4332308368108939, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.497910499572754, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6988133788108826, + "num_tokens": 101997943.0, + "step": 3945 + }, + { + "epoch": 0.43334065451350756, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.468350887298584, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7031798362731934, + "num_tokens": 102025822.0, + "step": 3946 + }, + { + "epoch": 0.43345047221612126, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.4346768856048584, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.686798095703125, + "num_tokens": 102053305.0, + "step": 3947 + }, + { + "epoch": 0.4335602899187349, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.61535382270813, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.7004685401916504, + "num_tokens": 102078836.0, + "step": 3948 + }, + { + "epoch": 0.43367010762134856, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5117757320404053, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7214518189430237, + "num_tokens": 102103001.0, + "step": 3949 + }, + { + "epoch": 0.4337799253239622, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.405914068222046, + "learning_rate": 1e-06, + "loss": 1.0929, + "mean_token_accuracy": 0.6802759170532227, + "num_tokens": 102131895.0, + "step": 3950 + }, + { + "epoch": 0.4338897430265759, + "ewc_loss": 6.4373016357421875e-06, + "grad_norm": 2.7254478931427, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7179018259048462, + "num_tokens": 102156023.0, + "step": 3951 + }, + { + "epoch": 0.43399956072918955, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.8366920948028564, + "learning_rate": 1e-06, + "loss": 1.1116, + "mean_token_accuracy": 0.6836076974868774, + "num_tokens": 102176947.0, + "step": 3952 + }, + { + "epoch": 0.4341093784318032, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6462697982788086, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6945216655731201, + "num_tokens": 102203157.0, + "step": 3953 + }, + { + "epoch": 0.43421919613441684, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.8052756786346436, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.6993160247802734, + "num_tokens": 102226106.0, + "step": 3954 + }, + { + "epoch": 0.43432901383703054, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.7209370136260986, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6912175416946411, + "num_tokens": 102250220.0, + "step": 3955 + }, + { + "epoch": 0.4344388315396442, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.560786485671997, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.676137387752533, + "num_tokens": 102276373.0, + "step": 3956 + }, + { + "epoch": 0.43454864924225783, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6949329376220703, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7054264545440674, + "num_tokens": 102298062.0, + "step": 3957 + }, + { + "epoch": 0.43465846694487154, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.553838014602661, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.6965542435646057, + "num_tokens": 102324012.0, + "step": 3958 + }, + { + "epoch": 0.4347682846474852, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5433595180511475, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6834853887557983, + "num_tokens": 102352815.0, + "step": 3959 + }, + { + "epoch": 0.4348781023500988, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6462841033935547, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6959509253501892, + "num_tokens": 102378014.0, + "step": 3960 + }, + { + "epoch": 0.4349879200527125, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.75581693649292, + "learning_rate": 1e-06, + "loss": 0.9251, + "mean_token_accuracy": 0.7258912920951843, + "num_tokens": 102402223.0, + "step": 3961 + }, + { + "epoch": 0.4350977377553262, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.569338321685791, + "learning_rate": 1e-06, + "loss": 0.8857, + "mean_token_accuracy": 0.7340830564498901, + "num_tokens": 102429827.0, + "step": 3962 + }, + { + "epoch": 0.4352075554579398, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.4918527603149414, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6918796896934509, + "num_tokens": 102458769.0, + "step": 3963 + }, + { + "epoch": 0.43531737316055347, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.8684000968933105, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6922717690467834, + "num_tokens": 102487677.0, + "step": 3964 + }, + { + "epoch": 0.43542719086316717, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6562016010284424, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.7031322717666626, + "num_tokens": 102516194.0, + "step": 3965 + }, + { + "epoch": 0.4355370085657808, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 3.2854952812194824, + "learning_rate": 1e-06, + "loss": 0.9324, + "mean_token_accuracy": 0.7172302007675171, + "num_tokens": 102534115.0, + "step": 3966 + }, + { + "epoch": 0.43564682626839446, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6741678714752197, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7117592096328735, + "num_tokens": 102557009.0, + "step": 3967 + }, + { + "epoch": 0.4357566439710081, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.781740427017212, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6929622888565063, + "num_tokens": 102581438.0, + "step": 3968 + }, + { + "epoch": 0.4358664616736218, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.755052328109741, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7135676145553589, + "num_tokens": 102603822.0, + "step": 3969 + }, + { + "epoch": 0.43597627937623545, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6939830780029297, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7032054662704468, + "num_tokens": 102625582.0, + "step": 3970 + }, + { + "epoch": 0.4360860970788491, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 9.693072319030762, + "learning_rate": 1e-06, + "loss": 1.0977, + "mean_token_accuracy": 0.6851131916046143, + "num_tokens": 102655237.0, + "step": 3971 + }, + { + "epoch": 0.43619591478146275, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.7240021228790283, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7023701667785645, + "num_tokens": 102681610.0, + "step": 3972 + }, + { + "epoch": 0.43630573248407645, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5523838996887207, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6781324148178101, + "num_tokens": 102711636.0, + "step": 3973 + }, + { + "epoch": 0.4364155501866901, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.682638645172119, + "learning_rate": 1e-06, + "loss": 1.1069, + "mean_token_accuracy": 0.6759167313575745, + "num_tokens": 102740269.0, + "step": 3974 + }, + { + "epoch": 0.43652536788930374, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.536357879638672, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.7260103225708008, + "num_tokens": 102765140.0, + "step": 3975 + }, + { + "epoch": 0.43663518559191744, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6460747718811035, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6920382976531982, + "num_tokens": 102790550.0, + "step": 3976 + }, + { + "epoch": 0.4367450032945311, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.665468215942383, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6884564757347107, + "num_tokens": 102818283.0, + "step": 3977 + }, + { + "epoch": 0.43685482099714473, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.7050323486328125, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6934760808944702, + "num_tokens": 102839697.0, + "step": 3978 + }, + { + "epoch": 0.4369646386997584, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6376893520355225, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.6903954744338989, + "num_tokens": 102868038.0, + "step": 3979 + }, + { + "epoch": 0.4370744564023721, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.905806541442871, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6939982771873474, + "num_tokens": 102890261.0, + "step": 3980 + }, + { + "epoch": 0.4371842741049857, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.796468734741211, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6806182861328125, + "num_tokens": 102912602.0, + "step": 3981 + }, + { + "epoch": 0.43729409180759937, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.6866061687469482, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7164795398712158, + "num_tokens": 102936652.0, + "step": 3982 + }, + { + "epoch": 0.4374039095102131, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.484149694442749, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6782208681106567, + "num_tokens": 102964564.0, + "step": 3983 + }, + { + "epoch": 0.4375137272128267, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5589427947998047, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7201657891273499, + "num_tokens": 102991026.0, + "step": 3984 + }, + { + "epoch": 0.43762354491544037, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5912368297576904, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.6915297508239746, + "num_tokens": 103017666.0, + "step": 3985 + }, + { + "epoch": 0.437733362618054, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.864494800567627, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.7157453894615173, + "num_tokens": 103038445.0, + "step": 3986 + }, + { + "epoch": 0.4378431803206677, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5802032947540283, + "learning_rate": 1e-06, + "loss": 1.0779, + "mean_token_accuracy": 0.6841244101524353, + "num_tokens": 103069019.0, + "step": 3987 + }, + { + "epoch": 0.43795299802328136, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.8529176712036133, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7048043012619019, + "num_tokens": 103090281.0, + "step": 3988 + }, + { + "epoch": 0.438062815725895, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.82373309135437, + "learning_rate": 1e-06, + "loss": 0.9066, + "mean_token_accuracy": 0.7284799218177795, + "num_tokens": 103118962.0, + "step": 3989 + }, + { + "epoch": 0.43817263342850865, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6526358127593994, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7139102220535278, + "num_tokens": 103145401.0, + "step": 3990 + }, + { + "epoch": 0.43828245113112235, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7794711589813232, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.694988489151001, + "num_tokens": 103172448.0, + "step": 3991 + }, + { + "epoch": 0.438392268833736, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.762108564376831, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7036877274513245, + "num_tokens": 103198006.0, + "step": 3992 + }, + { + "epoch": 0.43850208653634964, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5888569355010986, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7038706541061401, + "num_tokens": 103222454.0, + "step": 3993 + }, + { + "epoch": 0.43861190423896335, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.53438138961792, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7053425908088684, + "num_tokens": 103249089.0, + "step": 3994 + }, + { + "epoch": 0.438721721941577, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.711594581604004, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7260081768035889, + "num_tokens": 103270197.0, + "step": 3995 + }, + { + "epoch": 0.43883153964419064, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.4195172786712646, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6861804723739624, + "num_tokens": 103297499.0, + "step": 3996 + }, + { + "epoch": 0.4389413573468043, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.581913709640503, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7124953269958496, + "num_tokens": 103324855.0, + "step": 3997 + }, + { + "epoch": 0.439051175049418, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.521824836730957, + "learning_rate": 1e-06, + "loss": 1.142, + "mean_token_accuracy": 0.6765843629837036, + "num_tokens": 103352855.0, + "step": 3998 + }, + { + "epoch": 0.43916099275203163, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7003705501556396, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.6949104070663452, + "num_tokens": 103375974.0, + "step": 3999 + }, + { + "epoch": 0.4392708104546453, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.9213173389434814, + "learning_rate": 1e-06, + "loss": 1.1536, + "mean_token_accuracy": 0.6629263162612915, + "num_tokens": 103407738.0, + "step": 4000 + }, + { + "epoch": 0.4393806281572589, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.8258376121520996, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6939184069633484, + "num_tokens": 103437027.0, + "step": 4001 + }, + { + "epoch": 0.4394904458598726, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7136342525482178, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7018224000930786, + "num_tokens": 103460046.0, + "step": 4002 + }, + { + "epoch": 0.43960026356248627, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.595492362976074, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7009422779083252, + "num_tokens": 103485621.0, + "step": 4003 + }, + { + "epoch": 0.4397100812650999, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.193676233291626, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6813182830810547, + "num_tokens": 103518984.0, + "step": 4004 + }, + { + "epoch": 0.4398198989677136, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.6851654052734375, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6901983022689819, + "num_tokens": 103541595.0, + "step": 4005 + }, + { + "epoch": 0.43992971667032726, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7221434116363525, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6851162910461426, + "num_tokens": 103565995.0, + "step": 4006 + }, + { + "epoch": 0.4400395343729409, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.626859188079834, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6867994070053101, + "num_tokens": 103589961.0, + "step": 4007 + }, + { + "epoch": 0.44014935207555456, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.332139253616333, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6832643747329712, + "num_tokens": 103624015.0, + "step": 4008 + }, + { + "epoch": 0.44025916977816826, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.578148603439331, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.712761402130127, + "num_tokens": 103648981.0, + "step": 4009 + }, + { + "epoch": 0.4403689874807819, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.6383373737335205, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7110698223114014, + "num_tokens": 103672992.0, + "step": 4010 + }, + { + "epoch": 0.44047880518339555, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7223949432373047, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7066219449043274, + "num_tokens": 103693983.0, + "step": 4011 + }, + { + "epoch": 0.44058862288600925, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.783123016357422, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.6849415302276611, + "num_tokens": 103717413.0, + "step": 4012 + }, + { + "epoch": 0.4406984405886229, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7677154541015625, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.70485520362854, + "num_tokens": 103739793.0, + "step": 4013 + }, + { + "epoch": 0.44080825829123654, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.610903024673462, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7095587849617004, + "num_tokens": 103763425.0, + "step": 4014 + }, + { + "epoch": 0.4409180759938502, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.6665966510772705, + "learning_rate": 1e-06, + "loss": 0.9056, + "mean_token_accuracy": 0.7244687080383301, + "num_tokens": 103784225.0, + "step": 4015 + }, + { + "epoch": 0.4410278936964639, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.345461845397949, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7022327184677124, + "num_tokens": 103811751.0, + "step": 4016 + }, + { + "epoch": 0.44113771139907754, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.321915864944458, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.6869493722915649, + "num_tokens": 103843323.0, + "step": 4017 + }, + { + "epoch": 0.4412475291016912, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.5493404865264893, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.7043435573577881, + "num_tokens": 103869122.0, + "step": 4018 + }, + { + "epoch": 0.44135734680430483, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.6595137119293213, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.687716007232666, + "num_tokens": 103892813.0, + "step": 4019 + }, + { + "epoch": 0.44146716450691853, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.259582996368408, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.687605082988739, + "num_tokens": 103925498.0, + "step": 4020 + }, + { + "epoch": 0.4415769822095322, + "ewc_loss": 6.467103958129883e-06, + "grad_norm": 2.7563507556915283, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7109861373901367, + "num_tokens": 103948200.0, + "step": 4021 + }, + { + "epoch": 0.4416867999121458, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.565964937210083, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.7265386581420898, + "num_tokens": 103969306.0, + "step": 4022 + }, + { + "epoch": 0.4417966176147595, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.7020516395568848, + "learning_rate": 1e-06, + "loss": 0.979, + "mean_token_accuracy": 0.7084013223648071, + "num_tokens": 103992620.0, + "step": 4023 + }, + { + "epoch": 0.44190643531737317, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.566821813583374, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.6934534311294556, + "num_tokens": 104022069.0, + "step": 4024 + }, + { + "epoch": 0.4420162530199868, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.5263783931732178, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6965579390525818, + "num_tokens": 104052384.0, + "step": 4025 + }, + { + "epoch": 0.44212607072260046, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.503084659576416, + "learning_rate": 1e-06, + "loss": 1.1051, + "mean_token_accuracy": 0.6820266842842102, + "num_tokens": 104080332.0, + "step": 4026 + }, + { + "epoch": 0.44223588842521416, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.6631126403808594, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6881542205810547, + "num_tokens": 104106260.0, + "step": 4027 + }, + { + "epoch": 0.4423457061278278, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.346087694168091, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6944728493690491, + "num_tokens": 104135474.0, + "step": 4028 + }, + { + "epoch": 0.44245552383044146, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.7328498363494873, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6984906196594238, + "num_tokens": 104157966.0, + "step": 4029 + }, + { + "epoch": 0.4425653415330551, + "ewc_loss": 6.496906280517578e-06, + "grad_norm": 2.6967296600341797, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6894548535346985, + "num_tokens": 104180901.0, + "step": 4030 + }, + { + "epoch": 0.4426751592356688, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.3756442070007324, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6995922923088074, + "num_tokens": 104210427.0, + "step": 4031 + }, + { + "epoch": 0.44278497693828245, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.3847062587738037, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7073599100112915, + "num_tokens": 104240781.0, + "step": 4032 + }, + { + "epoch": 0.4428947946408961, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5152721405029297, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.6956380009651184, + "num_tokens": 104265444.0, + "step": 4033 + }, + { + "epoch": 0.4430046123435098, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.528493642807007, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7049524784088135, + "num_tokens": 104289851.0, + "step": 4034 + }, + { + "epoch": 0.44311443004612344, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.4698128700256348, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7038069367408752, + "num_tokens": 104318477.0, + "step": 4035 + }, + { + "epoch": 0.4432242477487371, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5897274017333984, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6873593926429749, + "num_tokens": 104344467.0, + "step": 4036 + }, + { + "epoch": 0.44333406545135073, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.567941904067993, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6848294734954834, + "num_tokens": 104370130.0, + "step": 4037 + }, + { + "epoch": 0.44344388315396444, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5603981018066406, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.719490647315979, + "num_tokens": 104393844.0, + "step": 4038 + }, + { + "epoch": 0.4435537008565781, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.94466495513916, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.6939821839332581, + "num_tokens": 104413570.0, + "step": 4039 + }, + { + "epoch": 0.4436635185591917, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5174386501312256, + "learning_rate": 1e-06, + "loss": 1.155, + "mean_token_accuracy": 0.6645037531852722, + "num_tokens": 104443227.0, + "step": 4040 + }, + { + "epoch": 0.44377333626180543, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5974667072296143, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.6978981494903564, + "num_tokens": 104467348.0, + "step": 4041 + }, + { + "epoch": 0.4438831539644191, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.8347232341766357, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.713845431804657, + "num_tokens": 104487199.0, + "step": 4042 + }, + { + "epoch": 0.4439929716670327, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.752990245819092, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.699101984500885, + "num_tokens": 104508389.0, + "step": 4043 + }, + { + "epoch": 0.44410278936964637, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.2932095527648926, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7056058645248413, + "num_tokens": 104541411.0, + "step": 4044 + }, + { + "epoch": 0.44421260707226007, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4325969219207764, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.7078818082809448, + "num_tokens": 104568153.0, + "step": 4045 + }, + { + "epoch": 0.4443224247748737, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6249442100524902, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6894898414611816, + "num_tokens": 104591936.0, + "step": 4046 + }, + { + "epoch": 0.44443224247748736, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.485694408416748, + "learning_rate": 1e-06, + "loss": 1.1007, + "mean_token_accuracy": 0.6829406023025513, + "num_tokens": 104621841.0, + "step": 4047 + }, + { + "epoch": 0.444542060180101, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4524543285369873, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6974061727523804, + "num_tokens": 104651647.0, + "step": 4048 + }, + { + "epoch": 0.4446518778827147, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.740168571472168, + "learning_rate": 1e-06, + "loss": 1.1045, + "mean_token_accuracy": 0.6846850514411926, + "num_tokens": 104677572.0, + "step": 4049 + }, + { + "epoch": 0.44476169558532835, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7189955711364746, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.711324155330658, + "num_tokens": 104706092.0, + "step": 4050 + }, + { + "epoch": 0.444871513287942, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5535433292388916, + "learning_rate": 1e-06, + "loss": 1.1059, + "mean_token_accuracy": 0.6822490096092224, + "num_tokens": 104732079.0, + "step": 4051 + }, + { + "epoch": 0.4449813309905557, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5374701023101807, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6825956106185913, + "num_tokens": 104760767.0, + "step": 4052 + }, + { + "epoch": 0.44509114869316935, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.516244888305664, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7123541831970215, + "num_tokens": 104788430.0, + "step": 4053 + }, + { + "epoch": 0.445200966395783, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4306154251098633, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6858199834823608, + "num_tokens": 104817342.0, + "step": 4054 + }, + { + "epoch": 0.44531078409839664, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.519179582595825, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.696787416934967, + "num_tokens": 104843385.0, + "step": 4055 + }, + { + "epoch": 0.44542060180101034, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.8100178241729736, + "learning_rate": 1e-06, + "loss": 1.1222, + "mean_token_accuracy": 0.6726869344711304, + "num_tokens": 104868452.0, + "step": 4056 + }, + { + "epoch": 0.445530419503624, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.769395112991333, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7185406684875488, + "num_tokens": 104890186.0, + "step": 4057 + }, + { + "epoch": 0.44564023720623763, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.682222604751587, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6907327175140381, + "num_tokens": 104914247.0, + "step": 4058 + }, + { + "epoch": 0.44575005490885133, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6089208126068115, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6916897296905518, + "num_tokens": 104939335.0, + "step": 4059 + }, + { + "epoch": 0.445859872611465, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7579517364501953, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7070471048355103, + "num_tokens": 104962635.0, + "step": 4060 + }, + { + "epoch": 0.4459696903140786, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.753422737121582, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7230173945426941, + "num_tokens": 104983584.0, + "step": 4061 + }, + { + "epoch": 0.44607950801669227, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.815647840499878, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7186673283576965, + "num_tokens": 105004819.0, + "step": 4062 + }, + { + "epoch": 0.446189325719306, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.645963430404663, + "learning_rate": 1e-06, + "loss": 0.9078, + "mean_token_accuracy": 0.7304322123527527, + "num_tokens": 105026420.0, + "step": 4063 + }, + { + "epoch": 0.4462991434219196, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.8379805088043213, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7020529508590698, + "num_tokens": 105048663.0, + "step": 4064 + }, + { + "epoch": 0.44640896112453327, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.828489303588867, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.6996413469314575, + "num_tokens": 105072117.0, + "step": 4065 + }, + { + "epoch": 0.4465187788271469, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5022783279418945, + "learning_rate": 1e-06, + "loss": 1.1636, + "mean_token_accuracy": 0.6676075458526611, + "num_tokens": 105100997.0, + "step": 4066 + }, + { + "epoch": 0.4466285965297606, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5053317546844482, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6863871812820435, + "num_tokens": 105126987.0, + "step": 4067 + }, + { + "epoch": 0.44673841423237426, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.573737621307373, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7070808410644531, + "num_tokens": 105151037.0, + "step": 4068 + }, + { + "epoch": 0.4468482319349879, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.567328691482544, + "learning_rate": 1e-06, + "loss": 1.113, + "mean_token_accuracy": 0.6784512996673584, + "num_tokens": 105179052.0, + "step": 4069 + }, + { + "epoch": 0.4469580496376016, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6393563747406006, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7067716121673584, + "num_tokens": 105202762.0, + "step": 4070 + }, + { + "epoch": 0.44706786734021525, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.471454620361328, + "learning_rate": 1e-06, + "loss": 1.1116, + "mean_token_accuracy": 0.6790350675582886, + "num_tokens": 105231743.0, + "step": 4071 + }, + { + "epoch": 0.4471776850428289, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3889174461364746, + "learning_rate": 1e-06, + "loss": 1.1124, + "mean_token_accuracy": 0.6741473078727722, + "num_tokens": 105258755.0, + "step": 4072 + }, + { + "epoch": 0.44728750274544254, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 3.1041979789733887, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6894630193710327, + "num_tokens": 105283134.0, + "step": 4073 + }, + { + "epoch": 0.44739732044805625, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 3.0873734951019287, + "learning_rate": 1e-06, + "loss": 0.8981, + "mean_token_accuracy": 0.7299916744232178, + "num_tokens": 105302217.0, + "step": 4074 + }, + { + "epoch": 0.4475071381506699, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7965681552886963, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.6974738836288452, + "num_tokens": 105324669.0, + "step": 4075 + }, + { + "epoch": 0.44761695585328354, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5433950424194336, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6856375932693481, + "num_tokens": 105351150.0, + "step": 4076 + }, + { + "epoch": 0.4477267735558972, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.2216684818267822, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.69691002368927, + "num_tokens": 105382344.0, + "step": 4077 + }, + { + "epoch": 0.4478365912585109, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4113106727600098, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7112480401992798, + "num_tokens": 105412565.0, + "step": 4078 + }, + { + "epoch": 0.44794640896112453, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.8503150939941406, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6957834959030151, + "num_tokens": 105434373.0, + "step": 4079 + }, + { + "epoch": 0.4480562266637382, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.717459201812744, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6896913051605225, + "num_tokens": 105457771.0, + "step": 4080 + }, + { + "epoch": 0.4481660443663519, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.418222427368164, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7158152461051941, + "num_tokens": 105483652.0, + "step": 4081 + }, + { + "epoch": 0.4482758620689655, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5449559688568115, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7170423865318298, + "num_tokens": 105506506.0, + "step": 4082 + }, + { + "epoch": 0.44838567977157917, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6255013942718506, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7108453512191772, + "num_tokens": 105535039.0, + "step": 4083 + }, + { + "epoch": 0.4484954974741928, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.467426300048828, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.6945821046829224, + "num_tokens": 105561456.0, + "step": 4084 + }, + { + "epoch": 0.4486053151768065, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5345706939697266, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6793988943099976, + "num_tokens": 105587627.0, + "step": 4085 + }, + { + "epoch": 0.44871513287942016, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6119043827056885, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7043731212615967, + "num_tokens": 105613040.0, + "step": 4086 + }, + { + "epoch": 0.4488249505820338, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5304439067840576, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6994110345840454, + "num_tokens": 105638821.0, + "step": 4087 + }, + { + "epoch": 0.4489347682846475, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6792232990264893, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.697676420211792, + "num_tokens": 105661894.0, + "step": 4088 + }, + { + "epoch": 0.44904458598726116, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.3851370811462402, + "learning_rate": 1e-06, + "loss": 1.0802, + "mean_token_accuracy": 0.6867997646331787, + "num_tokens": 105697736.0, + "step": 4089 + }, + { + "epoch": 0.4491544036898748, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7251946926116943, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7223984003067017, + "num_tokens": 105717630.0, + "step": 4090 + }, + { + "epoch": 0.44926422139248845, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6711385250091553, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7028094530105591, + "num_tokens": 105739780.0, + "step": 4091 + }, + { + "epoch": 0.44937403909510215, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.4735684394836426, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7118082046508789, + "num_tokens": 105767513.0, + "step": 4092 + }, + { + "epoch": 0.4494838567977158, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.5345497131347656, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6970174312591553, + "num_tokens": 105793045.0, + "step": 4093 + }, + { + "epoch": 0.44959367450032944, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.696096658706665, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6907083988189697, + "num_tokens": 105818510.0, + "step": 4094 + }, + { + "epoch": 0.4497034922029431, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.925079107284546, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7050654292106628, + "num_tokens": 105836643.0, + "step": 4095 + }, + { + "epoch": 0.4498133099055568, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.3100080490112305, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6811042428016663, + "num_tokens": 105865818.0, + "step": 4096 + }, + { + "epoch": 0.44992312760817044, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.581014633178711, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7090327143669128, + "num_tokens": 105891557.0, + "step": 4097 + }, + { + "epoch": 0.4500329453107841, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.734734058380127, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6871415376663208, + "num_tokens": 105914959.0, + "step": 4098 + }, + { + "epoch": 0.4501427630133978, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.636669397354126, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6955706477165222, + "num_tokens": 105940788.0, + "step": 4099 + }, + { + "epoch": 0.45025258071601143, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 3.3922278881073, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7240785360336304, + "num_tokens": 105960218.0, + "step": 4100 + }, + { + "epoch": 0.4503623984186251, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.991708993911743, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7052353024482727, + "num_tokens": 105985451.0, + "step": 4101 + }, + { + "epoch": 0.4504722161212387, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.392645835876465, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7062386274337769, + "num_tokens": 106015411.0, + "step": 4102 + }, + { + "epoch": 0.4505820338238524, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.5493316650390625, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6923546195030212, + "num_tokens": 106040921.0, + "step": 4103 + }, + { + "epoch": 0.45069185152646607, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.564908742904663, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7020125985145569, + "num_tokens": 106065656.0, + "step": 4104 + }, + { + "epoch": 0.4508016692290797, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.437859058380127, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7059484720230103, + "num_tokens": 106097280.0, + "step": 4105 + }, + { + "epoch": 0.45091148693169336, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.3734118938446045, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7269541025161743, + "num_tokens": 106123047.0, + "step": 4106 + }, + { + "epoch": 0.45102130463430706, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.4137890338897705, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7100644111633301, + "num_tokens": 106151119.0, + "step": 4107 + }, + { + "epoch": 0.4511311223369207, + "ewc_loss": 6.586313247680664e-06, + "grad_norm": 2.3162121772766113, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6799420118331909, + "num_tokens": 106183173.0, + "step": 4108 + }, + { + "epoch": 0.45124094003953436, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.8214755058288574, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7062812447547913, + "num_tokens": 106206388.0, + "step": 4109 + }, + { + "epoch": 0.45135075774214806, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3743810653686523, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7056145668029785, + "num_tokens": 106235476.0, + "step": 4110 + }, + { + "epoch": 0.4514605754447617, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.479142427444458, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7054524421691895, + "num_tokens": 106261962.0, + "step": 4111 + }, + { + "epoch": 0.45157039314737535, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5826542377471924, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6792969703674316, + "num_tokens": 106288609.0, + "step": 4112 + }, + { + "epoch": 0.451680210849989, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.481858491897583, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7098108530044556, + "num_tokens": 106316201.0, + "step": 4113 + }, + { + "epoch": 0.4517900285526027, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3443706035614014, + "learning_rate": 1e-06, + "loss": 1.1079, + "mean_token_accuracy": 0.6837380528450012, + "num_tokens": 106349835.0, + "step": 4114 + }, + { + "epoch": 0.45189984625521634, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.48544979095459, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6863700151443481, + "num_tokens": 106376203.0, + "step": 4115 + }, + { + "epoch": 0.45200966395783, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.4893453121185303, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6977633237838745, + "num_tokens": 106404493.0, + "step": 4116 + }, + { + "epoch": 0.4521194816604437, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.251575469970703, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.6978216171264648, + "num_tokens": 106432988.0, + "step": 4117 + }, + { + "epoch": 0.45222929936305734, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 3.132009506225586, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.69193434715271, + "num_tokens": 106453120.0, + "step": 4118 + }, + { + "epoch": 0.452339117065671, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.636256694793701, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.7110889554023743, + "num_tokens": 106475765.0, + "step": 4119 + }, + { + "epoch": 0.4524489347682846, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5995209217071533, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7085748314857483, + "num_tokens": 106501456.0, + "step": 4120 + }, + { + "epoch": 0.45255875247089833, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.302762508392334, + "learning_rate": 1e-06, + "loss": 1.1406, + "mean_token_accuracy": 0.6649976968765259, + "num_tokens": 106534135.0, + "step": 4121 + }, + { + "epoch": 0.452668570173512, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5431838035583496, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7130948305130005, + "num_tokens": 106558166.0, + "step": 4122 + }, + { + "epoch": 0.4527783878761256, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.4225895404815674, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6850378513336182, + "num_tokens": 106585004.0, + "step": 4123 + }, + { + "epoch": 0.45288820557873927, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5429744720458984, + "learning_rate": 1e-06, + "loss": 0.918, + "mean_token_accuracy": 0.7248636484146118, + "num_tokens": 106609383.0, + "step": 4124 + }, + { + "epoch": 0.45299802328135297, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.3342745304107666, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6939933896064758, + "num_tokens": 106641061.0, + "step": 4125 + }, + { + "epoch": 0.4531078409839666, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.2756967544555664, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7065452337265015, + "num_tokens": 106672179.0, + "step": 4126 + }, + { + "epoch": 0.45321765868658026, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.772900342941284, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6903491020202637, + "num_tokens": 106695512.0, + "step": 4127 + }, + { + "epoch": 0.45332747638919396, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 3.3603920936584473, + "learning_rate": 1e-06, + "loss": 0.932, + "mean_token_accuracy": 0.7241179347038269, + "num_tokens": 106723126.0, + "step": 4128 + }, + { + "epoch": 0.4534372940918076, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7339444160461426, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.7041788697242737, + "num_tokens": 106747737.0, + "step": 4129 + }, + { + "epoch": 0.45354711179442125, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.326314687728882, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6897739768028259, + "num_tokens": 106778830.0, + "step": 4130 + }, + { + "epoch": 0.4536569294970349, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5632073879241943, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6880139708518982, + "num_tokens": 106803825.0, + "step": 4131 + }, + { + "epoch": 0.4537667471996486, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3189384937286377, + "learning_rate": 1e-06, + "loss": 1.1449, + "mean_token_accuracy": 0.6741881966590881, + "num_tokens": 106836896.0, + "step": 4132 + }, + { + "epoch": 0.45387656490226225, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4966444969177246, + "learning_rate": 1e-06, + "loss": 1.124, + "mean_token_accuracy": 0.6730663776397705, + "num_tokens": 106862131.0, + "step": 4133 + }, + { + "epoch": 0.4539863826048759, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.560852527618408, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.7002876996994019, + "num_tokens": 106888132.0, + "step": 4134 + }, + { + "epoch": 0.4540962003074896, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.402953624725342, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7092610001564026, + "num_tokens": 106916637.0, + "step": 4135 + }, + { + "epoch": 0.45420601801010324, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4805908203125, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6968103647232056, + "num_tokens": 106943004.0, + "step": 4136 + }, + { + "epoch": 0.4543158357127169, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.513115882873535, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6922429800033569, + "num_tokens": 106972634.0, + "step": 4137 + }, + { + "epoch": 0.45442565341533053, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 3.040400743484497, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6851856708526611, + "num_tokens": 106995022.0, + "step": 4138 + }, + { + "epoch": 0.45453547111794423, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.8418848514556885, + "learning_rate": 1e-06, + "loss": 0.9399, + "mean_token_accuracy": 0.7151601314544678, + "num_tokens": 107014998.0, + "step": 4139 + }, + { + "epoch": 0.4546452888205579, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.229379415512085, + "learning_rate": 1e-06, + "loss": 1.1142, + "mean_token_accuracy": 0.6740360260009766, + "num_tokens": 107047636.0, + "step": 4140 + }, + { + "epoch": 0.4547551065231715, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 17.101572036743164, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7029519081115723, + "num_tokens": 107073075.0, + "step": 4141 + }, + { + "epoch": 0.45486492422578517, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 3.705047130584717, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.7052061557769775, + "num_tokens": 107095074.0, + "step": 4142 + }, + { + "epoch": 0.4549747419283989, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.6383113861083984, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.693554699420929, + "num_tokens": 107120478.0, + "step": 4143 + }, + { + "epoch": 0.4550845596310125, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.6042098999023438, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6981715559959412, + "num_tokens": 107149935.0, + "step": 4144 + }, + { + "epoch": 0.45519437733362617, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.707465887069702, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6977555751800537, + "num_tokens": 107176438.0, + "step": 4145 + }, + { + "epoch": 0.45530419503623987, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.47396183013916, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7035794258117676, + "num_tokens": 107202811.0, + "step": 4146 + }, + { + "epoch": 0.4554140127388535, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 3.2819161415100098, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6902387142181396, + "num_tokens": 107233423.0, + "step": 4147 + }, + { + "epoch": 0.45552383044146716, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.945335865020752, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6971395015716553, + "num_tokens": 107259979.0, + "step": 4148 + }, + { + "epoch": 0.4556336481440808, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.4943349361419678, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.6828896999359131, + "num_tokens": 107289683.0, + "step": 4149 + }, + { + "epoch": 0.4557434658466945, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6492159366607666, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7020754814147949, + "num_tokens": 107315025.0, + "step": 4150 + }, + { + "epoch": 0.45585328354930815, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.718967914581299, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.6977438926696777, + "num_tokens": 107337420.0, + "step": 4151 + }, + { + "epoch": 0.4559631012519218, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7075324058532715, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6940405964851379, + "num_tokens": 107360034.0, + "step": 4152 + }, + { + "epoch": 0.45607291895453544, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6544063091278076, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.682039737701416, + "num_tokens": 107389164.0, + "step": 4153 + }, + { + "epoch": 0.45618273665714915, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.787179708480835, + "learning_rate": 1e-06, + "loss": 1.1584, + "mean_token_accuracy": 0.6706169247627258, + "num_tokens": 107417239.0, + "step": 4154 + }, + { + "epoch": 0.4562925543597628, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6432178020477295, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6910793781280518, + "num_tokens": 107442241.0, + "step": 4155 + }, + { + "epoch": 0.45640237206237644, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.7144925594329834, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.700657844543457, + "num_tokens": 107464525.0, + "step": 4156 + }, + { + "epoch": 0.45651218976499014, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.74648118019104, + "learning_rate": 1e-06, + "loss": 0.9688, + "mean_token_accuracy": 0.7137224674224854, + "num_tokens": 107487863.0, + "step": 4157 + }, + { + "epoch": 0.4566220074676038, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.615478992462158, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6924831867218018, + "num_tokens": 107516463.0, + "step": 4158 + }, + { + "epoch": 0.45673182517021743, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4193460941314697, + "learning_rate": 1e-06, + "loss": 1.0991, + "mean_token_accuracy": 0.6785951852798462, + "num_tokens": 107547372.0, + "step": 4159 + }, + { + "epoch": 0.4568416428728311, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.648930072784424, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7025831937789917, + "num_tokens": 107570039.0, + "step": 4160 + }, + { + "epoch": 0.4569514605754448, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.5927350521087646, + "learning_rate": 1e-06, + "loss": 1.0845, + "mean_token_accuracy": 0.6892910003662109, + "num_tokens": 107595999.0, + "step": 4161 + }, + { + "epoch": 0.4570612782780584, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3900694847106934, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6942133903503418, + "num_tokens": 107624303.0, + "step": 4162 + }, + { + "epoch": 0.45717109598067207, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 3.03960919380188, + "learning_rate": 1e-06, + "loss": 1.1142, + "mean_token_accuracy": 0.683062732219696, + "num_tokens": 107654015.0, + "step": 4163 + }, + { + "epoch": 0.4572809136832858, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4398396015167236, + "learning_rate": 1e-06, + "loss": 0.9206, + "mean_token_accuracy": 0.7288307547569275, + "num_tokens": 107681066.0, + "step": 4164 + }, + { + "epoch": 0.4573907313858994, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6676931381225586, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7125980854034424, + "num_tokens": 107704738.0, + "step": 4165 + }, + { + "epoch": 0.45750054908851306, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.376586675643921, + "learning_rate": 1e-06, + "loss": 1.1195, + "mean_token_accuracy": 0.6733474135398865, + "num_tokens": 107738308.0, + "step": 4166 + }, + { + "epoch": 0.4576103667911267, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.2550137042999268, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6809048056602478, + "num_tokens": 107770284.0, + "step": 4167 + }, + { + "epoch": 0.4577201844937404, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6098952293395996, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6856452822685242, + "num_tokens": 107800283.0, + "step": 4168 + }, + { + "epoch": 0.45783000219635406, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.6428639888763428, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7159324884414673, + "num_tokens": 107824988.0, + "step": 4169 + }, + { + "epoch": 0.4579398198989677, + "ewc_loss": 6.5267086029052734e-06, + "grad_norm": 2.610032796859741, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7182844877243042, + "num_tokens": 107847507.0, + "step": 4170 + }, + { + "epoch": 0.45804963760158135, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3705434799194336, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6907357573509216, + "num_tokens": 107878874.0, + "step": 4171 + }, + { + "epoch": 0.45815945530419505, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.8548262119293213, + "learning_rate": 1e-06, + "loss": 1.1136, + "mean_token_accuracy": 0.6892300248146057, + "num_tokens": 107907987.0, + "step": 4172 + }, + { + "epoch": 0.4582692730068087, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.335608720779419, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6900925040245056, + "num_tokens": 107935921.0, + "step": 4173 + }, + { + "epoch": 0.45837909070942234, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.274434804916382, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6988388299942017, + "num_tokens": 107969167.0, + "step": 4174 + }, + { + "epoch": 0.45848890841203604, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.510185718536377, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.709142804145813, + "num_tokens": 107993529.0, + "step": 4175 + }, + { + "epoch": 0.4585987261146497, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.532783031463623, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6824732422828674, + "num_tokens": 108019805.0, + "step": 4176 + }, + { + "epoch": 0.45870854381726334, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.341643810272217, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.687942624092102, + "num_tokens": 108048592.0, + "step": 4177 + }, + { + "epoch": 0.458818361519877, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.3138210773468018, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7123762369155884, + "num_tokens": 108079410.0, + "step": 4178 + }, + { + "epoch": 0.4589281792224907, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5961503982543945, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.70833820104599, + "num_tokens": 108106842.0, + "step": 4179 + }, + { + "epoch": 0.45903799692510433, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6286873817443848, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6952885389328003, + "num_tokens": 108134819.0, + "step": 4180 + }, + { + "epoch": 0.459147814627718, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7163267135620117, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7083145380020142, + "num_tokens": 108160211.0, + "step": 4181 + }, + { + "epoch": 0.4592576323303316, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.608104944229126, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7037899494171143, + "num_tokens": 108184231.0, + "step": 4182 + }, + { + "epoch": 0.4593674500329453, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.6326541900634766, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6959757804870605, + "num_tokens": 108206916.0, + "step": 4183 + }, + { + "epoch": 0.45947726773555897, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.7260169982910156, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7101220488548279, + "num_tokens": 108227851.0, + "step": 4184 + }, + { + "epoch": 0.4595870854381726, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.5259976387023926, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6915326118469238, + "num_tokens": 108254622.0, + "step": 4185 + }, + { + "epoch": 0.4596969031407863, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.4212405681610107, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6961275339126587, + "num_tokens": 108282366.0, + "step": 4186 + }, + { + "epoch": 0.45980672084339996, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.3402750492095947, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6859514713287354, + "num_tokens": 108314359.0, + "step": 4187 + }, + { + "epoch": 0.4599165385460136, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.389455556869507, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.6969273686408997, + "num_tokens": 108342960.0, + "step": 4188 + }, + { + "epoch": 0.46002635624862726, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4462807178497314, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6851571202278137, + "num_tokens": 108369224.0, + "step": 4189 + }, + { + "epoch": 0.46013617395124096, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.47214674949646, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6875815391540527, + "num_tokens": 108397077.0, + "step": 4190 + }, + { + "epoch": 0.4602459916538546, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.506959915161133, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6927394270896912, + "num_tokens": 108421690.0, + "step": 4191 + }, + { + "epoch": 0.46035580935646825, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.4738597869873047, + "learning_rate": 1e-06, + "loss": 1.1121, + "mean_token_accuracy": 0.6747228503227234, + "num_tokens": 108449768.0, + "step": 4192 + }, + { + "epoch": 0.46046562705908195, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 3.271846055984497, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.701452374458313, + "num_tokens": 108466772.0, + "step": 4193 + }, + { + "epoch": 0.4605754447616956, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.6305463314056396, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7190840244293213, + "num_tokens": 108488184.0, + "step": 4194 + }, + { + "epoch": 0.46068526246430924, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.478764057159424, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6959750652313232, + "num_tokens": 108514984.0, + "step": 4195 + }, + { + "epoch": 0.4607950801669229, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 9.677952766418457, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7094407081604004, + "num_tokens": 108540998.0, + "step": 4196 + }, + { + "epoch": 0.4609048978695366, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.8978288173675537, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7038060426712036, + "num_tokens": 108570211.0, + "step": 4197 + }, + { + "epoch": 0.46101471557215024, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.506986379623413, + "learning_rate": 1e-06, + "loss": 1.1161, + "mean_token_accuracy": 0.6755243539810181, + "num_tokens": 108599725.0, + "step": 4198 + }, + { + "epoch": 0.4611245332747639, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.591042995452881, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6890866160392761, + "num_tokens": 108624771.0, + "step": 4199 + }, + { + "epoch": 0.4612343509773775, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.3146138191223145, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6872684955596924, + "num_tokens": 108655965.0, + "step": 4200 + }, + { + "epoch": 0.46134416867999123, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.595215320587158, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.6984677314758301, + "num_tokens": 108678648.0, + "step": 4201 + }, + { + "epoch": 0.4614539863826049, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.4408490657806396, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6883513927459717, + "num_tokens": 108708088.0, + "step": 4202 + }, + { + "epoch": 0.4615638040852185, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.3352200984954834, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7090091109275818, + "num_tokens": 108737497.0, + "step": 4203 + }, + { + "epoch": 0.4616736217878322, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.615093946456909, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7051423192024231, + "num_tokens": 108764609.0, + "step": 4204 + }, + { + "epoch": 0.46178343949044587, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.529468059539795, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6988409757614136, + "num_tokens": 108790774.0, + "step": 4205 + }, + { + "epoch": 0.4618932571930595, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.2432751655578613, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7128216624259949, + "num_tokens": 108820188.0, + "step": 4206 + }, + { + "epoch": 0.46200307489567316, + "ewc_loss": 6.556510925292969e-06, + "grad_norm": 2.8064537048339844, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7184518575668335, + "num_tokens": 108841518.0, + "step": 4207 + }, + { + "epoch": 0.46211289259828686, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.6453754901885986, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6922455430030823, + "num_tokens": 108865351.0, + "step": 4208 + }, + { + "epoch": 0.4622227103009005, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.7628672122955322, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6869732141494751, + "num_tokens": 108887902.0, + "step": 4209 + }, + { + "epoch": 0.46233252800351415, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.321493148803711, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7026799917221069, + "num_tokens": 108917553.0, + "step": 4210 + }, + { + "epoch": 0.46244234570612786, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.549028158187866, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6951181888580322, + "num_tokens": 108949999.0, + "step": 4211 + }, + { + "epoch": 0.4625521634087415, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.4637136459350586, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.7019084692001343, + "num_tokens": 108976054.0, + "step": 4212 + }, + { + "epoch": 0.46266198111135515, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.5117132663726807, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7133292555809021, + "num_tokens": 109001853.0, + "step": 4213 + }, + { + "epoch": 0.4627717988139688, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 3.0287868976593018, + "learning_rate": 1e-06, + "loss": 0.8913, + "mean_token_accuracy": 0.7304160594940186, + "num_tokens": 109018515.0, + "step": 4214 + }, + { + "epoch": 0.4628816165165825, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.8902738094329834, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7043433785438538, + "num_tokens": 109043524.0, + "step": 4215 + }, + { + "epoch": 0.46299143421919614, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.7886245250701904, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6907578110694885, + "num_tokens": 109065870.0, + "step": 4216 + }, + { + "epoch": 0.4631012519218098, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.423388957977295, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.6853687167167664, + "num_tokens": 109092819.0, + "step": 4217 + }, + { + "epoch": 0.46321106962442343, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.5827770233154297, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7143285274505615, + "num_tokens": 109119366.0, + "step": 4218 + }, + { + "epoch": 0.46332088732703713, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.835949182510376, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6931019425392151, + "num_tokens": 109142027.0, + "step": 4219 + }, + { + "epoch": 0.4634307050296508, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.64406156539917, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7001761198043823, + "num_tokens": 109164566.0, + "step": 4220 + }, + { + "epoch": 0.4635405227322644, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.99192214012146, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7215448617935181, + "num_tokens": 109185471.0, + "step": 4221 + }, + { + "epoch": 0.4636503404348781, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.745558500289917, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6990109086036682, + "num_tokens": 109209593.0, + "step": 4222 + }, + { + "epoch": 0.4637601581374918, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.631472110748291, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6822412014007568, + "num_tokens": 109233085.0, + "step": 4223 + }, + { + "epoch": 0.4638699758401054, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.704637050628662, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7063817381858826, + "num_tokens": 109255439.0, + "step": 4224 + }, + { + "epoch": 0.46397979354271907, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6069600582122803, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6878067255020142, + "num_tokens": 109285240.0, + "step": 4225 + }, + { + "epoch": 0.46408961124533277, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.791623830795288, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7161586880683899, + "num_tokens": 109306534.0, + "step": 4226 + }, + { + "epoch": 0.4641994289479464, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.504481077194214, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.689704179763794, + "num_tokens": 109335144.0, + "step": 4227 + }, + { + "epoch": 0.46430924665056006, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.457716464996338, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6902288198471069, + "num_tokens": 109363223.0, + "step": 4228 + }, + { + "epoch": 0.4644190643531737, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.7005836963653564, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7049271464347839, + "num_tokens": 109385976.0, + "step": 4229 + }, + { + "epoch": 0.4645288820557874, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.644563913345337, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.6899794340133667, + "num_tokens": 109413991.0, + "step": 4230 + }, + { + "epoch": 0.46463869975840105, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.6182596683502197, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7160214185714722, + "num_tokens": 109434389.0, + "step": 4231 + }, + { + "epoch": 0.4647485174610147, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.3159263134002686, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6867672801017761, + "num_tokens": 109466391.0, + "step": 4232 + }, + { + "epoch": 0.4648583351636284, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.6913692951202393, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7055948972702026, + "num_tokens": 109489623.0, + "step": 4233 + }, + { + "epoch": 0.46496815286624205, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.6493635177612305, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6879791617393494, + "num_tokens": 109515780.0, + "step": 4234 + }, + { + "epoch": 0.4650779705688557, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6419763565063477, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7030653953552246, + "num_tokens": 109539212.0, + "step": 4235 + }, + { + "epoch": 0.46518778827146934, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.826209783554077, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6845526099205017, + "num_tokens": 109560287.0, + "step": 4236 + }, + { + "epoch": 0.46529760597408304, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.487316846847534, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7016894817352295, + "num_tokens": 109588456.0, + "step": 4237 + }, + { + "epoch": 0.4654074236766967, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.571342706680298, + "learning_rate": 1e-06, + "loss": 1.1205, + "mean_token_accuracy": 0.6788958311080933, + "num_tokens": 109616581.0, + "step": 4238 + }, + { + "epoch": 0.46551724137931033, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5924882888793945, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7142277956008911, + "num_tokens": 109641331.0, + "step": 4239 + }, + { + "epoch": 0.46562705908192403, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.368663787841797, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6874053478240967, + "num_tokens": 109670886.0, + "step": 4240 + }, + { + "epoch": 0.4657368767845377, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.7958054542541504, + "learning_rate": 1e-06, + "loss": 1.0874, + "mean_token_accuracy": 0.6899368762969971, + "num_tokens": 109697801.0, + "step": 4241 + }, + { + "epoch": 0.4658466944871513, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.3848609924316406, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7006640434265137, + "num_tokens": 109724159.0, + "step": 4242 + }, + { + "epoch": 0.46595651218976497, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.684556007385254, + "learning_rate": 1e-06, + "loss": 1.1154, + "mean_token_accuracy": 0.6691617965698242, + "num_tokens": 109749383.0, + "step": 4243 + }, + { + "epoch": 0.4660663298923787, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5468051433563232, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7049227952957153, + "num_tokens": 109776098.0, + "step": 4244 + }, + { + "epoch": 0.4661761475949923, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.6854407787323, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7082452774047852, + "num_tokens": 109797559.0, + "step": 4245 + }, + { + "epoch": 0.46628596529760596, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.3782241344451904, + "learning_rate": 1e-06, + "loss": 0.9193, + "mean_token_accuracy": 0.7255748510360718, + "num_tokens": 109828678.0, + "step": 4246 + }, + { + "epoch": 0.4663957830002196, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.5159621238708496, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.6790832281112671, + "num_tokens": 109857652.0, + "step": 4247 + }, + { + "epoch": 0.4665056007028333, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.3889901638031006, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6945739984512329, + "num_tokens": 109884554.0, + "step": 4248 + }, + { + "epoch": 0.46661541840544696, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.412254571914673, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7141994833946228, + "num_tokens": 109911682.0, + "step": 4249 + }, + { + "epoch": 0.4667252361080606, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.649756908416748, + "learning_rate": 1e-06, + "loss": 1.1171, + "mean_token_accuracy": 0.6822754740715027, + "num_tokens": 109936157.0, + "step": 4250 + }, + { + "epoch": 0.4668350538106743, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.373481035232544, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.6927570104598999, + "num_tokens": 109963461.0, + "step": 4251 + }, + { + "epoch": 0.46694487151328795, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 6.720503330230713, + "learning_rate": 1e-06, + "loss": 0.9132, + "mean_token_accuracy": 0.7235736846923828, + "num_tokens": 109985239.0, + "step": 4252 + }, + { + "epoch": 0.4670546892159016, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7853548526763916, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6862266063690186, + "num_tokens": 110009774.0, + "step": 4253 + }, + { + "epoch": 0.46716450691851524, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.9156429767608643, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6979475021362305, + "num_tokens": 110030532.0, + "step": 4254 + }, + { + "epoch": 0.46727432462112894, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.350642681121826, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6839941143989563, + "num_tokens": 110064065.0, + "step": 4255 + }, + { + "epoch": 0.4673841423237426, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.562182903289795, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.7002617120742798, + "num_tokens": 110089461.0, + "step": 4256 + }, + { + "epoch": 0.46749396002635624, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.2044878005981445, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6931239366531372, + "num_tokens": 110124072.0, + "step": 4257 + }, + { + "epoch": 0.4676037777289699, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.722738027572632, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.7015928626060486, + "num_tokens": 110145663.0, + "step": 4258 + }, + { + "epoch": 0.4677135954315836, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.436164140701294, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6835989356040955, + "num_tokens": 110175157.0, + "step": 4259 + }, + { + "epoch": 0.46782341313419723, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 8.022193908691406, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7040865421295166, + "num_tokens": 110198393.0, + "step": 4260 + }, + { + "epoch": 0.4679332308368109, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.3357841968536377, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6916749477386475, + "num_tokens": 110231770.0, + "step": 4261 + }, + { + "epoch": 0.4680430485394246, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.5747580528259277, + "learning_rate": 1e-06, + "loss": 1.1172, + "mean_token_accuracy": 0.6805911064147949, + "num_tokens": 110261755.0, + "step": 4262 + }, + { + "epoch": 0.4681528662420382, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.8150389194488525, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.68375563621521, + "num_tokens": 110281687.0, + "step": 4263 + }, + { + "epoch": 0.46826268394465187, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.6955406665802, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7113932967185974, + "num_tokens": 110305069.0, + "step": 4264 + }, + { + "epoch": 0.4683725016472655, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.781658411026001, + "learning_rate": 1e-06, + "loss": 0.8914, + "mean_token_accuracy": 0.7352584004402161, + "num_tokens": 110324935.0, + "step": 4265 + }, + { + "epoch": 0.4684823193498792, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.711317539215088, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6893718838691711, + "num_tokens": 110346955.0, + "step": 4266 + }, + { + "epoch": 0.46859213705249286, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.35001540184021, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7111828327178955, + "num_tokens": 110373544.0, + "step": 4267 + }, + { + "epoch": 0.4687019547551065, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.482252597808838, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6809837818145752, + "num_tokens": 110405117.0, + "step": 4268 + }, + { + "epoch": 0.4688117724577202, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 8.084890365600586, + "learning_rate": 1e-06, + "loss": 1.0609, + "mean_token_accuracy": 0.6927961111068726, + "num_tokens": 110426767.0, + "step": 4269 + }, + { + "epoch": 0.46892159016033386, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 3.316423177719116, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.686638593673706, + "num_tokens": 110450394.0, + "step": 4270 + }, + { + "epoch": 0.4690314078629475, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5110886096954346, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6805525422096252, + "num_tokens": 110479240.0, + "step": 4271 + }, + { + "epoch": 0.46914122556556115, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.4957275390625, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7054376602172852, + "num_tokens": 110503201.0, + "step": 4272 + }, + { + "epoch": 0.46925104326817485, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 3.5847604274749756, + "learning_rate": 1e-06, + "loss": 1.0943, + "mean_token_accuracy": 0.6856076121330261, + "num_tokens": 110533873.0, + "step": 4273 + }, + { + "epoch": 0.4693608609707885, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 3.5582032203674316, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7101535797119141, + "num_tokens": 110553795.0, + "step": 4274 + }, + { + "epoch": 0.46947067867340214, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.5697295665740967, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.681362509727478, + "num_tokens": 110582884.0, + "step": 4275 + }, + { + "epoch": 0.4695804963760158, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.538769483566284, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7036285996437073, + "num_tokens": 110608619.0, + "step": 4276 + }, + { + "epoch": 0.4696903140786295, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.5511980056762695, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7047714591026306, + "num_tokens": 110633119.0, + "step": 4277 + }, + { + "epoch": 0.46980013178124314, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.4868710041046143, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6855998039245605, + "num_tokens": 110658497.0, + "step": 4278 + }, + { + "epoch": 0.4699099494838568, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.7041189670562744, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7082818746566772, + "num_tokens": 110682326.0, + "step": 4279 + }, + { + "epoch": 0.4700197671864705, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.494157552719116, + "learning_rate": 1e-06, + "loss": 1.1118, + "mean_token_accuracy": 0.685031533241272, + "num_tokens": 110712977.0, + "step": 4280 + }, + { + "epoch": 0.47012958488908413, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.670044183731079, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6903250813484192, + "num_tokens": 110735079.0, + "step": 4281 + }, + { + "epoch": 0.4702394025916978, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.3991289138793945, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6872695684432983, + "num_tokens": 110763270.0, + "step": 4282 + }, + { + "epoch": 0.4703492202943114, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.522994041442871, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6892151832580566, + "num_tokens": 110789102.0, + "step": 4283 + }, + { + "epoch": 0.4704590379969251, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.2938711643218994, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7155911922454834, + "num_tokens": 110820768.0, + "step": 4284 + }, + { + "epoch": 0.47056885569953877, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.25469708442688, + "learning_rate": 1e-06, + "loss": 1.1359, + "mean_token_accuracy": 0.6706975102424622, + "num_tokens": 110852042.0, + "step": 4285 + }, + { + "epoch": 0.4706786734021524, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.564481258392334, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7130403518676758, + "num_tokens": 110877114.0, + "step": 4286 + }, + { + "epoch": 0.4707884911047661, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.257471799850464, + "learning_rate": 1e-06, + "loss": 0.8849, + "mean_token_accuracy": 0.7360822558403015, + "num_tokens": 110905762.0, + "step": 4287 + }, + { + "epoch": 0.47089830880737976, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.6585893630981445, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.6938163042068481, + "num_tokens": 110926649.0, + "step": 4288 + }, + { + "epoch": 0.4710081265099934, + "ewc_loss": 6.616115570068359e-06, + "grad_norm": 2.65480899810791, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6894923448562622, + "num_tokens": 110952555.0, + "step": 4289 + }, + { + "epoch": 0.47111794421260705, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.5961413383483887, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6879973411560059, + "num_tokens": 110978068.0, + "step": 4290 + }, + { + "epoch": 0.47122776191522076, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.8073930740356445, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.7087539434432983, + "num_tokens": 111000492.0, + "step": 4291 + }, + { + "epoch": 0.4713375796178344, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.435321807861328, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6847304105758667, + "num_tokens": 111031966.0, + "step": 4292 + }, + { + "epoch": 0.47144739732044805, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.6563758850097656, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.6798169612884521, + "num_tokens": 111060055.0, + "step": 4293 + }, + { + "epoch": 0.4715572150230617, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.497946262359619, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.688313901424408, + "num_tokens": 111088646.0, + "step": 4294 + }, + { + "epoch": 0.4716670327256754, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.558734893798828, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7020633220672607, + "num_tokens": 111113189.0, + "step": 4295 + }, + { + "epoch": 0.47177685042828904, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.510762929916382, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7012664079666138, + "num_tokens": 111136436.0, + "step": 4296 + }, + { + "epoch": 0.4718866681309027, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.561596632003784, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7075097560882568, + "num_tokens": 111162183.0, + "step": 4297 + }, + { + "epoch": 0.4719964858335164, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.377741813659668, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6837508678436279, + "num_tokens": 111191417.0, + "step": 4298 + }, + { + "epoch": 0.47210630353613003, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.4352033138275146, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.699674129486084, + "num_tokens": 111219309.0, + "step": 4299 + }, + { + "epoch": 0.4722161212387437, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.417471170425415, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7129989266395569, + "num_tokens": 111247213.0, + "step": 4300 + }, + { + "epoch": 0.4723259389413573, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 3.0341153144836426, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6851162910461426, + "num_tokens": 111270004.0, + "step": 4301 + }, + { + "epoch": 0.472435756643971, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.63073468208313, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7069563269615173, + "num_tokens": 111294602.0, + "step": 4302 + }, + { + "epoch": 0.4725455743465847, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.6789731979370117, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.6864762306213379, + "num_tokens": 111321839.0, + "step": 4303 + }, + { + "epoch": 0.4726553920491983, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.517442226409912, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.684628963470459, + "num_tokens": 111348923.0, + "step": 4304 + }, + { + "epoch": 0.47276520975181197, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.4086086750030518, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6873264908790588, + "num_tokens": 111378979.0, + "step": 4305 + }, + { + "epoch": 0.47287502745442567, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.73134183883667, + "learning_rate": 1e-06, + "loss": 1.099, + "mean_token_accuracy": 0.6925715804100037, + "num_tokens": 111402072.0, + "step": 4306 + }, + { + "epoch": 0.4729848451570393, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7505950927734375, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7042679786682129, + "num_tokens": 111424366.0, + "step": 4307 + }, + { + "epoch": 0.47309466285965296, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.242727756500244, + "learning_rate": 1e-06, + "loss": 1.1208, + "mean_token_accuracy": 0.6770556569099426, + "num_tokens": 111457727.0, + "step": 4308 + }, + { + "epoch": 0.47320448056226666, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.6232528686523438, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7114604711532593, + "num_tokens": 111479473.0, + "step": 4309 + }, + { + "epoch": 0.4733142982648803, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.4768452644348145, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6830514669418335, + "num_tokens": 111505459.0, + "step": 4310 + }, + { + "epoch": 0.47342411596749395, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5020062923431396, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.6909117698669434, + "num_tokens": 111529820.0, + "step": 4311 + }, + { + "epoch": 0.4735339336701076, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.62935209274292, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6942726969718933, + "num_tokens": 111556109.0, + "step": 4312 + }, + { + "epoch": 0.4736437513727213, + "ewc_loss": 6.645917892456055e-06, + "grad_norm": 2.6017940044403076, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7062500715255737, + "num_tokens": 111580721.0, + "step": 4313 + }, + { + "epoch": 0.47375356907533495, + "ewc_loss": 6.67572021484375e-06, + "grad_norm": 2.5584664344787598, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7105550169944763, + "num_tokens": 111604459.0, + "step": 4314 + }, + { + "epoch": 0.4738633867779486, + "ewc_loss": 6.705522537231445e-06, + "grad_norm": 2.678110361099243, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.6970318555831909, + "num_tokens": 111629930.0, + "step": 4315 + }, + { + "epoch": 0.4739732044805623, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5432653427124023, + "learning_rate": 1e-06, + "loss": 1.1127, + "mean_token_accuracy": 0.6800103187561035, + "num_tokens": 111657613.0, + "step": 4316 + }, + { + "epoch": 0.47408302218317594, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 3.063751697540283, + "learning_rate": 1e-06, + "loss": 0.922, + "mean_token_accuracy": 0.7230884432792664, + "num_tokens": 111677344.0, + "step": 4317 + }, + { + "epoch": 0.4741928398857896, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.7982373237609863, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7111119031906128, + "num_tokens": 111701017.0, + "step": 4318 + }, + { + "epoch": 0.47430265758840323, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.7788188457489014, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7151615619659424, + "num_tokens": 111721446.0, + "step": 4319 + }, + { + "epoch": 0.47441247529101693, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.528900146484375, + "learning_rate": 1e-06, + "loss": 1.1059, + "mean_token_accuracy": 0.6817697286605835, + "num_tokens": 111749768.0, + "step": 4320 + }, + { + "epoch": 0.4745222929936306, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5064148902893066, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6958135366439819, + "num_tokens": 111780595.0, + "step": 4321 + }, + { + "epoch": 0.4746321106962442, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.383094549179077, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6937675476074219, + "num_tokens": 111808453.0, + "step": 4322 + }, + { + "epoch": 0.47474192839885787, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.472074270248413, + "learning_rate": 1e-06, + "loss": 1.1203, + "mean_token_accuracy": 0.6747840642929077, + "num_tokens": 111838333.0, + "step": 4323 + }, + { + "epoch": 0.4748517461014716, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.6014623641967773, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.700064480304718, + "num_tokens": 111862836.0, + "step": 4324 + }, + { + "epoch": 0.4749615638040852, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.653730630874634, + "learning_rate": 1e-06, + "loss": 1.1468, + "mean_token_accuracy": 0.6712005138397217, + "num_tokens": 111889887.0, + "step": 4325 + }, + { + "epoch": 0.47507138150669886, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.649493455886841, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6851112246513367, + "num_tokens": 111916363.0, + "step": 4326 + }, + { + "epoch": 0.47518119920931257, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.2860822677612305, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7101249098777771, + "num_tokens": 111947248.0, + "step": 4327 + }, + { + "epoch": 0.4752910169119262, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.34444260597229, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.6844212412834167, + "num_tokens": 111979006.0, + "step": 4328 + }, + { + "epoch": 0.47540083461453986, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.784383773803711, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7168837785720825, + "num_tokens": 112001590.0, + "step": 4329 + }, + { + "epoch": 0.4755106523171535, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7390716075897217, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7088443636894226, + "num_tokens": 112022137.0, + "step": 4330 + }, + { + "epoch": 0.4756204700197672, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5535128116607666, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7242478728294373, + "num_tokens": 112047019.0, + "step": 4331 + }, + { + "epoch": 0.47573028772238085, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.4883744716644287, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6889045238494873, + "num_tokens": 112073720.0, + "step": 4332 + }, + { + "epoch": 0.4758401054249945, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6475515365600586, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6981159448623657, + "num_tokens": 112100905.0, + "step": 4333 + }, + { + "epoch": 0.47594992312760814, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.4632997512817383, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.6995983123779297, + "num_tokens": 112127185.0, + "step": 4334 + }, + { + "epoch": 0.47605974083022184, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7239105701446533, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7130252718925476, + "num_tokens": 112148344.0, + "step": 4335 + }, + { + "epoch": 0.4761695585328355, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7598681449890137, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6924597024917603, + "num_tokens": 112172753.0, + "step": 4336 + }, + { + "epoch": 0.47627937623544914, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.611666679382324, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.7036206722259521, + "num_tokens": 112195835.0, + "step": 4337 + }, + { + "epoch": 0.47638919393806284, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.4192416667938232, + "learning_rate": 1e-06, + "loss": 1.1277, + "mean_token_accuracy": 0.6722307205200195, + "num_tokens": 112225281.0, + "step": 4338 + }, + { + "epoch": 0.4764990116406765, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.562757730484009, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.6983312964439392, + "num_tokens": 112249829.0, + "step": 4339 + }, + { + "epoch": 0.47660882934329013, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.294619560241699, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6881422400474548, + "num_tokens": 112279445.0, + "step": 4340 + }, + { + "epoch": 0.4767186470459038, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.455301284790039, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6902443170547485, + "num_tokens": 112309786.0, + "step": 4341 + }, + { + "epoch": 0.4768284647485175, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.566138744354248, + "learning_rate": 1e-06, + "loss": 1.1076, + "mean_token_accuracy": 0.6808954477310181, + "num_tokens": 112338737.0, + "step": 4342 + }, + { + "epoch": 0.4769382824511311, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.2426421642303467, + "learning_rate": 1e-06, + "loss": 1.0833, + "mean_token_accuracy": 0.6823358535766602, + "num_tokens": 112376012.0, + "step": 4343 + }, + { + "epoch": 0.47704810015374477, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.8791332244873047, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.6889681816101074, + "num_tokens": 112398477.0, + "step": 4344 + }, + { + "epoch": 0.47715791785635847, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.662214994430542, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7008066177368164, + "num_tokens": 112421259.0, + "step": 4345 + }, + { + "epoch": 0.4772677355589721, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 3.115222215652466, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6839948892593384, + "num_tokens": 112445827.0, + "step": 4346 + }, + { + "epoch": 0.47737755326158576, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 3.0697009563446045, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6816166639328003, + "num_tokens": 112468017.0, + "step": 4347 + }, + { + "epoch": 0.4774873709641994, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7725772857666016, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6779754161834717, + "num_tokens": 112493136.0, + "step": 4348 + }, + { + "epoch": 0.4775971886668131, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.499027729034424, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6876015663146973, + "num_tokens": 112517619.0, + "step": 4349 + }, + { + "epoch": 0.47770700636942676, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.3678877353668213, + "learning_rate": 1e-06, + "loss": 1.1715, + "mean_token_accuracy": 0.6677205562591553, + "num_tokens": 112553523.0, + "step": 4350 + }, + { + "epoch": 0.4778168240720404, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.555299997329712, + "learning_rate": 1e-06, + "loss": 1.209, + "mean_token_accuracy": 0.6545101404190063, + "num_tokens": 112581580.0, + "step": 4351 + }, + { + "epoch": 0.47792664177465405, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.52066707611084, + "learning_rate": 1e-06, + "loss": 1.1179, + "mean_token_accuracy": 0.681475818157196, + "num_tokens": 112612031.0, + "step": 4352 + }, + { + "epoch": 0.47803645947726775, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.2862648963928223, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7053741812705994, + "num_tokens": 112642814.0, + "step": 4353 + }, + { + "epoch": 0.4781462771798814, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7371504306793213, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7005497217178345, + "num_tokens": 112665706.0, + "step": 4354 + }, + { + "epoch": 0.47825609488249504, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5999298095703125, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.689703106880188, + "num_tokens": 112690358.0, + "step": 4355 + }, + { + "epoch": 0.47836591258510874, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.326542377471924, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6915968656539917, + "num_tokens": 112719946.0, + "step": 4356 + }, + { + "epoch": 0.4784757302877224, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6165292263031006, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6889979839324951, + "num_tokens": 112744322.0, + "step": 4357 + }, + { + "epoch": 0.47858554799033604, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.4806880950927734, + "learning_rate": 1e-06, + "loss": 1.1506, + "mean_token_accuracy": 0.6683356761932373, + "num_tokens": 112775838.0, + "step": 4358 + }, + { + "epoch": 0.4786953656929497, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.4746806621551514, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7084219455718994, + "num_tokens": 112803649.0, + "step": 4359 + }, + { + "epoch": 0.4788051833955634, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.674686908721924, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6810365915298462, + "num_tokens": 112831812.0, + "step": 4360 + }, + { + "epoch": 0.47891500109817703, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5313491821289062, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7232197523117065, + "num_tokens": 112858557.0, + "step": 4361 + }, + { + "epoch": 0.4790248188007907, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.496432304382324, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7025589942932129, + "num_tokens": 112884800.0, + "step": 4362 + }, + { + "epoch": 0.4791346365034044, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.470785617828369, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.6980412006378174, + "num_tokens": 112913848.0, + "step": 4363 + }, + { + "epoch": 0.479244454206018, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7435083389282227, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7023968696594238, + "num_tokens": 112939866.0, + "step": 4364 + }, + { + "epoch": 0.47935427190863167, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.603736639022827, + "learning_rate": 1e-06, + "loss": 0.8922, + "mean_token_accuracy": 0.7326617240905762, + "num_tokens": 112965056.0, + "step": 4365 + }, + { + "epoch": 0.4794640896112453, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.432368755340576, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6834177374839783, + "num_tokens": 112993346.0, + "step": 4366 + }, + { + "epoch": 0.479573907313859, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5750341415405273, + "learning_rate": 1e-06, + "loss": 1.1296, + "mean_token_accuracy": 0.6796693801879883, + "num_tokens": 113020155.0, + "step": 4367 + }, + { + "epoch": 0.47968372501647266, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6004180908203125, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.6917650699615479, + "num_tokens": 113044250.0, + "step": 4368 + }, + { + "epoch": 0.4797935427190863, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.584233283996582, + "learning_rate": 1e-06, + "loss": 1.0832, + "mean_token_accuracy": 0.6915771961212158, + "num_tokens": 113072237.0, + "step": 4369 + }, + { + "epoch": 0.47990336042169995, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.706472635269165, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.704438328742981, + "num_tokens": 113096600.0, + "step": 4370 + }, + { + "epoch": 0.48001317812431366, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.706258773803711, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6910191774368286, + "num_tokens": 113131476.0, + "step": 4371 + }, + { + "epoch": 0.4801229958269273, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.385246515274048, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.685406506061554, + "num_tokens": 113161614.0, + "step": 4372 + }, + { + "epoch": 0.48023281352954095, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7886247634887695, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.696427583694458, + "num_tokens": 113184723.0, + "step": 4373 + }, + { + "epoch": 0.48034263123215465, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6294713020324707, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.6919007897377014, + "num_tokens": 113207737.0, + "step": 4374 + }, + { + "epoch": 0.4804524489347683, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.481433391571045, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6931930184364319, + "num_tokens": 113236517.0, + "step": 4375 + }, + { + "epoch": 0.48056226663738194, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.6704249382019043, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7108250856399536, + "num_tokens": 113258955.0, + "step": 4376 + }, + { + "epoch": 0.4806720843399956, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.540623903274536, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6961439251899719, + "num_tokens": 113287178.0, + "step": 4377 + }, + { + "epoch": 0.4807819020426093, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5871617794036865, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7041488885879517, + "num_tokens": 113314384.0, + "step": 4378 + }, + { + "epoch": 0.48089171974522293, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.8312575817108154, + "learning_rate": 1e-06, + "loss": 1.1041, + "mean_token_accuracy": 0.6910190582275391, + "num_tokens": 113335624.0, + "step": 4379 + }, + { + "epoch": 0.4810015374478366, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5359320640563965, + "learning_rate": 1e-06, + "loss": 1.1121, + "mean_token_accuracy": 0.6795830726623535, + "num_tokens": 113364171.0, + "step": 4380 + }, + { + "epoch": 0.4811113551504502, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 3.070571184158325, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7085259556770325, + "num_tokens": 113385080.0, + "step": 4381 + }, + { + "epoch": 0.4812211728530639, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.4435012340545654, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6952959299087524, + "num_tokens": 113413851.0, + "step": 4382 + }, + { + "epoch": 0.4813309905556776, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.3628623485565186, + "learning_rate": 1e-06, + "loss": 1.1032, + "mean_token_accuracy": 0.6785934567451477, + "num_tokens": 113444131.0, + "step": 4383 + }, + { + "epoch": 0.4814408082582912, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5554797649383545, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6841867566108704, + "num_tokens": 113473839.0, + "step": 4384 + }, + { + "epoch": 0.4815506259609049, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.7372238636016846, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6897239685058594, + "num_tokens": 113496652.0, + "step": 4385 + }, + { + "epoch": 0.48166044366351857, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.8646154403686523, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6958281993865967, + "num_tokens": 113517813.0, + "step": 4386 + }, + { + "epoch": 0.4817702613661322, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.542015552520752, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6878117322921753, + "num_tokens": 113544589.0, + "step": 4387 + }, + { + "epoch": 0.48188007906874586, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.3361663818359375, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6929193735122681, + "num_tokens": 113575425.0, + "step": 4388 + }, + { + "epoch": 0.48198989677135956, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.6725237369537354, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7112661600112915, + "num_tokens": 113600542.0, + "step": 4389 + }, + { + "epoch": 0.4820997144739732, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5567686557769775, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7045498490333557, + "num_tokens": 113626938.0, + "step": 4390 + }, + { + "epoch": 0.48220953217658685, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.49018931388855, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6908679008483887, + "num_tokens": 113653288.0, + "step": 4391 + }, + { + "epoch": 0.48231934987920055, + "ewc_loss": 6.735324859619141e-06, + "grad_norm": 2.5030527114868164, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.6972460746765137, + "num_tokens": 113680097.0, + "step": 4392 + }, + { + "epoch": 0.4824291675818142, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.7602102756500244, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6905632019042969, + "num_tokens": 113703073.0, + "step": 4393 + }, + { + "epoch": 0.48253898528442785, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.6305627822875977, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7009766101837158, + "num_tokens": 113727831.0, + "step": 4394 + }, + { + "epoch": 0.4826488029870415, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.7298922538757324, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.699462354183197, + "num_tokens": 113750652.0, + "step": 4395 + }, + { + "epoch": 0.4827586206896552, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.473670482635498, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6972931623458862, + "num_tokens": 113777924.0, + "step": 4396 + }, + { + "epoch": 0.48286843839226884, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.7049648761749268, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.7099705934524536, + "num_tokens": 113801616.0, + "step": 4397 + }, + { + "epoch": 0.4829782560948825, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.8111186027526855, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6823444366455078, + "num_tokens": 113827240.0, + "step": 4398 + }, + { + "epoch": 0.48308807379749613, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 3.0822057723999023, + "learning_rate": 1e-06, + "loss": 0.9096, + "mean_token_accuracy": 0.729796290397644, + "num_tokens": 113847276.0, + "step": 4399 + }, + { + "epoch": 0.48319789150010983, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.578153371810913, + "learning_rate": 1e-06, + "loss": 1.0725, + "mean_token_accuracy": 0.6989308595657349, + "num_tokens": 113876130.0, + "step": 4400 + }, + { + "epoch": 0.4833077092027235, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.706697702407837, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.7015924453735352, + "num_tokens": 113900524.0, + "step": 4401 + }, + { + "epoch": 0.4834175269053371, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.932328462600708, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7127032279968262, + "num_tokens": 113920621.0, + "step": 4402 + }, + { + "epoch": 0.4835273446079508, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.7676446437835693, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7208167910575867, + "num_tokens": 113942610.0, + "step": 4403 + }, + { + "epoch": 0.4836371623105645, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4885106086730957, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6838017702102661, + "num_tokens": 113968029.0, + "step": 4404 + }, + { + "epoch": 0.4837469800131781, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.878170967102051, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7097319960594177, + "num_tokens": 113992927.0, + "step": 4405 + }, + { + "epoch": 0.48385679771579176, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.7923531532287598, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6934129595756531, + "num_tokens": 114018573.0, + "step": 4406 + }, + { + "epoch": 0.48396661541840547, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.5201542377471924, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.682310938835144, + "num_tokens": 114048655.0, + "step": 4407 + }, + { + "epoch": 0.4840764331210191, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.636392831802368, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6923732161521912, + "num_tokens": 114071771.0, + "step": 4408 + }, + { + "epoch": 0.48418625082363276, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.798434257507324, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.7013856768608093, + "num_tokens": 114094334.0, + "step": 4409 + }, + { + "epoch": 0.4842960685262464, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.627150535583496, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7077010273933411, + "num_tokens": 114118168.0, + "step": 4410 + }, + { + "epoch": 0.4844058862288601, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.8938651084899902, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6914677619934082, + "num_tokens": 114138611.0, + "step": 4411 + }, + { + "epoch": 0.48451570393147375, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.709019899368286, + "learning_rate": 1e-06, + "loss": 0.9417, + "mean_token_accuracy": 0.7161895036697388, + "num_tokens": 114160678.0, + "step": 4412 + }, + { + "epoch": 0.4846255216340874, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.8280551433563232, + "learning_rate": 1e-06, + "loss": 0.9081, + "mean_token_accuracy": 0.7270805835723877, + "num_tokens": 114180098.0, + "step": 4413 + }, + { + "epoch": 0.4847353393367011, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.7783071994781494, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7181721329689026, + "num_tokens": 114203210.0, + "step": 4414 + }, + { + "epoch": 0.48484515703931474, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.3936028480529785, + "learning_rate": 1e-06, + "loss": 1.0466, + "mean_token_accuracy": 0.7008171081542969, + "num_tokens": 114232636.0, + "step": 4415 + }, + { + "epoch": 0.4849549747419284, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6571872234344482, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6968852877616882, + "num_tokens": 114256563.0, + "step": 4416 + }, + { + "epoch": 0.48506479244454204, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6448328495025635, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.695410966873169, + "num_tokens": 114282338.0, + "step": 4417 + }, + { + "epoch": 0.48517461014715574, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6827027797698975, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6868957281112671, + "num_tokens": 114306573.0, + "step": 4418 + }, + { + "epoch": 0.4852844278497694, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.750457525253296, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7092591524124146, + "num_tokens": 114330804.0, + "step": 4419 + }, + { + "epoch": 0.48539424555238303, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.482670783996582, + "learning_rate": 1e-06, + "loss": 1.0844, + "mean_token_accuracy": 0.6794313192367554, + "num_tokens": 114362071.0, + "step": 4420 + }, + { + "epoch": 0.48550406325499673, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.674727201461792, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.680073618888855, + "num_tokens": 114386985.0, + "step": 4421 + }, + { + "epoch": 0.4856138809576104, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.738865852355957, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7087401151657104, + "num_tokens": 114408362.0, + "step": 4422 + }, + { + "epoch": 0.485723698660224, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.5053040981292725, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.691307544708252, + "num_tokens": 114437239.0, + "step": 4423 + }, + { + "epoch": 0.48583351636283767, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.534414291381836, + "learning_rate": 1e-06, + "loss": 1.1089, + "mean_token_accuracy": 0.674616277217865, + "num_tokens": 114466495.0, + "step": 4424 + }, + { + "epoch": 0.48594333406545137, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.8522613048553467, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7098670601844788, + "num_tokens": 114486989.0, + "step": 4425 + }, + { + "epoch": 0.486053151768065, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5927722454071045, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.6958091259002686, + "num_tokens": 114509544.0, + "step": 4426 + }, + { + "epoch": 0.48616296947067866, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.5465235710144043, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6852998733520508, + "num_tokens": 114533663.0, + "step": 4427 + }, + { + "epoch": 0.4862727871732923, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.5228638648986816, + "learning_rate": 1e-06, + "loss": 1.0998, + "mean_token_accuracy": 0.6790047287940979, + "num_tokens": 114562909.0, + "step": 4428 + }, + { + "epoch": 0.486382604875906, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.660816192626953, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6832133531570435, + "num_tokens": 114585352.0, + "step": 4429 + }, + { + "epoch": 0.48649242257851966, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.707712173461914, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7135757207870483, + "num_tokens": 114607829.0, + "step": 4430 + }, + { + "epoch": 0.4866022402811333, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.459200143814087, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7171531915664673, + "num_tokens": 114635523.0, + "step": 4431 + }, + { + "epoch": 0.486712057983747, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.541971445083618, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.716064453125, + "num_tokens": 114660260.0, + "step": 4432 + }, + { + "epoch": 0.48682187568636065, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.66487979888916, + "learning_rate": 1e-06, + "loss": 1.0994, + "mean_token_accuracy": 0.6806652545928955, + "num_tokens": 114688684.0, + "step": 4433 + }, + { + "epoch": 0.4869316933889743, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.649540662765503, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7142273783683777, + "num_tokens": 114711677.0, + "step": 4434 + }, + { + "epoch": 0.48704151109158794, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.583615303039551, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6959426403045654, + "num_tokens": 114735640.0, + "step": 4435 + }, + { + "epoch": 0.48715132879420164, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.514782428741455, + "learning_rate": 1e-06, + "loss": 1.0943, + "mean_token_accuracy": 0.6826076507568359, + "num_tokens": 114765799.0, + "step": 4436 + }, + { + "epoch": 0.4872611464968153, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.3952956199645996, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6832512021064758, + "num_tokens": 114795705.0, + "step": 4437 + }, + { + "epoch": 0.48737096419942894, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.5598018169403076, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6902310848236084, + "num_tokens": 114820558.0, + "step": 4438 + }, + { + "epoch": 0.48748078190204264, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.468808889389038, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6967109441757202, + "num_tokens": 114847518.0, + "step": 4439 + }, + { + "epoch": 0.4875905996046563, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.4261467456817627, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6893738508224487, + "num_tokens": 114875666.0, + "step": 4440 + }, + { + "epoch": 0.48770041730726993, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.447235584259033, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.6953226923942566, + "num_tokens": 114906232.0, + "step": 4441 + }, + { + "epoch": 0.4878102350098836, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.8689703941345215, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7027814984321594, + "num_tokens": 114926526.0, + "step": 4442 + }, + { + "epoch": 0.4879200527124973, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.570971727371216, + "learning_rate": 1e-06, + "loss": 1.1234, + "mean_token_accuracy": 0.6832852363586426, + "num_tokens": 114954475.0, + "step": 4443 + }, + { + "epoch": 0.4880298704151109, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.637340784072876, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6866898536682129, + "num_tokens": 114981638.0, + "step": 4444 + }, + { + "epoch": 0.48813968811772457, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.812729597091675, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7141467332839966, + "num_tokens": 115006462.0, + "step": 4445 + }, + { + "epoch": 0.4882495058203382, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.611431837081909, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7128633856773376, + "num_tokens": 115031636.0, + "step": 4446 + }, + { + "epoch": 0.4883593235229519, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 3.274806261062622, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.705495297908783, + "num_tokens": 115056335.0, + "step": 4447 + }, + { + "epoch": 0.48846914122556556, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.816981554031372, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6905924677848816, + "num_tokens": 115078790.0, + "step": 4448 + }, + { + "epoch": 0.4885789589281792, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.77298903465271, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7145056128501892, + "num_tokens": 115101699.0, + "step": 4449 + }, + { + "epoch": 0.4886887766307929, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6058497428894043, + "learning_rate": 1e-06, + "loss": 1.1363, + "mean_token_accuracy": 0.6727920770645142, + "num_tokens": 115128503.0, + "step": 4450 + }, + { + "epoch": 0.48879859433340656, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.320509195327759, + "learning_rate": 1e-06, + "loss": 1.13, + "mean_token_accuracy": 0.680134654045105, + "num_tokens": 115164634.0, + "step": 4451 + }, + { + "epoch": 0.4889084120360202, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.9358110427856445, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.685988187789917, + "num_tokens": 115187246.0, + "step": 4452 + }, + { + "epoch": 0.48901822973863385, + "ewc_loss": 6.794929504394531e-06, + "grad_norm": 2.808537244796753, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.7008221745491028, + "num_tokens": 115213586.0, + "step": 4453 + }, + { + "epoch": 0.48912804744124755, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.778831720352173, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7197306752204895, + "num_tokens": 115235227.0, + "step": 4454 + }, + { + "epoch": 0.4892378651438612, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.327956438064575, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7023615837097168, + "num_tokens": 115263710.0, + "step": 4455 + }, + { + "epoch": 0.48934768284647484, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.5022475719451904, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7087531089782715, + "num_tokens": 115289195.0, + "step": 4456 + }, + { + "epoch": 0.4894575005490885, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.545767307281494, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.6995719075202942, + "num_tokens": 115313974.0, + "step": 4457 + }, + { + "epoch": 0.4895673182517022, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6254122257232666, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6945421695709229, + "num_tokens": 115339114.0, + "step": 4458 + }, + { + "epoch": 0.48967713595431583, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4838860034942627, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6921782493591309, + "num_tokens": 115365289.0, + "step": 4459 + }, + { + "epoch": 0.4897869536569295, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.7041330337524414, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.704036295413971, + "num_tokens": 115387444.0, + "step": 4460 + }, + { + "epoch": 0.4898967713595432, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.798940658569336, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6934524774551392, + "num_tokens": 115409230.0, + "step": 4461 + }, + { + "epoch": 0.4900065890621568, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.708590507507324, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7142537832260132, + "num_tokens": 115430694.0, + "step": 4462 + }, + { + "epoch": 0.4901164067647705, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.673333168029785, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6829386949539185, + "num_tokens": 115456685.0, + "step": 4463 + }, + { + "epoch": 0.4902262244673841, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4506232738494873, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6873267889022827, + "num_tokens": 115483755.0, + "step": 4464 + }, + { + "epoch": 0.4903360421699978, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.88674259185791, + "learning_rate": 1e-06, + "loss": 0.9303, + "mean_token_accuracy": 0.7190826535224915, + "num_tokens": 115502535.0, + "step": 4465 + }, + { + "epoch": 0.49044585987261147, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.5668962001800537, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.697271466255188, + "num_tokens": 115526350.0, + "step": 4466 + }, + { + "epoch": 0.4905556775752251, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.725208282470703, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7102581858634949, + "num_tokens": 115548180.0, + "step": 4467 + }, + { + "epoch": 0.4906654952778388, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4967267513275146, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.691806435585022, + "num_tokens": 115575251.0, + "step": 4468 + }, + { + "epoch": 0.49077531298045246, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.682284355163574, + "learning_rate": 1e-06, + "loss": 0.9131, + "mean_token_accuracy": 0.7242274284362793, + "num_tokens": 115595870.0, + "step": 4469 + }, + { + "epoch": 0.4908851306830661, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5335886478424072, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6846543550491333, + "num_tokens": 115622500.0, + "step": 4470 + }, + { + "epoch": 0.49099494838567975, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.808380126953125, + "learning_rate": 1e-06, + "loss": 1.0927, + "mean_token_accuracy": 0.6853731274604797, + "num_tokens": 115644812.0, + "step": 4471 + }, + { + "epoch": 0.49110476608829345, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5945963859558105, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.710018515586853, + "num_tokens": 115668225.0, + "step": 4472 + }, + { + "epoch": 0.4912145837909071, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.741750955581665, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7184934020042419, + "num_tokens": 115690591.0, + "step": 4473 + }, + { + "epoch": 0.49132440149352075, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.781942367553711, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6826435327529907, + "num_tokens": 115715069.0, + "step": 4474 + }, + { + "epoch": 0.4914342191961344, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7118027210235596, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6986305713653564, + "num_tokens": 115737035.0, + "step": 4475 + }, + { + "epoch": 0.4915440368987481, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.581883668899536, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.7016075849533081, + "num_tokens": 115762455.0, + "step": 4476 + }, + { + "epoch": 0.49165385460136174, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6988308429718018, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7029262781143188, + "num_tokens": 115786117.0, + "step": 4477 + }, + { + "epoch": 0.4917636723039754, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.728713274002075, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7228058576583862, + "num_tokens": 115809341.0, + "step": 4478 + }, + { + "epoch": 0.4918734900065891, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7951364517211914, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7015244960784912, + "num_tokens": 115832269.0, + "step": 4479 + }, + { + "epoch": 0.49198330770920273, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5238213539123535, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6979883909225464, + "num_tokens": 115859033.0, + "step": 4480 + }, + { + "epoch": 0.4920931254118164, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7524213790893555, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7036548256874084, + "num_tokens": 115881543.0, + "step": 4481 + }, + { + "epoch": 0.49220294311443, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.514925479888916, + "learning_rate": 1e-06, + "loss": 1.136, + "mean_token_accuracy": 0.6793000102043152, + "num_tokens": 115908358.0, + "step": 4482 + }, + { + "epoch": 0.4923127608170437, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5028984546661377, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6998386979103088, + "num_tokens": 115934190.0, + "step": 4483 + }, + { + "epoch": 0.4924225785196574, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5957083702087402, + "learning_rate": 1e-06, + "loss": 1.1002, + "mean_token_accuracy": 0.6790791749954224, + "num_tokens": 115958843.0, + "step": 4484 + }, + { + "epoch": 0.492532396222271, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6152708530426025, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6820173263549805, + "num_tokens": 115984943.0, + "step": 4485 + }, + { + "epoch": 0.49264221392488466, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4152636528015137, + "learning_rate": 1e-06, + "loss": 1.1603, + "mean_token_accuracy": 0.6749252080917358, + "num_tokens": 116015263.0, + "step": 4486 + }, + { + "epoch": 0.49275203162749837, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5049140453338623, + "learning_rate": 1e-06, + "loss": 1.1055, + "mean_token_accuracy": 0.6872721314430237, + "num_tokens": 116042471.0, + "step": 4487 + }, + { + "epoch": 0.492861849330112, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.604022741317749, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.7011719942092896, + "num_tokens": 116067499.0, + "step": 4488 + }, + { + "epoch": 0.49297166703272566, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.758759021759033, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7147759199142456, + "num_tokens": 116088440.0, + "step": 4489 + }, + { + "epoch": 0.49308148473533936, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.820415496826172, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6833173036575317, + "num_tokens": 116111112.0, + "step": 4490 + }, + { + "epoch": 0.493191302437953, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.687870740890503, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6976696848869324, + "num_tokens": 116135455.0, + "step": 4491 + }, + { + "epoch": 0.49330112014056665, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6498923301696777, + "learning_rate": 1e-06, + "loss": 1.0799, + "mean_token_accuracy": 0.6849821209907532, + "num_tokens": 116160890.0, + "step": 4492 + }, + { + "epoch": 0.4934109378431803, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4514427185058594, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7105906009674072, + "num_tokens": 116187808.0, + "step": 4493 + }, + { + "epoch": 0.493520755545794, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.374544858932495, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6843301653862, + "num_tokens": 116217838.0, + "step": 4494 + }, + { + "epoch": 0.49363057324840764, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4495859146118164, + "learning_rate": 1e-06, + "loss": 1.1186, + "mean_token_accuracy": 0.6846365332603455, + "num_tokens": 116246981.0, + "step": 4495 + }, + { + "epoch": 0.4937403909510213, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5079174041748047, + "learning_rate": 1e-06, + "loss": 1.1262, + "mean_token_accuracy": 0.674972653388977, + "num_tokens": 116276532.0, + "step": 4496 + }, + { + "epoch": 0.493850208653635, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.7289178371429443, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7054879665374756, + "num_tokens": 116302544.0, + "step": 4497 + }, + { + "epoch": 0.49396002635624864, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.711927890777588, + "learning_rate": 1e-06, + "loss": 1.1555, + "mean_token_accuracy": 0.6705506443977356, + "num_tokens": 116326255.0, + "step": 4498 + }, + { + "epoch": 0.4940698440588623, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5740816593170166, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7015832662582397, + "num_tokens": 116351717.0, + "step": 4499 + }, + { + "epoch": 0.49417966176147593, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.422778606414795, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.6868720650672913, + "num_tokens": 116381049.0, + "step": 4500 + }, + { + "epoch": 0.49428947946408963, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6397430896759033, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.710573136806488, + "num_tokens": 116404558.0, + "step": 4501 + }, + { + "epoch": 0.4943992971667033, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5485587120056152, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6905783414840698, + "num_tokens": 116430821.0, + "step": 4502 + }, + { + "epoch": 0.4945091148693169, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5298163890838623, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7103188633918762, + "num_tokens": 116456843.0, + "step": 4503 + }, + { + "epoch": 0.49461893257193057, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.3474457263946533, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.698795735836029, + "num_tokens": 116488522.0, + "step": 4504 + }, + { + "epoch": 0.49472875027454427, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.3259434700012207, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6859235763549805, + "num_tokens": 116518860.0, + "step": 4505 + }, + { + "epoch": 0.4948385679771579, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6689682006835938, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.711439847946167, + "num_tokens": 116545374.0, + "step": 4506 + }, + { + "epoch": 0.49494838567977156, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.505396842956543, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6997262239456177, + "num_tokens": 116572772.0, + "step": 4507 + }, + { + "epoch": 0.49505820338238526, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4609577655792236, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6911938190460205, + "num_tokens": 116601968.0, + "step": 4508 + }, + { + "epoch": 0.4951680210849989, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.7515997886657715, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6960680484771729, + "num_tokens": 116624913.0, + "step": 4509 + }, + { + "epoch": 0.49527783878761256, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7881059646606445, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.694068431854248, + "num_tokens": 116646611.0, + "step": 4510 + }, + { + "epoch": 0.4953876564902262, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6805224418640137, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.691834568977356, + "num_tokens": 116669267.0, + "step": 4511 + }, + { + "epoch": 0.4954974741928399, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6186389923095703, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7058632373809814, + "num_tokens": 116694288.0, + "step": 4512 + }, + { + "epoch": 0.49560729189545355, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5962162017822266, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7098162174224854, + "num_tokens": 116719128.0, + "step": 4513 + }, + { + "epoch": 0.4957171095980672, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.698178291320801, + "learning_rate": 1e-06, + "loss": 1.1294, + "mean_token_accuracy": 0.6839249134063721, + "num_tokens": 116745096.0, + "step": 4514 + }, + { + "epoch": 0.4958269273006809, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.902764320373535, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7099928855895996, + "num_tokens": 116763826.0, + "step": 4515 + }, + { + "epoch": 0.49593674500329454, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4325625896453857, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6991997957229614, + "num_tokens": 116795641.0, + "step": 4516 + }, + { + "epoch": 0.4960465627059082, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6183457374572754, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6850451231002808, + "num_tokens": 116822911.0, + "step": 4517 + }, + { + "epoch": 0.49615638040852184, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6861181259155273, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.6984358429908752, + "num_tokens": 116846246.0, + "step": 4518 + }, + { + "epoch": 0.49626619811113554, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.358508348464966, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6967713832855225, + "num_tokens": 116874624.0, + "step": 4519 + }, + { + "epoch": 0.4963760158137492, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5034408569335938, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7120565176010132, + "num_tokens": 116899231.0, + "step": 4520 + }, + { + "epoch": 0.49648583351636283, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.676345109939575, + "learning_rate": 1e-06, + "loss": 1.1378, + "mean_token_accuracy": 0.6706727147102356, + "num_tokens": 116924433.0, + "step": 4521 + }, + { + "epoch": 0.4965956512189765, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.433450698852539, + "learning_rate": 1e-06, + "loss": 1.1132, + "mean_token_accuracy": 0.6817870140075684, + "num_tokens": 116954796.0, + "step": 4522 + }, + { + "epoch": 0.4967054689215902, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.3099923133850098, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.6999580264091492, + "num_tokens": 116984925.0, + "step": 4523 + }, + { + "epoch": 0.4968152866242038, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4439358711242676, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6926165819168091, + "num_tokens": 117012418.0, + "step": 4524 + }, + { + "epoch": 0.49692510432681747, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.3252475261688232, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7118384838104248, + "num_tokens": 117040062.0, + "step": 4525 + }, + { + "epoch": 0.49703492202943117, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.295239210128784, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6840680241584778, + "num_tokens": 117069854.0, + "step": 4526 + }, + { + "epoch": 0.4971447397320448, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.362337827682495, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.6958834528923035, + "num_tokens": 117098015.0, + "step": 4527 + }, + { + "epoch": 0.49725455743465846, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6945321559906006, + "learning_rate": 1e-06, + "loss": 1.1197, + "mean_token_accuracy": 0.6789696216583252, + "num_tokens": 117127797.0, + "step": 4528 + }, + { + "epoch": 0.4973643751372721, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.769785165786743, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6917160153388977, + "num_tokens": 117147759.0, + "step": 4529 + }, + { + "epoch": 0.4974741928398858, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.561103582382202, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6892656087875366, + "num_tokens": 117171512.0, + "step": 4530 + }, + { + "epoch": 0.49758401054249946, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6759719848632812, + "learning_rate": 1e-06, + "loss": 1.1119, + "mean_token_accuracy": 0.6861639022827148, + "num_tokens": 117196904.0, + "step": 4531 + }, + { + "epoch": 0.4976938282451131, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.920142889022827, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7025715112686157, + "num_tokens": 117218115.0, + "step": 4532 + }, + { + "epoch": 0.49780364594772675, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.460299253463745, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7071804404258728, + "num_tokens": 117244201.0, + "step": 4533 + }, + { + "epoch": 0.49791346365034045, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.566901922225952, + "learning_rate": 1e-06, + "loss": 0.9085, + "mean_token_accuracy": 0.7321751713752747, + "num_tokens": 117267186.0, + "step": 4534 + }, + { + "epoch": 0.4980232813529541, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4233806133270264, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6917718052864075, + "num_tokens": 117297522.0, + "step": 4535 + }, + { + "epoch": 0.49813309905556774, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6093413829803467, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6893110275268555, + "num_tokens": 117323164.0, + "step": 4536 + }, + { + "epoch": 0.49824291675818144, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7562642097473145, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6947790384292603, + "num_tokens": 117347079.0, + "step": 4537 + }, + { + "epoch": 0.4983527344607951, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5792760848999023, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7030759453773499, + "num_tokens": 117373678.0, + "step": 4538 + }, + { + "epoch": 0.49846255216340873, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.435969352722168, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.6874386668205261, + "num_tokens": 117400721.0, + "step": 4539 + }, + { + "epoch": 0.4985723698660224, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.555055856704712, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7094461917877197, + "num_tokens": 117425568.0, + "step": 4540 + }, + { + "epoch": 0.4986821875686361, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.777947425842285, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.6958357691764832, + "num_tokens": 117448642.0, + "step": 4541 + }, + { + "epoch": 0.4987920052712497, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.8437368869781494, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6988271474838257, + "num_tokens": 117471867.0, + "step": 4542 + }, + { + "epoch": 0.4989018229738634, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.2912399768829346, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6966186761856079, + "num_tokens": 117503893.0, + "step": 4543 + }, + { + "epoch": 0.4990116406764771, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4724464416503906, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.689734935760498, + "num_tokens": 117532300.0, + "step": 4544 + }, + { + "epoch": 0.4991214583790907, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6114983558654785, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.7082524299621582, + "num_tokens": 117557057.0, + "step": 4545 + }, + { + "epoch": 0.49923127608170437, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5855207443237305, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6907691955566406, + "num_tokens": 117586763.0, + "step": 4546 + }, + { + "epoch": 0.499341093784318, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5045230388641357, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.7033380270004272, + "num_tokens": 117611783.0, + "step": 4547 + }, + { + "epoch": 0.4994509114869317, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.513439178466797, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7234747409820557, + "num_tokens": 117638085.0, + "step": 4548 + }, + { + "epoch": 0.49956072918954536, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.447432279586792, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6914454698562622, + "num_tokens": 117669103.0, + "step": 4549 + }, + { + "epoch": 0.499670546892159, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6361477375030518, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6848383545875549, + "num_tokens": 117693830.0, + "step": 4550 + }, + { + "epoch": 0.49978036459477265, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.7805960178375244, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6894824504852295, + "num_tokens": 117714230.0, + "step": 4551 + }, + { + "epoch": 0.49989018229738635, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.3911221027374268, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.6893771290779114, + "num_tokens": 117747833.0, + "step": 4552 + }, + { + "epoch": 0.5, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.810631275177002, + "learning_rate": 1e-06, + "loss": 1.1067, + "mean_token_accuracy": 0.680982768535614, + "num_tokens": 117773052.0, + "step": 4553 + }, + { + "epoch": 0.5001098177026136, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6003968715667725, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7023947238922119, + "num_tokens": 117798234.0, + "step": 4554 + }, + { + "epoch": 0.5002196354052273, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 3.0220766067504883, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7047984600067139, + "num_tokens": 117820361.0, + "step": 4555 + }, + { + "epoch": 0.5003294531078409, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5499634742736816, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6851857900619507, + "num_tokens": 117846772.0, + "step": 4556 + }, + { + "epoch": 0.5004392708104547, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.2918317317962646, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.686922013759613, + "num_tokens": 117877864.0, + "step": 4557 + }, + { + "epoch": 0.5005490885130683, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5831127166748047, + "learning_rate": 1e-06, + "loss": 1.1065, + "mean_token_accuracy": 0.6775292158126831, + "num_tokens": 117905908.0, + "step": 4558 + }, + { + "epoch": 0.500658906215682, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.8546791076660156, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6959913969039917, + "num_tokens": 117931453.0, + "step": 4559 + }, + { + "epoch": 0.5007687239182956, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.891745090484619, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7046184539794922, + "num_tokens": 117953672.0, + "step": 4560 + }, + { + "epoch": 0.5008785416209093, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.493448495864868, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6831544637680054, + "num_tokens": 117979254.0, + "step": 4561 + }, + { + "epoch": 0.5009883593235229, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.508833169937134, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7016869783401489, + "num_tokens": 118007249.0, + "step": 4562 + }, + { + "epoch": 0.5010981770261366, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.457231044769287, + "learning_rate": 1e-06, + "loss": 1.1147, + "mean_token_accuracy": 0.6799542903900146, + "num_tokens": 118038228.0, + "step": 4563 + }, + { + "epoch": 0.5012079947287503, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4323172569274902, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.7007662057876587, + "num_tokens": 118064982.0, + "step": 4564 + }, + { + "epoch": 0.501317812431364, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6411967277526855, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.7009264826774597, + "num_tokens": 118088105.0, + "step": 4565 + }, + { + "epoch": 0.5014276301339776, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.404136896133423, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6927692294120789, + "num_tokens": 118123176.0, + "step": 4566 + }, + { + "epoch": 0.5015374478365913, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5532419681549072, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6968870162963867, + "num_tokens": 118150578.0, + "step": 4567 + }, + { + "epoch": 0.5016472655392049, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.8508002758026123, + "learning_rate": 1e-06, + "loss": 0.9555, + "mean_token_accuracy": 0.7201758027076721, + "num_tokens": 118177358.0, + "step": 4568 + }, + { + "epoch": 0.5017570832418186, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.4687271118164062, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6952335834503174, + "num_tokens": 118207315.0, + "step": 4569 + }, + { + "epoch": 0.5018669009444322, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.528505802154541, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.6975823640823364, + "num_tokens": 118231938.0, + "step": 4570 + }, + { + "epoch": 0.5019767186470458, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.5054290294647217, + "learning_rate": 1e-06, + "loss": 1.1508, + "mean_token_accuracy": 0.6672825217247009, + "num_tokens": 118264767.0, + "step": 4571 + }, + { + "epoch": 0.5020865363496596, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.7181131839752197, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6843450665473938, + "num_tokens": 118287546.0, + "step": 4572 + }, + { + "epoch": 0.5021963540522733, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.725795269012451, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7070838809013367, + "num_tokens": 118310519.0, + "step": 4573 + }, + { + "epoch": 0.5023061717548869, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5719146728515625, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.6977043747901917, + "num_tokens": 118335584.0, + "step": 4574 + }, + { + "epoch": 0.5024159894575005, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.9465789794921875, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7129853963851929, + "num_tokens": 118355882.0, + "step": 4575 + }, + { + "epoch": 0.5025258071601142, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7815840244293213, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6936842203140259, + "num_tokens": 118379758.0, + "step": 4576 + }, + { + "epoch": 0.5026356248627278, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4450018405914307, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6904523372650146, + "num_tokens": 118409998.0, + "step": 4577 + }, + { + "epoch": 0.5027454425653415, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4106075763702393, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.69733726978302, + "num_tokens": 118437895.0, + "step": 4578 + }, + { + "epoch": 0.5028552602679552, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4429752826690674, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6908499002456665, + "num_tokens": 118466254.0, + "step": 4579 + }, + { + "epoch": 0.5029650779705689, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.7092154026031494, + "learning_rate": 1e-06, + "loss": 1.1178, + "mean_token_accuracy": 0.682018518447876, + "num_tokens": 118491266.0, + "step": 4580 + }, + { + "epoch": 0.5030748956731825, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.866173505783081, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.6829910278320312, + "num_tokens": 118513453.0, + "step": 4581 + }, + { + "epoch": 0.5031847133757962, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.761324882507324, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6852121949195862, + "num_tokens": 118536240.0, + "step": 4582 + }, + { + "epoch": 0.5032945310784098, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.3487696647644043, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6916443705558777, + "num_tokens": 118566445.0, + "step": 4583 + }, + { + "epoch": 0.5034043487810235, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.552520513534546, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6902065277099609, + "num_tokens": 118591213.0, + "step": 4584 + }, + { + "epoch": 0.5035141664836371, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.505511999130249, + "learning_rate": 1e-06, + "loss": 1.1751, + "mean_token_accuracy": 0.6678117513656616, + "num_tokens": 118621747.0, + "step": 4585 + }, + { + "epoch": 0.5036239841862509, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.0452306270599365, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6919952630996704, + "num_tokens": 118660496.0, + "step": 4586 + }, + { + "epoch": 0.5037338018888645, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6376378536224365, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7091764211654663, + "num_tokens": 118682445.0, + "step": 4587 + }, + { + "epoch": 0.5038436195914782, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.6210176944732666, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6975570321083069, + "num_tokens": 118709958.0, + "step": 4588 + }, + { + "epoch": 0.5039534372940918, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4884698390960693, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.703590452671051, + "num_tokens": 118736291.0, + "step": 4589 + }, + { + "epoch": 0.5040632549967055, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4758617877960205, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6858199834823608, + "num_tokens": 118763755.0, + "step": 4590 + }, + { + "epoch": 0.5041730726993191, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6963393688201904, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7073805332183838, + "num_tokens": 118785592.0, + "step": 4591 + }, + { + "epoch": 0.5042828904019327, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.342545747756958, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6959929466247559, + "num_tokens": 118815809.0, + "step": 4592 + }, + { + "epoch": 0.5043927081045465, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4390885829925537, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.699302077293396, + "num_tokens": 118844118.0, + "step": 4593 + }, + { + "epoch": 0.5045025258071602, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.185262441635132, + "learning_rate": 1e-06, + "loss": 1.1865, + "mean_token_accuracy": 0.6567341089248657, + "num_tokens": 118878278.0, + "step": 4594 + }, + { + "epoch": 0.5046123435097738, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.4177041053771973, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6895297765731812, + "num_tokens": 118904888.0, + "step": 4595 + }, + { + "epoch": 0.5047221612123874, + "ewc_loss": 6.8247318267822266e-06, + "grad_norm": 2.3973259925842285, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.697137713432312, + "num_tokens": 118936516.0, + "step": 4596 + }, + { + "epoch": 0.5048319789150011, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.475949764251709, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6840850114822388, + "num_tokens": 118965982.0, + "step": 4597 + }, + { + "epoch": 0.5049417966176147, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.62756085395813, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7124732136726379, + "num_tokens": 118989660.0, + "step": 4598 + }, + { + "epoch": 0.5050516143202284, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6402077674865723, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7103357911109924, + "num_tokens": 119014512.0, + "step": 4599 + }, + { + "epoch": 0.505161432022842, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5943524837493896, + "learning_rate": 1e-06, + "loss": 1.1106, + "mean_token_accuracy": 0.6761634349822998, + "num_tokens": 119046114.0, + "step": 4600 + }, + { + "epoch": 0.5052712497254558, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.822469472885132, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6957314610481262, + "num_tokens": 119069967.0, + "step": 4601 + }, + { + "epoch": 0.5053810674280694, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4740264415740967, + "learning_rate": 1e-06, + "loss": 1.1287, + "mean_token_accuracy": 0.6781313419342041, + "num_tokens": 119102712.0, + "step": 4602 + }, + { + "epoch": 0.5054908851306831, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5370774269104004, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6941428184509277, + "num_tokens": 119131270.0, + "step": 4603 + }, + { + "epoch": 0.5056007028332967, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 3.7774150371551514, + "learning_rate": 1e-06, + "loss": 0.9365, + "mean_token_accuracy": 0.7140979766845703, + "num_tokens": 119147580.0, + "step": 4604 + }, + { + "epoch": 0.5057105205359104, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.45255184173584, + "learning_rate": 1e-06, + "loss": 1.1145, + "mean_token_accuracy": 0.6766326427459717, + "num_tokens": 119176404.0, + "step": 4605 + }, + { + "epoch": 0.505820338238524, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.556267738342285, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6938247680664062, + "num_tokens": 119203393.0, + "step": 4606 + }, + { + "epoch": 0.5059301559411377, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4201903343200684, + "learning_rate": 1e-06, + "loss": 1.1444, + "mean_token_accuracy": 0.6717945337295532, + "num_tokens": 119231833.0, + "step": 4607 + }, + { + "epoch": 0.5060399736437514, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6213724613189697, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6967417001724243, + "num_tokens": 119256386.0, + "step": 4608 + }, + { + "epoch": 0.5061497913463651, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5622377395629883, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6795555949211121, + "num_tokens": 119285024.0, + "step": 4609 + }, + { + "epoch": 0.5062596090489787, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.504472255706787, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7181235551834106, + "num_tokens": 119308733.0, + "step": 4610 + }, + { + "epoch": 0.5063694267515924, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.792970895767212, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6909064054489136, + "num_tokens": 119332389.0, + "step": 4611 + }, + { + "epoch": 0.506479244454206, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.587744951248169, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6903318762779236, + "num_tokens": 119360140.0, + "step": 4612 + }, + { + "epoch": 0.5065890621568196, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.917149066925049, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.7168409824371338, + "num_tokens": 119381098.0, + "step": 4613 + }, + { + "epoch": 0.5066988798594333, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5180418491363525, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7252299785614014, + "num_tokens": 119406309.0, + "step": 4614 + }, + { + "epoch": 0.506808697562047, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.778223991394043, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7199451923370361, + "num_tokens": 119429642.0, + "step": 4615 + }, + { + "epoch": 0.5069185152646607, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.2643423080444336, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6893464922904968, + "num_tokens": 119461475.0, + "step": 4616 + }, + { + "epoch": 0.5070283329672743, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.7227399349212646, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7075541019439697, + "num_tokens": 119485600.0, + "step": 4617 + }, + { + "epoch": 0.507138150669888, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.569613218307495, + "learning_rate": 1e-06, + "loss": 1.1343, + "mean_token_accuracy": 0.6790732741355896, + "num_tokens": 119511824.0, + "step": 4618 + }, + { + "epoch": 0.5072479683725016, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.716423273086548, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7057687044143677, + "num_tokens": 119533133.0, + "step": 4619 + }, + { + "epoch": 0.5073577860751153, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.624241828918457, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6983497142791748, + "num_tokens": 119557080.0, + "step": 4620 + }, + { + "epoch": 0.5074676037777289, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.8373265266418457, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6873271465301514, + "num_tokens": 119577569.0, + "step": 4621 + }, + { + "epoch": 0.5075774214803427, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.723288059234619, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6890105605125427, + "num_tokens": 119605645.0, + "step": 4622 + }, + { + "epoch": 0.5076872391829563, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.8341760635375977, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.704068660736084, + "num_tokens": 119626742.0, + "step": 4623 + }, + { + "epoch": 0.50779705688557, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.675292491912842, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.6968342065811157, + "num_tokens": 119648331.0, + "step": 4624 + }, + { + "epoch": 0.5079068745881836, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.635157585144043, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7029520273208618, + "num_tokens": 119671073.0, + "step": 4625 + }, + { + "epoch": 0.5080166922907973, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.9742355346679688, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7010034918785095, + "num_tokens": 119697145.0, + "step": 4626 + }, + { + "epoch": 0.5081265099934109, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.858443260192871, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6891452670097351, + "num_tokens": 119719279.0, + "step": 4627 + }, + { + "epoch": 0.5082363276960246, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.4761877059936523, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7043228149414062, + "num_tokens": 119746924.0, + "step": 4628 + }, + { + "epoch": 0.5083461453986382, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.9723739624023438, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.6778639554977417, + "num_tokens": 119769037.0, + "step": 4629 + }, + { + "epoch": 0.508455963101252, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 5.325829982757568, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6837462186813354, + "num_tokens": 119797640.0, + "step": 4630 + }, + { + "epoch": 0.5085657808038656, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5641586780548096, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7097357511520386, + "num_tokens": 119821460.0, + "step": 4631 + }, + { + "epoch": 0.5086755985064793, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.451198101043701, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6819056868553162, + "num_tokens": 119850542.0, + "step": 4632 + }, + { + "epoch": 0.5087854162090929, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.3776962757110596, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7029010057449341, + "num_tokens": 119877816.0, + "step": 4633 + }, + { + "epoch": 0.5088952339117065, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.333221912384033, + "learning_rate": 1e-06, + "loss": 1.0992, + "mean_token_accuracy": 0.6797635555267334, + "num_tokens": 119907607.0, + "step": 4634 + }, + { + "epoch": 0.5090050516143202, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.574275016784668, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6943299770355225, + "num_tokens": 119933810.0, + "step": 4635 + }, + { + "epoch": 0.5091148693169338, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.7472355365753174, + "learning_rate": 1e-06, + "loss": 1.1006, + "mean_token_accuracy": 0.6753753423690796, + "num_tokens": 119960256.0, + "step": 4636 + }, + { + "epoch": 0.5092246870195476, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.483522891998291, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6968604922294617, + "num_tokens": 119988437.0, + "step": 4637 + }, + { + "epoch": 0.5093345047221612, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.739089250564575, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.6996251344680786, + "num_tokens": 120012553.0, + "step": 4638 + }, + { + "epoch": 0.5094443224247749, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5899581909179688, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6957510709762573, + "num_tokens": 120037367.0, + "step": 4639 + }, + { + "epoch": 0.5095541401273885, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.9549753665924072, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7165318727493286, + "num_tokens": 120056816.0, + "step": 4640 + }, + { + "epoch": 0.5096639578300022, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.311286687850952, + "learning_rate": 1e-06, + "loss": 1.1111, + "mean_token_accuracy": 0.6788731813430786, + "num_tokens": 120090690.0, + "step": 4641 + }, + { + "epoch": 0.5097737755326158, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.566073417663574, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7016488313674927, + "num_tokens": 120114819.0, + "step": 4642 + }, + { + "epoch": 0.5098835932352295, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4622273445129395, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7033382058143616, + "num_tokens": 120144814.0, + "step": 4643 + }, + { + "epoch": 0.5099934109378432, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6940758228302, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7170357704162598, + "num_tokens": 120168876.0, + "step": 4644 + }, + { + "epoch": 0.5101032286404569, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.62882399559021, + "learning_rate": 1e-06, + "loss": 1.1132, + "mean_token_accuracy": 0.6778557300567627, + "num_tokens": 120196920.0, + "step": 4645 + }, + { + "epoch": 0.5102130463430705, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.666203260421753, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.6927975416183472, + "num_tokens": 120222102.0, + "step": 4646 + }, + { + "epoch": 0.5103228640456842, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.5336639881134033, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.699151337146759, + "num_tokens": 120247712.0, + "step": 4647 + }, + { + "epoch": 0.5104326817482978, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.8764939308166504, + "learning_rate": 1e-06, + "loss": 1.1034, + "mean_token_accuracy": 0.6856708526611328, + "num_tokens": 120281315.0, + "step": 4648 + }, + { + "epoch": 0.5105424994509115, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.428110122680664, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6992465257644653, + "num_tokens": 120309377.0, + "step": 4649 + }, + { + "epoch": 0.5106523171535251, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4385733604431152, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6839288473129272, + "num_tokens": 120337949.0, + "step": 4650 + }, + { + "epoch": 0.5107621348561389, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.48305344581604, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.702327311038971, + "num_tokens": 120362736.0, + "step": 4651 + }, + { + "epoch": 0.5108719525587525, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.3881888389587402, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.687142550945282, + "num_tokens": 120394003.0, + "step": 4652 + }, + { + "epoch": 0.5109817702613662, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.6014797687530518, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.690554141998291, + "num_tokens": 120417078.0, + "step": 4653 + }, + { + "epoch": 0.5110915879639798, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.606487512588501, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7189843654632568, + "num_tokens": 120440605.0, + "step": 4654 + }, + { + "epoch": 0.5112014056665934, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.937635660171509, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7027092576026917, + "num_tokens": 120463428.0, + "step": 4655 + }, + { + "epoch": 0.5113112233692071, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.454131841659546, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6987610459327698, + "num_tokens": 120493073.0, + "step": 4656 + }, + { + "epoch": 0.5114210410718207, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.599743604660034, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6936926245689392, + "num_tokens": 120518143.0, + "step": 4657 + }, + { + "epoch": 0.5115308587744345, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 3.356635570526123, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7004417181015015, + "num_tokens": 120542359.0, + "step": 4658 + }, + { + "epoch": 0.5116406764770481, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.473039150238037, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6995112299919128, + "num_tokens": 120571395.0, + "step": 4659 + }, + { + "epoch": 0.5117504941796618, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.207871913909912, + "learning_rate": 1e-06, + "loss": 1.1068, + "mean_token_accuracy": 0.678541362285614, + "num_tokens": 120607912.0, + "step": 4660 + }, + { + "epoch": 0.5118603118822754, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.328245162963867, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6972035765647888, + "num_tokens": 120639223.0, + "step": 4661 + }, + { + "epoch": 0.5119701295848891, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.8937668800354004, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7032944560050964, + "num_tokens": 120659595.0, + "step": 4662 + }, + { + "epoch": 0.5120799472875027, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.3925998210906982, + "learning_rate": 1e-06, + "loss": 1.1021, + "mean_token_accuracy": 0.6836991310119629, + "num_tokens": 120694465.0, + "step": 4663 + }, + { + "epoch": 0.5121897649901164, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5405611991882324, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.695379376411438, + "num_tokens": 120722417.0, + "step": 4664 + }, + { + "epoch": 0.51229958269273, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4139244556427, + "learning_rate": 1e-06, + "loss": 1.1246, + "mean_token_accuracy": 0.678048312664032, + "num_tokens": 120750601.0, + "step": 4665 + }, + { + "epoch": 0.5124094003953438, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.466053009033203, + "learning_rate": 1e-06, + "loss": 1.1165, + "mean_token_accuracy": 0.6742451190948486, + "num_tokens": 120781378.0, + "step": 4666 + }, + { + "epoch": 0.5125192180979574, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.589505672454834, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6896356344223022, + "num_tokens": 120807616.0, + "step": 4667 + }, + { + "epoch": 0.5126290358005711, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.6513590812683105, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6911183595657349, + "num_tokens": 120832422.0, + "step": 4668 + }, + { + "epoch": 0.5127388535031847, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.6592483520507812, + "learning_rate": 1e-06, + "loss": 1.1197, + "mean_token_accuracy": 0.672491192817688, + "num_tokens": 120859714.0, + "step": 4669 + }, + { + "epoch": 0.5128486712057984, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.7963709831237793, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7102929353713989, + "num_tokens": 120883835.0, + "step": 4670 + }, + { + "epoch": 0.512958488908412, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.666686773300171, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.6917560696601868, + "num_tokens": 120907148.0, + "step": 4671 + }, + { + "epoch": 0.5130683066110256, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.3494088649749756, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.698931097984314, + "num_tokens": 120937197.0, + "step": 4672 + }, + { + "epoch": 0.5131781243136394, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.6527438163757324, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7069479823112488, + "num_tokens": 120960227.0, + "step": 4673 + }, + { + "epoch": 0.513287942016253, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5250468254089355, + "learning_rate": 1e-06, + "loss": 1.1166, + "mean_token_accuracy": 0.6947803497314453, + "num_tokens": 120987671.0, + "step": 4674 + }, + { + "epoch": 0.5133977597188667, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.577960729598999, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7135735154151917, + "num_tokens": 121014518.0, + "step": 4675 + }, + { + "epoch": 0.5135075774214803, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 3.2367992401123047, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6867243647575378, + "num_tokens": 121046518.0, + "step": 4676 + }, + { + "epoch": 0.513617395124094, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.672671318054199, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7019527554512024, + "num_tokens": 121070381.0, + "step": 4677 + }, + { + "epoch": 0.5137272128267076, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5330512523651123, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7037765979766846, + "num_tokens": 121093702.0, + "step": 4678 + }, + { + "epoch": 0.5138370305293213, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 3.1636219024658203, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.691042423248291, + "num_tokens": 121114919.0, + "step": 4679 + }, + { + "epoch": 0.513946848231935, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.679845094680786, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6950345635414124, + "num_tokens": 121137886.0, + "step": 4680 + }, + { + "epoch": 0.5140566659345487, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5642473697662354, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6980422139167786, + "num_tokens": 121164767.0, + "step": 4681 + }, + { + "epoch": 0.5141664836371623, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.475923538208008, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6892167329788208, + "num_tokens": 121194080.0, + "step": 4682 + }, + { + "epoch": 0.514276301339776, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.310556650161743, + "learning_rate": 1e-06, + "loss": 1.1624, + "mean_token_accuracy": 0.6653686761856079, + "num_tokens": 121226531.0, + "step": 4683 + }, + { + "epoch": 0.5143861190423896, + "ewc_loss": 6.854534149169922e-06, + "grad_norm": 2.9189112186431885, + "learning_rate": 1e-06, + "loss": 1.1488, + "mean_token_accuracy": 0.6721704602241516, + "num_tokens": 121254072.0, + "step": 4684 + }, + { + "epoch": 0.5144959367450033, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 3.130255937576294, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6965276002883911, + "num_tokens": 121279731.0, + "step": 4685 + }, + { + "epoch": 0.5146057544476169, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.1923065185546875, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7152167558670044, + "num_tokens": 121312266.0, + "step": 4686 + }, + { + "epoch": 0.5147155721502307, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5777063369750977, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6945393681526184, + "num_tokens": 121337329.0, + "step": 4687 + }, + { + "epoch": 0.5148253898528443, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.2575488090515137, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6890866160392761, + "num_tokens": 121369123.0, + "step": 4688 + }, + { + "epoch": 0.514935207555458, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.665668487548828, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.7030329704284668, + "num_tokens": 121391602.0, + "step": 4689 + }, + { + "epoch": 0.5150450252580716, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4310050010681152, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6818112730979919, + "num_tokens": 121418720.0, + "step": 4690 + }, + { + "epoch": 0.5151548429606853, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.415099620819092, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7147308588027954, + "num_tokens": 121444116.0, + "step": 4691 + }, + { + "epoch": 0.5152646606632989, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4099323749542236, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6967654824256897, + "num_tokens": 121474478.0, + "step": 4692 + }, + { + "epoch": 0.5153744783659125, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.6644930839538574, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6926636695861816, + "num_tokens": 121502744.0, + "step": 4693 + }, + { + "epoch": 0.5154842960685262, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.4193344116210938, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7192959785461426, + "num_tokens": 121530615.0, + "step": 4694 + }, + { + "epoch": 0.51559411377114, + "ewc_loss": 6.884336471557617e-06, + "grad_norm": 2.3677256107330322, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.697431206703186, + "num_tokens": 121559108.0, + "step": 4695 + }, + { + "epoch": 0.5157039314737536, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.322756052017212, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.7035385966300964, + "num_tokens": 121588392.0, + "step": 4696 + }, + { + "epoch": 0.5158137491763672, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.2913782596588135, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7056875228881836, + "num_tokens": 121617083.0, + "step": 4697 + }, + { + "epoch": 0.5159235668789809, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.7057597637176514, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.7083539366722107, + "num_tokens": 121641301.0, + "step": 4698 + }, + { + "epoch": 0.5160333845815945, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.3122477531433105, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7020007371902466, + "num_tokens": 121671011.0, + "step": 4699 + }, + { + "epoch": 0.5161432022842082, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.7504537105560303, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7118772268295288, + "num_tokens": 121697278.0, + "step": 4700 + }, + { + "epoch": 0.5162530199868218, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.486082077026367, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.699047327041626, + "num_tokens": 121724693.0, + "step": 4701 + }, + { + "epoch": 0.5163628376894356, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4521822929382324, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6948184967041016, + "num_tokens": 121753616.0, + "step": 4702 + }, + { + "epoch": 0.5164726553920492, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4767916202545166, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7019407153129578, + "num_tokens": 121781118.0, + "step": 4703 + }, + { + "epoch": 0.5165824730946629, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.433544158935547, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7176071405410767, + "num_tokens": 121807666.0, + "step": 4704 + }, + { + "epoch": 0.5166922907972765, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.7196896076202393, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.7273427248001099, + "num_tokens": 121826186.0, + "step": 4705 + }, + { + "epoch": 0.5168021084998902, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.5364785194396973, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7049580216407776, + "num_tokens": 121853749.0, + "step": 4706 + }, + { + "epoch": 0.5169119262025038, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.686223030090332, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6941965818405151, + "num_tokens": 121879214.0, + "step": 4707 + }, + { + "epoch": 0.5170217439051175, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.515104055404663, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.7012208104133606, + "num_tokens": 121905782.0, + "step": 4708 + }, + { + "epoch": 0.5171315616077312, + "ewc_loss": 6.943941116333008e-06, + "grad_norm": 2.57688570022583, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6854474544525146, + "num_tokens": 121935314.0, + "step": 4709 + }, + { + "epoch": 0.5172413793103449, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.3841118812561035, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7110906839370728, + "num_tokens": 121965963.0, + "step": 4710 + }, + { + "epoch": 0.5173511970129585, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.3599801063537598, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7030961513519287, + "num_tokens": 121994548.0, + "step": 4711 + }, + { + "epoch": 0.5174610147155722, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.462636947631836, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7134462594985962, + "num_tokens": 122021964.0, + "step": 4712 + }, + { + "epoch": 0.5175708324181858, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 16.745403289794922, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7116771936416626, + "num_tokens": 122043758.0, + "step": 4713 + }, + { + "epoch": 0.5176806501207994, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.6055164337158203, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7097506523132324, + "num_tokens": 122068302.0, + "step": 4714 + }, + { + "epoch": 0.5177904678234131, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.5724103450775146, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7017766237258911, + "num_tokens": 122095220.0, + "step": 4715 + }, + { + "epoch": 0.5179002855260268, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.894599437713623, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6983263492584229, + "num_tokens": 122117264.0, + "step": 4716 + }, + { + "epoch": 0.5180101032286405, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.919398784637451, + "learning_rate": 1e-06, + "loss": 1.1645, + "mean_token_accuracy": 0.6561591625213623, + "num_tokens": 122140503.0, + "step": 4717 + }, + { + "epoch": 0.5181199209312541, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.590209722518921, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6943665742874146, + "num_tokens": 122162514.0, + "step": 4718 + }, + { + "epoch": 0.5182297386338678, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 3.072948694229126, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6920879483222961, + "num_tokens": 122184999.0, + "step": 4719 + }, + { + "epoch": 0.5183395563364814, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5241591930389404, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7081079483032227, + "num_tokens": 122213918.0, + "step": 4720 + }, + { + "epoch": 0.5184493740390951, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.9991729259490967, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.7029663920402527, + "num_tokens": 122237734.0, + "step": 4721 + }, + { + "epoch": 0.5185591917417087, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.7052268981933594, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.6950603723526001, + "num_tokens": 122260160.0, + "step": 4722 + }, + { + "epoch": 0.5186690094443224, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5574653148651123, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7166727781295776, + "num_tokens": 122286833.0, + "step": 4723 + }, + { + "epoch": 0.5187788271469361, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.4516139030456543, + "learning_rate": 1e-06, + "loss": 0.8813, + "mean_token_accuracy": 0.7296987771987915, + "num_tokens": 122310433.0, + "step": 4724 + }, + { + "epoch": 0.5188886448495498, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5969533920288086, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.6938345432281494, + "num_tokens": 122334006.0, + "step": 4725 + }, + { + "epoch": 0.5189984625521634, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.4158449172973633, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7017152905464172, + "num_tokens": 122361228.0, + "step": 4726 + }, + { + "epoch": 0.5191082802547771, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.623274564743042, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7191838026046753, + "num_tokens": 122386834.0, + "step": 4727 + }, + { + "epoch": 0.5192180979573907, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.607839822769165, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6952362060546875, + "num_tokens": 122411805.0, + "step": 4728 + }, + { + "epoch": 0.5193279156600044, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.355241298675537, + "learning_rate": 1e-06, + "loss": 1.1567, + "mean_token_accuracy": 0.6740354299545288, + "num_tokens": 122446482.0, + "step": 4729 + }, + { + "epoch": 0.519437733362618, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.630324125289917, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7032815217971802, + "num_tokens": 122473380.0, + "step": 4730 + }, + { + "epoch": 0.5195475510652318, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.7749946117401123, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7169604897499084, + "num_tokens": 122497542.0, + "step": 4731 + }, + { + "epoch": 0.5196573687678454, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.340172529220581, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6863856315612793, + "num_tokens": 122526094.0, + "step": 4732 + }, + { + "epoch": 0.519767186470459, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.7896223068237305, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.6856167316436768, + "num_tokens": 122555978.0, + "step": 4733 + }, + { + "epoch": 0.5198770041730727, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5015933513641357, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.6998913884162903, + "num_tokens": 122584360.0, + "step": 4734 + }, + { + "epoch": 0.5199868218756863, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.7849528789520264, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6922982931137085, + "num_tokens": 122606846.0, + "step": 4735 + }, + { + "epoch": 0.5200966395783, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.9328484535217285, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7141178846359253, + "num_tokens": 122626112.0, + "step": 4736 + }, + { + "epoch": 0.5202064572809136, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.6558473110198975, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.687187910079956, + "num_tokens": 122650732.0, + "step": 4737 + }, + { + "epoch": 0.5203162749835274, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.6422688961029053, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7091097235679626, + "num_tokens": 122674126.0, + "step": 4738 + }, + { + "epoch": 0.520426092686141, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.739645481109619, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.7059018611907959, + "num_tokens": 122697488.0, + "step": 4739 + }, + { + "epoch": 0.5205359103887547, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.5901107788085938, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6980544328689575, + "num_tokens": 122722311.0, + "step": 4740 + }, + { + "epoch": 0.5206457280913683, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.5390076637268066, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7043069005012512, + "num_tokens": 122748210.0, + "step": 4741 + }, + { + "epoch": 0.520755545793982, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.746650218963623, + "learning_rate": 1e-06, + "loss": 1.1147, + "mean_token_accuracy": 0.6833988428115845, + "num_tokens": 122772245.0, + "step": 4742 + }, + { + "epoch": 0.5208653634965956, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.7216920852661133, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6967502236366272, + "num_tokens": 122798903.0, + "step": 4743 + }, + { + "epoch": 0.5209751811992093, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.9247374534606934, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.695488452911377, + "num_tokens": 122818671.0, + "step": 4744 + }, + { + "epoch": 0.521084998901823, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.4480202198028564, + "learning_rate": 1e-06, + "loss": 1.1892, + "mean_token_accuracy": 0.6634269952774048, + "num_tokens": 122849140.0, + "step": 4745 + }, + { + "epoch": 0.5211948166044367, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.754436731338501, + "learning_rate": 1e-06, + "loss": 1.1244, + "mean_token_accuracy": 0.6907227039337158, + "num_tokens": 122879171.0, + "step": 4746 + }, + { + "epoch": 0.5213046343070503, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.392951726913452, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6933109164237976, + "num_tokens": 122908443.0, + "step": 4747 + }, + { + "epoch": 0.521414452009664, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 16.839445114135742, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7133520841598511, + "num_tokens": 122928803.0, + "step": 4748 + }, + { + "epoch": 0.5215242697122776, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.756524085998535, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6978367567062378, + "num_tokens": 122954480.0, + "step": 4749 + }, + { + "epoch": 0.5216340874148913, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.8062021732330322, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7081466913223267, + "num_tokens": 122982025.0, + "step": 4750 + }, + { + "epoch": 0.5217439051175049, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.6431617736816406, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7023555040359497, + "num_tokens": 123009122.0, + "step": 4751 + }, + { + "epoch": 0.5218537228201185, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.6494665145874023, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7224718332290649, + "num_tokens": 123034418.0, + "step": 4752 + }, + { + "epoch": 0.5219635405227323, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 3.2821242809295654, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7009692192077637, + "num_tokens": 123055576.0, + "step": 4753 + }, + { + "epoch": 0.522073358225346, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 3.099287509918213, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7004938125610352, + "num_tokens": 123074540.0, + "step": 4754 + }, + { + "epoch": 0.5221831759279596, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.8147668838500977, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.702907145023346, + "num_tokens": 123097449.0, + "step": 4755 + }, + { + "epoch": 0.5222929936305732, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5579237937927246, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.6885629892349243, + "num_tokens": 123125290.0, + "step": 4756 + }, + { + "epoch": 0.5224028113331869, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5189363956451416, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.6936308145523071, + "num_tokens": 123155920.0, + "step": 4757 + }, + { + "epoch": 0.5225126290358005, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.637171745300293, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.715118408203125, + "num_tokens": 123176959.0, + "step": 4758 + }, + { + "epoch": 0.5226224467384142, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.503016233444214, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6866110563278198, + "num_tokens": 123202287.0, + "step": 4759 + }, + { + "epoch": 0.5227322644410279, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.968719959259033, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.684420108795166, + "num_tokens": 123225282.0, + "step": 4760 + }, + { + "epoch": 0.5228420821436416, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.525338649749756, + "learning_rate": 1e-06, + "loss": 1.1186, + "mean_token_accuracy": 0.6771355867385864, + "num_tokens": 123251470.0, + "step": 4761 + }, + { + "epoch": 0.5229518998462552, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.3400659561157227, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6831163763999939, + "num_tokens": 123280914.0, + "step": 4762 + }, + { + "epoch": 0.5230617175488689, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5932586193084717, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7050846815109253, + "num_tokens": 123307373.0, + "step": 4763 + }, + { + "epoch": 0.5231715352514825, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.2702667713165283, + "learning_rate": 1e-06, + "loss": 1.1143, + "mean_token_accuracy": 0.677580714225769, + "num_tokens": 123341593.0, + "step": 4764 + }, + { + "epoch": 0.5232813529540962, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.556262731552124, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7077358961105347, + "num_tokens": 123367583.0, + "step": 4765 + }, + { + "epoch": 0.5233911706567098, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.1845967769622803, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.685776948928833, + "num_tokens": 123404988.0, + "step": 4766 + }, + { + "epoch": 0.5235009883593236, + "ewc_loss": 6.9141387939453125e-06, + "grad_norm": 2.4944519996643066, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7106670141220093, + "num_tokens": 123429995.0, + "step": 4767 + }, + { + "epoch": 0.5236108060619372, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.521820545196533, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6904312372207642, + "num_tokens": 123458296.0, + "step": 4768 + }, + { + "epoch": 0.5237206237645509, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.537696361541748, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6965930461883545, + "num_tokens": 123484696.0, + "step": 4769 + }, + { + "epoch": 0.5238304414671645, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.719475746154785, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.6943418979644775, + "num_tokens": 123511954.0, + "step": 4770 + }, + { + "epoch": 0.5239402591697782, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.926274061203003, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7134937047958374, + "num_tokens": 123532302.0, + "step": 4771 + }, + { + "epoch": 0.5240500768723918, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.709568500518799, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.7046526670455933, + "num_tokens": 123557394.0, + "step": 4772 + }, + { + "epoch": 0.5241598945750054, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.979224681854248, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6865434646606445, + "num_tokens": 123581035.0, + "step": 4773 + }, + { + "epoch": 0.5242697122776192, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.5249266624450684, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6982728242874146, + "num_tokens": 123609927.0, + "step": 4774 + }, + { + "epoch": 0.5243795299802329, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.661369800567627, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6796214580535889, + "num_tokens": 123636801.0, + "step": 4775 + }, + { + "epoch": 0.5244893476828465, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.478797435760498, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7217453718185425, + "num_tokens": 123662412.0, + "step": 4776 + }, + { + "epoch": 0.5245991653854601, + "ewc_loss": 6.973743438720703e-06, + "grad_norm": 2.5920486450195312, + "learning_rate": 1e-06, + "loss": 1.1088, + "mean_token_accuracy": 0.6859065294265747, + "num_tokens": 123688200.0, + "step": 4777 + }, + { + "epoch": 0.5247089830880738, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 3.044466495513916, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.6969020962715149, + "num_tokens": 123707255.0, + "step": 4778 + }, + { + "epoch": 0.5248188007906874, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.7164273262023926, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7073402404785156, + "num_tokens": 123730727.0, + "step": 4779 + }, + { + "epoch": 0.5249286184933011, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.773327112197876, + "learning_rate": 1e-06, + "loss": 0.8555, + "mean_token_accuracy": 0.7395171523094177, + "num_tokens": 123750769.0, + "step": 4780 + }, + { + "epoch": 0.5250384361959147, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.5924813747406006, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7138318419456482, + "num_tokens": 123774443.0, + "step": 4781 + }, + { + "epoch": 0.5251482538985285, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.889498710632324, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.6962718963623047, + "num_tokens": 123797174.0, + "step": 4782 + }, + { + "epoch": 0.5252580716011421, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.592123508453369, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6934143304824829, + "num_tokens": 123823711.0, + "step": 4783 + }, + { + "epoch": 0.5253678893037558, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 3.0906848907470703, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.686421811580658, + "num_tokens": 123847343.0, + "step": 4784 + }, + { + "epoch": 0.5254777070063694, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.6606028079986572, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6820424795150757, + "num_tokens": 123872690.0, + "step": 4785 + }, + { + "epoch": 0.5255875247089831, + "ewc_loss": 7.0035457611083984e-06, + "grad_norm": 2.733409881591797, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6980671882629395, + "num_tokens": 123897410.0, + "step": 4786 + }, + { + "epoch": 0.5256973424115967, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6691741943359375, + "learning_rate": 1e-06, + "loss": 1.0832, + "mean_token_accuracy": 0.6899142265319824, + "num_tokens": 123921194.0, + "step": 4787 + }, + { + "epoch": 0.5258071601142104, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6755051612854004, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7038446068763733, + "num_tokens": 123946528.0, + "step": 4788 + }, + { + "epoch": 0.5259169778168241, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6607866287231445, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7008798718452454, + "num_tokens": 123974449.0, + "step": 4789 + }, + { + "epoch": 0.5260267955194378, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7773053646087646, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7153623700141907, + "num_tokens": 123994838.0, + "step": 4790 + }, + { + "epoch": 0.5261366132220514, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4193549156188965, + "learning_rate": 1e-06, + "loss": 1.1728, + "mean_token_accuracy": 0.6607074737548828, + "num_tokens": 124025635.0, + "step": 4791 + }, + { + "epoch": 0.526246430924665, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.58111834526062, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6973003149032593, + "num_tokens": 124052257.0, + "step": 4792 + }, + { + "epoch": 0.5263562486272787, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.606207847595215, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.6997160911560059, + "num_tokens": 124076722.0, + "step": 4793 + }, + { + "epoch": 0.5264660663298923, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.9125962257385254, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6888061761856079, + "num_tokens": 124095514.0, + "step": 4794 + }, + { + "epoch": 0.526575884032506, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.519270181655884, + "learning_rate": 1e-06, + "loss": 1.1127, + "mean_token_accuracy": 0.6701406240463257, + "num_tokens": 124123063.0, + "step": 4795 + }, + { + "epoch": 0.5266857017351197, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5216498374938965, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6855900287628174, + "num_tokens": 124146795.0, + "step": 4796 + }, + { + "epoch": 0.5267955194377334, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5549821853637695, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7029253244400024, + "num_tokens": 124170669.0, + "step": 4797 + }, + { + "epoch": 0.526905337140347, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.469219207763672, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6854163408279419, + "num_tokens": 124195844.0, + "step": 4798 + }, + { + "epoch": 0.5270151548429607, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.900381326675415, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6808909773826599, + "num_tokens": 124223053.0, + "step": 4799 + }, + { + "epoch": 0.5271249725455743, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.629976987838745, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.693494439125061, + "num_tokens": 124248218.0, + "step": 4800 + }, + { + "epoch": 0.527234790248188, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5950207710266113, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7154563069343567, + "num_tokens": 124273859.0, + "step": 4801 + }, + { + "epoch": 0.5273446079508016, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.874140501022339, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7119228839874268, + "num_tokens": 124294213.0, + "step": 4802 + }, + { + "epoch": 0.5274544256534154, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.4909377098083496, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6956202387809753, + "num_tokens": 124319793.0, + "step": 4803 + }, + { + "epoch": 0.527564243356029, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.501526355743408, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7040020823478699, + "num_tokens": 124346866.0, + "step": 4804 + }, + { + "epoch": 0.5276740610586427, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 4.465545177459717, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7032475471496582, + "num_tokens": 124368693.0, + "step": 4805 + }, + { + "epoch": 0.5277838787612563, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.9429593086242676, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6815359592437744, + "num_tokens": 124388582.0, + "step": 4806 + }, + { + "epoch": 0.52789369646387, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.5499064922332764, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7099111080169678, + "num_tokens": 124413481.0, + "step": 4807 + }, + { + "epoch": 0.5280035141664836, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.710942506790161, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7059605717658997, + "num_tokens": 124436784.0, + "step": 4808 + }, + { + "epoch": 0.5281133318690973, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.675243377685547, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6817214488983154, + "num_tokens": 124460433.0, + "step": 4809 + }, + { + "epoch": 0.528223149571711, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 16.832223892211914, + "learning_rate": 1e-06, + "loss": 1.0979, + "mean_token_accuracy": 0.6857715845108032, + "num_tokens": 124484380.0, + "step": 4810 + }, + { + "epoch": 0.5283329672743247, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.707629919052124, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7053747773170471, + "num_tokens": 124509622.0, + "step": 4811 + }, + { + "epoch": 0.5284427849769383, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.5434768199920654, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6862396001815796, + "num_tokens": 124540590.0, + "step": 4812 + }, + { + "epoch": 0.528552602679552, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.9283008575439453, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7036617994308472, + "num_tokens": 124562283.0, + "step": 4813 + }, + { + "epoch": 0.5286624203821656, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.477487087249756, + "learning_rate": 1e-06, + "loss": 1.1199, + "mean_token_accuracy": 0.6783372163772583, + "num_tokens": 124594488.0, + "step": 4814 + }, + { + "epoch": 0.5287722380847792, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.490727186203003, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7082849740982056, + "num_tokens": 124621489.0, + "step": 4815 + }, + { + "epoch": 0.5288820557873929, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4833076000213623, + "learning_rate": 1e-06, + "loss": 0.8808, + "mean_token_accuracy": 0.7390382289886475, + "num_tokens": 124649436.0, + "step": 4816 + }, + { + "epoch": 0.5289918734900065, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 3.043114423751831, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7084757089614868, + "num_tokens": 124671012.0, + "step": 4817 + }, + { + "epoch": 0.5291016911926203, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6140780448913574, + "learning_rate": 1e-06, + "loss": 1.0753, + "mean_token_accuracy": 0.6908054351806641, + "num_tokens": 124694958.0, + "step": 4818 + }, + { + "epoch": 0.5292115088952339, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.9286386966705322, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.7027298212051392, + "num_tokens": 124715344.0, + "step": 4819 + }, + { + "epoch": 0.5293213265978476, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6293938159942627, + "learning_rate": 1e-06, + "loss": 1.1888, + "mean_token_accuracy": 0.6634234189987183, + "num_tokens": 124743751.0, + "step": 4820 + }, + { + "epoch": 0.5294311443004612, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6733033657073975, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7229515910148621, + "num_tokens": 124769264.0, + "step": 4821 + }, + { + "epoch": 0.5295409620030749, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.704791784286499, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6913001537322998, + "num_tokens": 124796623.0, + "step": 4822 + }, + { + "epoch": 0.5296507797056885, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4282820224761963, + "learning_rate": 1e-06, + "loss": 1.1459, + "mean_token_accuracy": 0.6709790229797363, + "num_tokens": 124825236.0, + "step": 4823 + }, + { + "epoch": 0.5297605974083022, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.68068790435791, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.689242959022522, + "num_tokens": 124849570.0, + "step": 4824 + }, + { + "epoch": 0.5298704151109159, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5587596893310547, + "learning_rate": 1e-06, + "loss": 1.1069, + "mean_token_accuracy": 0.6833594441413879, + "num_tokens": 124876140.0, + "step": 4825 + }, + { + "epoch": 0.5299802328135296, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.521545171737671, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6873632073402405, + "num_tokens": 124902546.0, + "step": 4826 + }, + { + "epoch": 0.5300900505161432, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.659369468688965, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6980467438697815, + "num_tokens": 124926641.0, + "step": 4827 + }, + { + "epoch": 0.5301998682187569, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.388697385787964, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6975929141044617, + "num_tokens": 124955451.0, + "step": 4828 + }, + { + "epoch": 0.5303096859213705, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.507988214492798, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7094667553901672, + "num_tokens": 124983402.0, + "step": 4829 + }, + { + "epoch": 0.5304195036239842, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6061723232269287, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6960395574569702, + "num_tokens": 125008433.0, + "step": 4830 + }, + { + "epoch": 0.5305293213265978, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4892988204956055, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.6962536573410034, + "num_tokens": 125036151.0, + "step": 4831 + }, + { + "epoch": 0.5306391390292116, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.685106039047241, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7168874740600586, + "num_tokens": 125058766.0, + "step": 4832 + }, + { + "epoch": 0.5307489567318252, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.752732753753662, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7107017040252686, + "num_tokens": 125084381.0, + "step": 4833 + }, + { + "epoch": 0.5308587744344389, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.501910924911499, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6834248304367065, + "num_tokens": 125108921.0, + "step": 4834 + }, + { + "epoch": 0.5309685921370525, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.2923991680145264, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6986244320869446, + "num_tokens": 125140054.0, + "step": 4835 + }, + { + "epoch": 0.5310784098396661, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5304088592529297, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6818675994873047, + "num_tokens": 125167042.0, + "step": 4836 + }, + { + "epoch": 0.5311882275422798, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.231489896774292, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7115232348442078, + "num_tokens": 125195924.0, + "step": 4837 + }, + { + "epoch": 0.5312980452448934, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.593768358230591, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6943941116333008, + "num_tokens": 125221340.0, + "step": 4838 + }, + { + "epoch": 0.5314078629475072, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7776248455047607, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6944887638092041, + "num_tokens": 125243894.0, + "step": 4839 + }, + { + "epoch": 0.5315176806501208, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7043943405151367, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7081881761550903, + "num_tokens": 125268549.0, + "step": 4840 + }, + { + "epoch": 0.5316274983527345, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.886124610900879, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6925972700119019, + "num_tokens": 125295550.0, + "step": 4841 + }, + { + "epoch": 0.5317373160553481, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5360381603240967, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6986100673675537, + "num_tokens": 125320744.0, + "step": 4842 + }, + { + "epoch": 0.5318471337579618, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.842623710632324, + "learning_rate": 1e-06, + "loss": 1.1057, + "mean_token_accuracy": 0.6841063499450684, + "num_tokens": 125348066.0, + "step": 4843 + }, + { + "epoch": 0.5319569514605754, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.3611583709716797, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6911810040473938, + "num_tokens": 125378429.0, + "step": 4844 + }, + { + "epoch": 0.5320667691631891, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.511411190032959, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6910902261734009, + "num_tokens": 125403980.0, + "step": 4845 + }, + { + "epoch": 0.5321765868658027, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.3131182193756104, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6904234886169434, + "num_tokens": 125434953.0, + "step": 4846 + }, + { + "epoch": 0.5322864045684165, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5272135734558105, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.6978725790977478, + "num_tokens": 125461614.0, + "step": 4847 + }, + { + "epoch": 0.5323962222710301, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4289262294769287, + "learning_rate": 1e-06, + "loss": 1.1365, + "mean_token_accuracy": 0.6755858659744263, + "num_tokens": 125494269.0, + "step": 4848 + }, + { + "epoch": 0.5325060399736438, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.610325813293457, + "learning_rate": 1e-06, + "loss": 1.1379, + "mean_token_accuracy": 0.6683934926986694, + "num_tokens": 125523086.0, + "step": 4849 + }, + { + "epoch": 0.5326158576762574, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6096572875976562, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.687616229057312, + "num_tokens": 125548116.0, + "step": 4850 + }, + { + "epoch": 0.532725675378871, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4684808254241943, + "learning_rate": 1e-06, + "loss": 1.134, + "mean_token_accuracy": 0.6698213219642639, + "num_tokens": 125576092.0, + "step": 4851 + }, + { + "epoch": 0.5328354930814847, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4137561321258545, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.699123203754425, + "num_tokens": 125606678.0, + "step": 4852 + }, + { + "epoch": 0.5329453107840983, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.750446081161499, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.703446626663208, + "num_tokens": 125628470.0, + "step": 4853 + }, + { + "epoch": 0.5330551284867121, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6799092292785645, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6872022747993469, + "num_tokens": 125649194.0, + "step": 4854 + }, + { + "epoch": 0.5331649461893258, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4470067024230957, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.684795618057251, + "num_tokens": 125681786.0, + "step": 4855 + }, + { + "epoch": 0.5332747638919394, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.360593318939209, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7024898529052734, + "num_tokens": 125708871.0, + "step": 4856 + }, + { + "epoch": 0.533384581594553, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.531707525253296, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7031670808792114, + "num_tokens": 125736407.0, + "step": 4857 + }, + { + "epoch": 0.5334943992971667, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4160218238830566, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.6929813623428345, + "num_tokens": 125764625.0, + "step": 4858 + }, + { + "epoch": 0.5336042169997803, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7649433612823486, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.6960850954055786, + "num_tokens": 125788811.0, + "step": 4859 + }, + { + "epoch": 0.533714034702394, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.480034828186035, + "learning_rate": 1e-06, + "loss": 1.0418, + "mean_token_accuracy": 0.6909149289131165, + "num_tokens": 125816450.0, + "step": 4860 + }, + { + "epoch": 0.5338238524050077, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5368399620056152, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6979718804359436, + "num_tokens": 125843742.0, + "step": 4861 + }, + { + "epoch": 0.5339336701076214, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 4.119115352630615, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6845381259918213, + "num_tokens": 125867806.0, + "step": 4862 + }, + { + "epoch": 0.534043487810235, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.4188146591186523, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6865548491477966, + "num_tokens": 125898897.0, + "step": 4863 + }, + { + "epoch": 0.5341533055128487, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 3.20928692817688, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7118890285491943, + "num_tokens": 125926333.0, + "step": 4864 + }, + { + "epoch": 0.5342631232154623, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.5759692192077637, + "learning_rate": 1e-06, + "loss": 1.1058, + "mean_token_accuracy": 0.6772621870040894, + "num_tokens": 125955239.0, + "step": 4865 + }, + { + "epoch": 0.534372940918076, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4100701808929443, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7076176404953003, + "num_tokens": 125984817.0, + "step": 4866 + }, + { + "epoch": 0.5344827586206896, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.758559226989746, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7018332481384277, + "num_tokens": 126007118.0, + "step": 4867 + }, + { + "epoch": 0.5345925763233034, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.459820032119751, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6976888179779053, + "num_tokens": 126033833.0, + "step": 4868 + }, + { + "epoch": 0.534702394025917, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.568758964538574, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6917896866798401, + "num_tokens": 126058969.0, + "step": 4869 + }, + { + "epoch": 0.5348122117285307, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.357167959213257, + "learning_rate": 1e-06, + "loss": 1.0995, + "mean_token_accuracy": 0.6787505745887756, + "num_tokens": 126087601.0, + "step": 4870 + }, + { + "epoch": 0.5349220294311443, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.814542055130005, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6960582137107849, + "num_tokens": 126108863.0, + "step": 4871 + }, + { + "epoch": 0.535031847133758, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4590539932250977, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7095305919647217, + "num_tokens": 126135520.0, + "step": 4872 + }, + { + "epoch": 0.5351416648363716, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6432509422302246, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6971592307090759, + "num_tokens": 126159142.0, + "step": 4873 + }, + { + "epoch": 0.5352514825389852, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.413261890411377, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7089238166809082, + "num_tokens": 126185981.0, + "step": 4874 + }, + { + "epoch": 0.5353613002415989, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.60866379737854, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6918710470199585, + "num_tokens": 126211810.0, + "step": 4875 + }, + { + "epoch": 0.5354711179442126, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.770444631576538, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7016154527664185, + "num_tokens": 126236974.0, + "step": 4876 + }, + { + "epoch": 0.5355809356468263, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7087254524230957, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.6993018388748169, + "num_tokens": 126261310.0, + "step": 4877 + }, + { + "epoch": 0.5356907533494399, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.873029947280884, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6888002753257751, + "num_tokens": 126281643.0, + "step": 4878 + }, + { + "epoch": 0.5358005710520536, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.3895914554595947, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6852613091468811, + "num_tokens": 126312906.0, + "step": 4879 + }, + { + "epoch": 0.5359103887546672, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.415574550628662, + "learning_rate": 1e-06, + "loss": 1.1621, + "mean_token_accuracy": 0.6649125218391418, + "num_tokens": 126345375.0, + "step": 4880 + }, + { + "epoch": 0.5360202064572809, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5242607593536377, + "learning_rate": 1e-06, + "loss": 1.1138, + "mean_token_accuracy": 0.6777185201644897, + "num_tokens": 126376856.0, + "step": 4881 + }, + { + "epoch": 0.5361300241598945, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.648353338241577, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6942365765571594, + "num_tokens": 126401481.0, + "step": 4882 + }, + { + "epoch": 0.5362398418625083, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 3.095978021621704, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6958106756210327, + "num_tokens": 126421653.0, + "step": 4883 + }, + { + "epoch": 0.5363496595651219, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.5813281536102295, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.6858327388763428, + "num_tokens": 126446537.0, + "step": 4884 + }, + { + "epoch": 0.5364594772677356, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.540043354034424, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7257846593856812, + "num_tokens": 126472324.0, + "step": 4885 + }, + { + "epoch": 0.5365692949703492, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.538205862045288, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7245180606842041, + "num_tokens": 126498138.0, + "step": 4886 + }, + { + "epoch": 0.5366791126729629, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.533121347427368, + "learning_rate": 1e-06, + "loss": 1.1178, + "mean_token_accuracy": 0.6711114645004272, + "num_tokens": 126524152.0, + "step": 4887 + }, + { + "epoch": 0.5367889303755765, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5641701221466064, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.6977176666259766, + "num_tokens": 126551253.0, + "step": 4888 + }, + { + "epoch": 0.5368987480781902, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.504530668258667, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.68690425157547, + "num_tokens": 126583556.0, + "step": 4889 + }, + { + "epoch": 0.5370085657808039, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.860130786895752, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.712785005569458, + "num_tokens": 126602628.0, + "step": 4890 + }, + { + "epoch": 0.5371183834834176, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.429898977279663, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7052541971206665, + "num_tokens": 126630945.0, + "step": 4891 + }, + { + "epoch": 0.5372282011860312, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4487898349761963, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6997565627098083, + "num_tokens": 126660562.0, + "step": 4892 + }, + { + "epoch": 0.5373380188886449, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.331779718399048, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.6997799277305603, + "num_tokens": 126691431.0, + "step": 4893 + }, + { + "epoch": 0.5374478365912585, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4646053314208984, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.7001218795776367, + "num_tokens": 126718942.0, + "step": 4894 + }, + { + "epoch": 0.5375576542938721, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 3.639036178588867, + "learning_rate": 1e-06, + "loss": 0.918, + "mean_token_accuracy": 0.7258122563362122, + "num_tokens": 126737774.0, + "step": 4895 + }, + { + "epoch": 0.5376674719964858, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.681792736053467, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.6927499771118164, + "num_tokens": 126764676.0, + "step": 4896 + }, + { + "epoch": 0.5377772896990995, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.66275691986084, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6960024237632751, + "num_tokens": 126790046.0, + "step": 4897 + }, + { + "epoch": 0.5378871074017132, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.8972363471984863, + "learning_rate": 1e-06, + "loss": 0.8984, + "mean_token_accuracy": 0.7320423722267151, + "num_tokens": 126808339.0, + "step": 4898 + }, + { + "epoch": 0.5379969251043268, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.570732831954956, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7202085256576538, + "num_tokens": 126832548.0, + "step": 4899 + }, + { + "epoch": 0.5381067428069405, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.2718863487243652, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6915030479431152, + "num_tokens": 126864910.0, + "step": 4900 + }, + { + "epoch": 0.5382165605095541, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.8803303241729736, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6893407106399536, + "num_tokens": 126893270.0, + "step": 4901 + }, + { + "epoch": 0.5383263782121678, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4353208541870117, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6957338452339172, + "num_tokens": 126921753.0, + "step": 4902 + }, + { + "epoch": 0.5384361959147814, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.431154727935791, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6832519769668579, + "num_tokens": 126955918.0, + "step": 4903 + }, + { + "epoch": 0.5385460136173951, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6506640911102295, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7056200504302979, + "num_tokens": 126978163.0, + "step": 4904 + }, + { + "epoch": 0.5386558313200088, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7859671115875244, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.7025977969169617, + "num_tokens": 127001935.0, + "step": 4905 + }, + { + "epoch": 0.5387656490226225, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.7024312019348145, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7017433047294617, + "num_tokens": 127026364.0, + "step": 4906 + }, + { + "epoch": 0.5388754667252361, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7872214317321777, + "learning_rate": 1e-06, + "loss": 1.0937, + "mean_token_accuracy": 0.6818289160728455, + "num_tokens": 127051022.0, + "step": 4907 + }, + { + "epoch": 0.5389852844278498, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6550700664520264, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6930654048919678, + "num_tokens": 127077302.0, + "step": 4908 + }, + { + "epoch": 0.5390951021304634, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.569145441055298, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6924538612365723, + "num_tokens": 127107594.0, + "step": 4909 + }, + { + "epoch": 0.5392049198330771, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.826967716217041, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7015326619148254, + "num_tokens": 127129814.0, + "step": 4910 + }, + { + "epoch": 0.5393147375356907, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 3.047085762023926, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.712417721748352, + "num_tokens": 127148691.0, + "step": 4911 + }, + { + "epoch": 0.5394245552383045, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.7836391925811768, + "learning_rate": 1e-06, + "loss": 0.8906, + "mean_token_accuracy": 0.732548177242279, + "num_tokens": 127168865.0, + "step": 4912 + }, + { + "epoch": 0.5395343729409181, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.323361873626709, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6937153935432434, + "num_tokens": 127199139.0, + "step": 4913 + }, + { + "epoch": 0.5396441906435318, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.8136043548583984, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7058750987052917, + "num_tokens": 127219551.0, + "step": 4914 + }, + { + "epoch": 0.5397540083461454, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.4936931133270264, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7037783861160278, + "num_tokens": 127247441.0, + "step": 4915 + }, + { + "epoch": 0.539863826048759, + "ewc_loss": 7.063150405883789e-06, + "grad_norm": 2.355865716934204, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7045632004737854, + "num_tokens": 127280922.0, + "step": 4916 + }, + { + "epoch": 0.5399736437513727, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.8806753158569336, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7192267775535583, + "num_tokens": 127303976.0, + "step": 4917 + }, + { + "epoch": 0.5400834614539863, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5482773780822754, + "learning_rate": 1e-06, + "loss": 0.9287, + "mean_token_accuracy": 0.719152569770813, + "num_tokens": 127329285.0, + "step": 4918 + }, + { + "epoch": 0.5401932791566001, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.719332456588745, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7163407802581787, + "num_tokens": 127352261.0, + "step": 4919 + }, + { + "epoch": 0.5403030968592137, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.86651611328125, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7006799578666687, + "num_tokens": 127380086.0, + "step": 4920 + }, + { + "epoch": 0.5404129145618274, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.755692720413208, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7123712301254272, + "num_tokens": 127402010.0, + "step": 4921 + }, + { + "epoch": 0.540522732264441, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6192667484283447, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6870467662811279, + "num_tokens": 127429070.0, + "step": 4922 + }, + { + "epoch": 0.5406325499670547, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.3933351039886475, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7085763216018677, + "num_tokens": 127458094.0, + "step": 4923 + }, + { + "epoch": 0.5407423676696683, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.380098581314087, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.697830319404602, + "num_tokens": 127487961.0, + "step": 4924 + }, + { + "epoch": 0.540852185372282, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5576958656311035, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.6985475420951843, + "num_tokens": 127514020.0, + "step": 4925 + }, + { + "epoch": 0.5409620030748957, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.391226291656494, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.6890318393707275, + "num_tokens": 127542668.0, + "step": 4926 + }, + { + "epoch": 0.5410718207775094, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.77057147026062, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7008081674575806, + "num_tokens": 127562630.0, + "step": 4927 + }, + { + "epoch": 0.541181638480123, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.3611347675323486, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7020419836044312, + "num_tokens": 127592931.0, + "step": 4928 + }, + { + "epoch": 0.5412914561827367, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.383680582046509, + "learning_rate": 1e-06, + "loss": 1.0736, + "mean_token_accuracy": 0.6894816756248474, + "num_tokens": 127621487.0, + "step": 4929 + }, + { + "epoch": 0.5414012738853503, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.492725372314453, + "learning_rate": 1e-06, + "loss": 0.9324, + "mean_token_accuracy": 0.7258807420730591, + "num_tokens": 127650196.0, + "step": 4930 + }, + { + "epoch": 0.541511091587964, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.710129499435425, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7140067219734192, + "num_tokens": 127673970.0, + "step": 4931 + }, + { + "epoch": 0.5416209092905776, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.3391411304473877, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.683263897895813, + "num_tokens": 127703486.0, + "step": 4932 + }, + { + "epoch": 0.5417307269931912, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.439128875732422, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7020657062530518, + "num_tokens": 127731693.0, + "step": 4933 + }, + { + "epoch": 0.541840544695805, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.551699161529541, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7200632095336914, + "num_tokens": 127754534.0, + "step": 4934 + }, + { + "epoch": 0.5419503623984187, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.309993028640747, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.693203866481781, + "num_tokens": 127785708.0, + "step": 4935 + }, + { + "epoch": 0.5420601801010323, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.679103136062622, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7204051613807678, + "num_tokens": 127807870.0, + "step": 4936 + }, + { + "epoch": 0.5421699978036459, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.706070899963379, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6904494762420654, + "num_tokens": 127830538.0, + "step": 4937 + }, + { + "epoch": 0.5422798155062596, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4303126335144043, + "learning_rate": 1e-06, + "loss": 1.1254, + "mean_token_accuracy": 0.6717143058776855, + "num_tokens": 127861733.0, + "step": 4938 + }, + { + "epoch": 0.5423896332088732, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6708195209503174, + "learning_rate": 1e-06, + "loss": 1.1006, + "mean_token_accuracy": 0.6839950084686279, + "num_tokens": 127886190.0, + "step": 4939 + }, + { + "epoch": 0.5424994509114869, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.7649617195129395, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6938138008117676, + "num_tokens": 127914203.0, + "step": 4940 + }, + { + "epoch": 0.5426092686141006, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.4526662826538086, + "learning_rate": 1e-06, + "loss": 1.1157, + "mean_token_accuracy": 0.6798841953277588, + "num_tokens": 127942783.0, + "step": 4941 + }, + { + "epoch": 0.5427190863167143, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6261909008026123, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6941866874694824, + "num_tokens": 127967032.0, + "step": 4942 + }, + { + "epoch": 0.5428289040193279, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.78085994720459, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7125582098960876, + "num_tokens": 127988188.0, + "step": 4943 + }, + { + "epoch": 0.5429387217219416, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.659991979598999, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6864420771598816, + "num_tokens": 128012088.0, + "step": 4944 + }, + { + "epoch": 0.5430485394245552, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.687052011489868, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6953409910202026, + "num_tokens": 128037208.0, + "step": 4945 + }, + { + "epoch": 0.5431583571271689, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6240572929382324, + "learning_rate": 1e-06, + "loss": 0.9141, + "mean_token_accuracy": 0.727712869644165, + "num_tokens": 128062799.0, + "step": 4946 + }, + { + "epoch": 0.5432681748297825, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.6403753757476807, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7119656801223755, + "num_tokens": 128093529.0, + "step": 4947 + }, + { + "epoch": 0.5433779925323963, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.3715412616729736, + "learning_rate": 1e-06, + "loss": 1.1083, + "mean_token_accuracy": 0.6740866899490356, + "num_tokens": 128123690.0, + "step": 4948 + }, + { + "epoch": 0.5434878102350099, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.8492889404296875, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.688956618309021, + "num_tokens": 128148433.0, + "step": 4949 + }, + { + "epoch": 0.5435976279376236, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.685121774673462, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6831422448158264, + "num_tokens": 128173520.0, + "step": 4950 + }, + { + "epoch": 0.5437074456402372, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.4725348949432373, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6979894638061523, + "num_tokens": 128205469.0, + "step": 4951 + }, + { + "epoch": 0.5438172633428509, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.647688150405884, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7236589193344116, + "num_tokens": 128228251.0, + "step": 4952 + }, + { + "epoch": 0.5439270810454645, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.3378279209136963, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7079668045043945, + "num_tokens": 128255630.0, + "step": 4953 + }, + { + "epoch": 0.5440368987480781, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.7022690773010254, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7042741179466248, + "num_tokens": 128278625.0, + "step": 4954 + }, + { + "epoch": 0.5441467164506919, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.6789417266845703, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7135664820671082, + "num_tokens": 128299970.0, + "step": 4955 + }, + { + "epoch": 0.5442565341533055, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.5393054485321045, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7082229852676392, + "num_tokens": 128325057.0, + "step": 4956 + }, + { + "epoch": 0.5443663518559192, + "ewc_loss": 7.033348083496094e-06, + "grad_norm": 2.6809911727905273, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6973775625228882, + "num_tokens": 128348579.0, + "step": 4957 + }, + { + "epoch": 0.5444761695585328, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.5084280967712402, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7103564739227295, + "num_tokens": 128374047.0, + "step": 4958 + }, + { + "epoch": 0.5445859872611465, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.9251370429992676, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7022587656974792, + "num_tokens": 128394609.0, + "step": 4959 + }, + { + "epoch": 0.5446958049637601, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.5209643840789795, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.696530818939209, + "num_tokens": 128419846.0, + "step": 4960 + }, + { + "epoch": 0.5448056226663738, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.26387882232666, + "learning_rate": 1e-06, + "loss": 1.1335, + "mean_token_accuracy": 0.6715710163116455, + "num_tokens": 128453276.0, + "step": 4961 + }, + { + "epoch": 0.5449154403689875, + "ewc_loss": 7.092952728271484e-06, + "grad_norm": 2.61344575881958, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6941492557525635, + "num_tokens": 128476588.0, + "step": 4962 + }, + { + "epoch": 0.5450252580716012, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.4589297771453857, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6970357894897461, + "num_tokens": 128503964.0, + "step": 4963 + }, + { + "epoch": 0.5451350757742148, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.370051383972168, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7118939161300659, + "num_tokens": 128531796.0, + "step": 4964 + }, + { + "epoch": 0.5452448934768285, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.733201026916504, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7057193517684937, + "num_tokens": 128554987.0, + "step": 4965 + }, + { + "epoch": 0.5453547111794421, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.652951717376709, + "learning_rate": 1e-06, + "loss": 1.0997, + "mean_token_accuracy": 0.6860308647155762, + "num_tokens": 128580111.0, + "step": 4966 + }, + { + "epoch": 0.5454645288820558, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.2290220260620117, + "learning_rate": 1e-06, + "loss": 1.1296, + "mean_token_accuracy": 0.6743766069412231, + "num_tokens": 128612332.0, + "step": 4967 + }, + { + "epoch": 0.5455743465846694, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.732602596282959, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6840652227401733, + "num_tokens": 128636853.0, + "step": 4968 + }, + { + "epoch": 0.5456841642872831, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.514143943786621, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7180564403533936, + "num_tokens": 128664436.0, + "step": 4969 + }, + { + "epoch": 0.5457939819898968, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.469783067703247, + "learning_rate": 1e-06, + "loss": 1.1148, + "mean_token_accuracy": 0.6806575059890747, + "num_tokens": 128694240.0, + "step": 4970 + }, + { + "epoch": 0.5459037996925105, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.586544990539551, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6944604516029358, + "num_tokens": 128719182.0, + "step": 4971 + }, + { + "epoch": 0.5460136173951241, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5356414318084717, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.6973894834518433, + "num_tokens": 128746729.0, + "step": 4972 + }, + { + "epoch": 0.5461234350977378, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6590309143066406, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6917988061904907, + "num_tokens": 128772278.0, + "step": 4973 + }, + { + "epoch": 0.5462332528003514, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.4637227058410645, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6864926815032959, + "num_tokens": 128799048.0, + "step": 4974 + }, + { + "epoch": 0.546343070502965, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.427022933959961, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.6997814774513245, + "num_tokens": 128825845.0, + "step": 4975 + }, + { + "epoch": 0.5464528882055787, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.543109655380249, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7024509310722351, + "num_tokens": 128854929.0, + "step": 4976 + }, + { + "epoch": 0.5465627059081924, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.7195539474487305, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.700922966003418, + "num_tokens": 128878255.0, + "step": 4977 + }, + { + "epoch": 0.5466725236108061, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.2581512928009033, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6877416968345642, + "num_tokens": 128910671.0, + "step": 4978 + }, + { + "epoch": 0.5467823413134197, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.30110764503479, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.6880729794502258, + "num_tokens": 128941001.0, + "step": 4979 + }, + { + "epoch": 0.5468921590160334, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.5592684745788574, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7148294448852539, + "num_tokens": 128965677.0, + "step": 4980 + }, + { + "epoch": 0.547001976718647, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.534716844558716, + "learning_rate": 1e-06, + "loss": 1.1247, + "mean_token_accuracy": 0.6757591962814331, + "num_tokens": 128992259.0, + "step": 4981 + }, + { + "epoch": 0.5471117944212607, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.1964175701141357, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.711560070514679, + "num_tokens": 129026500.0, + "step": 4982 + }, + { + "epoch": 0.5472216121238743, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.503272533416748, + "learning_rate": 1e-06, + "loss": 1.134, + "mean_token_accuracy": 0.6737033724784851, + "num_tokens": 129055142.0, + "step": 4983 + }, + { + "epoch": 0.5473314298264881, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.3069677352905273, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7019222378730774, + "num_tokens": 129087650.0, + "step": 4984 + }, + { + "epoch": 0.5474412475291017, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.32184100151062, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6949249505996704, + "num_tokens": 129117894.0, + "step": 4985 + }, + { + "epoch": 0.5475510652317154, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.414473056793213, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.692704439163208, + "num_tokens": 129144524.0, + "step": 4986 + }, + { + "epoch": 0.547660882934329, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.8004519939422607, + "learning_rate": 1e-06, + "loss": 1.0997, + "mean_token_accuracy": 0.6869351863861084, + "num_tokens": 129167161.0, + "step": 4987 + }, + { + "epoch": 0.5477707006369427, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6476874351501465, + "learning_rate": 1e-06, + "loss": 1.1464, + "mean_token_accuracy": 0.675183892250061, + "num_tokens": 129192187.0, + "step": 4988 + }, + { + "epoch": 0.5478805183395563, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.5025699138641357, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6954866647720337, + "num_tokens": 129220666.0, + "step": 4989 + }, + { + "epoch": 0.54799033604217, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.7487518787384033, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7041191458702087, + "num_tokens": 129242823.0, + "step": 4990 + }, + { + "epoch": 0.5481001537447837, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5647780895233154, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6867427229881287, + "num_tokens": 129267928.0, + "step": 4991 + }, + { + "epoch": 0.5482099714473974, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5946788787841797, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6856997013092041, + "num_tokens": 129292838.0, + "step": 4992 + }, + { + "epoch": 0.548319789150011, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.6083688735961914, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7066093683242798, + "num_tokens": 129314953.0, + "step": 4993 + }, + { + "epoch": 0.5484296068526247, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6829659938812256, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7160887122154236, + "num_tokens": 129339660.0, + "step": 4994 + }, + { + "epoch": 0.5485394245552383, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.7720422744750977, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7111575603485107, + "num_tokens": 129363111.0, + "step": 4995 + }, + { + "epoch": 0.5486492422578519, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5294063091278076, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.683281660079956, + "num_tokens": 129390318.0, + "step": 4996 + }, + { + "epoch": 0.5487590599604656, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5422801971435547, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6897200345993042, + "num_tokens": 129416078.0, + "step": 4997 + }, + { + "epoch": 0.5488688776630792, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.9235801696777344, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6847765445709229, + "num_tokens": 129437106.0, + "step": 4998 + }, + { + "epoch": 0.548978695365693, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.589467763900757, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6994320154190063, + "num_tokens": 129469652.0, + "step": 4999 + }, + { + "epoch": 0.5490885130683066, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.54091477394104, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7058655023574829, + "num_tokens": 129496588.0, + "step": 5000 + }, + { + "epoch": 0.5491983307709203, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.5452871322631836, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.70197594165802, + "num_tokens": 129521502.0, + "step": 5001 + }, + { + "epoch": 0.5493081484735339, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.6611337661743164, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.714809000492096, + "num_tokens": 129544748.0, + "step": 5002 + }, + { + "epoch": 0.5494179661761476, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.393808126449585, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6986546516418457, + "num_tokens": 129574117.0, + "step": 5003 + }, + { + "epoch": 0.5495277838787612, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 3.24289608001709, + "learning_rate": 1e-06, + "loss": 1.163, + "mean_token_accuracy": 0.6824312210083008, + "num_tokens": 129595211.0, + "step": 5004 + }, + { + "epoch": 0.5496376015813749, + "ewc_loss": 7.12275505065918e-06, + "grad_norm": 2.7216036319732666, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.697310209274292, + "num_tokens": 129619967.0, + "step": 5005 + }, + { + "epoch": 0.5497474192839886, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.611199378967285, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7081935405731201, + "num_tokens": 129645294.0, + "step": 5006 + }, + { + "epoch": 0.5498572369866023, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.885117530822754, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7000799775123596, + "num_tokens": 129664580.0, + "step": 5007 + }, + { + "epoch": 0.5499670546892159, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.637897253036499, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6889481544494629, + "num_tokens": 129691874.0, + "step": 5008 + }, + { + "epoch": 0.5500768723918296, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.592745542526245, + "learning_rate": 1e-06, + "loss": 1.1137, + "mean_token_accuracy": 0.683244526386261, + "num_tokens": 129719829.0, + "step": 5009 + }, + { + "epoch": 0.5501866900944432, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5640347003936768, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7237750291824341, + "num_tokens": 129743967.0, + "step": 5010 + }, + { + "epoch": 0.5502965077970569, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.680168867111206, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7132351994514465, + "num_tokens": 129767756.0, + "step": 5011 + }, + { + "epoch": 0.5504063254996705, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5524356365203857, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7057431936264038, + "num_tokens": 129790423.0, + "step": 5012 + }, + { + "epoch": 0.5505161432022843, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6157119274139404, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6896535158157349, + "num_tokens": 129815734.0, + "step": 5013 + }, + { + "epoch": 0.5506259609048979, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.4471163749694824, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7080754041671753, + "num_tokens": 129842341.0, + "step": 5014 + }, + { + "epoch": 0.5507357786075116, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.3949081897735596, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.699255108833313, + "num_tokens": 129873370.0, + "step": 5015 + }, + { + "epoch": 0.5508455963101252, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.711848735809326, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6973461508750916, + "num_tokens": 129897220.0, + "step": 5016 + }, + { + "epoch": 0.5509554140127388, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6052300930023193, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6958560347557068, + "num_tokens": 129922265.0, + "step": 5017 + }, + { + "epoch": 0.5510652317153525, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.7046351432800293, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6985672116279602, + "num_tokens": 129946103.0, + "step": 5018 + }, + { + "epoch": 0.5511750494179661, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.452387809753418, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6951882243156433, + "num_tokens": 129973655.0, + "step": 5019 + }, + { + "epoch": 0.5512848671205799, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.4768996238708496, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7217841744422913, + "num_tokens": 129998339.0, + "step": 5020 + }, + { + "epoch": 0.5513946848231935, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.3635663986206055, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.6867960691452026, + "num_tokens": 130031600.0, + "step": 5021 + }, + { + "epoch": 0.5515045025258072, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.701040029525757, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6925944089889526, + "num_tokens": 130055600.0, + "step": 5022 + }, + { + "epoch": 0.5516143202284208, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.747279405593872, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7156234979629517, + "num_tokens": 130077860.0, + "step": 5023 + }, + { + "epoch": 0.5517241379310345, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6842713356018066, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7089803814888, + "num_tokens": 130102686.0, + "step": 5024 + }, + { + "epoch": 0.5518339556336481, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.7866241931915283, + "learning_rate": 1e-06, + "loss": 1.0903, + "mean_token_accuracy": 0.688098669052124, + "num_tokens": 130125007.0, + "step": 5025 + }, + { + "epoch": 0.5519437733362618, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.3060033321380615, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.697937548160553, + "num_tokens": 130154385.0, + "step": 5026 + }, + { + "epoch": 0.5520535910388754, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6615960597991943, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7131661176681519, + "num_tokens": 130179220.0, + "step": 5027 + }, + { + "epoch": 0.5521634087414892, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.6239817142486572, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.6982816457748413, + "num_tokens": 130204245.0, + "step": 5028 + }, + { + "epoch": 0.5522732264441028, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.333010196685791, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7157262563705444, + "num_tokens": 130237468.0, + "step": 5029 + }, + { + "epoch": 0.5523830441467165, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.73574161529541, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6936060190200806, + "num_tokens": 130261337.0, + "step": 5030 + }, + { + "epoch": 0.5524928618493301, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.6260626316070557, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6855915784835815, + "num_tokens": 130292001.0, + "step": 5031 + }, + { + "epoch": 0.5526026795519438, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.739443778991699, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6961604356765747, + "num_tokens": 130315471.0, + "step": 5032 + }, + { + "epoch": 0.5527124972545574, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5597450733184814, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6930639147758484, + "num_tokens": 130343018.0, + "step": 5033 + }, + { + "epoch": 0.552822314957171, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.613358497619629, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.688299834728241, + "num_tokens": 130368875.0, + "step": 5034 + }, + { + "epoch": 0.5529321326597848, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.3781137466430664, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.6970311403274536, + "num_tokens": 130396092.0, + "step": 5035 + }, + { + "epoch": 0.5530419503623984, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.660752058029175, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6894208192825317, + "num_tokens": 130420468.0, + "step": 5036 + }, + { + "epoch": 0.5531517680650121, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.911273241043091, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7011733055114746, + "num_tokens": 130440923.0, + "step": 5037 + }, + { + "epoch": 0.5532615857676257, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.6412193775177, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7251958847045898, + "num_tokens": 130465065.0, + "step": 5038 + }, + { + "epoch": 0.5533714034702394, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.840467691421509, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6815951466560364, + "num_tokens": 130487004.0, + "step": 5039 + }, + { + "epoch": 0.553481221172853, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5857431888580322, + "learning_rate": 1e-06, + "loss": 1.108, + "mean_token_accuracy": 0.6723871231079102, + "num_tokens": 130514627.0, + "step": 5040 + }, + { + "epoch": 0.5535910388754667, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 3.008604049682617, + "learning_rate": 1e-06, + "loss": 1.0833, + "mean_token_accuracy": 0.6822580099105835, + "num_tokens": 130536856.0, + "step": 5041 + }, + { + "epoch": 0.5537008565780804, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.587829828262329, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7035290002822876, + "num_tokens": 130562912.0, + "step": 5042 + }, + { + "epoch": 0.5538106742806941, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5517728328704834, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7041774392127991, + "num_tokens": 130589178.0, + "step": 5043 + }, + { + "epoch": 0.5539204919833077, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5423176288604736, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.6993356943130493, + "num_tokens": 130617557.0, + "step": 5044 + }, + { + "epoch": 0.5540303096859214, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.7707631587982178, + "learning_rate": 1e-06, + "loss": 1.1062, + "mean_token_accuracy": 0.6788450479507446, + "num_tokens": 130645531.0, + "step": 5045 + }, + { + "epoch": 0.554140127388535, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.879952907562256, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.6996657848358154, + "num_tokens": 130666322.0, + "step": 5046 + }, + { + "epoch": 0.5542499450911487, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.8066625595092773, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.712588369846344, + "num_tokens": 130689313.0, + "step": 5047 + }, + { + "epoch": 0.5543597627937623, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.611027956008911, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7032532095909119, + "num_tokens": 130717803.0, + "step": 5048 + }, + { + "epoch": 0.5544695804963761, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.3715763092041016, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6879976391792297, + "num_tokens": 130747161.0, + "step": 5049 + }, + { + "epoch": 0.5545793981989897, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.8143391609191895, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.696083664894104, + "num_tokens": 130768925.0, + "step": 5050 + }, + { + "epoch": 0.5546892159016034, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5857279300689697, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6995879411697388, + "num_tokens": 130792964.0, + "step": 5051 + }, + { + "epoch": 0.554799033604217, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.890501022338867, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7133771181106567, + "num_tokens": 130813028.0, + "step": 5052 + }, + { + "epoch": 0.5549088513068307, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5488221645355225, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6972200870513916, + "num_tokens": 130838883.0, + "step": 5053 + }, + { + "epoch": 0.5550186690094443, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.555110216140747, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6933431625366211, + "num_tokens": 130866937.0, + "step": 5054 + }, + { + "epoch": 0.555128486712058, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5710132122039795, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6866415739059448, + "num_tokens": 130891268.0, + "step": 5055 + }, + { + "epoch": 0.5552383044146716, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.535884380340576, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6938861608505249, + "num_tokens": 130920607.0, + "step": 5056 + }, + { + "epoch": 0.5553481221172853, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.6780667304992676, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.699751615524292, + "num_tokens": 130947053.0, + "step": 5057 + }, + { + "epoch": 0.555457939819899, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 3.2878122329711914, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6906512975692749, + "num_tokens": 130968392.0, + "step": 5058 + }, + { + "epoch": 0.5555677575225126, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.799755573272705, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.7251603603363037, + "num_tokens": 130990351.0, + "step": 5059 + }, + { + "epoch": 0.5556775752251263, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.654466390609741, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7047550678253174, + "num_tokens": 131016161.0, + "step": 5060 + }, + { + "epoch": 0.5557873929277399, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.8176023960113525, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7042967081069946, + "num_tokens": 131037117.0, + "step": 5061 + }, + { + "epoch": 0.5558972106303536, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.7832562923431396, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.6935948133468628, + "num_tokens": 131060013.0, + "step": 5062 + }, + { + "epoch": 0.5560070283329672, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.884129047393799, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6912730932235718, + "num_tokens": 131084337.0, + "step": 5063 + }, + { + "epoch": 0.556116846035581, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.433335304260254, + "learning_rate": 1e-06, + "loss": 1.1614, + "mean_token_accuracy": 0.6659689545631409, + "num_tokens": 131116657.0, + "step": 5064 + }, + { + "epoch": 0.5562266637381946, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5405964851379395, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7009637951850891, + "num_tokens": 131140951.0, + "step": 5065 + }, + { + "epoch": 0.5563364814408083, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.6360154151916504, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.6996344327926636, + "num_tokens": 131166105.0, + "step": 5066 + }, + { + "epoch": 0.5564462991434219, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5399909019470215, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6898860335350037, + "num_tokens": 131197081.0, + "step": 5067 + }, + { + "epoch": 0.5565561168460356, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.8819949626922607, + "learning_rate": 1e-06, + "loss": 0.947, + "mean_token_accuracy": 0.7204922437667847, + "num_tokens": 131218222.0, + "step": 5068 + }, + { + "epoch": 0.5566659345486492, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.6754019260406494, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6911046504974365, + "num_tokens": 131242354.0, + "step": 5069 + }, + { + "epoch": 0.5567757522512629, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5114457607269287, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7042372822761536, + "num_tokens": 131267390.0, + "step": 5070 + }, + { + "epoch": 0.5568855699538766, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.515589714050293, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6979671120643616, + "num_tokens": 131295486.0, + "step": 5071 + }, + { + "epoch": 0.5569953876564903, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.7819390296936035, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7027173638343811, + "num_tokens": 131317985.0, + "step": 5072 + }, + { + "epoch": 0.5571052053591039, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 3.0299859046936035, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7159503698348999, + "num_tokens": 131336640.0, + "step": 5073 + }, + { + "epoch": 0.5572150230617176, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5940418243408203, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.6853551268577576, + "num_tokens": 131361817.0, + "step": 5074 + }, + { + "epoch": 0.5573248407643312, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.6008453369140625, + "learning_rate": 1e-06, + "loss": 0.9267, + "mean_token_accuracy": 0.7233002185821533, + "num_tokens": 131388140.0, + "step": 5075 + }, + { + "epoch": 0.5574346584669448, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.8736572265625, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.6804527044296265, + "num_tokens": 131414579.0, + "step": 5076 + }, + { + "epoch": 0.5575444761695585, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.4445576667785645, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7125420570373535, + "num_tokens": 131445760.0, + "step": 5077 + }, + { + "epoch": 0.5576542938721722, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.4972453117370605, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6951205730438232, + "num_tokens": 131472746.0, + "step": 5078 + }, + { + "epoch": 0.5577641115747859, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.650758981704712, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7094658017158508, + "num_tokens": 131494842.0, + "step": 5079 + }, + { + "epoch": 0.5578739292773995, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.8494691848754883, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.6963812708854675, + "num_tokens": 131517296.0, + "step": 5080 + }, + { + "epoch": 0.5579837469800132, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.603219509124756, + "learning_rate": 1e-06, + "loss": 1.1101, + "mean_token_accuracy": 0.6771268248558044, + "num_tokens": 131546387.0, + "step": 5081 + }, + { + "epoch": 0.5580935646826268, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 3.0883588790893555, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7081037759780884, + "num_tokens": 131564509.0, + "step": 5082 + }, + { + "epoch": 0.5582033823852405, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.7315659523010254, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6965572834014893, + "num_tokens": 131588213.0, + "step": 5083 + }, + { + "epoch": 0.5583132000878541, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.4775493144989014, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.6898686289787292, + "num_tokens": 131616361.0, + "step": 5084 + }, + { + "epoch": 0.5584230177904678, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.3714587688446045, + "learning_rate": 1e-06, + "loss": 1.0916, + "mean_token_accuracy": 0.684515118598938, + "num_tokens": 131646562.0, + "step": 5085 + }, + { + "epoch": 0.5585328354930815, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5584471225738525, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7063059210777283, + "num_tokens": 131670099.0, + "step": 5086 + }, + { + "epoch": 0.5586426531956952, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5271923542022705, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6867566108703613, + "num_tokens": 131697822.0, + "step": 5087 + }, + { + "epoch": 0.5587524708983088, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.7014448642730713, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7075486183166504, + "num_tokens": 131721692.0, + "step": 5088 + }, + { + "epoch": 0.5588622886009225, + "ewc_loss": 7.241964340209961e-06, + "grad_norm": 2.8548479080200195, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.700627326965332, + "num_tokens": 131743548.0, + "step": 5089 + }, + { + "epoch": 0.5589721063035361, + "ewc_loss": 7.241964340209961e-06, + "grad_norm": 2.5970616340637207, + "learning_rate": 1e-06, + "loss": 1.1907, + "mean_token_accuracy": 0.6644237041473389, + "num_tokens": 131769790.0, + "step": 5090 + }, + { + "epoch": 0.5590819240061498, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.402320146560669, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6915819644927979, + "num_tokens": 131799241.0, + "step": 5091 + }, + { + "epoch": 0.5591917417087634, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.0913403034210205, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6836349964141846, + "num_tokens": 131839024.0, + "step": 5092 + }, + { + "epoch": 0.5593015594113772, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.4313805103302, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6803457736968994, + "num_tokens": 131866780.0, + "step": 5093 + }, + { + "epoch": 0.5594113771139908, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.3796324729919434, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6949648261070251, + "num_tokens": 131894181.0, + "step": 5094 + }, + { + "epoch": 0.5595211948166045, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.9914400577545166, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6909487843513489, + "num_tokens": 131914061.0, + "step": 5095 + }, + { + "epoch": 0.5596310125192181, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.4123475551605225, + "learning_rate": 1e-06, + "loss": 1.1487, + "mean_token_accuracy": 0.670881986618042, + "num_tokens": 131944439.0, + "step": 5096 + }, + { + "epoch": 0.5597408302218317, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.433361053466797, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.719550609588623, + "num_tokens": 131969212.0, + "step": 5097 + }, + { + "epoch": 0.5598506479244454, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.5400142669677734, + "learning_rate": 1e-06, + "loss": 1.1991, + "mean_token_accuracy": 0.6553202867507935, + "num_tokens": 131998108.0, + "step": 5098 + }, + { + "epoch": 0.559960465627059, + "ewc_loss": 7.152557373046875e-06, + "grad_norm": 2.2955052852630615, + "learning_rate": 1e-06, + "loss": 1.0969, + "mean_token_accuracy": 0.6741923093795776, + "num_tokens": 132031655.0, + "step": 5099 + }, + { + "epoch": 0.5600702833296728, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.733041763305664, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7087606191635132, + "num_tokens": 132051282.0, + "step": 5100 + }, + { + "epoch": 0.5601801010322864, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.3927502632141113, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6931443214416504, + "num_tokens": 132075903.0, + "step": 5101 + }, + { + "epoch": 0.5602899187349001, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.6404049396514893, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6943125128746033, + "num_tokens": 132101575.0, + "step": 5102 + }, + { + "epoch": 0.5603997364375137, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.773833990097046, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6964274048805237, + "num_tokens": 132127711.0, + "step": 5103 + }, + { + "epoch": 0.5605095541401274, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.7818753719329834, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7132638692855835, + "num_tokens": 132150607.0, + "step": 5104 + }, + { + "epoch": 0.560619371842741, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.696000337600708, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.6987413763999939, + "num_tokens": 132174942.0, + "step": 5105 + }, + { + "epoch": 0.5607291895453547, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.6346168518066406, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7154284715652466, + "num_tokens": 132197226.0, + "step": 5106 + }, + { + "epoch": 0.5608390072479684, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.4539270401000977, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7117902040481567, + "num_tokens": 132224700.0, + "step": 5107 + }, + { + "epoch": 0.5609488249505821, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5971062183380127, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6866719722747803, + "num_tokens": 132249511.0, + "step": 5108 + }, + { + "epoch": 0.5610586426531957, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.454038143157959, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6818133592605591, + "num_tokens": 132276127.0, + "step": 5109 + }, + { + "epoch": 0.5611684603558094, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.3547024726867676, + "learning_rate": 1e-06, + "loss": 1.1265, + "mean_token_accuracy": 0.6724298000335693, + "num_tokens": 132306672.0, + "step": 5110 + }, + { + "epoch": 0.561278278058423, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5549962520599365, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.7006628513336182, + "num_tokens": 132333910.0, + "step": 5111 + }, + { + "epoch": 0.5613880957610367, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.6654088497161865, + "learning_rate": 1e-06, + "loss": 1.1262, + "mean_token_accuracy": 0.6815363764762878, + "num_tokens": 132363479.0, + "step": 5112 + }, + { + "epoch": 0.5614979134636503, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.460660934448242, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7005059719085693, + "num_tokens": 132390880.0, + "step": 5113 + }, + { + "epoch": 0.5616077311662641, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5667030811309814, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.6994625329971313, + "num_tokens": 132416496.0, + "step": 5114 + }, + { + "epoch": 0.5617175488688777, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.525946617126465, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7055588960647583, + "num_tokens": 132441545.0, + "step": 5115 + }, + { + "epoch": 0.5618273665714913, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5717382431030273, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7161564230918884, + "num_tokens": 132466047.0, + "step": 5116 + }, + { + "epoch": 0.561937184274105, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.612433671951294, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7191654443740845, + "num_tokens": 132487881.0, + "step": 5117 + }, + { + "epoch": 0.5620470019767186, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.583019495010376, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.713976263999939, + "num_tokens": 132515071.0, + "step": 5118 + }, + { + "epoch": 0.5621568196793323, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.599353313446045, + "learning_rate": 1e-06, + "loss": 1.1075, + "mean_token_accuracy": 0.6791973114013672, + "num_tokens": 132539234.0, + "step": 5119 + }, + { + "epoch": 0.5622666373819459, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.4278945922851562, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6852438449859619, + "num_tokens": 132569368.0, + "step": 5120 + }, + { + "epoch": 0.5623764550845596, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5412726402282715, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6985477209091187, + "num_tokens": 132595696.0, + "step": 5121 + }, + { + "epoch": 0.5624862727871733, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5001425743103027, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6887818574905396, + "num_tokens": 132627634.0, + "step": 5122 + }, + { + "epoch": 0.562596090489787, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.676086664199829, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.692916750907898, + "num_tokens": 132652802.0, + "step": 5123 + }, + { + "epoch": 0.5627059081924006, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.4736218452453613, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.722068190574646, + "num_tokens": 132677330.0, + "step": 5124 + }, + { + "epoch": 0.5628157258950143, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.529092788696289, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.7011418342590332, + "num_tokens": 132703704.0, + "step": 5125 + }, + { + "epoch": 0.5629255435976279, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.580284833908081, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7093300819396973, + "num_tokens": 132731686.0, + "step": 5126 + }, + { + "epoch": 0.5630353613002416, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.4251515865325928, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.7085164189338684, + "num_tokens": 132761771.0, + "step": 5127 + }, + { + "epoch": 0.5631451790028552, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5664546489715576, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6863370537757874, + "num_tokens": 132786439.0, + "step": 5128 + }, + { + "epoch": 0.563254996705469, + "ewc_loss": 7.212162017822266e-06, + "grad_norm": 2.5295538902282715, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7189960479736328, + "num_tokens": 132811005.0, + "step": 5129 + }, + { + "epoch": 0.5633648144080826, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5014400482177734, + "learning_rate": 1e-06, + "loss": 1.1651, + "mean_token_accuracy": 0.6649864315986633, + "num_tokens": 132839399.0, + "step": 5130 + }, + { + "epoch": 0.5634746321106963, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.5502285957336426, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6844040155410767, + "num_tokens": 132867618.0, + "step": 5131 + }, + { + "epoch": 0.5635844498133099, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.2387351989746094, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6846401691436768, + "num_tokens": 132901567.0, + "step": 5132 + }, + { + "epoch": 0.5636942675159236, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.6555700302124023, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7123085856437683, + "num_tokens": 132925044.0, + "step": 5133 + }, + { + "epoch": 0.5638040852185372, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.8864831924438477, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7189091444015503, + "num_tokens": 132944209.0, + "step": 5134 + }, + { + "epoch": 0.5639139029211508, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.354808807373047, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6968085169792175, + "num_tokens": 132973354.0, + "step": 5135 + }, + { + "epoch": 0.5640237206237646, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.9454424381256104, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7284737825393677, + "num_tokens": 132992898.0, + "step": 5136 + }, + { + "epoch": 0.5641335383263782, + "ewc_loss": 7.18235969543457e-06, + "grad_norm": 2.7319939136505127, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6947610378265381, + "num_tokens": 133013935.0, + "step": 5137 + }, + { + "epoch": 0.5642433560289919, + "ewc_loss": 7.241964340209961e-06, + "grad_norm": 2.9192023277282715, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7195607423782349, + "num_tokens": 133033719.0, + "step": 5138 + }, + { + "epoch": 0.5643531737316055, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.4935462474823, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6943032741546631, + "num_tokens": 133062200.0, + "step": 5139 + }, + { + "epoch": 0.5644629914342192, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5010907649993896, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6927095651626587, + "num_tokens": 133090360.0, + "step": 5140 + }, + { + "epoch": 0.5645728091368328, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.587599277496338, + "learning_rate": 1e-06, + "loss": 1.1015, + "mean_token_accuracy": 0.691344678401947, + "num_tokens": 133117850.0, + "step": 5141 + }, + { + "epoch": 0.5646826268394465, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5858418941497803, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7111871838569641, + "num_tokens": 133142324.0, + "step": 5142 + }, + { + "epoch": 0.5647924445420602, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.691347360610962, + "learning_rate": 1e-06, + "loss": 0.9044, + "mean_token_accuracy": 0.7363081574440002, + "num_tokens": 133163820.0, + "step": 5143 + }, + { + "epoch": 0.5649022622446739, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.755681037902832, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7100646495819092, + "num_tokens": 133190171.0, + "step": 5144 + }, + { + "epoch": 0.5650120799472875, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.8301444053649902, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7095932364463806, + "num_tokens": 133211488.0, + "step": 5145 + }, + { + "epoch": 0.5651218976499012, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.7781221866607666, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7019312381744385, + "num_tokens": 133232643.0, + "step": 5146 + }, + { + "epoch": 0.5652317153525148, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.946594476699829, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7069092988967896, + "num_tokens": 133250041.0, + "step": 5147 + }, + { + "epoch": 0.5653415330551285, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6810312271118164, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6874152421951294, + "num_tokens": 133274638.0, + "step": 5148 + }, + { + "epoch": 0.5654513507577421, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.463975667953491, + "learning_rate": 1e-06, + "loss": 1.1081, + "mean_token_accuracy": 0.6869556903839111, + "num_tokens": 133307670.0, + "step": 5149 + }, + { + "epoch": 0.5655611684603558, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5755198001861572, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.6997685432434082, + "num_tokens": 133332989.0, + "step": 5150 + }, + { + "epoch": 0.5656709861629695, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4459354877471924, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7031481266021729, + "num_tokens": 133360939.0, + "step": 5151 + }, + { + "epoch": 0.5657808038655832, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4408013820648193, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7068367004394531, + "num_tokens": 133388730.0, + "step": 5152 + }, + { + "epoch": 0.5658906215681968, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.439789295196533, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6959773302078247, + "num_tokens": 133416971.0, + "step": 5153 + }, + { + "epoch": 0.5660004392708105, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.523430585861206, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6842858791351318, + "num_tokens": 133444888.0, + "step": 5154 + }, + { + "epoch": 0.5661102569734241, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.889740228652954, + "learning_rate": 1e-06, + "loss": 1.13, + "mean_token_accuracy": 0.6788058876991272, + "num_tokens": 133471556.0, + "step": 5155 + }, + { + "epoch": 0.5662200746760377, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.3512847423553467, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6900810599327087, + "num_tokens": 133504244.0, + "step": 5156 + }, + { + "epoch": 0.5663298923786514, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.469665288925171, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6808191537857056, + "num_tokens": 133533015.0, + "step": 5157 + }, + { + "epoch": 0.5664397100812651, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.846914768218994, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6883491277694702, + "num_tokens": 133553724.0, + "step": 5158 + }, + { + "epoch": 0.5665495277838788, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.2361457347869873, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6817832589149475, + "num_tokens": 133586088.0, + "step": 5159 + }, + { + "epoch": 0.5666593454864924, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.607182741165161, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6945756673812866, + "num_tokens": 133610842.0, + "step": 5160 + }, + { + "epoch": 0.5667691631891061, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6957249641418457, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6962276101112366, + "num_tokens": 133632718.0, + "step": 5161 + }, + { + "epoch": 0.5668789808917197, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.4311909675598145, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6903256177902222, + "num_tokens": 133657697.0, + "step": 5162 + }, + { + "epoch": 0.5669887985943334, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5462019443511963, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6983543634414673, + "num_tokens": 133680641.0, + "step": 5163 + }, + { + "epoch": 0.567098616296947, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.477968454360962, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.718294620513916, + "num_tokens": 133707898.0, + "step": 5164 + }, + { + "epoch": 0.5672084339995608, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.731785535812378, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6992210745811462, + "num_tokens": 133729956.0, + "step": 5165 + }, + { + "epoch": 0.5673182517021744, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.650235176086426, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.727668821811676, + "num_tokens": 133756836.0, + "step": 5166 + }, + { + "epoch": 0.5674280694047881, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.7358999252319336, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7124552130699158, + "num_tokens": 133778886.0, + "step": 5167 + }, + { + "epoch": 0.5675378871074017, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5791258811950684, + "learning_rate": 1e-06, + "loss": 1.1079, + "mean_token_accuracy": 0.6801697611808777, + "num_tokens": 133809059.0, + "step": 5168 + }, + { + "epoch": 0.5676477048100154, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.3297231197357178, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7015950679779053, + "num_tokens": 133839123.0, + "step": 5169 + }, + { + "epoch": 0.567757522512629, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.260305643081665, + "learning_rate": 1e-06, + "loss": 0.9059, + "mean_token_accuracy": 0.7300723195075989, + "num_tokens": 133866533.0, + "step": 5170 + }, + { + "epoch": 0.5678673402152427, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.3759472370147705, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.684745728969574, + "num_tokens": 133897919.0, + "step": 5171 + }, + { + "epoch": 0.5679771579178564, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.375410795211792, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6932336688041687, + "num_tokens": 133928244.0, + "step": 5172 + }, + { + "epoch": 0.5680869756204701, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.495197057723999, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6864590644836426, + "num_tokens": 133956029.0, + "step": 5173 + }, + { + "epoch": 0.5681967933230837, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.526890516281128, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6775314211845398, + "num_tokens": 133986082.0, + "step": 5174 + }, + { + "epoch": 0.5683066110256974, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5333027839660645, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6952141523361206, + "num_tokens": 134011899.0, + "step": 5175 + }, + { + "epoch": 0.568416428728311, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 3.344226837158203, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7035305500030518, + "num_tokens": 134032637.0, + "step": 5176 + }, + { + "epoch": 0.5685262464309246, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6263208389282227, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.7018972635269165, + "num_tokens": 134056987.0, + "step": 5177 + }, + { + "epoch": 0.5686360641335383, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6328558921813965, + "learning_rate": 1e-06, + "loss": 1.0829, + "mean_token_accuracy": 0.686845064163208, + "num_tokens": 134081097.0, + "step": 5178 + }, + { + "epoch": 0.5687458818361519, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.7265360355377197, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7035256624221802, + "num_tokens": 134104706.0, + "step": 5179 + }, + { + "epoch": 0.5688556995387657, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5327486991882324, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6817741394042969, + "num_tokens": 134131050.0, + "step": 5180 + }, + { + "epoch": 0.5689655172413793, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5455527305603027, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6889338493347168, + "num_tokens": 134157262.0, + "step": 5181 + }, + { + "epoch": 0.569075334943993, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.458238124847412, + "learning_rate": 1e-06, + "loss": 1.1893, + "mean_token_accuracy": 0.6644434928894043, + "num_tokens": 134188743.0, + "step": 5182 + }, + { + "epoch": 0.5691851526466066, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.536837339401245, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6997324228286743, + "num_tokens": 134215418.0, + "step": 5183 + }, + { + "epoch": 0.5692949703492203, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.3838584423065186, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7041130661964417, + "num_tokens": 134248000.0, + "step": 5184 + }, + { + "epoch": 0.5694047880518339, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.7078135013580322, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7069119215011597, + "num_tokens": 134271086.0, + "step": 5185 + }, + { + "epoch": 0.5695146057544476, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6962099075317383, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7117022275924683, + "num_tokens": 134293514.0, + "step": 5186 + }, + { + "epoch": 0.5696244234570613, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.502392053604126, + "learning_rate": 1e-06, + "loss": 1.1041, + "mean_token_accuracy": 0.6810492277145386, + "num_tokens": 134324497.0, + "step": 5187 + }, + { + "epoch": 0.569734241159675, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6087429523468018, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.704820990562439, + "num_tokens": 134348726.0, + "step": 5188 + }, + { + "epoch": 0.5698440588622886, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5398733615875244, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7125064730644226, + "num_tokens": 134375115.0, + "step": 5189 + }, + { + "epoch": 0.5699538765649023, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.740802049636841, + "learning_rate": 1e-06, + "loss": 1.1261, + "mean_token_accuracy": 0.6802225112915039, + "num_tokens": 134397922.0, + "step": 5190 + }, + { + "epoch": 0.5700636942675159, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.388418436050415, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7174252271652222, + "num_tokens": 134425011.0, + "step": 5191 + }, + { + "epoch": 0.5701735119701296, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.697315216064453, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6914740800857544, + "num_tokens": 134452007.0, + "step": 5192 + }, + { + "epoch": 0.5702833296727432, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5044822692871094, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6952752470970154, + "num_tokens": 134480460.0, + "step": 5193 + }, + { + "epoch": 0.570393147375357, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5135087966918945, + "learning_rate": 1e-06, + "loss": 1.1449, + "mean_token_accuracy": 0.6727377772331238, + "num_tokens": 134507478.0, + "step": 5194 + }, + { + "epoch": 0.5705029650779706, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6699516773223877, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6849657297134399, + "num_tokens": 134531741.0, + "step": 5195 + }, + { + "epoch": 0.5706127827805842, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.404613494873047, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7048112154006958, + "num_tokens": 134557636.0, + "step": 5196 + }, + { + "epoch": 0.5707226004831979, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6641008853912354, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.6993765830993652, + "num_tokens": 134579402.0, + "step": 5197 + }, + { + "epoch": 0.5708324181858115, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.601177930831909, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7002601027488708, + "num_tokens": 134606547.0, + "step": 5198 + }, + { + "epoch": 0.5709422358884252, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.695939064025879, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7153171300888062, + "num_tokens": 134630304.0, + "step": 5199 + }, + { + "epoch": 0.5710520535910388, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.5157268047332764, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.6960105299949646, + "num_tokens": 134654483.0, + "step": 5200 + }, + { + "epoch": 0.5711618712936526, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.4090378284454346, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.6770702004432678, + "num_tokens": 134687064.0, + "step": 5201 + }, + { + "epoch": 0.5712716889962662, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.454073905944824, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7050338983535767, + "num_tokens": 134713343.0, + "step": 5202 + }, + { + "epoch": 0.5713815066988799, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.7663300037384033, + "learning_rate": 1e-06, + "loss": 1.1009, + "mean_token_accuracy": 0.6832139492034912, + "num_tokens": 134736151.0, + "step": 5203 + }, + { + "epoch": 0.5714913244014935, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.575672149658203, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7171427011489868, + "num_tokens": 134760650.0, + "step": 5204 + }, + { + "epoch": 0.5716011421041072, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.9091928005218506, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7035301327705383, + "num_tokens": 134780341.0, + "step": 5205 + }, + { + "epoch": 0.5717109598067208, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.59665846824646, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.6972429752349854, + "num_tokens": 134804164.0, + "step": 5206 + }, + { + "epoch": 0.5718207775093345, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6126129627227783, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6984581351280212, + "num_tokens": 134830883.0, + "step": 5207 + }, + { + "epoch": 0.5719305952119481, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6662704944610596, + "learning_rate": 1e-06, + "loss": 1.1161, + "mean_token_accuracy": 0.6804027557373047, + "num_tokens": 134857646.0, + "step": 5208 + }, + { + "epoch": 0.5720404129145619, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.481062889099121, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7029589414596558, + "num_tokens": 134886570.0, + "step": 5209 + }, + { + "epoch": 0.5721502306171755, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.7800655364990234, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7156413793563843, + "num_tokens": 134909165.0, + "step": 5210 + }, + { + "epoch": 0.5722600483197892, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.528587818145752, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.697917640209198, + "num_tokens": 134935723.0, + "step": 5211 + }, + { + "epoch": 0.5723698660224028, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 3.09995698928833, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7064037919044495, + "num_tokens": 134952760.0, + "step": 5212 + }, + { + "epoch": 0.5724796837250165, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5613887310028076, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.701665997505188, + "num_tokens": 134978712.0, + "step": 5213 + }, + { + "epoch": 0.5725895014276301, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.3816521167755127, + "learning_rate": 1e-06, + "loss": 1.1112, + "mean_token_accuracy": 0.6786983609199524, + "num_tokens": 135007948.0, + "step": 5214 + }, + { + "epoch": 0.5726993191302437, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.335353136062622, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6834255456924438, + "num_tokens": 135039118.0, + "step": 5215 + }, + { + "epoch": 0.5728091368328575, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.423750400543213, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.6989283561706543, + "num_tokens": 135076922.0, + "step": 5216 + }, + { + "epoch": 0.5729189545354711, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4498884677886963, + "learning_rate": 1e-06, + "loss": 1.1202, + "mean_token_accuracy": 0.6803604960441589, + "num_tokens": 135106676.0, + "step": 5217 + }, + { + "epoch": 0.5730287722380848, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6871659755706787, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7111169099807739, + "num_tokens": 135127502.0, + "step": 5218 + }, + { + "epoch": 0.5731385899406984, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5982015132904053, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.6776382923126221, + "num_tokens": 135152895.0, + "step": 5219 + }, + { + "epoch": 0.5732484076433121, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.228447437286377, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7031774520874023, + "num_tokens": 135185791.0, + "step": 5220 + }, + { + "epoch": 0.5733582253459257, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.3754003047943115, + "learning_rate": 1e-06, + "loss": 1.0996, + "mean_token_accuracy": 0.6842215657234192, + "num_tokens": 135216237.0, + "step": 5221 + }, + { + "epoch": 0.5734680430485394, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6465137004852295, + "learning_rate": 1e-06, + "loss": 1.126, + "mean_token_accuracy": 0.6769711971282959, + "num_tokens": 135241407.0, + "step": 5222 + }, + { + "epoch": 0.5735778607511531, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6961538791656494, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7168883681297302, + "num_tokens": 135262741.0, + "step": 5223 + }, + { + "epoch": 0.5736876784537668, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.6573715209960938, + "learning_rate": 1e-06, + "loss": 0.91, + "mean_token_accuracy": 0.7282605171203613, + "num_tokens": 135285964.0, + "step": 5224 + }, + { + "epoch": 0.5737974961563804, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6193668842315674, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6869968175888062, + "num_tokens": 135314741.0, + "step": 5225 + }, + { + "epoch": 0.5739073138589941, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.686771869659424, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6975919008255005, + "num_tokens": 135336909.0, + "step": 5226 + }, + { + "epoch": 0.5740171315616077, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.486703395843506, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7226329445838928, + "num_tokens": 135364698.0, + "step": 5227 + }, + { + "epoch": 0.5741269492642214, + "ewc_loss": 7.271766662597656e-06, + "grad_norm": 2.745311737060547, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.6984118819236755, + "num_tokens": 135389786.0, + "step": 5228 + }, + { + "epoch": 0.574236766966835, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5650453567504883, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7123942375183105, + "num_tokens": 135413067.0, + "step": 5229 + }, + { + "epoch": 0.5743465846694488, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.858560562133789, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7006743550300598, + "num_tokens": 135436186.0, + "step": 5230 + }, + { + "epoch": 0.5744564023720624, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.7236874103546143, + "learning_rate": 1e-06, + "loss": 0.9282, + "mean_token_accuracy": 0.7273669838905334, + "num_tokens": 135457201.0, + "step": 5231 + }, + { + "epoch": 0.5745662200746761, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.8764612674713135, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7062440514564514, + "num_tokens": 135481982.0, + "step": 5232 + }, + { + "epoch": 0.5746760377772897, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.7608985900878906, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6880950927734375, + "num_tokens": 135505097.0, + "step": 5233 + }, + { + "epoch": 0.5747858554799034, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4408187866210938, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.690721333026886, + "num_tokens": 135533256.0, + "step": 5234 + }, + { + "epoch": 0.574895673182517, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.565107583999634, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7140979170799255, + "num_tokens": 135559313.0, + "step": 5235 + }, + { + "epoch": 0.5750054908851306, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.996220350265503, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7069029211997986, + "num_tokens": 135583174.0, + "step": 5236 + }, + { + "epoch": 0.5751153085877443, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5887551307678223, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.699090301990509, + "num_tokens": 135610717.0, + "step": 5237 + }, + { + "epoch": 0.575225126290358, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5871493816375732, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.693519115447998, + "num_tokens": 135635496.0, + "step": 5238 + }, + { + "epoch": 0.5753349439929717, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.385392665863037, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7173610925674438, + "num_tokens": 135662797.0, + "step": 5239 + }, + { + "epoch": 0.5754447616955853, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.8494513034820557, + "learning_rate": 1e-06, + "loss": 0.9212, + "mean_token_accuracy": 0.7348921895027161, + "num_tokens": 135681463.0, + "step": 5240 + }, + { + "epoch": 0.575554579398199, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.303281307220459, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6922801733016968, + "num_tokens": 135712559.0, + "step": 5241 + }, + { + "epoch": 0.5756643971008126, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 3.0162649154663086, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7106416821479797, + "num_tokens": 135735649.0, + "step": 5242 + }, + { + "epoch": 0.5757742148034263, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5229783058166504, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6933588981628418, + "num_tokens": 135763058.0, + "step": 5243 + }, + { + "epoch": 0.5758840325060399, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6899516582489014, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7082690000534058, + "num_tokens": 135785873.0, + "step": 5244 + }, + { + "epoch": 0.5759938502086537, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.77650785446167, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7130423188209534, + "num_tokens": 135809402.0, + "step": 5245 + }, + { + "epoch": 0.5761036679112673, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.748950958251953, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6821125745773315, + "num_tokens": 135832842.0, + "step": 5246 + }, + { + "epoch": 0.576213485613881, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6718909740448, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6937394738197327, + "num_tokens": 135859345.0, + "step": 5247 + }, + { + "epoch": 0.5763233033164946, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4700446128845215, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.7018637657165527, + "num_tokens": 135885536.0, + "step": 5248 + }, + { + "epoch": 0.5764331210191083, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.589167356491089, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6767650842666626, + "num_tokens": 135912140.0, + "step": 5249 + }, + { + "epoch": 0.5765429387217219, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.494067430496216, + "learning_rate": 1e-06, + "loss": 1.1565, + "mean_token_accuracy": 0.6746364235877991, + "num_tokens": 135942276.0, + "step": 5250 + }, + { + "epoch": 0.5766527564243356, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.379026412963867, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7040297389030457, + "num_tokens": 135969380.0, + "step": 5251 + }, + { + "epoch": 0.5767625741269493, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.3253531455993652, + "learning_rate": 1e-06, + "loss": 0.9597, + "mean_token_accuracy": 0.7213889360427856, + "num_tokens": 135997935.0, + "step": 5252 + }, + { + "epoch": 0.576872391829563, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.699678421020508, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6892082691192627, + "num_tokens": 136022158.0, + "step": 5253 + }, + { + "epoch": 0.5769822095321766, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5133039951324463, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7040777206420898, + "num_tokens": 136048515.0, + "step": 5254 + }, + { + "epoch": 0.5770920272347903, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.54095458984375, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7051268219947815, + "num_tokens": 136072347.0, + "step": 5255 + }, + { + "epoch": 0.5772018449374039, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5764758586883545, + "learning_rate": 1e-06, + "loss": 1.1119, + "mean_token_accuracy": 0.6760647296905518, + "num_tokens": 136098672.0, + "step": 5256 + }, + { + "epoch": 0.5773116626400175, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.605391025543213, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.693663477897644, + "num_tokens": 136123095.0, + "step": 5257 + }, + { + "epoch": 0.5774214803426312, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.473928689956665, + "learning_rate": 1e-06, + "loss": 1.0983, + "mean_token_accuracy": 0.6774167418479919, + "num_tokens": 136150673.0, + "step": 5258 + }, + { + "epoch": 0.577531298045245, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.2617294788360596, + "learning_rate": 1e-06, + "loss": 1.111, + "mean_token_accuracy": 0.6838889122009277, + "num_tokens": 136185183.0, + "step": 5259 + }, + { + "epoch": 0.5776411157478586, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4091238975524902, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7053014039993286, + "num_tokens": 136217660.0, + "step": 5260 + }, + { + "epoch": 0.5777509334504722, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.580725908279419, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7094020247459412, + "num_tokens": 136240480.0, + "step": 5261 + }, + { + "epoch": 0.5778607511530859, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6344029903411865, + "learning_rate": 1e-06, + "loss": 1.101, + "mean_token_accuracy": 0.6842278242111206, + "num_tokens": 136265258.0, + "step": 5262 + }, + { + "epoch": 0.5779705688556995, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.682157278060913, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.6992553472518921, + "num_tokens": 136291840.0, + "step": 5263 + }, + { + "epoch": 0.5780803865583132, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4550387859344482, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7082546949386597, + "num_tokens": 136320057.0, + "step": 5264 + }, + { + "epoch": 0.5781902042609268, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.3894569873809814, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.6780548691749573, + "num_tokens": 136350046.0, + "step": 5265 + }, + { + "epoch": 0.5783000219635406, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.514418125152588, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.7063621282577515, + "num_tokens": 136378683.0, + "step": 5266 + }, + { + "epoch": 0.5784098396661542, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.401733160018921, + "learning_rate": 1e-06, + "loss": 1.113, + "mean_token_accuracy": 0.6759921312332153, + "num_tokens": 136408297.0, + "step": 5267 + }, + { + "epoch": 0.5785196573687679, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.405010461807251, + "learning_rate": 1e-06, + "loss": 1.1128, + "mean_token_accuracy": 0.6746421456336975, + "num_tokens": 136438358.0, + "step": 5268 + }, + { + "epoch": 0.5786294750713815, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4862608909606934, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6993579268455505, + "num_tokens": 136464614.0, + "step": 5269 + }, + { + "epoch": 0.5787392927739952, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.519111394882202, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.6988280415534973, + "num_tokens": 136487921.0, + "step": 5270 + }, + { + "epoch": 0.5788491104766088, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.523444414138794, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6948070526123047, + "num_tokens": 136512898.0, + "step": 5271 + }, + { + "epoch": 0.5789589281792225, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.473803758621216, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.7223111391067505, + "num_tokens": 136539048.0, + "step": 5272 + }, + { + "epoch": 0.5790687458818361, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.505568265914917, + "learning_rate": 1e-06, + "loss": 1.123, + "mean_token_accuracy": 0.6720569133758545, + "num_tokens": 136567431.0, + "step": 5273 + }, + { + "epoch": 0.5791785635844499, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.524935245513916, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7081789374351501, + "num_tokens": 136593366.0, + "step": 5274 + }, + { + "epoch": 0.5792883812870635, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 3.1490321159362793, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6858999729156494, + "num_tokens": 136614886.0, + "step": 5275 + }, + { + "epoch": 0.5793981989896771, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4900460243225098, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6833083629608154, + "num_tokens": 136644948.0, + "step": 5276 + }, + { + "epoch": 0.5795080166922908, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.815608501434326, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.7003320455551147, + "num_tokens": 136667190.0, + "step": 5277 + }, + { + "epoch": 0.5796178343949044, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4226086139678955, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7140876650810242, + "num_tokens": 136691470.0, + "step": 5278 + }, + { + "epoch": 0.5797276520975181, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.346890687942505, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6970352530479431, + "num_tokens": 136720589.0, + "step": 5279 + }, + { + "epoch": 0.5798374698001317, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.4308643341064453, + "learning_rate": 1e-06, + "loss": 1.0992, + "mean_token_accuracy": 0.6831130981445312, + "num_tokens": 136751274.0, + "step": 5280 + }, + { + "epoch": 0.5799472875027455, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 3.3597123622894287, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6928661465644836, + "num_tokens": 136772880.0, + "step": 5281 + }, + { + "epoch": 0.5800571052053591, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.6476666927337646, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6975215077400208, + "num_tokens": 136800349.0, + "step": 5282 + }, + { + "epoch": 0.5801669229079728, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.9282279014587402, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7154412865638733, + "num_tokens": 136819943.0, + "step": 5283 + }, + { + "epoch": 0.5802767406105864, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.446350574493408, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6974027752876282, + "num_tokens": 136849882.0, + "step": 5284 + }, + { + "epoch": 0.5803865583132001, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.525994300842285, + "learning_rate": 1e-06, + "loss": 1.1223, + "mean_token_accuracy": 0.6712450981140137, + "num_tokens": 136878293.0, + "step": 5285 + }, + { + "epoch": 0.5804963760158137, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.457422971725464, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6914598941802979, + "num_tokens": 136906377.0, + "step": 5286 + }, + { + "epoch": 0.5806061937184274, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.423816680908203, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6900179982185364, + "num_tokens": 136937276.0, + "step": 5287 + }, + { + "epoch": 0.5807160114210411, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5696353912353516, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.696152925491333, + "num_tokens": 136962371.0, + "step": 5288 + }, + { + "epoch": 0.5808258291236548, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.5922625064849854, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7125604748725891, + "num_tokens": 136989852.0, + "step": 5289 + }, + { + "epoch": 0.5809356468262684, + "ewc_loss": 7.3015689849853516e-06, + "grad_norm": 2.9039454460144043, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7057546377182007, + "num_tokens": 137008127.0, + "step": 5290 + }, + { + "epoch": 0.5810454645288821, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5522971153259277, + "learning_rate": 1e-06, + "loss": 0.8894, + "mean_token_accuracy": 0.738408088684082, + "num_tokens": 137031663.0, + "step": 5291 + }, + { + "epoch": 0.5811552822314957, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.566621780395508, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6937627792358398, + "num_tokens": 137059073.0, + "step": 5292 + }, + { + "epoch": 0.5812650999341094, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 3.269278049468994, + "learning_rate": 1e-06, + "loss": 0.927, + "mean_token_accuracy": 0.7242254614830017, + "num_tokens": 137079263.0, + "step": 5293 + }, + { + "epoch": 0.581374917636723, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.3661274909973145, + "learning_rate": 1e-06, + "loss": 1.119, + "mean_token_accuracy": 0.6762223243713379, + "num_tokens": 137111579.0, + "step": 5294 + }, + { + "epoch": 0.5814847353393368, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5159730911254883, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7173944115638733, + "num_tokens": 137136444.0, + "step": 5295 + }, + { + "epoch": 0.5815945530419504, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.6309561729431152, + "learning_rate": 1e-06, + "loss": 1.1853, + "mean_token_accuracy": 0.6563832759857178, + "num_tokens": 137165499.0, + "step": 5296 + }, + { + "epoch": 0.581704370744564, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.346721649169922, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6932896375656128, + "num_tokens": 137194506.0, + "step": 5297 + }, + { + "epoch": 0.5818141884471777, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4598588943481445, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6961802244186401, + "num_tokens": 137222222.0, + "step": 5298 + }, + { + "epoch": 0.5819240061497913, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.699127197265625, + "learning_rate": 1e-06, + "loss": 1.0915, + "mean_token_accuracy": 0.685276210308075, + "num_tokens": 137248256.0, + "step": 5299 + }, + { + "epoch": 0.582033823852405, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6236515045166016, + "learning_rate": 1e-06, + "loss": 0.9146, + "mean_token_accuracy": 0.7317598462104797, + "num_tokens": 137272022.0, + "step": 5300 + }, + { + "epoch": 0.5821436415550186, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.7158942222595215, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7000465989112854, + "num_tokens": 137294545.0, + "step": 5301 + }, + { + "epoch": 0.5822534592576323, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.933511734008789, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7063740491867065, + "num_tokens": 137314838.0, + "step": 5302 + }, + { + "epoch": 0.582363276960246, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6200125217437744, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6943800449371338, + "num_tokens": 137340879.0, + "step": 5303 + }, + { + "epoch": 0.5824730946628597, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.632986307144165, + "learning_rate": 1e-06, + "loss": 1.1312, + "mean_token_accuracy": 0.6885440349578857, + "num_tokens": 137368376.0, + "step": 5304 + }, + { + "epoch": 0.5825829123654733, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.441361427307129, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6932594776153564, + "num_tokens": 137396813.0, + "step": 5305 + }, + { + "epoch": 0.582692730068087, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.5028016567230225, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.6990605592727661, + "num_tokens": 137421814.0, + "step": 5306 + }, + { + "epoch": 0.5828025477707006, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4636294841766357, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7007744312286377, + "num_tokens": 137448987.0, + "step": 5307 + }, + { + "epoch": 0.5829123654733143, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4161367416381836, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6986496448516846, + "num_tokens": 137478806.0, + "step": 5308 + }, + { + "epoch": 0.5830221831759279, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 20.66619300842285, + "learning_rate": 1e-06, + "loss": 1.0854, + "mean_token_accuracy": 0.6804816722869873, + "num_tokens": 137504325.0, + "step": 5309 + }, + { + "epoch": 0.5831320008785417, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.9686291217803955, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.6994475722312927, + "num_tokens": 137525204.0, + "step": 5310 + }, + { + "epoch": 0.5832418185811553, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.6720221042633057, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7062833309173584, + "num_tokens": 137551618.0, + "step": 5311 + }, + { + "epoch": 0.583351636283769, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.6125717163085938, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7025694251060486, + "num_tokens": 137577095.0, + "step": 5312 + }, + { + "epoch": 0.5834614539863826, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.437770366668701, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7099331617355347, + "num_tokens": 137606454.0, + "step": 5313 + }, + { + "epoch": 0.5835712716889963, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.377983808517456, + "learning_rate": 1e-06, + "loss": 1.1348, + "mean_token_accuracy": 0.6677794456481934, + "num_tokens": 137634449.0, + "step": 5314 + }, + { + "epoch": 0.5836810893916099, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.544879674911499, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6993471384048462, + "num_tokens": 137658097.0, + "step": 5315 + }, + { + "epoch": 0.5837909070942235, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.5260422229766846, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6825278997421265, + "num_tokens": 137683043.0, + "step": 5316 + }, + { + "epoch": 0.5839007247968373, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.831958532333374, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7156213521957397, + "num_tokens": 137703335.0, + "step": 5317 + }, + { + "epoch": 0.584010542499451, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.9023032188415527, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7050856351852417, + "num_tokens": 137722898.0, + "step": 5318 + }, + { + "epoch": 0.5841203602020646, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.290724992752075, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.7035558223724365, + "num_tokens": 137755357.0, + "step": 5319 + }, + { + "epoch": 0.5842301779046782, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.6054763793945312, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.6978437900543213, + "num_tokens": 137785996.0, + "step": 5320 + }, + { + "epoch": 0.5843399956072919, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4966835975646973, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.700951099395752, + "num_tokens": 137812226.0, + "step": 5321 + }, + { + "epoch": 0.5844498133099055, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.2031922340393066, + "learning_rate": 1e-06, + "loss": 1.1879, + "mean_token_accuracy": 0.6607050895690918, + "num_tokens": 137845770.0, + "step": 5322 + }, + { + "epoch": 0.5845596310125192, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4772589206695557, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6940145492553711, + "num_tokens": 137874955.0, + "step": 5323 + }, + { + "epoch": 0.5846694487151329, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.316648006439209, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7062022686004639, + "num_tokens": 137908104.0, + "step": 5324 + }, + { + "epoch": 0.5847792664177466, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5185458660125732, + "learning_rate": 1e-06, + "loss": 1.1241, + "mean_token_accuracy": 0.6782079935073853, + "num_tokens": 137933143.0, + "step": 5325 + }, + { + "epoch": 0.5848890841203602, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.271548271179199, + "learning_rate": 1e-06, + "loss": 1.1102, + "mean_token_accuracy": 0.6783410310745239, + "num_tokens": 137963416.0, + "step": 5326 + }, + { + "epoch": 0.5849989018229739, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4975011348724365, + "learning_rate": 1e-06, + "loss": 1.1428, + "mean_token_accuracy": 0.6657278537750244, + "num_tokens": 137990308.0, + "step": 5327 + }, + { + "epoch": 0.5851087195255875, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5993289947509766, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7204814553260803, + "num_tokens": 138012509.0, + "step": 5328 + }, + { + "epoch": 0.5852185372282012, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.9564058780670166, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7166505455970764, + "num_tokens": 138032504.0, + "step": 5329 + }, + { + "epoch": 0.5853283549308148, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5277116298675537, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.720984935760498, + "num_tokens": 138057651.0, + "step": 5330 + }, + { + "epoch": 0.5854381726334285, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.4536755084991455, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7133808732032776, + "num_tokens": 138086199.0, + "step": 5331 + }, + { + "epoch": 0.5855479903360422, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.629713535308838, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7112262845039368, + "num_tokens": 138113881.0, + "step": 5332 + }, + { + "epoch": 0.5856578080386559, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.883312702178955, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7166577577590942, + "num_tokens": 138135806.0, + "step": 5333 + }, + { + "epoch": 0.5857676257412695, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.600207805633545, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7071378231048584, + "num_tokens": 138159020.0, + "step": 5334 + }, + { + "epoch": 0.5858774434438832, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5410687923431396, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6982667446136475, + "num_tokens": 138184601.0, + "step": 5335 + }, + { + "epoch": 0.5859872611464968, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.5218324661254883, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.6811191439628601, + "num_tokens": 138212102.0, + "step": 5336 + }, + { + "epoch": 0.5860970788491104, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.559171199798584, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6809470653533936, + "num_tokens": 138238813.0, + "step": 5337 + }, + { + "epoch": 0.5862068965517241, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.7239739894866943, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6949814558029175, + "num_tokens": 138260141.0, + "step": 5338 + }, + { + "epoch": 0.5863167142543378, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.5971386432647705, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.722409188747406, + "num_tokens": 138283651.0, + "step": 5339 + }, + { + "epoch": 0.5864265319569515, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.3331620693206787, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.6933674812316895, + "num_tokens": 138315435.0, + "step": 5340 + }, + { + "epoch": 0.5865363496595651, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4020190238952637, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7052907943725586, + "num_tokens": 138342803.0, + "step": 5341 + }, + { + "epoch": 0.5866461673621788, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.7767457962036133, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6910091638565063, + "num_tokens": 138365491.0, + "step": 5342 + }, + { + "epoch": 0.5867559850647924, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.296713352203369, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7135390639305115, + "num_tokens": 138395065.0, + "step": 5343 + }, + { + "epoch": 0.5868658027674061, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.6644744873046875, + "learning_rate": 1e-06, + "loss": 1.1089, + "mean_token_accuracy": 0.6777403354644775, + "num_tokens": 138420604.0, + "step": 5344 + }, + { + "epoch": 0.5869756204700197, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.428358316421509, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6992106437683105, + "num_tokens": 138448634.0, + "step": 5345 + }, + { + "epoch": 0.5870854381726335, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.4567313194274902, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.694062352180481, + "num_tokens": 138477350.0, + "step": 5346 + }, + { + "epoch": 0.5871952558752471, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.784043788909912, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7022629976272583, + "num_tokens": 138499062.0, + "step": 5347 + }, + { + "epoch": 0.5873050735778608, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.704808473587036, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.698097825050354, + "num_tokens": 138520145.0, + "step": 5348 + }, + { + "epoch": 0.5874148912804744, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.9157304763793945, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6965417861938477, + "num_tokens": 138541908.0, + "step": 5349 + }, + { + "epoch": 0.5875247089830881, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.5698091983795166, + "learning_rate": 1e-06, + "loss": 0.8973, + "mean_token_accuracy": 0.7354864478111267, + "num_tokens": 138566648.0, + "step": 5350 + }, + { + "epoch": 0.5876345266857017, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.481293201446533, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7125307321548462, + "num_tokens": 138593706.0, + "step": 5351 + }, + { + "epoch": 0.5877443443883154, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.4557037353515625, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7012398838996887, + "num_tokens": 138622965.0, + "step": 5352 + }, + { + "epoch": 0.5878541620909291, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.5007619857788086, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6867662668228149, + "num_tokens": 138649693.0, + "step": 5353 + }, + { + "epoch": 0.5879639797935428, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.2353146076202393, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6896920204162598, + "num_tokens": 138681641.0, + "step": 5354 + }, + { + "epoch": 0.5880737974961564, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.7764179706573486, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6939419507980347, + "num_tokens": 138705096.0, + "step": 5355 + }, + { + "epoch": 0.58818361519877, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.900480270385742, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7061067223548889, + "num_tokens": 138725318.0, + "step": 5356 + }, + { + "epoch": 0.5882934329013837, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.612516164779663, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7059088945388794, + "num_tokens": 138752583.0, + "step": 5357 + }, + { + "epoch": 0.5884032506039973, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7030084133148193, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7084895968437195, + "num_tokens": 138775165.0, + "step": 5358 + }, + { + "epoch": 0.588513068306611, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.4172933101654053, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7059476375579834, + "num_tokens": 138803464.0, + "step": 5359 + }, + { + "epoch": 0.5886228860092246, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.649986982345581, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.689953088760376, + "num_tokens": 138830517.0, + "step": 5360 + }, + { + "epoch": 0.5887327037118384, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.8993887901306152, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6979795694351196, + "num_tokens": 138852743.0, + "step": 5361 + }, + { + "epoch": 0.588842521414452, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5871520042419434, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.7006484866142273, + "num_tokens": 138877770.0, + "step": 5362 + }, + { + "epoch": 0.5889523391170657, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.3670225143432617, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6944054961204529, + "num_tokens": 138907114.0, + "step": 5363 + }, + { + "epoch": 0.5890621568196793, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.5049498081207275, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7090946435928345, + "num_tokens": 138934234.0, + "step": 5364 + }, + { + "epoch": 0.589171974522293, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.8799004554748535, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7037855386734009, + "num_tokens": 138953750.0, + "step": 5365 + }, + { + "epoch": 0.5892817922249066, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.320685625076294, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.701999843120575, + "num_tokens": 138985148.0, + "step": 5366 + }, + { + "epoch": 0.5893916099275203, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.6385064125061035, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.716639518737793, + "num_tokens": 139009650.0, + "step": 5367 + }, + { + "epoch": 0.589501427630134, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6408634185791016, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7226040363311768, + "num_tokens": 139030423.0, + "step": 5368 + }, + { + "epoch": 0.5896112453327477, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.4884073734283447, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6987268924713135, + "num_tokens": 139056646.0, + "step": 5369 + }, + { + "epoch": 0.5897210630353613, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.900381565093994, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7146978378295898, + "num_tokens": 139076987.0, + "step": 5370 + }, + { + "epoch": 0.589830880737975, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5817902088165283, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7213904857635498, + "num_tokens": 139101805.0, + "step": 5371 + }, + { + "epoch": 0.5899406984405886, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5349183082580566, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6946403980255127, + "num_tokens": 139129225.0, + "step": 5372 + }, + { + "epoch": 0.5900505161432023, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5013210773468018, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7059868574142456, + "num_tokens": 139155600.0, + "step": 5373 + }, + { + "epoch": 0.5901603338458159, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.5037760734558105, + "learning_rate": 1e-06, + "loss": 1.1065, + "mean_token_accuracy": 0.6869981288909912, + "num_tokens": 139183045.0, + "step": 5374 + }, + { + "epoch": 0.5902701515484297, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.307363510131836, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.6984566450119019, + "num_tokens": 139216042.0, + "step": 5375 + }, + { + "epoch": 0.5903799692510433, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.665041208267212, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7025831341743469, + "num_tokens": 139240011.0, + "step": 5376 + }, + { + "epoch": 0.590489786953657, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.5163469314575195, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7097398042678833, + "num_tokens": 139268069.0, + "step": 5377 + }, + { + "epoch": 0.5905996046562706, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.796013832092285, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7186498045921326, + "num_tokens": 139289186.0, + "step": 5378 + }, + { + "epoch": 0.5907094223588842, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.4555699825286865, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.7036401033401489, + "num_tokens": 139318862.0, + "step": 5379 + }, + { + "epoch": 0.5908192400614979, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6355249881744385, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.6985889673233032, + "num_tokens": 139342120.0, + "step": 5380 + }, + { + "epoch": 0.5909290577641115, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.458702325820923, + "learning_rate": 1e-06, + "loss": 1.1757, + "mean_token_accuracy": 0.6595970392227173, + "num_tokens": 139374511.0, + "step": 5381 + }, + { + "epoch": 0.5910388754667253, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.8155317306518555, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7218467593193054, + "num_tokens": 139395596.0, + "step": 5382 + }, + { + "epoch": 0.5911486931693389, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.485976457595825, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6911964416503906, + "num_tokens": 139420950.0, + "step": 5383 + }, + { + "epoch": 0.5912585108719526, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6483569145202637, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7096685171127319, + "num_tokens": 139449011.0, + "step": 5384 + }, + { + "epoch": 0.5913683285745662, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.2554280757904053, + "learning_rate": 1e-06, + "loss": 1.1099, + "mean_token_accuracy": 0.6875131130218506, + "num_tokens": 139480828.0, + "step": 5385 + }, + { + "epoch": 0.5914781462771799, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.503758668899536, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6976344585418701, + "num_tokens": 139507020.0, + "step": 5386 + }, + { + "epoch": 0.5915879639797935, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.4912643432617188, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.696786642074585, + "num_tokens": 139534335.0, + "step": 5387 + }, + { + "epoch": 0.5916977816824072, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.568833112716675, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7030727863311768, + "num_tokens": 139561396.0, + "step": 5388 + }, + { + "epoch": 0.5918075993850208, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.58762788772583, + "learning_rate": 1e-06, + "loss": 1.0766, + "mean_token_accuracy": 0.683771014213562, + "num_tokens": 139588667.0, + "step": 5389 + }, + { + "epoch": 0.5919174170876346, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.4097208976745605, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.6833460330963135, + "num_tokens": 139621085.0, + "step": 5390 + }, + { + "epoch": 0.5920272347902482, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6106128692626953, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7207137942314148, + "num_tokens": 139644535.0, + "step": 5391 + }, + { + "epoch": 0.5921370524928619, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.2676050662994385, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.7020071148872375, + "num_tokens": 139675145.0, + "step": 5392 + }, + { + "epoch": 0.5922468701954755, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.6859824657440186, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7106676697731018, + "num_tokens": 139698271.0, + "step": 5393 + }, + { + "epoch": 0.5923566878980892, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.267669677734375, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7027072310447693, + "num_tokens": 139728259.0, + "step": 5394 + }, + { + "epoch": 0.5924665056007028, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.393660068511963, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7148926854133606, + "num_tokens": 139751081.0, + "step": 5395 + }, + { + "epoch": 0.5925763233033164, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.306934356689453, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6941682696342468, + "num_tokens": 139783392.0, + "step": 5396 + }, + { + "epoch": 0.5926861410059302, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.442014217376709, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.7016503810882568, + "num_tokens": 139809224.0, + "step": 5397 + }, + { + "epoch": 0.5927959587085438, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 2.4466679096221924, + "learning_rate": 1e-06, + "loss": 0.9152, + "mean_token_accuracy": 0.7268503904342651, + "num_tokens": 139835349.0, + "step": 5398 + }, + { + "epoch": 0.5929057764111575, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.8055877685546875, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7174316644668579, + "num_tokens": 139858710.0, + "step": 5399 + }, + { + "epoch": 0.5930155941137711, + "ewc_loss": 7.331371307373047e-06, + "grad_norm": 2.8212382793426514, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6900967359542847, + "num_tokens": 139883838.0, + "step": 5400 + }, + { + "epoch": 0.5931254118163848, + "ewc_loss": 7.361173629760742e-06, + "grad_norm": 3.10103440284729, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6938613653182983, + "num_tokens": 139908596.0, + "step": 5401 + }, + { + "epoch": 0.5932352295189984, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.6266582012176514, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.7021425366401672, + "num_tokens": 139937133.0, + "step": 5402 + }, + { + "epoch": 0.5933450472216121, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.8037092685699463, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6849414110183716, + "num_tokens": 139962948.0, + "step": 5403 + }, + { + "epoch": 0.5934548649242258, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.6143879890441895, + "learning_rate": 1e-06, + "loss": 1.1687, + "mean_token_accuracy": 0.659740149974823, + "num_tokens": 139993617.0, + "step": 5404 + }, + { + "epoch": 0.5935646826268395, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5025317668914795, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6920740604400635, + "num_tokens": 140021652.0, + "step": 5405 + }, + { + "epoch": 0.5936745003294531, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.4425876140594482, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6916625499725342, + "num_tokens": 140049633.0, + "step": 5406 + }, + { + "epoch": 0.5937843180320668, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7158782482147217, + "learning_rate": 1e-06, + "loss": 0.8936, + "mean_token_accuracy": 0.7335491180419922, + "num_tokens": 140070854.0, + "step": 5407 + }, + { + "epoch": 0.5938941357346804, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5071170330047607, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7027739882469177, + "num_tokens": 140094943.0, + "step": 5408 + }, + { + "epoch": 0.5940039534372941, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.821732521057129, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.6996639966964722, + "num_tokens": 140114169.0, + "step": 5409 + }, + { + "epoch": 0.5941137711399077, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.558748960494995, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6970549821853638, + "num_tokens": 140139128.0, + "step": 5410 + }, + { + "epoch": 0.5942235888425215, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.459019184112549, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7140190005302429, + "num_tokens": 140165992.0, + "step": 5411 + }, + { + "epoch": 0.5943334065451351, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.71885085105896, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6850224137306213, + "num_tokens": 140190420.0, + "step": 5412 + }, + { + "epoch": 0.5944432242477488, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.476663112640381, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7127645015716553, + "num_tokens": 140218466.0, + "step": 5413 + }, + { + "epoch": 0.5945530419503624, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7879116535186768, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7153708338737488, + "num_tokens": 140241243.0, + "step": 5414 + }, + { + "epoch": 0.594662859652976, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.4392054080963135, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7052264213562012, + "num_tokens": 140269198.0, + "step": 5415 + }, + { + "epoch": 0.5947726773555897, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.703871011734009, + "learning_rate": 1e-06, + "loss": 0.9508, + "mean_token_accuracy": 0.7224953174591064, + "num_tokens": 140291919.0, + "step": 5416 + }, + { + "epoch": 0.5948824950582033, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.644831418991089, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7126854658126831, + "num_tokens": 140314477.0, + "step": 5417 + }, + { + "epoch": 0.5949923127608171, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7479653358459473, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.7028732299804688, + "num_tokens": 140337521.0, + "step": 5418 + }, + { + "epoch": 0.5951021304634307, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.664698839187622, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6951805353164673, + "num_tokens": 140360715.0, + "step": 5419 + }, + { + "epoch": 0.5952119481660444, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.8545355796813965, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.7004490494728088, + "num_tokens": 140384016.0, + "step": 5420 + }, + { + "epoch": 0.595321765868658, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.8645706176757812, + "learning_rate": 1e-06, + "loss": 0.9227, + "mean_token_accuracy": 0.7234658002853394, + "num_tokens": 140408011.0, + "step": 5421 + }, + { + "epoch": 0.5954315835712717, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3886189460754395, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7000402212142944, + "num_tokens": 140439623.0, + "step": 5422 + }, + { + "epoch": 0.5955414012738853, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5110185146331787, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6843359470367432, + "num_tokens": 140463762.0, + "step": 5423 + }, + { + "epoch": 0.595651218976499, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.6482677459716797, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6834741830825806, + "num_tokens": 140489649.0, + "step": 5424 + }, + { + "epoch": 0.5957610366791126, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8653619289398193, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7046292424201965, + "num_tokens": 140509941.0, + "step": 5425 + }, + { + "epoch": 0.5958708543817264, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.577340602874756, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6920778155326843, + "num_tokens": 140535871.0, + "step": 5426 + }, + { + "epoch": 0.59598067208434, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5024940967559814, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7054716348648071, + "num_tokens": 140562930.0, + "step": 5427 + }, + { + "epoch": 0.5960904897869537, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.4785869121551514, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7088741660118103, + "num_tokens": 140590643.0, + "step": 5428 + }, + { + "epoch": 0.5962003074895673, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 7.970820903778076, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6903003454208374, + "num_tokens": 140615200.0, + "step": 5429 + }, + { + "epoch": 0.596310125192181, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3468101024627686, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7123735547065735, + "num_tokens": 140645253.0, + "step": 5430 + }, + { + "epoch": 0.5964199428947946, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.764075517654419, + "learning_rate": 1e-06, + "loss": 0.942, + "mean_token_accuracy": 0.7165586948394775, + "num_tokens": 140667899.0, + "step": 5431 + }, + { + "epoch": 0.5965297605974083, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.694338321685791, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6878930330276489, + "num_tokens": 140691109.0, + "step": 5432 + }, + { + "epoch": 0.596639578300022, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.538567304611206, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.690058171749115, + "num_tokens": 140717798.0, + "step": 5433 + }, + { + "epoch": 0.5967493960026357, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.29736065864563, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6874547004699707, + "num_tokens": 140749119.0, + "step": 5434 + }, + { + "epoch": 0.5968592137052493, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.6178040504455566, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7095314264297485, + "num_tokens": 140774407.0, + "step": 5435 + }, + { + "epoch": 0.596969031407863, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.4999642372131348, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6917915344238281, + "num_tokens": 140800384.0, + "step": 5436 + }, + { + "epoch": 0.5970788491104766, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7318344116210938, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7137939929962158, + "num_tokens": 140823824.0, + "step": 5437 + }, + { + "epoch": 0.5971886668130902, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.6527037620544434, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7010987997055054, + "num_tokens": 140845377.0, + "step": 5438 + }, + { + "epoch": 0.5972984845157039, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 3.4522242546081543, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.712386965751648, + "num_tokens": 140869853.0, + "step": 5439 + }, + { + "epoch": 0.5974083022183176, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 7.906757831573486, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7193788290023804, + "num_tokens": 140893376.0, + "step": 5440 + }, + { + "epoch": 0.5975181199209313, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3718128204345703, + "learning_rate": 1e-06, + "loss": 1.129, + "mean_token_accuracy": 0.6683240532875061, + "num_tokens": 140930049.0, + "step": 5441 + }, + { + "epoch": 0.5976279376235449, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.3507964611053467, + "learning_rate": 1e-06, + "loss": 1.1301, + "mean_token_accuracy": 0.6756109595298767, + "num_tokens": 140960774.0, + "step": 5442 + }, + { + "epoch": 0.5977377553261586, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.2279250621795654, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.693705141544342, + "num_tokens": 140991688.0, + "step": 5443 + }, + { + "epoch": 0.5978475730287722, + "ewc_loss": 7.3909759521484375e-06, + "grad_norm": 2.487496852874756, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.706585705280304, + "num_tokens": 141018534.0, + "step": 5444 + }, + { + "epoch": 0.5979573907313859, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.692807674407959, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6925979852676392, + "num_tokens": 141040847.0, + "step": 5445 + }, + { + "epoch": 0.5980672084339995, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.383704900741577, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.690066933631897, + "num_tokens": 141068941.0, + "step": 5446 + }, + { + "epoch": 0.5981770261366133, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.926150321960449, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7252649664878845, + "num_tokens": 141088818.0, + "step": 5447 + }, + { + "epoch": 0.5982868438392269, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.456434726715088, + "learning_rate": 1e-06, + "loss": 0.8828, + "mean_token_accuracy": 0.7382259368896484, + "num_tokens": 141111050.0, + "step": 5448 + }, + { + "epoch": 0.5983966615418406, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.495443344116211, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7217782735824585, + "num_tokens": 141134434.0, + "step": 5449 + }, + { + "epoch": 0.5985064792444542, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.880714178085327, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7129016518592834, + "num_tokens": 141156074.0, + "step": 5450 + }, + { + "epoch": 0.5986162969470679, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5306787490844727, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.694624125957489, + "num_tokens": 141184073.0, + "step": 5451 + }, + { + "epoch": 0.5987261146496815, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5622825622558594, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7076405882835388, + "num_tokens": 141207261.0, + "step": 5452 + }, + { + "epoch": 0.5988359323522952, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.929187059402466, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7123944163322449, + "num_tokens": 141227943.0, + "step": 5453 + }, + { + "epoch": 0.5989457500549088, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.338069200515747, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6936076283454895, + "num_tokens": 141259938.0, + "step": 5454 + }, + { + "epoch": 0.5990555677575226, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7613203525543213, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.700319230556488, + "num_tokens": 141281372.0, + "step": 5455 + }, + { + "epoch": 0.5991653854601362, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.7012507915496826, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.714135468006134, + "num_tokens": 141301792.0, + "step": 5456 + }, + { + "epoch": 0.5992752031627498, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.50083327293396, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6988258361816406, + "num_tokens": 141330241.0, + "step": 5457 + }, + { + "epoch": 0.5993850208653635, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.560903549194336, + "learning_rate": 1e-06, + "loss": 1.1109, + "mean_token_accuracy": 0.6877419352531433, + "num_tokens": 141360496.0, + "step": 5458 + }, + { + "epoch": 0.5994948385679771, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5542986392974854, + "learning_rate": 1e-06, + "loss": 1.0943, + "mean_token_accuracy": 0.6857292056083679, + "num_tokens": 141386431.0, + "step": 5459 + }, + { + "epoch": 0.5996046562705908, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.556224822998047, + "learning_rate": 1e-06, + "loss": 1.1606, + "mean_token_accuracy": 0.6625608205795288, + "num_tokens": 141415944.0, + "step": 5460 + }, + { + "epoch": 0.5997144739732044, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.4372966289520264, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6965757012367249, + "num_tokens": 141444005.0, + "step": 5461 + }, + { + "epoch": 0.5998242916758182, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5042707920074463, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6873160004615784, + "num_tokens": 141470558.0, + "step": 5462 + }, + { + "epoch": 0.5999341093784318, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.404484987258911, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.684857964515686, + "num_tokens": 141499619.0, + "step": 5463 + }, + { + "epoch": 0.6000439270810455, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.8527510166168213, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7186375856399536, + "num_tokens": 141519971.0, + "step": 5464 + }, + { + "epoch": 0.6001537447836591, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.520362615585327, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7186988592147827, + "num_tokens": 141547413.0, + "step": 5465 + }, + { + "epoch": 0.6002635624862728, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.29436993598938, + "learning_rate": 1e-06, + "loss": 1.1241, + "mean_token_accuracy": 0.6736796498298645, + "num_tokens": 141582883.0, + "step": 5466 + }, + { + "epoch": 0.6003733801888864, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5782203674316406, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7003391981124878, + "num_tokens": 141609111.0, + "step": 5467 + }, + { + "epoch": 0.6004831978915001, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.644106388092041, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7188341617584229, + "num_tokens": 141632586.0, + "step": 5468 + }, + { + "epoch": 0.6005930155941138, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.804100513458252, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7091726660728455, + "num_tokens": 141653030.0, + "step": 5469 + }, + { + "epoch": 0.6007028332967275, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5573952198028564, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6881031394004822, + "num_tokens": 141680762.0, + "step": 5470 + }, + { + "epoch": 0.6008126509993411, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.6475987434387207, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.685706615447998, + "num_tokens": 141705139.0, + "step": 5471 + }, + { + "epoch": 0.6009224687019548, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.4897196292877197, + "learning_rate": 1e-06, + "loss": 1.1208, + "mean_token_accuracy": 0.6763980388641357, + "num_tokens": 141732248.0, + "step": 5472 + }, + { + "epoch": 0.6010322864045684, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.8366506099700928, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6832559704780579, + "num_tokens": 141756347.0, + "step": 5473 + }, + { + "epoch": 0.601142104107182, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.2262041568756104, + "learning_rate": 1e-06, + "loss": 1.0984, + "mean_token_accuracy": 0.6854329109191895, + "num_tokens": 141790860.0, + "step": 5474 + }, + { + "epoch": 0.6012519218097957, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.4669220447540283, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7003481388092041, + "num_tokens": 141818768.0, + "step": 5475 + }, + { + "epoch": 0.6013617395124095, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.4226486682891846, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7118567228317261, + "num_tokens": 141846167.0, + "step": 5476 + }, + { + "epoch": 0.6014715572150231, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.807121992111206, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7038706541061401, + "num_tokens": 141867109.0, + "step": 5477 + }, + { + "epoch": 0.6015813749176367, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.7011194229125977, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6857468485832214, + "num_tokens": 141894244.0, + "step": 5478 + }, + { + "epoch": 0.6016911926202504, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.7079036235809326, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7087249159812927, + "num_tokens": 141918529.0, + "step": 5479 + }, + { + "epoch": 0.601801010322864, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.612867832183838, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7067346572875977, + "num_tokens": 141941287.0, + "step": 5480 + }, + { + "epoch": 0.6019108280254777, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.6714444160461426, + "learning_rate": 1e-06, + "loss": 1.0832, + "mean_token_accuracy": 0.6792284250259399, + "num_tokens": 141965353.0, + "step": 5481 + }, + { + "epoch": 0.6020206457280913, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.6660025119781494, + "learning_rate": 1e-06, + "loss": 1.1606, + "mean_token_accuracy": 0.6724550724029541, + "num_tokens": 141992824.0, + "step": 5482 + }, + { + "epoch": 0.602130463430705, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.736642837524414, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6861778497695923, + "num_tokens": 142016379.0, + "step": 5483 + }, + { + "epoch": 0.6022402811333187, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.3576366901397705, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6792447566986084, + "num_tokens": 142051236.0, + "step": 5484 + }, + { + "epoch": 0.6023500988359324, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.635533332824707, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6901279091835022, + "num_tokens": 142076418.0, + "step": 5485 + }, + { + "epoch": 0.602459916538546, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.834775686264038, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.716091513633728, + "num_tokens": 142096091.0, + "step": 5486 + }, + { + "epoch": 0.6025697342411597, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6692702770233154, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.703822135925293, + "num_tokens": 142124345.0, + "step": 5487 + }, + { + "epoch": 0.6026795519437733, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.645583391189575, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6834023594856262, + "num_tokens": 142154203.0, + "step": 5488 + }, + { + "epoch": 0.602789369646387, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6272366046905518, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.695799708366394, + "num_tokens": 142178440.0, + "step": 5489 + }, + { + "epoch": 0.6028991873490006, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.597182512283325, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7044222950935364, + "num_tokens": 142203260.0, + "step": 5490 + }, + { + "epoch": 0.6030090050516144, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0890841484069824, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7100673913955688, + "num_tokens": 142222642.0, + "step": 5491 + }, + { + "epoch": 0.603118822754228, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6427714824676514, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7078366875648499, + "num_tokens": 142248504.0, + "step": 5492 + }, + { + "epoch": 0.6032286404568417, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5740911960601807, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7034873962402344, + "num_tokens": 142275888.0, + "step": 5493 + }, + { + "epoch": 0.6033384581594553, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.567885160446167, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6919541358947754, + "num_tokens": 142302175.0, + "step": 5494 + }, + { + "epoch": 0.603448275862069, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.593897581100464, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7050555348396301, + "num_tokens": 142326863.0, + "step": 5495 + }, + { + "epoch": 0.6035580935646826, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.327679395675659, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7007176876068115, + "num_tokens": 142355356.0, + "step": 5496 + }, + { + "epoch": 0.6036679112672962, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6529388427734375, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7028632164001465, + "num_tokens": 142379442.0, + "step": 5497 + }, + { + "epoch": 0.60377772896991, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6290957927703857, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.7000862956047058, + "num_tokens": 142402486.0, + "step": 5498 + }, + { + "epoch": 0.6038875466725236, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6833112239837646, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6890507936477661, + "num_tokens": 142427756.0, + "step": 5499 + }, + { + "epoch": 0.6039973643751373, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.531710624694824, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7010776996612549, + "num_tokens": 142456504.0, + "step": 5500 + }, + { + "epoch": 0.6041071820777509, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.459599256515503, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.715494692325592, + "num_tokens": 142481999.0, + "step": 5501 + }, + { + "epoch": 0.6042169997803646, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 8.77280330657959, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7108505964279175, + "num_tokens": 142508085.0, + "step": 5502 + }, + { + "epoch": 0.6043268174829782, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7664711475372314, + "learning_rate": 1e-06, + "loss": 1.1097, + "mean_token_accuracy": 0.6785088777542114, + "num_tokens": 142533502.0, + "step": 5503 + }, + { + "epoch": 0.6044366351855919, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.437141180038452, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7045154571533203, + "num_tokens": 142562232.0, + "step": 5504 + }, + { + "epoch": 0.6045464528882056, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5766661167144775, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7143545746803284, + "num_tokens": 142584535.0, + "step": 5505 + }, + { + "epoch": 0.6046562705908193, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4063098430633545, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6948932409286499, + "num_tokens": 142613548.0, + "step": 5506 + }, + { + "epoch": 0.6047660882934329, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6905577182769775, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6997823715209961, + "num_tokens": 142634501.0, + "step": 5507 + }, + { + "epoch": 0.6048759059960466, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5178911685943604, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7076883316040039, + "num_tokens": 142660053.0, + "step": 5508 + }, + { + "epoch": 0.6049857236986602, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.49768328666687, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6970027685165405, + "num_tokens": 142686587.0, + "step": 5509 + }, + { + "epoch": 0.6050955414012739, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.606290340423584, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6962310075759888, + "num_tokens": 142710842.0, + "step": 5510 + }, + { + "epoch": 0.6052053591038875, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.243037223815918, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6935526132583618, + "num_tokens": 142744140.0, + "step": 5511 + }, + { + "epoch": 0.6053151768065012, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.6057701110839844, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7011024355888367, + "num_tokens": 142769131.0, + "step": 5512 + }, + { + "epoch": 0.6054249945091149, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.5482916831970215, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.7052789926528931, + "num_tokens": 142798422.0, + "step": 5513 + }, + { + "epoch": 0.6055348122117286, + "ewc_loss": 7.420778274536133e-06, + "grad_norm": 2.704887628555298, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7072886228561401, + "num_tokens": 142820626.0, + "step": 5514 + }, + { + "epoch": 0.6056446299143422, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.644047975540161, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7093888521194458, + "num_tokens": 142842981.0, + "step": 5515 + }, + { + "epoch": 0.6057544476169558, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.9053914546966553, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.704696774482727, + "num_tokens": 142862928.0, + "step": 5516 + }, + { + "epoch": 0.6058642653195695, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.621312379837036, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6910231113433838, + "num_tokens": 142891872.0, + "step": 5517 + }, + { + "epoch": 0.6059740830221831, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5772523880004883, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6973251104354858, + "num_tokens": 142919028.0, + "step": 5518 + }, + { + "epoch": 0.6060839007247968, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.2523834705352783, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.714226484298706, + "num_tokens": 142947374.0, + "step": 5519 + }, + { + "epoch": 0.6061937184274105, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.4870221614837646, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6846723556518555, + "num_tokens": 142974095.0, + "step": 5520 + }, + { + "epoch": 0.6063035361300242, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.7222726345062256, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6842155456542969, + "num_tokens": 142995967.0, + "step": 5521 + }, + { + "epoch": 0.6064133538326378, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.4111247062683105, + "learning_rate": 1e-06, + "loss": 1.1161, + "mean_token_accuracy": 0.6738722920417786, + "num_tokens": 143025221.0, + "step": 5522 + }, + { + "epoch": 0.6065231715352515, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.453909397125244, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.6987426280975342, + "num_tokens": 143051527.0, + "step": 5523 + }, + { + "epoch": 0.6066329892378651, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.732928514480591, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6921702027320862, + "num_tokens": 143075544.0, + "step": 5524 + }, + { + "epoch": 0.6067428069404788, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.506319761276245, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7150835990905762, + "num_tokens": 143099855.0, + "step": 5525 + }, + { + "epoch": 0.6068526246430924, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.6071343421936035, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7088623046875, + "num_tokens": 143129817.0, + "step": 5526 + }, + { + "epoch": 0.6069624423457062, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.801806926727295, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.693933367729187, + "num_tokens": 143150698.0, + "step": 5527 + }, + { + "epoch": 0.6070722600483198, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8398921489715576, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7019802331924438, + "num_tokens": 143171043.0, + "step": 5528 + }, + { + "epoch": 0.6071820777509335, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6426568031311035, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7114288210868835, + "num_tokens": 143193800.0, + "step": 5529 + }, + { + "epoch": 0.6072918954535471, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.468045711517334, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6923598051071167, + "num_tokens": 143222251.0, + "step": 5530 + }, + { + "epoch": 0.6074017131561608, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9005424976348877, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7077860832214355, + "num_tokens": 143245623.0, + "step": 5531 + }, + { + "epoch": 0.6075115308587744, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.594536542892456, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.7164725065231323, + "num_tokens": 143271328.0, + "step": 5532 + }, + { + "epoch": 0.607621348561388, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9880552291870117, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.713387668132782, + "num_tokens": 143290653.0, + "step": 5533 + }, + { + "epoch": 0.6077311662640018, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.433570146560669, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7203664779663086, + "num_tokens": 143316964.0, + "step": 5534 + }, + { + "epoch": 0.6078409839666155, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7699034214019775, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.7189531922340393, + "num_tokens": 143339087.0, + "step": 5535 + }, + { + "epoch": 0.6079508016692291, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.229818105697632, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6969120502471924, + "num_tokens": 143372999.0, + "step": 5536 + }, + { + "epoch": 0.6080606193718427, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5997989177703857, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6986208558082581, + "num_tokens": 143398034.0, + "step": 5537 + }, + { + "epoch": 0.6081704370744564, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6321840286254883, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.714605450630188, + "num_tokens": 143419511.0, + "step": 5538 + }, + { + "epoch": 0.60828025477707, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3622324466705322, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6864213943481445, + "num_tokens": 143447620.0, + "step": 5539 + }, + { + "epoch": 0.6083900724796837, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.678283214569092, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6915205717086792, + "num_tokens": 143470927.0, + "step": 5540 + }, + { + "epoch": 0.6084998901822973, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4826674461364746, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.700593113899231, + "num_tokens": 143495362.0, + "step": 5541 + }, + { + "epoch": 0.6086097078849111, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3979971408843994, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7096223831176758, + "num_tokens": 143522298.0, + "step": 5542 + }, + { + "epoch": 0.6087195255875247, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0384562015533447, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.7138034701347351, + "num_tokens": 143540762.0, + "step": 5543 + }, + { + "epoch": 0.6088293432901384, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.943784713745117, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.7255057096481323, + "num_tokens": 143562935.0, + "step": 5544 + }, + { + "epoch": 0.608939160992752, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.671330690383911, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6927953958511353, + "num_tokens": 143586717.0, + "step": 5545 + }, + { + "epoch": 0.6090489786953657, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.551523208618164, + "learning_rate": 1e-06, + "loss": 1.0765, + "mean_token_accuracy": 0.6852371692657471, + "num_tokens": 143613936.0, + "step": 5546 + }, + { + "epoch": 0.6091587963979793, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5559794902801514, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7060911655426025, + "num_tokens": 143641181.0, + "step": 5547 + }, + { + "epoch": 0.609268614100593, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4270873069763184, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7011232376098633, + "num_tokens": 143666141.0, + "step": 5548 + }, + { + "epoch": 0.6093784318032067, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6515750885009766, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6941862106323242, + "num_tokens": 143694096.0, + "step": 5549 + }, + { + "epoch": 0.6094882495058204, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.348538398742676, + "learning_rate": 1e-06, + "loss": 1.0915, + "mean_token_accuracy": 0.6924866437911987, + "num_tokens": 143724483.0, + "step": 5550 + }, + { + "epoch": 0.609598067208434, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5658066272735596, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7067230939865112, + "num_tokens": 143748850.0, + "step": 5551 + }, + { + "epoch": 0.6097078849110477, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.506011486053467, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6932536363601685, + "num_tokens": 143776175.0, + "step": 5552 + }, + { + "epoch": 0.6098177026136613, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5196146965026855, + "learning_rate": 1e-06, + "loss": 1.1233, + "mean_token_accuracy": 0.6852185130119324, + "num_tokens": 143804042.0, + "step": 5553 + }, + { + "epoch": 0.609927520316275, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4840409755706787, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.7025967836380005, + "num_tokens": 143831258.0, + "step": 5554 + }, + { + "epoch": 0.6100373380188886, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3106369972229004, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.6895252466201782, + "num_tokens": 143861686.0, + "step": 5555 + }, + { + "epoch": 0.6101471557215024, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9008941650390625, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.6956793665885925, + "num_tokens": 143883454.0, + "step": 5556 + }, + { + "epoch": 0.610256973424116, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6264774799346924, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.7000269889831543, + "num_tokens": 143907384.0, + "step": 5557 + }, + { + "epoch": 0.6103667911267296, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5316007137298584, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6885495185852051, + "num_tokens": 143934663.0, + "step": 5558 + }, + { + "epoch": 0.6104766088293433, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.388050079345703, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7071253061294556, + "num_tokens": 143962719.0, + "step": 5559 + }, + { + "epoch": 0.6105864265319569, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.887268543243408, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7071380019187927, + "num_tokens": 143983202.0, + "step": 5560 + }, + { + "epoch": 0.6106962442345706, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7670578956604004, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7097867131233215, + "num_tokens": 144004714.0, + "step": 5561 + }, + { + "epoch": 0.6108060619371842, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9183053970336914, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.6918683648109436, + "num_tokens": 144025861.0, + "step": 5562 + }, + { + "epoch": 0.610915879639798, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6789958477020264, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7103372812271118, + "num_tokens": 144048018.0, + "step": 5563 + }, + { + "epoch": 0.6110256973424116, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6729745864868164, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7220684289932251, + "num_tokens": 144073098.0, + "step": 5564 + }, + { + "epoch": 0.6111355150450253, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.371555805206299, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6897168159484863, + "num_tokens": 144103692.0, + "step": 5565 + }, + { + "epoch": 0.6112453327476389, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5583441257476807, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6951492428779602, + "num_tokens": 144130071.0, + "step": 5566 + }, + { + "epoch": 0.6113551504502526, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.1672205924987793, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7074269652366638, + "num_tokens": 144160736.0, + "step": 5567 + }, + { + "epoch": 0.6114649681528662, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5014214515686035, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7116038799285889, + "num_tokens": 144187160.0, + "step": 5568 + }, + { + "epoch": 0.6115747858554799, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4988150596618652, + "learning_rate": 1e-06, + "loss": 1.0766, + "mean_token_accuracy": 0.686328649520874, + "num_tokens": 144215373.0, + "step": 5569 + }, + { + "epoch": 0.6116846035580935, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3314599990844727, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7173048257827759, + "num_tokens": 144246010.0, + "step": 5570 + }, + { + "epoch": 0.6117944212607073, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6329894065856934, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6846678853034973, + "num_tokens": 144271534.0, + "step": 5571 + }, + { + "epoch": 0.6119042389633209, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5115952491760254, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.691432535648346, + "num_tokens": 144296962.0, + "step": 5572 + }, + { + "epoch": 0.6120140566659346, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.798807144165039, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7139172554016113, + "num_tokens": 144319660.0, + "step": 5573 + }, + { + "epoch": 0.6121238743685482, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4432685375213623, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7017244100570679, + "num_tokens": 144345996.0, + "step": 5574 + }, + { + "epoch": 0.6122336920711619, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6791324615478516, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7140152454376221, + "num_tokens": 144370199.0, + "step": 5575 + }, + { + "epoch": 0.6123435097737755, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3786628246307373, + "learning_rate": 1e-06, + "loss": 1.0855, + "mean_token_accuracy": 0.6783163547515869, + "num_tokens": 144398358.0, + "step": 5576 + }, + { + "epoch": 0.6124533274763891, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6635236740112305, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7079226970672607, + "num_tokens": 144422666.0, + "step": 5577 + }, + { + "epoch": 0.6125631451790029, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3663320541381836, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6864548325538635, + "num_tokens": 144453770.0, + "step": 5578 + }, + { + "epoch": 0.6126729628816165, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6012024879455566, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6920725703239441, + "num_tokens": 144487826.0, + "step": 5579 + }, + { + "epoch": 0.6127827805842302, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9051408767700195, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7158121466636658, + "num_tokens": 144510583.0, + "step": 5580 + }, + { + "epoch": 0.6128925982868438, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5458898544311523, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.702674150466919, + "num_tokens": 144537090.0, + "step": 5581 + }, + { + "epoch": 0.6130024159894575, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.128671884536743, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7033222317695618, + "num_tokens": 144573721.0, + "step": 5582 + }, + { + "epoch": 0.6131122336920711, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4182512760162354, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.6965398788452148, + "num_tokens": 144600228.0, + "step": 5583 + }, + { + "epoch": 0.6132220513946848, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4886932373046875, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6925690770149231, + "num_tokens": 144624825.0, + "step": 5584 + }, + { + "epoch": 0.6133318690972985, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3212623596191406, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7020043730735779, + "num_tokens": 144653870.0, + "step": 5585 + }, + { + "epoch": 0.6134416867999122, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.502995014190674, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.694907546043396, + "num_tokens": 144682126.0, + "step": 5586 + }, + { + "epoch": 0.6135515045025258, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.593451499938965, + "learning_rate": 1e-06, + "loss": 1.1225, + "mean_token_accuracy": 0.6731388568878174, + "num_tokens": 144711108.0, + "step": 5587 + }, + { + "epoch": 0.6136613222051395, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7242801189422607, + "learning_rate": 1e-06, + "loss": 0.8937, + "mean_token_accuracy": 0.7326905727386475, + "num_tokens": 144732059.0, + "step": 5588 + }, + { + "epoch": 0.6137711399077531, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5499727725982666, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7132748961448669, + "num_tokens": 144756402.0, + "step": 5589 + }, + { + "epoch": 0.6138809576103668, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7641139030456543, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6967776417732239, + "num_tokens": 144778478.0, + "step": 5590 + }, + { + "epoch": 0.6139907753129804, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7543575763702393, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6913650631904602, + "num_tokens": 144803171.0, + "step": 5591 + }, + { + "epoch": 0.6141005930155942, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5955557823181152, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.693038821220398, + "num_tokens": 144829122.0, + "step": 5592 + }, + { + "epoch": 0.6142104107182078, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.693110704421997, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.6985933184623718, + "num_tokens": 144853840.0, + "step": 5593 + }, + { + "epoch": 0.6143202284208215, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4894638061523438, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7248971462249756, + "num_tokens": 144879799.0, + "step": 5594 + }, + { + "epoch": 0.6144300461234351, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4815866947174072, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7238509654998779, + "num_tokens": 144903876.0, + "step": 5595 + }, + { + "epoch": 0.6145398638260487, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.548865795135498, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6842501759529114, + "num_tokens": 144932095.0, + "step": 5596 + }, + { + "epoch": 0.6146496815286624, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6668343544006348, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7099504470825195, + "num_tokens": 144955621.0, + "step": 5597 + }, + { + "epoch": 0.614759499231276, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.383767604827881, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6976814270019531, + "num_tokens": 144985577.0, + "step": 5598 + }, + { + "epoch": 0.6148693169338898, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7853639125823975, + "learning_rate": 1e-06, + "loss": 0.9117, + "mean_token_accuracy": 0.7240373492240906, + "num_tokens": 145005958.0, + "step": 5599 + }, + { + "epoch": 0.6149791346365034, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.893946647644043, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7134740352630615, + "num_tokens": 145025793.0, + "step": 5600 + }, + { + "epoch": 0.6150889523391171, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.395742893218994, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.6951225399971008, + "num_tokens": 145056561.0, + "step": 5601 + }, + { + "epoch": 0.6151987700417307, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.469069480895996, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7160940766334534, + "num_tokens": 145084877.0, + "step": 5602 + }, + { + "epoch": 0.6153085877443444, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4982855319976807, + "learning_rate": 1e-06, + "loss": 0.9443, + "mean_token_accuracy": 0.7221370935440063, + "num_tokens": 145108660.0, + "step": 5603 + }, + { + "epoch": 0.615418405446958, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.541759729385376, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7041288614273071, + "num_tokens": 145132828.0, + "step": 5604 + }, + { + "epoch": 0.6155282231495717, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8278281688690186, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6964046955108643, + "num_tokens": 145154962.0, + "step": 5605 + }, + { + "epoch": 0.6156380408521853, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7269864082336426, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6781144142150879, + "num_tokens": 145182406.0, + "step": 5606 + }, + { + "epoch": 0.6157478585547991, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.560995101928711, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6826395988464355, + "num_tokens": 145208430.0, + "step": 5607 + }, + { + "epoch": 0.6158576762574127, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5427467823028564, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7043558955192566, + "num_tokens": 145236623.0, + "step": 5608 + }, + { + "epoch": 0.6159674939600264, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.641087532043457, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7038752436637878, + "num_tokens": 145261618.0, + "step": 5609 + }, + { + "epoch": 0.61607731166264, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.531787633895874, + "learning_rate": 1e-06, + "loss": 1.1051, + "mean_token_accuracy": 0.6924152970314026, + "num_tokens": 145291071.0, + "step": 5610 + }, + { + "epoch": 0.6161871293652537, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.270705223083496, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7311512231826782, + "num_tokens": 145323857.0, + "step": 5611 + }, + { + "epoch": 0.6162969470678673, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4328525066375732, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.6858162879943848, + "num_tokens": 145353905.0, + "step": 5612 + }, + { + "epoch": 0.616406764770481, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.2919650077819824, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7169074416160583, + "num_tokens": 145384056.0, + "step": 5613 + }, + { + "epoch": 0.6165165824730947, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.426799774169922, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7227640748023987, + "num_tokens": 145410449.0, + "step": 5614 + }, + { + "epoch": 0.6166264001757084, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.7002692222595215, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7219959497451782, + "num_tokens": 145433886.0, + "step": 5615 + }, + { + "epoch": 0.616736217878322, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.917513370513916, + "learning_rate": 1e-06, + "loss": 1.1052, + "mean_token_accuracy": 0.6845943927764893, + "num_tokens": 145455812.0, + "step": 5616 + }, + { + "epoch": 0.6168460355809356, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.517576217651367, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6850905418395996, + "num_tokens": 145483198.0, + "step": 5617 + }, + { + "epoch": 0.6169558532835493, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.640742540359497, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7042743563652039, + "num_tokens": 145508062.0, + "step": 5618 + }, + { + "epoch": 0.6170656709861629, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.526538610458374, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6922009587287903, + "num_tokens": 145533803.0, + "step": 5619 + }, + { + "epoch": 0.6171754886887766, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6638453006744385, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.6860119104385376, + "num_tokens": 145561163.0, + "step": 5620 + }, + { + "epoch": 0.6172853063913903, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.560983419418335, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7012670636177063, + "num_tokens": 145587924.0, + "step": 5621 + }, + { + "epoch": 0.617395124094004, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.4852538108825684, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7088907957077026, + "num_tokens": 145614186.0, + "step": 5622 + }, + { + "epoch": 0.6175049417966176, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.555168867111206, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7116689682006836, + "num_tokens": 145636444.0, + "step": 5623 + }, + { + "epoch": 0.6176147594992313, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.382638692855835, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.698949933052063, + "num_tokens": 145665371.0, + "step": 5624 + }, + { + "epoch": 0.6177245772018449, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.496885299682617, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.6978124380111694, + "num_tokens": 145689331.0, + "step": 5625 + }, + { + "epoch": 0.6178343949044586, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.859761953353882, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6961506605148315, + "num_tokens": 145709752.0, + "step": 5626 + }, + { + "epoch": 0.6179442126070722, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.638913154602051, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7115629315376282, + "num_tokens": 145732554.0, + "step": 5627 + }, + { + "epoch": 0.618054030309686, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.247647762298584, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6789673566818237, + "num_tokens": 145768213.0, + "step": 5628 + }, + { + "epoch": 0.6181638480122996, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7247979640960693, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7118320465087891, + "num_tokens": 145788714.0, + "step": 5629 + }, + { + "epoch": 0.6182736657149133, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.594222068786621, + "learning_rate": 1e-06, + "loss": 0.9688, + "mean_token_accuracy": 0.713628888130188, + "num_tokens": 145811217.0, + "step": 5630 + }, + { + "epoch": 0.6183834834175269, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.669863700866699, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6927326321601868, + "num_tokens": 145834633.0, + "step": 5631 + }, + { + "epoch": 0.6184933011201406, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.95346736907959, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7010037899017334, + "num_tokens": 145855250.0, + "step": 5632 + }, + { + "epoch": 0.6186031188227542, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.341496467590332, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6919252276420593, + "num_tokens": 145888420.0, + "step": 5633 + }, + { + "epoch": 0.6187129365253679, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.682988405227661, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7053722143173218, + "num_tokens": 145911814.0, + "step": 5634 + }, + { + "epoch": 0.6188227542279815, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3929083347320557, + "learning_rate": 1e-06, + "loss": 0.8773, + "mean_token_accuracy": 0.7399677038192749, + "num_tokens": 145937059.0, + "step": 5635 + }, + { + "epoch": 0.6189325719305953, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5993151664733887, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6975101232528687, + "num_tokens": 145964732.0, + "step": 5636 + }, + { + "epoch": 0.6190423896332089, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.553398609161377, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6957489252090454, + "num_tokens": 145990592.0, + "step": 5637 + }, + { + "epoch": 0.6191522073358225, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.385607957839966, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6884934902191162, + "num_tokens": 146020581.0, + "step": 5638 + }, + { + "epoch": 0.6192620250384362, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4947142601013184, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6883288621902466, + "num_tokens": 146045852.0, + "step": 5639 + }, + { + "epoch": 0.6193718427410498, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5647218227386475, + "learning_rate": 1e-06, + "loss": 0.908, + "mean_token_accuracy": 0.7382515668869019, + "num_tokens": 146069988.0, + "step": 5640 + }, + { + "epoch": 0.6194816604436635, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9019417762756348, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.6996865272521973, + "num_tokens": 146090334.0, + "step": 5641 + }, + { + "epoch": 0.6195914781462771, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 3.2243688106536865, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7082650661468506, + "num_tokens": 146113724.0, + "step": 5642 + }, + { + "epoch": 0.6197012958488909, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.541346549987793, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7096683979034424, + "num_tokens": 146138973.0, + "step": 5643 + }, + { + "epoch": 0.6198111135515045, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.2769548892974854, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7261488437652588, + "num_tokens": 146168630.0, + "step": 5644 + }, + { + "epoch": 0.6199209312541182, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.492238759994507, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6856903433799744, + "num_tokens": 146198815.0, + "step": 5645 + }, + { + "epoch": 0.6200307489567318, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.881096124649048, + "learning_rate": 1e-06, + "loss": 0.9288, + "mean_token_accuracy": 0.7213214635848999, + "num_tokens": 146219566.0, + "step": 5646 + }, + { + "epoch": 0.6201405666593455, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6373085975646973, + "learning_rate": 1e-06, + "loss": 1.1191, + "mean_token_accuracy": 0.6785322427749634, + "num_tokens": 146243718.0, + "step": 5647 + }, + { + "epoch": 0.6202503843619591, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7414917945861816, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.693848729133606, + "num_tokens": 146268636.0, + "step": 5648 + }, + { + "epoch": 0.6203602020645728, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.1429944038391113, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7108234167098999, + "num_tokens": 146291335.0, + "step": 5649 + }, + { + "epoch": 0.6204700197671865, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.86745548248291, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7109490633010864, + "num_tokens": 146311255.0, + "step": 5650 + }, + { + "epoch": 0.6205798374698002, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6279006004333496, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6946832537651062, + "num_tokens": 146335164.0, + "step": 5651 + }, + { + "epoch": 0.6206896551724138, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8497211933135986, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.721016526222229, + "num_tokens": 146354287.0, + "step": 5652 + }, + { + "epoch": 0.6207994728750275, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7417495250701904, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7157224416732788, + "num_tokens": 146380335.0, + "step": 5653 + }, + { + "epoch": 0.6209092905776411, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7028720378875732, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6949600577354431, + "num_tokens": 146401453.0, + "step": 5654 + }, + { + "epoch": 0.6210191082802548, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3373215198516846, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.6935656070709229, + "num_tokens": 146431847.0, + "step": 5655 + }, + { + "epoch": 0.6211289259828684, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5782973766326904, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6992038488388062, + "num_tokens": 146456537.0, + "step": 5656 + }, + { + "epoch": 0.6212387436854822, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0105645656585693, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.702406644821167, + "num_tokens": 146479809.0, + "step": 5657 + }, + { + "epoch": 0.6213485613880958, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5857415199279785, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.690787672996521, + "num_tokens": 146505965.0, + "step": 5658 + }, + { + "epoch": 0.6214583790907094, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5810556411743164, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7034796476364136, + "num_tokens": 146531551.0, + "step": 5659 + }, + { + "epoch": 0.6215681967933231, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.334665060043335, + "learning_rate": 1e-06, + "loss": 1.1043, + "mean_token_accuracy": 0.6760421991348267, + "num_tokens": 146563551.0, + "step": 5660 + }, + { + "epoch": 0.6216780144959367, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.779815673828125, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6853678226470947, + "num_tokens": 146591120.0, + "step": 5661 + }, + { + "epoch": 0.6217878321985504, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9532339572906494, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7124767303466797, + "num_tokens": 146611965.0, + "step": 5662 + }, + { + "epoch": 0.621897649901164, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.145927667617798, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.7009866237640381, + "num_tokens": 146648461.0, + "step": 5663 + }, + { + "epoch": 0.6220074676037777, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4947824478149414, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6961356997489929, + "num_tokens": 146676427.0, + "step": 5664 + }, + { + "epoch": 0.6221172853063914, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5715670585632324, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7150972485542297, + "num_tokens": 146700165.0, + "step": 5665 + }, + { + "epoch": 0.6222271030090051, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6471893787384033, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.69623202085495, + "num_tokens": 146724996.0, + "step": 5666 + }, + { + "epoch": 0.6223369207116187, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8626747131347656, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.707909882068634, + "num_tokens": 146750905.0, + "step": 5667 + }, + { + "epoch": 0.6224467384142324, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0273425579071045, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6995974183082581, + "num_tokens": 146776383.0, + "step": 5668 + }, + { + "epoch": 0.622556556116846, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5387370586395264, + "learning_rate": 1e-06, + "loss": 0.942, + "mean_token_accuracy": 0.7277671098709106, + "num_tokens": 146800241.0, + "step": 5669 + }, + { + "epoch": 0.6226663738194597, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3801276683807373, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.704055666923523, + "num_tokens": 146831318.0, + "step": 5670 + }, + { + "epoch": 0.6227761915220733, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.941981554031372, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.69663405418396, + "num_tokens": 146853834.0, + "step": 5671 + }, + { + "epoch": 0.6228860092246871, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.856496572494507, + "learning_rate": 1e-06, + "loss": 0.902, + "mean_token_accuracy": 0.7300994992256165, + "num_tokens": 146872499.0, + "step": 5672 + }, + { + "epoch": 0.6229958269273007, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6128311157226562, + "learning_rate": 1e-06, + "loss": 1.1233, + "mean_token_accuracy": 0.6786950826644897, + "num_tokens": 146897374.0, + "step": 5673 + }, + { + "epoch": 0.6231056446299144, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.337355375289917, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6885840892791748, + "num_tokens": 146927243.0, + "step": 5674 + }, + { + "epoch": 0.623215462332528, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6137888431549072, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7109807729721069, + "num_tokens": 146949952.0, + "step": 5675 + }, + { + "epoch": 0.6233252800351416, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4629008769989014, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6844528317451477, + "num_tokens": 146979182.0, + "step": 5676 + }, + { + "epoch": 0.6234350977377553, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.387660026550293, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6928039193153381, + "num_tokens": 147009131.0, + "step": 5677 + }, + { + "epoch": 0.6235449154403689, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5642595291137695, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6949930191040039, + "num_tokens": 147038561.0, + "step": 5678 + }, + { + "epoch": 0.6236547331429827, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9644510746002197, + "learning_rate": 1e-06, + "loss": 0.9284, + "mean_token_accuracy": 0.725901186466217, + "num_tokens": 147055735.0, + "step": 5679 + }, + { + "epoch": 0.6237645508455963, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7806923389434814, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7091400623321533, + "num_tokens": 147076421.0, + "step": 5680 + }, + { + "epoch": 0.62387436854821, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6750893592834473, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7021967768669128, + "num_tokens": 147099457.0, + "step": 5681 + }, + { + "epoch": 0.6239841862508236, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6204731464385986, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6896367073059082, + "num_tokens": 147125823.0, + "step": 5682 + }, + { + "epoch": 0.6240940039534373, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.485978841781616, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7022625207901001, + "num_tokens": 147154551.0, + "step": 5683 + }, + { + "epoch": 0.6242038216560509, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7546091079711914, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6982067823410034, + "num_tokens": 147177596.0, + "step": 5684 + }, + { + "epoch": 0.6243136393586646, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.768669605255127, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6941640377044678, + "num_tokens": 147199678.0, + "step": 5685 + }, + { + "epoch": 0.6244234570612783, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5291049480438232, + "learning_rate": 1e-06, + "loss": 0.934, + "mean_token_accuracy": 0.724589467048645, + "num_tokens": 147223851.0, + "step": 5686 + }, + { + "epoch": 0.624533274763892, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3091814517974854, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7157763242721558, + "num_tokens": 147251227.0, + "step": 5687 + }, + { + "epoch": 0.6246430924665056, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4628031253814697, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6829463243484497, + "num_tokens": 147277361.0, + "step": 5688 + }, + { + "epoch": 0.6247529101691193, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.683401346206665, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.6992560625076294, + "num_tokens": 147297440.0, + "step": 5689 + }, + { + "epoch": 0.6248627278717329, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.584726095199585, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.704039454460144, + "num_tokens": 147324500.0, + "step": 5690 + }, + { + "epoch": 0.6249725455743466, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.483243942260742, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6889196634292603, + "num_tokens": 147352353.0, + "step": 5691 + }, + { + "epoch": 0.6250823632769602, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.094641923904419, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7053951621055603, + "num_tokens": 147371328.0, + "step": 5692 + }, + { + "epoch": 0.6251921809795739, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.825740337371826, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7147269248962402, + "num_tokens": 147390884.0, + "step": 5693 + }, + { + "epoch": 0.6253019986821876, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.7998218536376953, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7166515588760376, + "num_tokens": 147410305.0, + "step": 5694 + }, + { + "epoch": 0.6254118163848013, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7159111499786377, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7010616660118103, + "num_tokens": 147434813.0, + "step": 5695 + }, + { + "epoch": 0.6255216340874149, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4716434478759766, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6965005993843079, + "num_tokens": 147466456.0, + "step": 5696 + }, + { + "epoch": 0.6256314517900285, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.01733660697937, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7098077535629272, + "num_tokens": 147485784.0, + "step": 5697 + }, + { + "epoch": 0.6257412694926422, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.848264694213867, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7216071486473083, + "num_tokens": 147509484.0, + "step": 5698 + }, + { + "epoch": 0.6258510871952558, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.764132261276245, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.702350914478302, + "num_tokens": 147530235.0, + "step": 5699 + }, + { + "epoch": 0.6259609048978695, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5823864936828613, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.6845766305923462, + "num_tokens": 147556317.0, + "step": 5700 + }, + { + "epoch": 0.6260707226004832, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6855032444000244, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.7049265503883362, + "num_tokens": 147580805.0, + "step": 5701 + }, + { + "epoch": 0.6261805403030969, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5873172283172607, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7099776864051819, + "num_tokens": 147605713.0, + "step": 5702 + }, + { + "epoch": 0.6262903580057105, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.490049123764038, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6905263662338257, + "num_tokens": 147632335.0, + "step": 5703 + }, + { + "epoch": 0.6264001757083242, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3863022327423096, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6944983005523682, + "num_tokens": 147658970.0, + "step": 5704 + }, + { + "epoch": 0.6265099934109378, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.421560764312744, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6972532868385315, + "num_tokens": 147686015.0, + "step": 5705 + }, + { + "epoch": 0.6266198111135515, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0035290718078613, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.700258195400238, + "num_tokens": 147708507.0, + "step": 5706 + }, + { + "epoch": 0.6267296288161651, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5790815353393555, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6786308288574219, + "num_tokens": 147736217.0, + "step": 5707 + }, + { + "epoch": 0.6268394465187789, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5834431648254395, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.7002004981040955, + "num_tokens": 147760731.0, + "step": 5708 + }, + { + "epoch": 0.6269492642213925, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5631346702575684, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6852755546569824, + "num_tokens": 147786980.0, + "step": 5709 + }, + { + "epoch": 0.6270590819240062, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.810790777206421, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7160816788673401, + "num_tokens": 147807755.0, + "step": 5710 + }, + { + "epoch": 0.6271688996266198, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7717554569244385, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6927366852760315, + "num_tokens": 147834274.0, + "step": 5711 + }, + { + "epoch": 0.6272787173292335, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.655750274658203, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7147566080093384, + "num_tokens": 147860854.0, + "step": 5712 + }, + { + "epoch": 0.6273885350318471, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.708526134490967, + "learning_rate": 1e-06, + "loss": 1.0917, + "mean_token_accuracy": 0.6786431670188904, + "num_tokens": 147887369.0, + "step": 5713 + }, + { + "epoch": 0.6274983527344608, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3747782707214355, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.718242883682251, + "num_tokens": 147915582.0, + "step": 5714 + }, + { + "epoch": 0.6276081704370745, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4746429920196533, + "learning_rate": 1e-06, + "loss": 1.097, + "mean_token_accuracy": 0.6889816522598267, + "num_tokens": 147943778.0, + "step": 5715 + }, + { + "epoch": 0.6277179881396882, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.62558913230896, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6902740001678467, + "num_tokens": 147967687.0, + "step": 5716 + }, + { + "epoch": 0.6278278058423018, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6450767517089844, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6933558583259583, + "num_tokens": 147996893.0, + "step": 5717 + }, + { + "epoch": 0.6279376235449154, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.327890396118164, + "learning_rate": 1e-06, + "loss": 1.121, + "mean_token_accuracy": 0.6761237978935242, + "num_tokens": 148030636.0, + "step": 5718 + }, + { + "epoch": 0.6280474412475291, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5256507396698, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7102006673812866, + "num_tokens": 148054012.0, + "step": 5719 + }, + { + "epoch": 0.6281572589501427, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6838855743408203, + "learning_rate": 1e-06, + "loss": 1.1315, + "mean_token_accuracy": 0.6746165156364441, + "num_tokens": 148085309.0, + "step": 5720 + }, + { + "epoch": 0.6282670766527564, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0147764682769775, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6904273629188538, + "num_tokens": 148109119.0, + "step": 5721 + }, + { + "epoch": 0.62837689435537, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.868650197982788, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6935036182403564, + "num_tokens": 148132385.0, + "step": 5722 + }, + { + "epoch": 0.6284867120579838, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.346175193786621, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.6976787447929382, + "num_tokens": 148165336.0, + "step": 5723 + }, + { + "epoch": 0.6285965297605974, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5115809440612793, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6950502991676331, + "num_tokens": 148195345.0, + "step": 5724 + }, + { + "epoch": 0.6287063474632111, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6515932083129883, + "learning_rate": 1e-06, + "loss": 1.1234, + "mean_token_accuracy": 0.6815745830535889, + "num_tokens": 148218406.0, + "step": 5725 + }, + { + "epoch": 0.6288161651658247, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.28063702583313, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7034170627593994, + "num_tokens": 148249170.0, + "step": 5726 + }, + { + "epoch": 0.6289259828684384, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7848165035247803, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.7158681750297546, + "num_tokens": 148269680.0, + "step": 5727 + }, + { + "epoch": 0.629035800571052, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7164206504821777, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7211013436317444, + "num_tokens": 148291251.0, + "step": 5728 + }, + { + "epoch": 0.6291456182736657, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.621020793914795, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7084332704544067, + "num_tokens": 148318381.0, + "step": 5729 + }, + { + "epoch": 0.6292554359762794, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.37664532661438, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7080817818641663, + "num_tokens": 148345808.0, + "step": 5730 + }, + { + "epoch": 0.6293652536788931, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.807379722595215, + "learning_rate": 1e-06, + "loss": 0.9269, + "mean_token_accuracy": 0.7199429273605347, + "num_tokens": 148367241.0, + "step": 5731 + }, + { + "epoch": 0.6294750713815067, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.499859094619751, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6819891929626465, + "num_tokens": 148394134.0, + "step": 5732 + }, + { + "epoch": 0.6295848890841204, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4835824966430664, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7085385322570801, + "num_tokens": 148418898.0, + "step": 5733 + }, + { + "epoch": 0.629694706786734, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.67539381980896, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6981385946273804, + "num_tokens": 148445447.0, + "step": 5734 + }, + { + "epoch": 0.6298045244893477, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7461657524108887, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.695367693901062, + "num_tokens": 148468281.0, + "step": 5735 + }, + { + "epoch": 0.6299143421919613, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.472841262817383, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7025294303894043, + "num_tokens": 148494389.0, + "step": 5736 + }, + { + "epoch": 0.630024159894575, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5655486583709717, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7016788125038147, + "num_tokens": 148521662.0, + "step": 5737 + }, + { + "epoch": 0.6301339775971887, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.03808856010437, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7124412655830383, + "num_tokens": 148541457.0, + "step": 5738 + }, + { + "epoch": 0.6302437952998023, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3889856338500977, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6840760111808777, + "num_tokens": 148569634.0, + "step": 5739 + }, + { + "epoch": 0.630353613002416, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.2672295570373535, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6960979700088501, + "num_tokens": 148600795.0, + "step": 5740 + }, + { + "epoch": 0.6304634307050296, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.474717855453491, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7063726186752319, + "num_tokens": 148629434.0, + "step": 5741 + }, + { + "epoch": 0.6305732484076433, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3392887115478516, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7006109356880188, + "num_tokens": 148658548.0, + "step": 5742 + }, + { + "epoch": 0.6306830661102569, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.650108575820923, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7093876600265503, + "num_tokens": 148685449.0, + "step": 5743 + }, + { + "epoch": 0.6307928838128707, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.577509641647339, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.692332923412323, + "num_tokens": 148709970.0, + "step": 5744 + }, + { + "epoch": 0.6309027015154843, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3957407474517822, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7007694244384766, + "num_tokens": 148737062.0, + "step": 5745 + }, + { + "epoch": 0.631012519218098, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.442347526550293, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7019162774085999, + "num_tokens": 148767671.0, + "step": 5746 + }, + { + "epoch": 0.6311223369207116, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4272871017456055, + "learning_rate": 1e-06, + "loss": 1.1278, + "mean_token_accuracy": 0.6709752082824707, + "num_tokens": 148798121.0, + "step": 5747 + }, + { + "epoch": 0.6312321546233253, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.559657096862793, + "learning_rate": 1e-06, + "loss": 0.8831, + "mean_token_accuracy": 0.7394327521324158, + "num_tokens": 148822303.0, + "step": 5748 + }, + { + "epoch": 0.6313419723259389, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.685786724090576, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.704398512840271, + "num_tokens": 148846091.0, + "step": 5749 + }, + { + "epoch": 0.6314517900285526, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4056639671325684, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.7014052867889404, + "num_tokens": 148875284.0, + "step": 5750 + }, + { + "epoch": 0.6315616077311663, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4660773277282715, + "learning_rate": 1e-06, + "loss": 1.128, + "mean_token_accuracy": 0.6902731657028198, + "num_tokens": 148903453.0, + "step": 5751 + }, + { + "epoch": 0.63167142543378, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4817304611206055, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6992346048355103, + "num_tokens": 148930804.0, + "step": 5752 + }, + { + "epoch": 0.6317812431363936, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3191378116607666, + "learning_rate": 1e-06, + "loss": 1.1208, + "mean_token_accuracy": 0.6772294044494629, + "num_tokens": 148961324.0, + "step": 5753 + }, + { + "epoch": 0.6318910608390073, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5289723873138428, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7034175395965576, + "num_tokens": 148989833.0, + "step": 5754 + }, + { + "epoch": 0.6320008785416209, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.3496198654174805, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6949724555015564, + "num_tokens": 149021071.0, + "step": 5755 + }, + { + "epoch": 0.6321106962442345, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.528573513031006, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7122669219970703, + "num_tokens": 149044826.0, + "step": 5756 + }, + { + "epoch": 0.6322205139468482, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.435823678970337, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6948134303092957, + "num_tokens": 149071338.0, + "step": 5757 + }, + { + "epoch": 0.6323303316494618, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.4918570518493652, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.7031363844871521, + "num_tokens": 149100176.0, + "step": 5758 + }, + { + "epoch": 0.6324401493520756, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.729822874069214, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6880418062210083, + "num_tokens": 149126871.0, + "step": 5759 + }, + { + "epoch": 0.6325499670546892, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.701463460922241, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7012141942977905, + "num_tokens": 149149003.0, + "step": 5760 + }, + { + "epoch": 0.6326597847573029, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 3.126957416534424, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.6758864521980286, + "num_tokens": 149172255.0, + "step": 5761 + }, + { + "epoch": 0.6327696024599165, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.726545572280884, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6889815926551819, + "num_tokens": 149194487.0, + "step": 5762 + }, + { + "epoch": 0.6328794201625302, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5171172618865967, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6972833871841431, + "num_tokens": 149219638.0, + "step": 5763 + }, + { + "epoch": 0.6329892378651438, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.071955442428589, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6935317516326904, + "num_tokens": 149238091.0, + "step": 5764 + }, + { + "epoch": 0.6330990555677575, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.617774486541748, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7023593187332153, + "num_tokens": 149261916.0, + "step": 5765 + }, + { + "epoch": 0.6332088732703712, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8558757305145264, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.721028208732605, + "num_tokens": 149283449.0, + "step": 5766 + }, + { + "epoch": 0.6333186909729849, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4108543395996094, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7070527672767639, + "num_tokens": 149309221.0, + "step": 5767 + }, + { + "epoch": 0.6334285086755985, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4305872917175293, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6855934262275696, + "num_tokens": 149338800.0, + "step": 5768 + }, + { + "epoch": 0.6335383263782122, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.40757417678833, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6998425126075745, + "num_tokens": 149367427.0, + "step": 5769 + }, + { + "epoch": 0.6336481440808258, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6908230781555176, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.7027965784072876, + "num_tokens": 149391851.0, + "step": 5770 + }, + { + "epoch": 0.6337579617834395, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8495664596557617, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.709979772567749, + "num_tokens": 149415251.0, + "step": 5771 + }, + { + "epoch": 0.6338677794860531, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.572622060775757, + "learning_rate": 1e-06, + "loss": 1.1009, + "mean_token_accuracy": 0.6794096827507019, + "num_tokens": 149444650.0, + "step": 5772 + }, + { + "epoch": 0.6339775971886669, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3622004985809326, + "learning_rate": 1e-06, + "loss": 1.1373, + "mean_token_accuracy": 0.66944420337677, + "num_tokens": 149475849.0, + "step": 5773 + }, + { + "epoch": 0.6340874148912805, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.407381534576416, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6879832744598389, + "num_tokens": 149503813.0, + "step": 5774 + }, + { + "epoch": 0.6341972325938942, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.622122287750244, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6953718662261963, + "num_tokens": 149527726.0, + "step": 5775 + }, + { + "epoch": 0.6343070502965078, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3477513790130615, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6873645186424255, + "num_tokens": 149560705.0, + "step": 5776 + }, + { + "epoch": 0.6344168679991214, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3853209018707275, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.700701892375946, + "num_tokens": 149587557.0, + "step": 5777 + }, + { + "epoch": 0.6345266857017351, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.441688060760498, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6926050186157227, + "num_tokens": 149613308.0, + "step": 5778 + }, + { + "epoch": 0.6346365034043487, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6968705654144287, + "learning_rate": 1e-06, + "loss": 1.1207, + "mean_token_accuracy": 0.6725186705589294, + "num_tokens": 149639459.0, + "step": 5779 + }, + { + "epoch": 0.6347463211069625, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4369008541107178, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6975346803665161, + "num_tokens": 149669146.0, + "step": 5780 + }, + { + "epoch": 0.6348561388095761, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7890920639038086, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7047306895256042, + "num_tokens": 149695518.0, + "step": 5781 + }, + { + "epoch": 0.6349659565121898, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.898942708969116, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.711239218711853, + "num_tokens": 149718383.0, + "step": 5782 + }, + { + "epoch": 0.6350757742148034, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.756606340408325, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7011444568634033, + "num_tokens": 149741179.0, + "step": 5783 + }, + { + "epoch": 0.6351855919174171, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.688866138458252, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6914027333259583, + "num_tokens": 149763981.0, + "step": 5784 + }, + { + "epoch": 0.6352954096200307, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.352426290512085, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6932646036148071, + "num_tokens": 149794146.0, + "step": 5785 + }, + { + "epoch": 0.6354052273226444, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5782687664031982, + "learning_rate": 1e-06, + "loss": 1.1132, + "mean_token_accuracy": 0.6903504133224487, + "num_tokens": 149819811.0, + "step": 5786 + }, + { + "epoch": 0.635515045025258, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3970608711242676, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6923394203186035, + "num_tokens": 149848238.0, + "step": 5787 + }, + { + "epoch": 0.6356248627278718, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6766700744628906, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7019516825675964, + "num_tokens": 149874801.0, + "step": 5788 + }, + { + "epoch": 0.6357346804304854, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.76419734954834, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7066159248352051, + "num_tokens": 149897048.0, + "step": 5789 + }, + { + "epoch": 0.6358444981330991, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6530332565307617, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.687009334564209, + "num_tokens": 149919727.0, + "step": 5790 + }, + { + "epoch": 0.6359543158357127, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.566243886947632, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.6994728446006775, + "num_tokens": 149945061.0, + "step": 5791 + }, + { + "epoch": 0.6360641335383264, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7878029346466064, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7141886949539185, + "num_tokens": 149966169.0, + "step": 5792 + }, + { + "epoch": 0.63617395124094, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5497422218322754, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7182134389877319, + "num_tokens": 149992607.0, + "step": 5793 + }, + { + "epoch": 0.6362837689435537, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6481010913848877, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6794034242630005, + "num_tokens": 150019636.0, + "step": 5794 + }, + { + "epoch": 0.6363935866461674, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5976641178131104, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.694298505783081, + "num_tokens": 150046268.0, + "step": 5795 + }, + { + "epoch": 0.636503404348781, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.487943649291992, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6846358776092529, + "num_tokens": 150072283.0, + "step": 5796 + }, + { + "epoch": 0.6366132220513947, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5061702728271484, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6981913447380066, + "num_tokens": 150099010.0, + "step": 5797 + }, + { + "epoch": 0.6367230397540083, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5080156326293945, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7051334977149963, + "num_tokens": 150124807.0, + "step": 5798 + }, + { + "epoch": 0.636832857456622, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6046712398529053, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7056249976158142, + "num_tokens": 150153282.0, + "step": 5799 + }, + { + "epoch": 0.6369426751592356, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.821672201156616, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6894217133522034, + "num_tokens": 150175704.0, + "step": 5800 + }, + { + "epoch": 0.6370524928618493, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.518984794616699, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7102248668670654, + "num_tokens": 150201266.0, + "step": 5801 + }, + { + "epoch": 0.637162310564463, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5726821422576904, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7097731232643127, + "num_tokens": 150223486.0, + "step": 5802 + }, + { + "epoch": 0.6372721282670767, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6521122455596924, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6988308429718018, + "num_tokens": 150249127.0, + "step": 5803 + }, + { + "epoch": 0.6373819459696903, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.62156081199646, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7181957960128784, + "num_tokens": 150272652.0, + "step": 5804 + }, + { + "epoch": 0.637491763672304, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6037116050720215, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7145956754684448, + "num_tokens": 150296163.0, + "step": 5805 + }, + { + "epoch": 0.6376015813749176, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4718759059906006, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6875250935554504, + "num_tokens": 150326248.0, + "step": 5806 + }, + { + "epoch": 0.6377113990775313, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.426079034805298, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.710243284702301, + "num_tokens": 150350743.0, + "step": 5807 + }, + { + "epoch": 0.6378212167801449, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6705806255340576, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7197698354721069, + "num_tokens": 150374351.0, + "step": 5808 + }, + { + "epoch": 0.6379310344827587, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3410427570343018, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6796305775642395, + "num_tokens": 150403427.0, + "step": 5809 + }, + { + "epoch": 0.6380408521853723, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5921566486358643, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7146113514900208, + "num_tokens": 150425915.0, + "step": 5810 + }, + { + "epoch": 0.638150669887986, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0608572959899902, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7093808650970459, + "num_tokens": 150448372.0, + "step": 5811 + }, + { + "epoch": 0.6382604875905996, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.571733236312866, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7167184352874756, + "num_tokens": 150470554.0, + "step": 5812 + }, + { + "epoch": 0.6383703052932133, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7076804637908936, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.6802796721458435, + "num_tokens": 150493751.0, + "step": 5813 + }, + { + "epoch": 0.6384801229958269, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.755690336227417, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6891663670539856, + "num_tokens": 150518023.0, + "step": 5814 + }, + { + "epoch": 0.6385899406984406, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.2275497913360596, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7186987400054932, + "num_tokens": 150549241.0, + "step": 5815 + }, + { + "epoch": 0.6386997584010542, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3864214420318604, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6994947195053101, + "num_tokens": 150580556.0, + "step": 5816 + }, + { + "epoch": 0.638809576103668, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3504323959350586, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.6870136260986328, + "num_tokens": 150611486.0, + "step": 5817 + }, + { + "epoch": 0.6389193938062816, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.612288236618042, + "learning_rate": 1e-06, + "loss": 1.123, + "mean_token_accuracy": 0.6721155047416687, + "num_tokens": 150636022.0, + "step": 5818 + }, + { + "epoch": 0.6390292115088952, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.653085947036743, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6943018436431885, + "num_tokens": 150659070.0, + "step": 5819 + }, + { + "epoch": 0.6391390292115089, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4599075317382812, + "learning_rate": 1e-06, + "loss": 0.924, + "mean_token_accuracy": 0.7272626757621765, + "num_tokens": 150685946.0, + "step": 5820 + }, + { + "epoch": 0.6392488469141225, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.631527900695801, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7078176736831665, + "num_tokens": 150710552.0, + "step": 5821 + }, + { + "epoch": 0.6393586646167362, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4506640434265137, + "learning_rate": 1e-06, + "loss": 1.1301, + "mean_token_accuracy": 0.6818221807479858, + "num_tokens": 150739809.0, + "step": 5822 + }, + { + "epoch": 0.6394684823193498, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8002302646636963, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7033679485321045, + "num_tokens": 150761127.0, + "step": 5823 + }, + { + "epoch": 0.6395783000219636, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.562009811401367, + "learning_rate": 1e-06, + "loss": 0.8305, + "mean_token_accuracy": 0.7546646595001221, + "num_tokens": 150786097.0, + "step": 5824 + }, + { + "epoch": 0.6396881177245772, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.1323516368865967, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6896282434463501, + "num_tokens": 150807894.0, + "step": 5825 + }, + { + "epoch": 0.6397979354271909, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6486451625823975, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7043638229370117, + "num_tokens": 150831323.0, + "step": 5826 + }, + { + "epoch": 0.6399077531298045, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4666354656219482, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6929436922073364, + "num_tokens": 150858194.0, + "step": 5827 + }, + { + "epoch": 0.6400175708324182, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5620276927948, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7045285701751709, + "num_tokens": 150882564.0, + "step": 5828 + }, + { + "epoch": 0.6401273885350318, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6835620403289795, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.7184613943099976, + "num_tokens": 150907187.0, + "step": 5829 + }, + { + "epoch": 0.6402372062376455, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6499948501586914, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6817325353622437, + "num_tokens": 150934895.0, + "step": 5830 + }, + { + "epoch": 0.6403470239402592, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6420369148254395, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.6916742324829102, + "num_tokens": 150958210.0, + "step": 5831 + }, + { + "epoch": 0.6404568416428729, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.818225145339966, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7299633026123047, + "num_tokens": 150980567.0, + "step": 5832 + }, + { + "epoch": 0.6405666593454865, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4172146320343018, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6883819103240967, + "num_tokens": 151009372.0, + "step": 5833 + }, + { + "epoch": 0.6406764770481002, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5606484413146973, + "learning_rate": 1e-06, + "loss": 1.0837, + "mean_token_accuracy": 0.6920830011367798, + "num_tokens": 151035712.0, + "step": 5834 + }, + { + "epoch": 0.6407862947507138, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4798123836517334, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7182700037956238, + "num_tokens": 151061318.0, + "step": 5835 + }, + { + "epoch": 0.6408961124533274, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3334314823150635, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6996920704841614, + "num_tokens": 151087793.0, + "step": 5836 + }, + { + "epoch": 0.6410059301559411, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.1471660137176514, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6903368234634399, + "num_tokens": 151111689.0, + "step": 5837 + }, + { + "epoch": 0.6411157478585549, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.312654972076416, + "learning_rate": 1e-06, + "loss": 1.1102, + "mean_token_accuracy": 0.6833269596099854, + "num_tokens": 151144752.0, + "step": 5838 + }, + { + "epoch": 0.6412255655611685, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3626532554626465, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.7051644325256348, + "num_tokens": 151175374.0, + "step": 5839 + }, + { + "epoch": 0.6413353832637821, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3554272651672363, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6926133036613464, + "num_tokens": 151207581.0, + "step": 5840 + }, + { + "epoch": 0.6414452009663958, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.599299669265747, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7173279523849487, + "num_tokens": 151231272.0, + "step": 5841 + }, + { + "epoch": 0.6415550186690094, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.567298650741577, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.6929052472114563, + "num_tokens": 151259848.0, + "step": 5842 + }, + { + "epoch": 0.6416648363716231, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7013182640075684, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7131207585334778, + "num_tokens": 151283535.0, + "step": 5843 + }, + { + "epoch": 0.6417746540742367, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9388041496276855, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7137485146522522, + "num_tokens": 151307785.0, + "step": 5844 + }, + { + "epoch": 0.6418844717768504, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.529937267303467, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6864862442016602, + "num_tokens": 151334314.0, + "step": 5845 + }, + { + "epoch": 0.6419942894794641, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7806875705718994, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7131170630455017, + "num_tokens": 151355616.0, + "step": 5846 + }, + { + "epoch": 0.6421041071820778, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.630913496017456, + "learning_rate": 1e-06, + "loss": 0.9287, + "mean_token_accuracy": 0.7251104116439819, + "num_tokens": 151377054.0, + "step": 5847 + }, + { + "epoch": 0.6422139248846914, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4486453533172607, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7075531482696533, + "num_tokens": 151402658.0, + "step": 5848 + }, + { + "epoch": 0.6423237425873051, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.495314598083496, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7005801200866699, + "num_tokens": 151429171.0, + "step": 5849 + }, + { + "epoch": 0.6424335602899187, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4708845615386963, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6989843249320984, + "num_tokens": 151458553.0, + "step": 5850 + }, + { + "epoch": 0.6425433779925324, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.662989616394043, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7119592428207397, + "num_tokens": 151481152.0, + "step": 5851 + }, + { + "epoch": 0.642653195695146, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4719252586364746, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7039196491241455, + "num_tokens": 151511318.0, + "step": 5852 + }, + { + "epoch": 0.6427630133977598, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.662959098815918, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7071230411529541, + "num_tokens": 151534627.0, + "step": 5853 + }, + { + "epoch": 0.6428728311003734, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.633411169052124, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6941548585891724, + "num_tokens": 151560458.0, + "step": 5854 + }, + { + "epoch": 0.6429826488029871, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.838257312774658, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.6983029842376709, + "num_tokens": 151582610.0, + "step": 5855 + }, + { + "epoch": 0.6430924665056007, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.681014060974121, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7159112691879272, + "num_tokens": 151608983.0, + "step": 5856 + }, + { + "epoch": 0.6432022842082143, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.888415575027466, + "learning_rate": 1e-06, + "loss": 0.9428, + "mean_token_accuracy": 0.7200408577919006, + "num_tokens": 151627493.0, + "step": 5857 + }, + { + "epoch": 0.643312101910828, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.1582324504852295, + "learning_rate": 1e-06, + "loss": 0.9063, + "mean_token_accuracy": 0.7377581596374512, + "num_tokens": 151661357.0, + "step": 5858 + }, + { + "epoch": 0.6434219196134416, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.240243911743164, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.692695140838623, + "num_tokens": 151696267.0, + "step": 5859 + }, + { + "epoch": 0.6435317373160554, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.412227153778076, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.69640052318573, + "num_tokens": 151724986.0, + "step": 5860 + }, + { + "epoch": 0.643641555018669, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5673530101776123, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7171270847320557, + "num_tokens": 151752918.0, + "step": 5861 + }, + { + "epoch": 0.6437513727212827, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9268670082092285, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7202936410903931, + "num_tokens": 151777243.0, + "step": 5862 + }, + { + "epoch": 0.6438611904238963, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7004380226135254, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6956698894500732, + "num_tokens": 151805063.0, + "step": 5863 + }, + { + "epoch": 0.64397100812651, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7541701793670654, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7074580788612366, + "num_tokens": 151831224.0, + "step": 5864 + }, + { + "epoch": 0.6440808258291236, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.433631420135498, + "learning_rate": 1e-06, + "loss": 1.1016, + "mean_token_accuracy": 0.6745748519897461, + "num_tokens": 151862421.0, + "step": 5865 + }, + { + "epoch": 0.6441906435317373, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.690678596496582, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6917426586151123, + "num_tokens": 151888055.0, + "step": 5866 + }, + { + "epoch": 0.644300461234351, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7340505123138428, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7222048044204712, + "num_tokens": 151907627.0, + "step": 5867 + }, + { + "epoch": 0.6444102789369647, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9604194164276123, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7014148831367493, + "num_tokens": 151931472.0, + "step": 5868 + }, + { + "epoch": 0.6445200966395783, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.681495428085327, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7122061252593994, + "num_tokens": 151958960.0, + "step": 5869 + }, + { + "epoch": 0.644629914342192, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.373345136642456, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7049541473388672, + "num_tokens": 151986428.0, + "step": 5870 + }, + { + "epoch": 0.6447397320448056, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6758713722229004, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7023037075996399, + "num_tokens": 152009759.0, + "step": 5871 + }, + { + "epoch": 0.6448495497474193, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5053598880767822, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.6825827360153198, + "num_tokens": 152039419.0, + "step": 5872 + }, + { + "epoch": 0.6449593674500329, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.639842987060547, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7065427899360657, + "num_tokens": 152061413.0, + "step": 5873 + }, + { + "epoch": 0.6450691851526466, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6770262718200684, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6820181012153625, + "num_tokens": 152086909.0, + "step": 5874 + }, + { + "epoch": 0.6451790028552603, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.760190486907959, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7071611881256104, + "num_tokens": 152110852.0, + "step": 5875 + }, + { + "epoch": 0.645288820557874, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.593787670135498, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6866379380226135, + "num_tokens": 152139170.0, + "step": 5876 + }, + { + "epoch": 0.6453986382604876, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7008275985717773, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6917794942855835, + "num_tokens": 152162709.0, + "step": 5877 + }, + { + "epoch": 0.6455084559631012, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5387377738952637, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6899789571762085, + "num_tokens": 152190478.0, + "step": 5878 + }, + { + "epoch": 0.6456182736657149, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4762706756591797, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.6992204189300537, + "num_tokens": 152218556.0, + "step": 5879 + }, + { + "epoch": 0.6457280913683285, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9291207790374756, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.7058367133140564, + "num_tokens": 152242928.0, + "step": 5880 + }, + { + "epoch": 0.6458379090709422, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.651765823364258, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7055951356887817, + "num_tokens": 152266912.0, + "step": 5881 + }, + { + "epoch": 0.6459477267735559, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3491554260253906, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.6933550834655762, + "num_tokens": 152297859.0, + "step": 5882 + }, + { + "epoch": 0.6460575444761696, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7844526767730713, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7030113339424133, + "num_tokens": 152320490.0, + "step": 5883 + }, + { + "epoch": 0.6461673621787832, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5450026988983154, + "learning_rate": 1e-06, + "loss": 1.0946, + "mean_token_accuracy": 0.687067449092865, + "num_tokens": 152346999.0, + "step": 5884 + }, + { + "epoch": 0.6462771798813969, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.470698595046997, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7011879682540894, + "num_tokens": 152377770.0, + "step": 5885 + }, + { + "epoch": 0.6463869975840105, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.713754653930664, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.6787829995155334, + "num_tokens": 152408079.0, + "step": 5886 + }, + { + "epoch": 0.6464968152866242, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.995039701461792, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6908569931983948, + "num_tokens": 152430651.0, + "step": 5887 + }, + { + "epoch": 0.6466066329892378, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6767916679382324, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6917169094085693, + "num_tokens": 152459557.0, + "step": 5888 + }, + { + "epoch": 0.6467164506918516, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.051720142364502, + "learning_rate": 1e-06, + "loss": 0.9086, + "mean_token_accuracy": 0.7271568775177002, + "num_tokens": 152478085.0, + "step": 5889 + }, + { + "epoch": 0.6468262683944652, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5984749794006348, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7206786870956421, + "num_tokens": 152506522.0, + "step": 5890 + }, + { + "epoch": 0.6469360860970789, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8138461112976074, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6773015260696411, + "num_tokens": 152535998.0, + "step": 5891 + }, + { + "epoch": 0.6470459037996925, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6021690368652344, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6886736750602722, + "num_tokens": 152562365.0, + "step": 5892 + }, + { + "epoch": 0.6471557215023062, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4659409523010254, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6808860898017883, + "num_tokens": 152591858.0, + "step": 5893 + }, + { + "epoch": 0.6472655392049198, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8041372299194336, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7244417071342468, + "num_tokens": 152616172.0, + "step": 5894 + }, + { + "epoch": 0.6473753569075335, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7169687747955322, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6941169500350952, + "num_tokens": 152639601.0, + "step": 5895 + }, + { + "epoch": 0.6474851746101472, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4900379180908203, + "learning_rate": 1e-06, + "loss": 1.12, + "mean_token_accuracy": 0.6734118461608887, + "num_tokens": 152667239.0, + "step": 5896 + }, + { + "epoch": 0.6475949923127609, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.338684320449829, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6825205683708191, + "num_tokens": 152697231.0, + "step": 5897 + }, + { + "epoch": 0.6477048100153745, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.330907106399536, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6917574405670166, + "num_tokens": 152728080.0, + "step": 5898 + }, + { + "epoch": 0.6478146277179881, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3404688835144043, + "learning_rate": 1e-06, + "loss": 1.0921, + "mean_token_accuracy": 0.6813318729400635, + "num_tokens": 152762345.0, + "step": 5899 + }, + { + "epoch": 0.6479244454206018, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.49446702003479, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6899288892745972, + "num_tokens": 152789650.0, + "step": 5900 + }, + { + "epoch": 0.6480342631232154, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.390714406967163, + "learning_rate": 1e-06, + "loss": 1.1476, + "mean_token_accuracy": 0.6750912666320801, + "num_tokens": 152821126.0, + "step": 5901 + }, + { + "epoch": 0.6481440808258291, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7275073528289795, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6954919099807739, + "num_tokens": 152846609.0, + "step": 5902 + }, + { + "epoch": 0.6482538985284428, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4809536933898926, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6946594715118408, + "num_tokens": 152872815.0, + "step": 5903 + }, + { + "epoch": 0.6483637162310565, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4255988597869873, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6858310699462891, + "num_tokens": 152900870.0, + "step": 5904 + }, + { + "epoch": 0.6484735339336701, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.560370922088623, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7162529230117798, + "num_tokens": 152926475.0, + "step": 5905 + }, + { + "epoch": 0.6485833516362838, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6485021114349365, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6867467761039734, + "num_tokens": 152957867.0, + "step": 5906 + }, + { + "epoch": 0.6486931693388974, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6830127239227295, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7108442783355713, + "num_tokens": 152982578.0, + "step": 5907 + }, + { + "epoch": 0.6488029870415111, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6321098804473877, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7133225202560425, + "num_tokens": 153006120.0, + "step": 5908 + }, + { + "epoch": 0.6489128047441247, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4726297855377197, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7234190702438354, + "num_tokens": 153032777.0, + "step": 5909 + }, + { + "epoch": 0.6490226224467384, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.597484588623047, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6845786571502686, + "num_tokens": 153061135.0, + "step": 5910 + }, + { + "epoch": 0.6491324401493521, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3368794918060303, + "learning_rate": 1e-06, + "loss": 1.1561, + "mean_token_accuracy": 0.6567562818527222, + "num_tokens": 153094626.0, + "step": 5911 + }, + { + "epoch": 0.6492422578519658, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.675964593887329, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6965024471282959, + "num_tokens": 153120192.0, + "step": 5912 + }, + { + "epoch": 0.6493520755545794, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5486481189727783, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7028102278709412, + "num_tokens": 153146784.0, + "step": 5913 + }, + { + "epoch": 0.6494618932571931, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8499255180358887, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7161140441894531, + "num_tokens": 153170056.0, + "step": 5914 + }, + { + "epoch": 0.6495717109598067, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5020759105682373, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.716254711151123, + "num_tokens": 153195030.0, + "step": 5915 + }, + { + "epoch": 0.6496815286624203, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5628745555877686, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7104653120040894, + "num_tokens": 153221898.0, + "step": 5916 + }, + { + "epoch": 0.649791346365034, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4099502563476562, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6969946622848511, + "num_tokens": 153255269.0, + "step": 5917 + }, + { + "epoch": 0.6499011640676478, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3734235763549805, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.7161343097686768, + "num_tokens": 153281596.0, + "step": 5918 + }, + { + "epoch": 0.6500109817702614, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9698147773742676, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7160145044326782, + "num_tokens": 153299996.0, + "step": 5919 + }, + { + "epoch": 0.650120799472875, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9662225246429443, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7009444832801819, + "num_tokens": 153324129.0, + "step": 5920 + }, + { + "epoch": 0.6502306171754887, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7717983722686768, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6960885524749756, + "num_tokens": 153348145.0, + "step": 5921 + }, + { + "epoch": 0.6503404348781023, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.755922555923462, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7086306810379028, + "num_tokens": 153371737.0, + "step": 5922 + }, + { + "epoch": 0.650450252580716, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.73439359664917, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6981367468833923, + "num_tokens": 153402899.0, + "step": 5923 + }, + { + "epoch": 0.6505600702833296, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4731879234313965, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7239030599594116, + "num_tokens": 153428385.0, + "step": 5924 + }, + { + "epoch": 0.6506698879859434, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.417548418045044, + "learning_rate": 1e-06, + "loss": 1.1065, + "mean_token_accuracy": 0.6830787062644958, + "num_tokens": 153461679.0, + "step": 5925 + }, + { + "epoch": 0.650779705688557, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3719897270202637, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6947285532951355, + "num_tokens": 153491875.0, + "step": 5926 + }, + { + "epoch": 0.6508895233911707, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8015170097351074, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6942081451416016, + "num_tokens": 153512525.0, + "step": 5927 + }, + { + "epoch": 0.6509993410937843, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.556678056716919, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7018996477127075, + "num_tokens": 153537574.0, + "step": 5928 + }, + { + "epoch": 0.651109158796398, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9474942684173584, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.709443211555481, + "num_tokens": 153561400.0, + "step": 5929 + }, + { + "epoch": 0.6512189764990116, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.507751941680908, + "learning_rate": 1e-06, + "loss": 1.0753, + "mean_token_accuracy": 0.6898221373558044, + "num_tokens": 153588151.0, + "step": 5930 + }, + { + "epoch": 0.6513287942016253, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6278135776519775, + "learning_rate": 1e-06, + "loss": 1.1686, + "mean_token_accuracy": 0.6648062467575073, + "num_tokens": 153619776.0, + "step": 5931 + }, + { + "epoch": 0.651438611904239, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9264395236968994, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7007753849029541, + "num_tokens": 153639527.0, + "step": 5932 + }, + { + "epoch": 0.6515484296068527, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3281030654907227, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7191856503486633, + "num_tokens": 153667505.0, + "step": 5933 + }, + { + "epoch": 0.6516582473094663, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4472646713256836, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6936434507369995, + "num_tokens": 153695390.0, + "step": 5934 + }, + { + "epoch": 0.65176806501208, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.758784055709839, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7088886499404907, + "num_tokens": 153717638.0, + "step": 5935 + }, + { + "epoch": 0.6518778827146936, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.782721757888794, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7084364891052246, + "num_tokens": 153738948.0, + "step": 5936 + }, + { + "epoch": 0.6519877004173072, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 3.0062355995178223, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7247625589370728, + "num_tokens": 153755757.0, + "step": 5937 + }, + { + "epoch": 0.6520975181199209, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.605996608734131, + "learning_rate": 1e-06, + "loss": 1.1037, + "mean_token_accuracy": 0.6803319454193115, + "num_tokens": 153782883.0, + "step": 5938 + }, + { + "epoch": 0.6522073358225345, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.574158191680908, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7132328152656555, + "num_tokens": 153809530.0, + "step": 5939 + }, + { + "epoch": 0.6523171535251483, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.6053483486175537, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7301926612854004, + "num_tokens": 153833445.0, + "step": 5940 + }, + { + "epoch": 0.6524269712277619, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.657665252685547, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6896066665649414, + "num_tokens": 153863494.0, + "step": 5941 + }, + { + "epoch": 0.6525367889303756, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.566859245300293, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6978926658630371, + "num_tokens": 153889125.0, + "step": 5942 + }, + { + "epoch": 0.6526466066329892, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.771609306335449, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6980100870132446, + "num_tokens": 153911542.0, + "step": 5943 + }, + { + "epoch": 0.6527564243356029, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 3.854166030883789, + "learning_rate": 1e-06, + "loss": 1.1228, + "mean_token_accuracy": 0.6826819181442261, + "num_tokens": 153942533.0, + "step": 5944 + }, + { + "epoch": 0.6528662420382165, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.6020493507385254, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6923060417175293, + "num_tokens": 153967534.0, + "step": 5945 + }, + { + "epoch": 0.6529760597408302, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.532723903656006, + "learning_rate": 1e-06, + "loss": 1.1438, + "mean_token_accuracy": 0.6723044514656067, + "num_tokens": 153998781.0, + "step": 5946 + }, + { + "epoch": 0.6530858774434439, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.9849138259887695, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.6911789178848267, + "num_tokens": 154019262.0, + "step": 5947 + }, + { + "epoch": 0.6531956951460576, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3803181648254395, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6949936151504517, + "num_tokens": 154047477.0, + "step": 5948 + }, + { + "epoch": 0.6533055128486712, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.565587043762207, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7169742584228516, + "num_tokens": 154070596.0, + "step": 5949 + }, + { + "epoch": 0.6534153305512849, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.077781915664673, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7112054824829102, + "num_tokens": 154088080.0, + "step": 5950 + }, + { + "epoch": 0.6535251482538985, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9824259281158447, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.715336799621582, + "num_tokens": 154111183.0, + "step": 5951 + }, + { + "epoch": 0.6536349659565122, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6140506267547607, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7007261514663696, + "num_tokens": 154133988.0, + "step": 5952 + }, + { + "epoch": 0.6537447836591258, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.3077025413513184, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7027338743209839, + "num_tokens": 154162949.0, + "step": 5953 + }, + { + "epoch": 0.6538546013617396, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4342145919799805, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7001721858978271, + "num_tokens": 154191471.0, + "step": 5954 + }, + { + "epoch": 0.6539644190643532, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.47251033782959, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7044682502746582, + "num_tokens": 154214934.0, + "step": 5955 + }, + { + "epoch": 0.6540742367669669, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4583585262298584, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.6981768608093262, + "num_tokens": 154242775.0, + "step": 5956 + }, + { + "epoch": 0.6541840544695805, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.440201759338379, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.6794642210006714, + "num_tokens": 154271659.0, + "step": 5957 + }, + { + "epoch": 0.6542938721721941, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9468955993652344, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.732006311416626, + "num_tokens": 154291248.0, + "step": 5958 + }, + { + "epoch": 0.6544036898748078, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.44796085357666, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6967861652374268, + "num_tokens": 154319496.0, + "step": 5959 + }, + { + "epoch": 0.6545135075774214, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3637523651123047, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.704032838344574, + "num_tokens": 154348439.0, + "step": 5960 + }, + { + "epoch": 0.6546233252800352, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.732590913772583, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7236173748970032, + "num_tokens": 154372375.0, + "step": 5961 + }, + { + "epoch": 0.6547331429826488, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4832921028137207, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.7146157026290894, + "num_tokens": 154397081.0, + "step": 5962 + }, + { + "epoch": 0.6548429606852625, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.562133312225342, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.6957549452781677, + "num_tokens": 154425629.0, + "step": 5963 + }, + { + "epoch": 0.6549527783878761, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7983508110046387, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7248682379722595, + "num_tokens": 154447369.0, + "step": 5964 + }, + { + "epoch": 0.6550625960904898, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8681623935699463, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6945558786392212, + "num_tokens": 154469352.0, + "step": 5965 + }, + { + "epoch": 0.6551724137931034, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4933254718780518, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6868694424629211, + "num_tokens": 154498264.0, + "step": 5966 + }, + { + "epoch": 0.6552822314957171, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9794199466705322, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7258157730102539, + "num_tokens": 154518816.0, + "step": 5967 + }, + { + "epoch": 0.6553920491983307, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3361082077026367, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.6782556772232056, + "num_tokens": 154553216.0, + "step": 5968 + }, + { + "epoch": 0.6555018669009445, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.748657703399658, + "learning_rate": 1e-06, + "loss": 0.9235, + "mean_token_accuracy": 0.7316854000091553, + "num_tokens": 154573981.0, + "step": 5969 + }, + { + "epoch": 0.6556116846035581, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3697831630706787, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6961902379989624, + "num_tokens": 154602589.0, + "step": 5970 + }, + { + "epoch": 0.6557215023061718, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.0838639736175537, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.6981411576271057, + "num_tokens": 154621052.0, + "step": 5971 + }, + { + "epoch": 0.6558313200087854, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.550229549407959, + "learning_rate": 1e-06, + "loss": 1.1025, + "mean_token_accuracy": 0.680852472782135, + "num_tokens": 154647699.0, + "step": 5972 + }, + { + "epoch": 0.6559411377113991, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.607491970062256, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7141092419624329, + "num_tokens": 154669577.0, + "step": 5973 + }, + { + "epoch": 0.6560509554140127, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5416173934936523, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7095046043395996, + "num_tokens": 154695662.0, + "step": 5974 + }, + { + "epoch": 0.6561607731166264, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.592231035232544, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.688629150390625, + "num_tokens": 154723545.0, + "step": 5975 + }, + { + "epoch": 0.6562705908192401, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.477665662765503, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6835935711860657, + "num_tokens": 154755534.0, + "step": 5976 + }, + { + "epoch": 0.6563804085218538, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.648282051086426, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6924970149993896, + "num_tokens": 154781348.0, + "step": 5977 + }, + { + "epoch": 0.6564902262244674, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5067718029022217, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.693735659122467, + "num_tokens": 154810297.0, + "step": 5978 + }, + { + "epoch": 0.656600043927081, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7837400436401367, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7076467275619507, + "num_tokens": 154831098.0, + "step": 5979 + }, + { + "epoch": 0.6567098616296947, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.6780154705047607, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6787474751472473, + "num_tokens": 154857877.0, + "step": 5980 + }, + { + "epoch": 0.6568196793323083, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.639126777648926, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7089143991470337, + "num_tokens": 154883350.0, + "step": 5981 + }, + { + "epoch": 0.656929497034922, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5740976333618164, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.689612627029419, + "num_tokens": 154909297.0, + "step": 5982 + }, + { + "epoch": 0.6570393147375357, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.3966684341430664, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7063912153244019, + "num_tokens": 154936292.0, + "step": 5983 + }, + { + "epoch": 0.6571491324401494, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5002944469451904, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6791666150093079, + "num_tokens": 154970211.0, + "step": 5984 + }, + { + "epoch": 0.657258950142763, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.8747718334198, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7132263779640198, + "num_tokens": 154989482.0, + "step": 5985 + }, + { + "epoch": 0.6573687678453767, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.461575508117676, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7010335326194763, + "num_tokens": 155017921.0, + "step": 5986 + }, + { + "epoch": 0.6574785855479903, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.605191946029663, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6901384592056274, + "num_tokens": 155042883.0, + "step": 5987 + }, + { + "epoch": 0.657588403250604, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.288789749145508, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6893423795700073, + "num_tokens": 155071816.0, + "step": 5988 + }, + { + "epoch": 0.6576982209532176, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6730144023895264, + "learning_rate": 1e-06, + "loss": 1.1074, + "mean_token_accuracy": 0.6911976933479309, + "num_tokens": 155098747.0, + "step": 5989 + }, + { + "epoch": 0.6578080386558314, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7938356399536133, + "learning_rate": 1e-06, + "loss": 0.9536, + "mean_token_accuracy": 0.7148005962371826, + "num_tokens": 155119942.0, + "step": 5990 + }, + { + "epoch": 0.657917856358445, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5741472244262695, + "learning_rate": 1e-06, + "loss": 1.1097, + "mean_token_accuracy": 0.678787112236023, + "num_tokens": 155146995.0, + "step": 5991 + }, + { + "epoch": 0.6580276740610587, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3922061920166016, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.687964916229248, + "num_tokens": 155179232.0, + "step": 5992 + }, + { + "epoch": 0.6581374917636723, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 3.2545673847198486, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.6958611011505127, + "num_tokens": 155202130.0, + "step": 5993 + }, + { + "epoch": 0.658247309466286, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.677335262298584, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6839299201965332, + "num_tokens": 155227196.0, + "step": 5994 + }, + { + "epoch": 0.6583571271688996, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6207146644592285, + "learning_rate": 1e-06, + "loss": 1.0979, + "mean_token_accuracy": 0.6827741265296936, + "num_tokens": 155254350.0, + "step": 5995 + }, + { + "epoch": 0.6584669448715132, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8488306999206543, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7040090560913086, + "num_tokens": 155274055.0, + "step": 5996 + }, + { + "epoch": 0.6585767625741269, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5424387454986572, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7011206150054932, + "num_tokens": 155299575.0, + "step": 5997 + }, + { + "epoch": 0.6586865802767407, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8962793350219727, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7020131349563599, + "num_tokens": 155322020.0, + "step": 5998 + }, + { + "epoch": 0.6587963979793543, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.404226303100586, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6889485120773315, + "num_tokens": 155351256.0, + "step": 5999 + }, + { + "epoch": 0.6589062156819679, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4740917682647705, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6841482520103455, + "num_tokens": 155380192.0, + "step": 6000 + }, + { + "epoch": 0.6590160333845816, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.602529287338257, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.6928566694259644, + "num_tokens": 155404549.0, + "step": 6001 + }, + { + "epoch": 0.6591258510871952, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5631048679351807, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6978557705879211, + "num_tokens": 155432136.0, + "step": 6002 + }, + { + "epoch": 0.6592356687898089, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.64255428314209, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6968135833740234, + "num_tokens": 155462102.0, + "step": 6003 + }, + { + "epoch": 0.6593454864924225, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.635747194290161, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.701322078704834, + "num_tokens": 155486045.0, + "step": 6004 + }, + { + "epoch": 0.6594553041950363, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.68882417678833, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.694159209728241, + "num_tokens": 155510272.0, + "step": 6005 + }, + { + "epoch": 0.6595651218976499, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4919660091400146, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7190021276473999, + "num_tokens": 155541765.0, + "step": 6006 + }, + { + "epoch": 0.6596749396002636, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.2688450813293457, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6997350454330444, + "num_tokens": 155574579.0, + "step": 6007 + }, + { + "epoch": 0.6597847573028772, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.31520414352417, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6886005401611328, + "num_tokens": 155603237.0, + "step": 6008 + }, + { + "epoch": 0.6598945750054909, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.885727882385254, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7040905952453613, + "num_tokens": 155623349.0, + "step": 6009 + }, + { + "epoch": 0.6600043927081045, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.459334373474121, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6814833879470825, + "num_tokens": 155651660.0, + "step": 6010 + }, + { + "epoch": 0.6601142104107182, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.606424570083618, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6989248394966125, + "num_tokens": 155676112.0, + "step": 6011 + }, + { + "epoch": 0.6602240281133319, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4170758724212646, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7169880867004395, + "num_tokens": 155706650.0, + "step": 6012 + }, + { + "epoch": 0.6603338458159456, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.197993040084839, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7014432549476624, + "num_tokens": 155742884.0, + "step": 6013 + }, + { + "epoch": 0.6604436635185592, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.4138453006744385, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7234743237495422, + "num_tokens": 155769598.0, + "step": 6014 + }, + { + "epoch": 0.6605534812211729, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.59769344329834, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6932458877563477, + "num_tokens": 155792342.0, + "step": 6015 + }, + { + "epoch": 0.6606632989237865, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.422661304473877, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7012933492660522, + "num_tokens": 155820917.0, + "step": 6016 + }, + { + "epoch": 0.6607731166264001, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5937440395355225, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7120017409324646, + "num_tokens": 155846192.0, + "step": 6017 + }, + { + "epoch": 0.6608829343290138, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9057512283325195, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6887589693069458, + "num_tokens": 155869072.0, + "step": 6018 + }, + { + "epoch": 0.6609927520316276, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8282182216644287, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6890585422515869, + "num_tokens": 155898142.0, + "step": 6019 + }, + { + "epoch": 0.6611025697342412, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6259467601776123, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7149059772491455, + "num_tokens": 155925699.0, + "step": 6020 + }, + { + "epoch": 0.6612123874368548, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6416139602661133, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6886815428733826, + "num_tokens": 155950313.0, + "step": 6021 + }, + { + "epoch": 0.6613222051394685, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.35117506980896, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7150714993476868, + "num_tokens": 155977689.0, + "step": 6022 + }, + { + "epoch": 0.6614320228420821, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6256704330444336, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.701145589351654, + "num_tokens": 156003254.0, + "step": 6023 + }, + { + "epoch": 0.6615418405446958, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6083483695983887, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7154480218887329, + "num_tokens": 156027004.0, + "step": 6024 + }, + { + "epoch": 0.6616516582473094, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.477923631668091, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6781488060951233, + "num_tokens": 156057555.0, + "step": 6025 + }, + { + "epoch": 0.6617614759499231, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5607664585113525, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6890385150909424, + "num_tokens": 156083938.0, + "step": 6026 + }, + { + "epoch": 0.6618712936525368, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.405477523803711, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7044464945793152, + "num_tokens": 156115291.0, + "step": 6027 + }, + { + "epoch": 0.6619811113551505, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4826090335845947, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7128199934959412, + "num_tokens": 156141710.0, + "step": 6028 + }, + { + "epoch": 0.6620909290577641, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.549757480621338, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6938705444335938, + "num_tokens": 156167447.0, + "step": 6029 + }, + { + "epoch": 0.6622007467603778, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.3505988121032715, + "learning_rate": 1e-06, + "loss": 1.1254, + "mean_token_accuracy": 0.6772543787956238, + "num_tokens": 156197415.0, + "step": 6030 + }, + { + "epoch": 0.6623105644629914, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.144912004470825, + "learning_rate": 1e-06, + "loss": 1.1361, + "mean_token_accuracy": 0.6751141548156738, + "num_tokens": 156235692.0, + "step": 6031 + }, + { + "epoch": 0.6624203821656051, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.5459389686584473, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7121649980545044, + "num_tokens": 156263420.0, + "step": 6032 + }, + { + "epoch": 0.6625301998682187, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.905822277069092, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7168028354644775, + "num_tokens": 156285855.0, + "step": 6033 + }, + { + "epoch": 0.6626400175708325, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.410135269165039, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.697909951210022, + "num_tokens": 156315182.0, + "step": 6034 + }, + { + "epoch": 0.6627498352734461, + "ewc_loss": 7.450580596923828e-06, + "grad_norm": 2.787322759628296, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7111219167709351, + "num_tokens": 156340739.0, + "step": 6035 + }, + { + "epoch": 0.6628596529760598, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.494992971420288, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6926438808441162, + "num_tokens": 156367369.0, + "step": 6036 + }, + { + "epoch": 0.6629694706786734, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6474666595458984, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6900045275688171, + "num_tokens": 156393412.0, + "step": 6037 + }, + { + "epoch": 0.663079288381287, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.563669204711914, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7064180374145508, + "num_tokens": 156419102.0, + "step": 6038 + }, + { + "epoch": 0.6631891060839007, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6458213329315186, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6901490092277527, + "num_tokens": 156446862.0, + "step": 6039 + }, + { + "epoch": 0.6632989237865143, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6098649501800537, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.702017068862915, + "num_tokens": 156471500.0, + "step": 6040 + }, + { + "epoch": 0.6634087414891281, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5063841342926025, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7039021253585815, + "num_tokens": 156497443.0, + "step": 6041 + }, + { + "epoch": 0.6635185591917417, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5094642639160156, + "learning_rate": 1e-06, + "loss": 1.0991, + "mean_token_accuracy": 0.6850967407226562, + "num_tokens": 156527999.0, + "step": 6042 + }, + { + "epoch": 0.6636283768943554, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5652198791503906, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6946719884872437, + "num_tokens": 156551836.0, + "step": 6043 + }, + { + "epoch": 0.663738194596969, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.47817325592041, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6820995211601257, + "num_tokens": 156580317.0, + "step": 6044 + }, + { + "epoch": 0.6638480122995827, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6834945678710938, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7095351815223694, + "num_tokens": 156603323.0, + "step": 6045 + }, + { + "epoch": 0.6639578300021963, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.434006452560425, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6993907690048218, + "num_tokens": 156630293.0, + "step": 6046 + }, + { + "epoch": 0.66406764770481, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6632909774780273, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6865668296813965, + "num_tokens": 156658155.0, + "step": 6047 + }, + { + "epoch": 0.6641774654074237, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.496814012527466, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6922581195831299, + "num_tokens": 156685519.0, + "step": 6048 + }, + { + "epoch": 0.6642872831100374, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4041967391967773, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7098252773284912, + "num_tokens": 156712257.0, + "step": 6049 + }, + { + "epoch": 0.664397100812651, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5566205978393555, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.6986252069473267, + "num_tokens": 156735594.0, + "step": 6050 + }, + { + "epoch": 0.6645069185152647, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6058900356292725, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.6948504447937012, + "num_tokens": 156760714.0, + "step": 6051 + }, + { + "epoch": 0.6646167362178783, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.444582939147949, + "learning_rate": 1e-06, + "loss": 1.1568, + "mean_token_accuracy": 0.6758575439453125, + "num_tokens": 156791882.0, + "step": 6052 + }, + { + "epoch": 0.664726553920492, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6305088996887207, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6928919553756714, + "num_tokens": 156816170.0, + "step": 6053 + }, + { + "epoch": 0.6648363716231056, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.640068292617798, + "learning_rate": 1e-06, + "loss": 0.9161, + "mean_token_accuracy": 0.7305984497070312, + "num_tokens": 156838617.0, + "step": 6054 + }, + { + "epoch": 0.6649461893257194, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5897529125213623, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.6831657886505127, + "num_tokens": 156870103.0, + "step": 6055 + }, + { + "epoch": 0.665056007028333, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5927376747131348, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.703129768371582, + "num_tokens": 156895180.0, + "step": 6056 + }, + { + "epoch": 0.6651658247309467, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6098411083221436, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.7023155689239502, + "num_tokens": 156919471.0, + "step": 6057 + }, + { + "epoch": 0.6652756424335603, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.51729154586792, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.7008754014968872, + "num_tokens": 156947861.0, + "step": 6058 + }, + { + "epoch": 0.665385460136174, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9648849964141846, + "learning_rate": 1e-06, + "loss": 0.9367, + "mean_token_accuracy": 0.717657208442688, + "num_tokens": 156966379.0, + "step": 6059 + }, + { + "epoch": 0.6654952778387876, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6202526092529297, + "learning_rate": 1e-06, + "loss": 0.8949, + "mean_token_accuracy": 0.7322259545326233, + "num_tokens": 156988824.0, + "step": 6060 + }, + { + "epoch": 0.6656050955414012, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.7952234745025635, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.6879552006721497, + "num_tokens": 157012835.0, + "step": 6061 + }, + { + "epoch": 0.6657149132440149, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.9851222038269043, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7090227603912354, + "num_tokens": 157041090.0, + "step": 6062 + }, + { + "epoch": 0.6658247309466286, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.4970786571502686, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6852424144744873, + "num_tokens": 157069696.0, + "step": 6063 + }, + { + "epoch": 0.6659345486492423, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.988308906555176, + "learning_rate": 1e-06, + "loss": 0.8986, + "mean_token_accuracy": 0.7325792908668518, + "num_tokens": 157088566.0, + "step": 6064 + }, + { + "epoch": 0.6660443663518559, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5845284461975098, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7035477161407471, + "num_tokens": 157114916.0, + "step": 6065 + }, + { + "epoch": 0.6661541840544696, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5042076110839844, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.6746421456336975, + "num_tokens": 157146212.0, + "step": 6066 + }, + { + "epoch": 0.6662640017570832, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.6870033740997314, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6975830793380737, + "num_tokens": 157168951.0, + "step": 6067 + }, + { + "epoch": 0.6663738194596969, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.882006883621216, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7054285407066345, + "num_tokens": 157189670.0, + "step": 6068 + }, + { + "epoch": 0.6664836371623105, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.9495129585266113, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7047311067581177, + "num_tokens": 157214174.0, + "step": 6069 + }, + { + "epoch": 0.6665934548649243, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.46637225151062, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6818211674690247, + "num_tokens": 157242797.0, + "step": 6070 + }, + { + "epoch": 0.6667032725675379, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.9374594688415527, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.6997114419937134, + "num_tokens": 157262190.0, + "step": 6071 + }, + { + "epoch": 0.6668130902701516, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.824903726577759, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7103086113929749, + "num_tokens": 157283822.0, + "step": 6072 + }, + { + "epoch": 0.6669229079727652, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.694683074951172, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6891494989395142, + "num_tokens": 157308337.0, + "step": 6073 + }, + { + "epoch": 0.6670327256753789, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 3.5597341060638428, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7067660689353943, + "num_tokens": 157332784.0, + "step": 6074 + }, + { + "epoch": 0.6671425433779925, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.515031337738037, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6843613386154175, + "num_tokens": 157360868.0, + "step": 6075 + }, + { + "epoch": 0.6672523610806061, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.8832526206970215, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6933996677398682, + "num_tokens": 157382569.0, + "step": 6076 + }, + { + "epoch": 0.6673621787832199, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5149078369140625, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6895426511764526, + "num_tokens": 157409460.0, + "step": 6077 + }, + { + "epoch": 0.6674719964858336, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.8646886348724365, + "learning_rate": 1e-06, + "loss": 1.1002, + "mean_token_accuracy": 0.6827903985977173, + "num_tokens": 157440052.0, + "step": 6078 + }, + { + "epoch": 0.6675818141884472, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.654982566833496, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6917877197265625, + "num_tokens": 157464038.0, + "step": 6079 + }, + { + "epoch": 0.6676916318910608, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.293125867843628, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6932203769683838, + "num_tokens": 157492927.0, + "step": 6080 + }, + { + "epoch": 0.6678014495936745, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.2967689037323, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.71065354347229, + "num_tokens": 157522518.0, + "step": 6081 + }, + { + "epoch": 0.6679112672962881, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5555365085601807, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7025174498558044, + "num_tokens": 157547170.0, + "step": 6082 + }, + { + "epoch": 0.6680210849989018, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.5905561447143555, + "learning_rate": 1e-06, + "loss": 1.1915, + "mean_token_accuracy": 0.6653757095336914, + "num_tokens": 157575044.0, + "step": 6083 + }, + { + "epoch": 0.6681309027015155, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.1402182579040527, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6830286383628845, + "num_tokens": 157614962.0, + "step": 6084 + }, + { + "epoch": 0.6682407204041292, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6493606567382812, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.699495792388916, + "num_tokens": 157638797.0, + "step": 6085 + }, + { + "epoch": 0.6683505381067428, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5957655906677246, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.6985197067260742, + "num_tokens": 157663629.0, + "step": 6086 + }, + { + "epoch": 0.6684603558093565, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.566561222076416, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7167824506759644, + "num_tokens": 157686851.0, + "step": 6087 + }, + { + "epoch": 0.6685701735119701, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9218244552612305, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7183602452278137, + "num_tokens": 157706324.0, + "step": 6088 + }, + { + "epoch": 0.6686799912145838, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.844794988632202, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6890461444854736, + "num_tokens": 157730442.0, + "step": 6089 + }, + { + "epoch": 0.6687898089171974, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.9783012866973877, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7161766290664673, + "num_tokens": 157753769.0, + "step": 6090 + }, + { + "epoch": 0.6688996266198111, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.441755533218384, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6920613050460815, + "num_tokens": 157783437.0, + "step": 6091 + }, + { + "epoch": 0.6690094443224248, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.363149642944336, + "learning_rate": 1e-06, + "loss": 1.1306, + "mean_token_accuracy": 0.6686211824417114, + "num_tokens": 157813602.0, + "step": 6092 + }, + { + "epoch": 0.6691192620250385, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.472576141357422, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7088863849639893, + "num_tokens": 157841665.0, + "step": 6093 + }, + { + "epoch": 0.6692290797276521, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 3.6292724609375, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7015372514724731, + "num_tokens": 157864011.0, + "step": 6094 + }, + { + "epoch": 0.6693388974302658, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5511422157287598, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7045606374740601, + "num_tokens": 157893302.0, + "step": 6095 + }, + { + "epoch": 0.6694487151328794, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.6342673301696777, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7190994024276733, + "num_tokens": 157917252.0, + "step": 6096 + }, + { + "epoch": 0.669558532835493, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.447431802749634, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.702312707901001, + "num_tokens": 157946229.0, + "step": 6097 + }, + { + "epoch": 0.6696683505381067, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5035488605499268, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6848074793815613, + "num_tokens": 157971749.0, + "step": 6098 + }, + { + "epoch": 0.6697781682407205, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.2462875843048096, + "learning_rate": 1e-06, + "loss": 1.1105, + "mean_token_accuracy": 0.6780905723571777, + "num_tokens": 158004797.0, + "step": 6099 + }, + { + "epoch": 0.6698879859433341, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.462169885635376, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6971134543418884, + "num_tokens": 158032561.0, + "step": 6100 + }, + { + "epoch": 0.6699978036459477, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.652594804763794, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7226077318191528, + "num_tokens": 158054892.0, + "step": 6101 + }, + { + "epoch": 0.6701076213485614, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5832529067993164, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6853009462356567, + "num_tokens": 158083531.0, + "step": 6102 + }, + { + "epoch": 0.670217439051175, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.889512777328491, + "learning_rate": 1e-06, + "loss": 1.0829, + "mean_token_accuracy": 0.6884881258010864, + "num_tokens": 158105142.0, + "step": 6103 + }, + { + "epoch": 0.6703272567537887, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.8863344192504883, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7002101540565491, + "num_tokens": 158124184.0, + "step": 6104 + }, + { + "epoch": 0.6704370744564023, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.819763422012329, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7148154973983765, + "num_tokens": 158144656.0, + "step": 6105 + }, + { + "epoch": 0.6705468921590161, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.8495194911956787, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7023640275001526, + "num_tokens": 158168309.0, + "step": 6106 + }, + { + "epoch": 0.6706567098616297, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.9170637130737305, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.714321494102478, + "num_tokens": 158189567.0, + "step": 6107 + }, + { + "epoch": 0.6707665275642434, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.765329122543335, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.7051364183425903, + "num_tokens": 158210476.0, + "step": 6108 + }, + { + "epoch": 0.670876345266857, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.8942408561706543, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7163715362548828, + "num_tokens": 158237155.0, + "step": 6109 + }, + { + "epoch": 0.6709861629694707, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.757746934890747, + "learning_rate": 1e-06, + "loss": 1.0997, + "mean_token_accuracy": 0.6847456693649292, + "num_tokens": 158260792.0, + "step": 6110 + }, + { + "epoch": 0.6710959806720843, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.7630741596221924, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7173569798469543, + "num_tokens": 158280874.0, + "step": 6111 + }, + { + "epoch": 0.671205798374698, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5707380771636963, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7161629796028137, + "num_tokens": 158304659.0, + "step": 6112 + }, + { + "epoch": 0.6713156160773117, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.6454317569732666, + "learning_rate": 1e-06, + "loss": 1.1151, + "mean_token_accuracy": 0.680384635925293, + "num_tokens": 158334472.0, + "step": 6113 + }, + { + "epoch": 0.6714254337799254, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.4457991123199463, + "learning_rate": 1e-06, + "loss": 1.0982, + "mean_token_accuracy": 0.6824002265930176, + "num_tokens": 158367217.0, + "step": 6114 + }, + { + "epoch": 0.671535251482539, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.4275455474853516, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6918065547943115, + "num_tokens": 158396304.0, + "step": 6115 + }, + { + "epoch": 0.6716450691851527, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.4391276836395264, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6905032992362976, + "num_tokens": 158422496.0, + "step": 6116 + }, + { + "epoch": 0.6717548868877663, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.6159114837646484, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7013372182846069, + "num_tokens": 158445506.0, + "step": 6117 + }, + { + "epoch": 0.67186470459038, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.2742345333099365, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7083653211593628, + "num_tokens": 158473721.0, + "step": 6118 + }, + { + "epoch": 0.6719745222929936, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.5990917682647705, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7101948857307434, + "num_tokens": 158496941.0, + "step": 6119 + }, + { + "epoch": 0.6720843399956072, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.3383121490478516, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7169791460037231, + "num_tokens": 158525505.0, + "step": 6120 + }, + { + "epoch": 0.672194157698221, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.7526772022247314, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.6993774175643921, + "num_tokens": 158548342.0, + "step": 6121 + }, + { + "epoch": 0.6723039754008346, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.4121029376983643, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6895281672477722, + "num_tokens": 158577693.0, + "step": 6122 + }, + { + "epoch": 0.6724137931034483, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5240566730499268, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7139604687690735, + "num_tokens": 158602480.0, + "step": 6123 + }, + { + "epoch": 0.6725236108060619, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.527290105819702, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.6998215913772583, + "num_tokens": 158631339.0, + "step": 6124 + }, + { + "epoch": 0.6726334285086756, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 3.0002548694610596, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7061530947685242, + "num_tokens": 158651126.0, + "step": 6125 + }, + { + "epoch": 0.6727432462112892, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.678393602371216, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7149732112884521, + "num_tokens": 158674262.0, + "step": 6126 + }, + { + "epoch": 0.6728530639139029, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.850041151046753, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6920607089996338, + "num_tokens": 158695067.0, + "step": 6127 + }, + { + "epoch": 0.6729628816165166, + "ewc_loss": 7.510185241699219e-06, + "grad_norm": 2.7807085514068604, + "learning_rate": 1e-06, + "loss": 0.9223, + "mean_token_accuracy": 0.7244727611541748, + "num_tokens": 158716763.0, + "step": 6128 + }, + { + "epoch": 0.6730726993191303, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5633792877197266, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.6997201442718506, + "num_tokens": 158743441.0, + "step": 6129 + }, + { + "epoch": 0.6731825170217439, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.6777455806732178, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7080936431884766, + "num_tokens": 158768551.0, + "step": 6130 + }, + { + "epoch": 0.6732923347243576, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 3.020214319229126, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6959105134010315, + "num_tokens": 158793558.0, + "step": 6131 + }, + { + "epoch": 0.6734021524269712, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.604484796524048, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7228840589523315, + "num_tokens": 158818430.0, + "step": 6132 + }, + { + "epoch": 0.6735119701295849, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.676696538925171, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6900272369384766, + "num_tokens": 158843138.0, + "step": 6133 + }, + { + "epoch": 0.6736217878321985, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4673521518707275, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.7014077305793762, + "num_tokens": 158874477.0, + "step": 6134 + }, + { + "epoch": 0.6737316055348123, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.6631062030792236, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7047350406646729, + "num_tokens": 158899004.0, + "step": 6135 + }, + { + "epoch": 0.6738414232374259, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.550041913986206, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.700703501701355, + "num_tokens": 158924939.0, + "step": 6136 + }, + { + "epoch": 0.6739512409400396, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.440098524093628, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7128545641899109, + "num_tokens": 158950178.0, + "step": 6137 + }, + { + "epoch": 0.6740610586426532, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.339184522628784, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.696963906288147, + "num_tokens": 158979755.0, + "step": 6138 + }, + { + "epoch": 0.6741708763452668, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.545222759246826, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7028466463088989, + "num_tokens": 159007944.0, + "step": 6139 + }, + { + "epoch": 0.6742806940478805, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.261594772338867, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7153460383415222, + "num_tokens": 159037619.0, + "step": 6140 + }, + { + "epoch": 0.6743905117504941, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.503976345062256, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7098232507705688, + "num_tokens": 159065207.0, + "step": 6141 + }, + { + "epoch": 0.6745003294531079, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4994161128997803, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6949739456176758, + "num_tokens": 159091637.0, + "step": 6142 + }, + { + "epoch": 0.6746101471557215, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.9015018939971924, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7052905559539795, + "num_tokens": 159114372.0, + "step": 6143 + }, + { + "epoch": 0.6747199648583352, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.7928121089935303, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6906744241714478, + "num_tokens": 159141680.0, + "step": 6144 + }, + { + "epoch": 0.6748297825609488, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.854198932647705, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7092558741569519, + "num_tokens": 159162816.0, + "step": 6145 + }, + { + "epoch": 0.6749396002635625, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.562110185623169, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6932448744773865, + "num_tokens": 159188403.0, + "step": 6146 + }, + { + "epoch": 0.6750494179661761, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.487473964691162, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.69352126121521, + "num_tokens": 159218711.0, + "step": 6147 + }, + { + "epoch": 0.6751592356687898, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.704918384552002, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6846135854721069, + "num_tokens": 159246877.0, + "step": 6148 + }, + { + "epoch": 0.6752690533714034, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.384141683578491, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7057924866676331, + "num_tokens": 159275042.0, + "step": 6149 + }, + { + "epoch": 0.6753788710740172, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.1913015842437744, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7102802991867065, + "num_tokens": 159307958.0, + "step": 6150 + }, + { + "epoch": 0.6754886887766308, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.7342069149017334, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.7016184329986572, + "num_tokens": 159330280.0, + "step": 6151 + }, + { + "epoch": 0.6755985064792445, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4517016410827637, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6980594992637634, + "num_tokens": 159358692.0, + "step": 6152 + }, + { + "epoch": 0.6757083241818581, + "ewc_loss": 7.4803829193115234e-06, + "grad_norm": 2.4105124473571777, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7033231854438782, + "num_tokens": 159387486.0, + "step": 6153 + }, + { + "epoch": 0.6758181418844718, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.6368162631988525, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7054911255836487, + "num_tokens": 159411417.0, + "step": 6154 + }, + { + "epoch": 0.6759279595870854, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.39001727104187, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.7069080471992493, + "num_tokens": 159442444.0, + "step": 6155 + }, + { + "epoch": 0.676037777289699, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.6385498046875, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6939024329185486, + "num_tokens": 159466545.0, + "step": 6156 + }, + { + "epoch": 0.6761475949923128, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 3.5052998065948486, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6940779685974121, + "num_tokens": 159491951.0, + "step": 6157 + }, + { + "epoch": 0.6762574126949265, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.3448739051818848, + "learning_rate": 1e-06, + "loss": 1.166, + "mean_token_accuracy": 0.6696889400482178, + "num_tokens": 159524107.0, + "step": 6158 + }, + { + "epoch": 0.6763672303975401, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.2529098987579346, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7085075378417969, + "num_tokens": 159560268.0, + "step": 6159 + }, + { + "epoch": 0.6764770481001537, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.633979082107544, + "learning_rate": 1e-06, + "loss": 0.8952, + "mean_token_accuracy": 0.7360267639160156, + "num_tokens": 159582247.0, + "step": 6160 + }, + { + "epoch": 0.6765868658027674, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5689847469329834, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7072186470031738, + "num_tokens": 159608502.0, + "step": 6161 + }, + { + "epoch": 0.676696683505381, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.691420793533325, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7171096801757812, + "num_tokens": 159629373.0, + "step": 6162 + }, + { + "epoch": 0.6768065012079947, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.57173752784729, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6875841617584229, + "num_tokens": 159654888.0, + "step": 6163 + }, + { + "epoch": 0.6769163189106084, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5380661487579346, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7095194458961487, + "num_tokens": 159678909.0, + "step": 6164 + }, + { + "epoch": 0.6770261366132221, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.9570260047912598, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7116398811340332, + "num_tokens": 159699296.0, + "step": 6165 + }, + { + "epoch": 0.6771359543158357, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5453670024871826, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6877713203430176, + "num_tokens": 159728051.0, + "step": 6166 + }, + { + "epoch": 0.6772457720184494, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.573296308517456, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7143442630767822, + "num_tokens": 159752609.0, + "step": 6167 + }, + { + "epoch": 0.677355589721063, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.7652225494384766, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6945700645446777, + "num_tokens": 159775222.0, + "step": 6168 + }, + { + "epoch": 0.6774654074236767, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.817499876022339, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6927205324172974, + "num_tokens": 159795594.0, + "step": 6169 + }, + { + "epoch": 0.6775752251262903, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.674826145172119, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6956924200057983, + "num_tokens": 159821063.0, + "step": 6170 + }, + { + "epoch": 0.6776850428289041, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.425856590270996, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7070186138153076, + "num_tokens": 159850148.0, + "step": 6171 + }, + { + "epoch": 0.6777948605315177, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.546541452407837, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7096377611160278, + "num_tokens": 159875384.0, + "step": 6172 + }, + { + "epoch": 0.6779046782341314, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.784359931945801, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7031220197677612, + "num_tokens": 159898946.0, + "step": 6173 + }, + { + "epoch": 0.678014495936745, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5676939487457275, + "learning_rate": 1e-06, + "loss": 1.1112, + "mean_token_accuracy": 0.6794080138206482, + "num_tokens": 159925065.0, + "step": 6174 + }, + { + "epoch": 0.6781243136393587, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5001001358032227, + "learning_rate": 1e-06, + "loss": 1.098, + "mean_token_accuracy": 0.6790861487388611, + "num_tokens": 159950968.0, + "step": 6175 + }, + { + "epoch": 0.6782341313419723, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.661695718765259, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6909853219985962, + "num_tokens": 159972968.0, + "step": 6176 + }, + { + "epoch": 0.678343949044586, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4545342922210693, + "learning_rate": 1e-06, + "loss": 0.9124, + "mean_token_accuracy": 0.736054539680481, + "num_tokens": 160000824.0, + "step": 6177 + }, + { + "epoch": 0.6784537667471996, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.570925712585449, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.6794611811637878, + "num_tokens": 160026215.0, + "step": 6178 + }, + { + "epoch": 0.6785635844498134, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5822949409484863, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.6988523006439209, + "num_tokens": 160048907.0, + "step": 6179 + }, + { + "epoch": 0.678673402152427, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.351419687271118, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.690194308757782, + "num_tokens": 160078943.0, + "step": 6180 + }, + { + "epoch": 0.6787832198550406, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4913628101348877, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7064046263694763, + "num_tokens": 160105582.0, + "step": 6181 + }, + { + "epoch": 0.6788930375576543, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5886878967285156, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7020512819290161, + "num_tokens": 160128859.0, + "step": 6182 + }, + { + "epoch": 0.6790028552602679, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.371859550476074, + "learning_rate": 1e-06, + "loss": 1.1636, + "mean_token_accuracy": 0.6660693883895874, + "num_tokens": 160161771.0, + "step": 6183 + }, + { + "epoch": 0.6791126729628816, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.8400535583496094, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7084372043609619, + "num_tokens": 160185315.0, + "step": 6184 + }, + { + "epoch": 0.6792224906654952, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5447134971618652, + "learning_rate": 1e-06, + "loss": 0.921, + "mean_token_accuracy": 0.7297760248184204, + "num_tokens": 160210093.0, + "step": 6185 + }, + { + "epoch": 0.679332308368109, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.629056692123413, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6945002675056458, + "num_tokens": 160235947.0, + "step": 6186 + }, + { + "epoch": 0.6794421260707226, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.508749485015869, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.7100486159324646, + "num_tokens": 160262839.0, + "step": 6187 + }, + { + "epoch": 0.6795519437733363, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.465527057647705, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.69886314868927, + "num_tokens": 160291066.0, + "step": 6188 + }, + { + "epoch": 0.6796617614759499, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5799994468688965, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6940882802009583, + "num_tokens": 160316101.0, + "step": 6189 + }, + { + "epoch": 0.6797715791785636, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4475762844085693, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7174713611602783, + "num_tokens": 160342531.0, + "step": 6190 + }, + { + "epoch": 0.6798813968811772, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.908872365951538, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7042909264564514, + "num_tokens": 160361907.0, + "step": 6191 + }, + { + "epoch": 0.6799912145837909, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.598755121231079, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.7225027084350586, + "num_tokens": 160385012.0, + "step": 6192 + }, + { + "epoch": 0.6801010322864046, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.2491843700408936, + "learning_rate": 1e-06, + "loss": 1.1361, + "mean_token_accuracy": 0.6727805137634277, + "num_tokens": 160418417.0, + "step": 6193 + }, + { + "epoch": 0.6802108499890183, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.9322543144226074, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7159837484359741, + "num_tokens": 160436585.0, + "step": 6194 + }, + { + "epoch": 0.6803206676916319, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.5265610218048096, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6924337148666382, + "num_tokens": 160465543.0, + "step": 6195 + }, + { + "epoch": 0.6804304853942456, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6258223056793213, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6859431266784668, + "num_tokens": 160494434.0, + "step": 6196 + }, + { + "epoch": 0.6805403030968592, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.328601837158203, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6949814558029175, + "num_tokens": 160522554.0, + "step": 6197 + }, + { + "epoch": 0.6806501207994728, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.731121063232422, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.6986414194107056, + "num_tokens": 160545157.0, + "step": 6198 + }, + { + "epoch": 0.6807599385020865, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.316298484802246, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.689989447593689, + "num_tokens": 160573827.0, + "step": 6199 + }, + { + "epoch": 0.6808697562047002, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 3.805647850036621, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7180356979370117, + "num_tokens": 160592697.0, + "step": 6200 + }, + { + "epoch": 0.6809795739073139, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4693751335144043, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7023336887359619, + "num_tokens": 160621152.0, + "step": 6201 + }, + { + "epoch": 0.6810893916099275, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4320008754730225, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7030057907104492, + "num_tokens": 160649670.0, + "step": 6202 + }, + { + "epoch": 0.6811992093125412, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.680867910385132, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.6997703313827515, + "num_tokens": 160672004.0, + "step": 6203 + }, + { + "epoch": 0.6813090270151548, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.4936838150024414, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6890390515327454, + "num_tokens": 160701671.0, + "step": 6204 + }, + { + "epoch": 0.6814188447177685, + "ewc_loss": 7.539987564086914e-06, + "grad_norm": 2.7563536167144775, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.6875827312469482, + "num_tokens": 160724240.0, + "step": 6205 + }, + { + "epoch": 0.6815286624203821, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.512115001678467, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6909812688827515, + "num_tokens": 160750043.0, + "step": 6206 + }, + { + "epoch": 0.6816384801229959, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6503148078918457, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6940402984619141, + "num_tokens": 160775094.0, + "step": 6207 + }, + { + "epoch": 0.6817482978256095, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.29780912399292, + "learning_rate": 1e-06, + "loss": 0.901, + "mean_token_accuracy": 0.729265570640564, + "num_tokens": 160804642.0, + "step": 6208 + }, + { + "epoch": 0.6818581155282232, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.61678409576416, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7099630832672119, + "num_tokens": 160828357.0, + "step": 6209 + }, + { + "epoch": 0.6819679332308368, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.689969062805176, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7228236198425293, + "num_tokens": 160852946.0, + "step": 6210 + }, + { + "epoch": 0.6820777509334505, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6408352851867676, + "learning_rate": 1e-06, + "loss": 1.1051, + "mean_token_accuracy": 0.6821739077568054, + "num_tokens": 160880009.0, + "step": 6211 + }, + { + "epoch": 0.6821875686360641, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.3775222301483154, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7036309838294983, + "num_tokens": 160910735.0, + "step": 6212 + }, + { + "epoch": 0.6822973863386778, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5551607608795166, + "learning_rate": 1e-06, + "loss": 1.1323, + "mean_token_accuracy": 0.6752719879150391, + "num_tokens": 160940187.0, + "step": 6213 + }, + { + "epoch": 0.6824072040412914, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.471571207046509, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7126057147979736, + "num_tokens": 160967733.0, + "step": 6214 + }, + { + "epoch": 0.6825170217439052, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6904966831207275, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7089540362358093, + "num_tokens": 160991485.0, + "step": 6215 + }, + { + "epoch": 0.6826268394465188, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.468518018722534, + "learning_rate": 1e-06, + "loss": 1.1069, + "mean_token_accuracy": 0.6912999749183655, + "num_tokens": 161022484.0, + "step": 6216 + }, + { + "epoch": 0.6827366571491325, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.9930341243743896, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7058621644973755, + "num_tokens": 161041677.0, + "step": 6217 + }, + { + "epoch": 0.6828464748517461, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.2733957767486572, + "learning_rate": 1e-06, + "loss": 1.1548, + "mean_token_accuracy": 0.6765361428260803, + "num_tokens": 161078083.0, + "step": 6218 + }, + { + "epoch": 0.6829562925543597, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.477999687194824, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7046212553977966, + "num_tokens": 161105164.0, + "step": 6219 + }, + { + "epoch": 0.6830661102569734, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.474560260772705, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.695572555065155, + "num_tokens": 161134797.0, + "step": 6220 + }, + { + "epoch": 0.683175927959587, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.341296672821045, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7069929838180542, + "num_tokens": 161163605.0, + "step": 6221 + }, + { + "epoch": 0.6832857456622008, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.475634813308716, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7034438252449036, + "num_tokens": 161189308.0, + "step": 6222 + }, + { + "epoch": 0.6833955633648144, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.8395867347717285, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7022390365600586, + "num_tokens": 161212337.0, + "step": 6223 + }, + { + "epoch": 0.6835053810674281, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 3.035898208618164, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7045674324035645, + "num_tokens": 161234858.0, + "step": 6224 + }, + { + "epoch": 0.6836151987700417, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5085294246673584, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7051804661750793, + "num_tokens": 161261948.0, + "step": 6225 + }, + { + "epoch": 0.6837250164726554, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7216837406158447, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7028903961181641, + "num_tokens": 161285655.0, + "step": 6226 + }, + { + "epoch": 0.683834834175269, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.621257781982422, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7029978036880493, + "num_tokens": 161313042.0, + "step": 6227 + }, + { + "epoch": 0.6839446518778827, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6577508449554443, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6943949460983276, + "num_tokens": 161339421.0, + "step": 6228 + }, + { + "epoch": 0.6840544695804964, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.38373064994812, + "learning_rate": 1e-06, + "loss": 1.1212, + "mean_token_accuracy": 0.6823647022247314, + "num_tokens": 161374361.0, + "step": 6229 + }, + { + "epoch": 0.6841642872831101, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.577071189880371, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7014370560646057, + "num_tokens": 161400878.0, + "step": 6230 + }, + { + "epoch": 0.6842741049857237, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5193841457366943, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.7023663520812988, + "num_tokens": 161428464.0, + "step": 6231 + }, + { + "epoch": 0.6843839226883374, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7623980045318604, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7267887592315674, + "num_tokens": 161450688.0, + "step": 6232 + }, + { + "epoch": 0.684493740390951, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.625483751296997, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.700685977935791, + "num_tokens": 161472646.0, + "step": 6233 + }, + { + "epoch": 0.6846035580935647, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6646430492401123, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6930378675460815, + "num_tokens": 161495999.0, + "step": 6234 + }, + { + "epoch": 0.6847133757961783, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5727508068084717, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7153620719909668, + "num_tokens": 161518854.0, + "step": 6235 + }, + { + "epoch": 0.6848231934987921, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7508299350738525, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6961121559143066, + "num_tokens": 161541927.0, + "step": 6236 + }, + { + "epoch": 0.6849330112014057, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7245676517486572, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7024345397949219, + "num_tokens": 161568560.0, + "step": 6237 + }, + { + "epoch": 0.6850428289040194, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.83941388130188, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7019089460372925, + "num_tokens": 161592384.0, + "step": 6238 + }, + { + "epoch": 0.685152646606633, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.526550769805908, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.685204029083252, + "num_tokens": 161620897.0, + "step": 6239 + }, + { + "epoch": 0.6852624643092466, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7157156467437744, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7137895226478577, + "num_tokens": 161644189.0, + "step": 6240 + }, + { + "epoch": 0.6853722820118603, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.601440906524658, + "learning_rate": 1e-06, + "loss": 1.0766, + "mean_token_accuracy": 0.6866776943206787, + "num_tokens": 161668350.0, + "step": 6241 + }, + { + "epoch": 0.6854820997144739, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.8238906860351562, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7167810201644897, + "num_tokens": 161691163.0, + "step": 6242 + }, + { + "epoch": 0.6855919174170876, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.650179147720337, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7161833047866821, + "num_tokens": 161715801.0, + "step": 6243 + }, + { + "epoch": 0.6857017351197013, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6320176124572754, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.6878851652145386, + "num_tokens": 161743236.0, + "step": 6244 + }, + { + "epoch": 0.685811552822315, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.504701614379883, + "learning_rate": 1e-06, + "loss": 1.1504, + "mean_token_accuracy": 0.6715276837348938, + "num_tokens": 161772576.0, + "step": 6245 + }, + { + "epoch": 0.6859213705249286, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5495176315307617, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7022929191589355, + "num_tokens": 161798925.0, + "step": 6246 + }, + { + "epoch": 0.6860311882275423, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 3.067694902420044, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7056258916854858, + "num_tokens": 161820941.0, + "step": 6247 + }, + { + "epoch": 0.6861410059301559, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7025437355041504, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.692516565322876, + "num_tokens": 161846081.0, + "step": 6248 + }, + { + "epoch": 0.6862508236327696, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.412012815475464, + "learning_rate": 1e-06, + "loss": 1.1632, + "mean_token_accuracy": 0.6658296585083008, + "num_tokens": 161875326.0, + "step": 6249 + }, + { + "epoch": 0.6863606413353832, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4875712394714355, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7026233077049255, + "num_tokens": 161902576.0, + "step": 6250 + }, + { + "epoch": 0.686470459037997, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 3.0555191040039062, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7128167748451233, + "num_tokens": 161920422.0, + "step": 6251 + }, + { + "epoch": 0.6865802767406106, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7111988067626953, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7133060097694397, + "num_tokens": 161943729.0, + "step": 6252 + }, + { + "epoch": 0.6866900944432243, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6361606121063232, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6890735030174255, + "num_tokens": 161970782.0, + "step": 6253 + }, + { + "epoch": 0.6867999121458379, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 4.9369659423828125, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6856111288070679, + "num_tokens": 162007033.0, + "step": 6254 + }, + { + "epoch": 0.6869097298484516, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7670178413391113, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7090544700622559, + "num_tokens": 162027881.0, + "step": 6255 + }, + { + "epoch": 0.6870195475510652, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4613447189331055, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6973199844360352, + "num_tokens": 162058289.0, + "step": 6256 + }, + { + "epoch": 0.6871293652536788, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7415454387664795, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.72015380859375, + "num_tokens": 162079047.0, + "step": 6257 + }, + { + "epoch": 0.6872391829562926, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6423933506011963, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6999585628509521, + "num_tokens": 162100921.0, + "step": 6258 + }, + { + "epoch": 0.6873490006589063, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.669915199279785, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6964900493621826, + "num_tokens": 162125491.0, + "step": 6259 + }, + { + "epoch": 0.6874588183615199, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6487069129943848, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6904768347740173, + "num_tokens": 162150785.0, + "step": 6260 + }, + { + "epoch": 0.6875686360641335, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7212295532226562, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6892670392990112, + "num_tokens": 162173188.0, + "step": 6261 + }, + { + "epoch": 0.6876784537667472, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.375744104385376, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7190825939178467, + "num_tokens": 162203116.0, + "step": 6262 + }, + { + "epoch": 0.6877882714693608, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4075987339019775, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6936256885528564, + "num_tokens": 162230643.0, + "step": 6263 + }, + { + "epoch": 0.6878980891719745, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4462838172912598, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6888539791107178, + "num_tokens": 162259943.0, + "step": 6264 + }, + { + "epoch": 0.6880079068745882, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5555450916290283, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7038624882698059, + "num_tokens": 162284553.0, + "step": 6265 + }, + { + "epoch": 0.6881177245772019, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 3.28193998336792, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6914398074150085, + "num_tokens": 162309858.0, + "step": 6266 + }, + { + "epoch": 0.6882275422798155, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4869425296783447, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6962122917175293, + "num_tokens": 162337622.0, + "step": 6267 + }, + { + "epoch": 0.6883373599824292, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4116125106811523, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6932865977287292, + "num_tokens": 162366547.0, + "step": 6268 + }, + { + "epoch": 0.6884471776850428, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 3.405700206756592, + "learning_rate": 1e-06, + "loss": 1.1091, + "mean_token_accuracy": 0.6762407422065735, + "num_tokens": 162396141.0, + "step": 6269 + }, + { + "epoch": 0.6885569953876565, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6097235679626465, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.712895393371582, + "num_tokens": 162420169.0, + "step": 6270 + }, + { + "epoch": 0.6886668130902701, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.55985951423645, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6883562207221985, + "num_tokens": 162445872.0, + "step": 6271 + }, + { + "epoch": 0.6887766307928838, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5465927124023438, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.6998677253723145, + "num_tokens": 162473104.0, + "step": 6272 + }, + { + "epoch": 0.6888864484954975, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 3.1412370204925537, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7175365090370178, + "num_tokens": 162498431.0, + "step": 6273 + }, + { + "epoch": 0.6889962661981112, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.3620169162750244, + "learning_rate": 1e-06, + "loss": 1.1295, + "mean_token_accuracy": 0.6747327446937561, + "num_tokens": 162528782.0, + "step": 6274 + }, + { + "epoch": 0.6891060839007248, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5232768058776855, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6985915899276733, + "num_tokens": 162553110.0, + "step": 6275 + }, + { + "epoch": 0.6892159016033385, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.3911824226379395, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6853857040405273, + "num_tokens": 162581838.0, + "step": 6276 + }, + { + "epoch": 0.6893257193059521, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.679307460784912, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.6851880550384521, + "num_tokens": 162604965.0, + "step": 6277 + }, + { + "epoch": 0.6894355370085657, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.425447463989258, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7063093185424805, + "num_tokens": 162634386.0, + "step": 6278 + }, + { + "epoch": 0.6895453547111794, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.555208444595337, + "learning_rate": 1e-06, + "loss": 1.1318, + "mean_token_accuracy": 0.6789690852165222, + "num_tokens": 162660520.0, + "step": 6279 + }, + { + "epoch": 0.6896551724137931, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.8697545528411865, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7141062021255493, + "num_tokens": 162683482.0, + "step": 6280 + }, + { + "epoch": 0.6897649901164068, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4635956287384033, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7029509544372559, + "num_tokens": 162712341.0, + "step": 6281 + }, + { + "epoch": 0.6898748078190204, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.564331293106079, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7159183025360107, + "num_tokens": 162735854.0, + "step": 6282 + }, + { + "epoch": 0.6899846255216341, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.3704373836517334, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6984544396400452, + "num_tokens": 162767452.0, + "step": 6283 + }, + { + "epoch": 0.6900944432242477, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.798393487930298, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7175314426422119, + "num_tokens": 162788127.0, + "step": 6284 + }, + { + "epoch": 0.6902042609268614, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5168678760528564, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7025003433227539, + "num_tokens": 162817399.0, + "step": 6285 + }, + { + "epoch": 0.690314078629475, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7081711292266846, + "learning_rate": 1e-06, + "loss": 1.0983, + "mean_token_accuracy": 0.6802403926849365, + "num_tokens": 162842073.0, + "step": 6286 + }, + { + "epoch": 0.6904238963320888, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.504747152328491, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7265141010284424, + "num_tokens": 162865066.0, + "step": 6287 + }, + { + "epoch": 0.6905337140347024, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.589498519897461, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7069109678268433, + "num_tokens": 162889656.0, + "step": 6288 + }, + { + "epoch": 0.6906435317373161, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.6133310794830322, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7067039012908936, + "num_tokens": 162913759.0, + "step": 6289 + }, + { + "epoch": 0.6907533494399297, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5015721321105957, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.708208441734314, + "num_tokens": 162939871.0, + "step": 6290 + }, + { + "epoch": 0.6908631671425434, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4423210620880127, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.7009929418563843, + "num_tokens": 162968388.0, + "step": 6291 + }, + { + "epoch": 0.690972984845157, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.719144105911255, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7209724187850952, + "num_tokens": 162991016.0, + "step": 6292 + }, + { + "epoch": 0.6910828025477707, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5237386226654053, + "learning_rate": 1e-06, + "loss": 0.9044, + "mean_token_accuracy": 0.7311409115791321, + "num_tokens": 163014896.0, + "step": 6293 + }, + { + "epoch": 0.6911926202503844, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.8220467567443848, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.716538667678833, + "num_tokens": 163035705.0, + "step": 6294 + }, + { + "epoch": 0.6913024379529981, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.451662540435791, + "learning_rate": 1e-06, + "loss": 1.1097, + "mean_token_accuracy": 0.6759417653083801, + "num_tokens": 163064708.0, + "step": 6295 + }, + { + "epoch": 0.6914122556556117, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.461315393447876, + "learning_rate": 1e-06, + "loss": 1.113, + "mean_token_accuracy": 0.6736928820610046, + "num_tokens": 163092567.0, + "step": 6296 + }, + { + "epoch": 0.6915220733582254, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.6525936126708984, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7111183404922485, + "num_tokens": 163117372.0, + "step": 6297 + }, + { + "epoch": 0.691631891060839, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7686238288879395, + "learning_rate": 1e-06, + "loss": 1.1417, + "mean_token_accuracy": 0.6707282066345215, + "num_tokens": 163142418.0, + "step": 6298 + }, + { + "epoch": 0.6917417087634526, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.422734022140503, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6895609498023987, + "num_tokens": 163172080.0, + "step": 6299 + }, + { + "epoch": 0.6918515264660663, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5713086128234863, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6935724020004272, + "num_tokens": 163197984.0, + "step": 6300 + }, + { + "epoch": 0.6919613441686799, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.438002824783325, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.689206600189209, + "num_tokens": 163226464.0, + "step": 6301 + }, + { + "epoch": 0.6920711618712937, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5741889476776123, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7061493396759033, + "num_tokens": 163256158.0, + "step": 6302 + }, + { + "epoch": 0.6921809795739073, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7172129154205322, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6923556327819824, + "num_tokens": 163277834.0, + "step": 6303 + }, + { + "epoch": 0.692290797276521, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7274508476257324, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7092483043670654, + "num_tokens": 163299028.0, + "step": 6304 + }, + { + "epoch": 0.6924006149791346, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 3.212240695953369, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6851742267608643, + "num_tokens": 163329095.0, + "step": 6305 + }, + { + "epoch": 0.6925104326817483, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.494439125061035, + "learning_rate": 1e-06, + "loss": 1.1226, + "mean_token_accuracy": 0.6749569177627563, + "num_tokens": 163357278.0, + "step": 6306 + }, + { + "epoch": 0.6926202503843619, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5369725227355957, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7055891752243042, + "num_tokens": 163381779.0, + "step": 6307 + }, + { + "epoch": 0.6927300680869756, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.2350964546203613, + "learning_rate": 1e-06, + "loss": 1.1484, + "mean_token_accuracy": 0.6709157824516296, + "num_tokens": 163415146.0, + "step": 6308 + }, + { + "epoch": 0.6928398857895893, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.4657199382781982, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6896668076515198, + "num_tokens": 163440947.0, + "step": 6309 + }, + { + "epoch": 0.692949703492203, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.735027551651001, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7121176719665527, + "num_tokens": 163460802.0, + "step": 6310 + }, + { + "epoch": 0.6930595211948166, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.2038137912750244, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.6970328688621521, + "num_tokens": 163496352.0, + "step": 6311 + }, + { + "epoch": 0.6931693388974303, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.3158771991729736, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.7033029198646545, + "num_tokens": 163526044.0, + "step": 6312 + }, + { + "epoch": 0.6932791566000439, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7940125465393066, + "learning_rate": 1e-06, + "loss": 1.0932, + "mean_token_accuracy": 0.6859791874885559, + "num_tokens": 163550244.0, + "step": 6313 + }, + { + "epoch": 0.6933889743026576, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6218326091766357, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6860120296478271, + "num_tokens": 163576341.0, + "step": 6314 + }, + { + "epoch": 0.6934987920052712, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.526095151901245, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6984114646911621, + "num_tokens": 163604671.0, + "step": 6315 + }, + { + "epoch": 0.693608609707885, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.646815776824951, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6927322745323181, + "num_tokens": 163629666.0, + "step": 6316 + }, + { + "epoch": 0.6937184274104986, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.3590149879455566, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7064108848571777, + "num_tokens": 163658034.0, + "step": 6317 + }, + { + "epoch": 0.6938282451131123, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5147340297698975, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7079504132270813, + "num_tokens": 163683740.0, + "step": 6318 + }, + { + "epoch": 0.6939380628157259, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5813419818878174, + "learning_rate": 1e-06, + "loss": 0.9178, + "mean_token_accuracy": 0.7205517292022705, + "num_tokens": 163706240.0, + "step": 6319 + }, + { + "epoch": 0.6940478805183395, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7513089179992676, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7153328657150269, + "num_tokens": 163726391.0, + "step": 6320 + }, + { + "epoch": 0.6941576982209532, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5876917839050293, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7104092836380005, + "num_tokens": 163754046.0, + "step": 6321 + }, + { + "epoch": 0.6942675159235668, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.6173527240753174, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7003044486045837, + "num_tokens": 163779277.0, + "step": 6322 + }, + { + "epoch": 0.6943773336261806, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.397050619125366, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6830346584320068, + "num_tokens": 163811487.0, + "step": 6323 + }, + { + "epoch": 0.6944871513287942, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.5060174465179443, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6924504637718201, + "num_tokens": 163837897.0, + "step": 6324 + }, + { + "epoch": 0.6945969690314079, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.364973783493042, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6884561777114868, + "num_tokens": 163866188.0, + "step": 6325 + }, + { + "epoch": 0.6947067867340215, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4129526615142822, + "learning_rate": 1e-06, + "loss": 1.1527, + "mean_token_accuracy": 0.6759060621261597, + "num_tokens": 163897427.0, + "step": 6326 + }, + { + "epoch": 0.6948166044366352, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.494623899459839, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.714962363243103, + "num_tokens": 163926436.0, + "step": 6327 + }, + { + "epoch": 0.6949264221392488, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4361867904663086, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.6836032867431641, + "num_tokens": 163956225.0, + "step": 6328 + }, + { + "epoch": 0.6950362398418625, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.7937252521514893, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.7031064033508301, + "num_tokens": 163982151.0, + "step": 6329 + }, + { + "epoch": 0.6951460575444761, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.601181983947754, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.696675181388855, + "num_tokens": 164007634.0, + "step": 6330 + }, + { + "epoch": 0.6952558752470899, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4579787254333496, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7087919116020203, + "num_tokens": 164035131.0, + "step": 6331 + }, + { + "epoch": 0.6953656929497035, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.4041078090667725, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.6963455080986023, + "num_tokens": 164063187.0, + "step": 6332 + }, + { + "epoch": 0.6954755106523172, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.74459171295166, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.716166079044342, + "num_tokens": 164085148.0, + "step": 6333 + }, + { + "epoch": 0.6955853283549308, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.783405303955078, + "learning_rate": 1e-06, + "loss": 1.0831, + "mean_token_accuracy": 0.6853082776069641, + "num_tokens": 164108696.0, + "step": 6334 + }, + { + "epoch": 0.6956951460575445, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7053120136260986, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7119302153587341, + "num_tokens": 164132496.0, + "step": 6335 + }, + { + "epoch": 0.6958049637601581, + "ewc_loss": 7.569789886474609e-06, + "grad_norm": 2.590313673019409, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7025636434555054, + "num_tokens": 164158020.0, + "step": 6336 + }, + { + "epoch": 0.6959147814627717, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7297275066375732, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7138648629188538, + "num_tokens": 164179998.0, + "step": 6337 + }, + { + "epoch": 0.6960245991653855, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 3.0476856231689453, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.7245516777038574, + "num_tokens": 164197247.0, + "step": 6338 + }, + { + "epoch": 0.6961344168679992, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5849390029907227, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7039794325828552, + "num_tokens": 164222013.0, + "step": 6339 + }, + { + "epoch": 0.6962442345706128, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7051846981048584, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7165619134902954, + "num_tokens": 164244311.0, + "step": 6340 + }, + { + "epoch": 0.6963540522732264, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5573055744171143, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.6989046335220337, + "num_tokens": 164272798.0, + "step": 6341 + }, + { + "epoch": 0.6964638699758401, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5578343868255615, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7011536359786987, + "num_tokens": 164298367.0, + "step": 6342 + }, + { + "epoch": 0.6965736876784537, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.4035325050354004, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6801594495773315, + "num_tokens": 164328980.0, + "step": 6343 + }, + { + "epoch": 0.6966835053810674, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.6130385398864746, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6988253593444824, + "num_tokens": 164354971.0, + "step": 6344 + }, + { + "epoch": 0.6967933230836811, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.450636386871338, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7049901485443115, + "num_tokens": 164382484.0, + "step": 6345 + }, + { + "epoch": 0.6969031407862948, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7737979888916016, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.7085217237472534, + "num_tokens": 164406329.0, + "step": 6346 + }, + { + "epoch": 0.6970129584889084, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.2869274616241455, + "learning_rate": 1e-06, + "loss": 1.1593, + "mean_token_accuracy": 0.6729159951210022, + "num_tokens": 164438435.0, + "step": 6347 + }, + { + "epoch": 0.6971227761915221, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.768632650375366, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6960358619689941, + "num_tokens": 164458132.0, + "step": 6348 + }, + { + "epoch": 0.6972325938941357, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.584317922592163, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7144149541854858, + "num_tokens": 164483148.0, + "step": 6349 + }, + { + "epoch": 0.6973424115967494, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5459461212158203, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7126675844192505, + "num_tokens": 164507414.0, + "step": 6350 + }, + { + "epoch": 0.697452229299363, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.459116220474243, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7017894983291626, + "num_tokens": 164535855.0, + "step": 6351 + }, + { + "epoch": 0.6975620470019768, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.7356135845184326, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7094744443893433, + "num_tokens": 164558183.0, + "step": 6352 + }, + { + "epoch": 0.6976718647045904, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.5139951705932617, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7039502263069153, + "num_tokens": 164584290.0, + "step": 6353 + }, + { + "epoch": 0.6977816824072041, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 2.984896421432495, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7105929851531982, + "num_tokens": 164601322.0, + "step": 6354 + }, + { + "epoch": 0.6978915001098177, + "ewc_loss": 7.599592208862305e-06, + "grad_norm": 3.104076385498047, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7274397611618042, + "num_tokens": 164618813.0, + "step": 6355 + }, + { + "epoch": 0.6980013178124314, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7828245162963867, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7108576893806458, + "num_tokens": 164642036.0, + "step": 6356 + }, + { + "epoch": 0.698111135515045, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4575588703155518, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.7054443359375, + "num_tokens": 164674367.0, + "step": 6357 + }, + { + "epoch": 0.6982209532176586, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.8714759349823, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.708580493927002, + "num_tokens": 164696694.0, + "step": 6358 + }, + { + "epoch": 0.6983307709202724, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5860846042633057, + "learning_rate": 1e-06, + "loss": 0.9166, + "mean_token_accuracy": 0.7257567048072815, + "num_tokens": 164722792.0, + "step": 6359 + }, + { + "epoch": 0.698440588622886, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.59920597076416, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7072756290435791, + "num_tokens": 164748014.0, + "step": 6360 + }, + { + "epoch": 0.6985504063254997, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.617158889770508, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6912968754768372, + "num_tokens": 164772019.0, + "step": 6361 + }, + { + "epoch": 0.6986602240281133, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.407912015914917, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6941050291061401, + "num_tokens": 164800535.0, + "step": 6362 + }, + { + "epoch": 0.698770041730727, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.590141773223877, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6794794797897339, + "num_tokens": 164827307.0, + "step": 6363 + }, + { + "epoch": 0.6988798594333406, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 3.006077527999878, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7141172289848328, + "num_tokens": 164850156.0, + "step": 6364 + }, + { + "epoch": 0.6989896771359543, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.689811944961548, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.7071715593338013, + "num_tokens": 164876107.0, + "step": 6365 + }, + { + "epoch": 0.6990994948385679, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.396002769470215, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.697234034538269, + "num_tokens": 164905536.0, + "step": 6366 + }, + { + "epoch": 0.6992093125411817, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4553089141845703, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7142372131347656, + "num_tokens": 164930711.0, + "step": 6367 + }, + { + "epoch": 0.6993191302437953, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5050320625305176, + "learning_rate": 1e-06, + "loss": 1.1313, + "mean_token_accuracy": 0.6727959513664246, + "num_tokens": 164965215.0, + "step": 6368 + }, + { + "epoch": 0.699428947946409, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7231650352478027, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6976857781410217, + "num_tokens": 164987382.0, + "step": 6369 + }, + { + "epoch": 0.6995387656490226, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.503347158432007, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6823420524597168, + "num_tokens": 165016576.0, + "step": 6370 + }, + { + "epoch": 0.6996485833516363, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.76733660697937, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6903098821640015, + "num_tokens": 165043318.0, + "step": 6371 + }, + { + "epoch": 0.6997584010542499, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.9332687854766846, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6964749693870544, + "num_tokens": 165063912.0, + "step": 6372 + }, + { + "epoch": 0.6998682187568636, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4136366844177246, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6948280334472656, + "num_tokens": 165091183.0, + "step": 6373 + }, + { + "epoch": 0.6999780364594773, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4377124309539795, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7161717414855957, + "num_tokens": 165117875.0, + "step": 6374 + }, + { + "epoch": 0.700087854162091, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.839099645614624, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.717157244682312, + "num_tokens": 165139314.0, + "step": 6375 + }, + { + "epoch": 0.7001976718647046, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6422553062438965, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7114900350570679, + "num_tokens": 165164950.0, + "step": 6376 + }, + { + "epoch": 0.7003074895673183, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.767237424850464, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.706730842590332, + "num_tokens": 165190399.0, + "step": 6377 + }, + { + "epoch": 0.7004173072699319, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.897199869155884, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.7031163573265076, + "num_tokens": 165213424.0, + "step": 6378 + }, + { + "epoch": 0.7005271249725455, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5950725078582764, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7092801928520203, + "num_tokens": 165239322.0, + "step": 6379 + }, + { + "epoch": 0.7006369426751592, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.313344955444336, + "learning_rate": 1e-06, + "loss": 1.1112, + "mean_token_accuracy": 0.6725969314575195, + "num_tokens": 165276415.0, + "step": 6380 + }, + { + "epoch": 0.700746760377773, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5204269886016846, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6815384030342102, + "num_tokens": 165304440.0, + "step": 6381 + }, + { + "epoch": 0.7008565780803866, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5464744567871094, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7125908136367798, + "num_tokens": 165329692.0, + "step": 6382 + }, + { + "epoch": 0.7009663957830002, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.227482795715332, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7047731876373291, + "num_tokens": 165361978.0, + "step": 6383 + }, + { + "epoch": 0.7010762134856139, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.3890936374664307, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7109377384185791, + "num_tokens": 165391661.0, + "step": 6384 + }, + { + "epoch": 0.7011860311882275, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 3.755504608154297, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7028589844703674, + "num_tokens": 165412857.0, + "step": 6385 + }, + { + "epoch": 0.7012958488908412, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 3.085470199584961, + "learning_rate": 1e-06, + "loss": 0.9341, + "mean_token_accuracy": 0.7219499945640564, + "num_tokens": 165434553.0, + "step": 6386 + }, + { + "epoch": 0.7014056665934548, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.700906276702881, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7154785394668579, + "num_tokens": 165458779.0, + "step": 6387 + }, + { + "epoch": 0.7015154842960686, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4009504318237305, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7222981452941895, + "num_tokens": 165484387.0, + "step": 6388 + }, + { + "epoch": 0.7016253019986822, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.627250909805298, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7076926231384277, + "num_tokens": 165508058.0, + "step": 6389 + }, + { + "epoch": 0.7017351197012959, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4923105239868164, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.6923968195915222, + "num_tokens": 165534481.0, + "step": 6390 + }, + { + "epoch": 0.7018449374039095, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.3779804706573486, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7021923661231995, + "num_tokens": 165562167.0, + "step": 6391 + }, + { + "epoch": 0.7019547551065232, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.984264612197876, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.7238070964813232, + "num_tokens": 165580369.0, + "step": 6392 + }, + { + "epoch": 0.7020645728091368, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.50042724609375, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6974884271621704, + "num_tokens": 165606836.0, + "step": 6393 + }, + { + "epoch": 0.7021743905117505, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.737196683883667, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.6982473731040955, + "num_tokens": 165629360.0, + "step": 6394 + }, + { + "epoch": 0.7022842082143641, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5096662044525146, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6882734894752502, + "num_tokens": 165653418.0, + "step": 6395 + }, + { + "epoch": 0.7023940259169779, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.3945164680480957, + "learning_rate": 1e-06, + "loss": 0.9152, + "mean_token_accuracy": 0.7287989854812622, + "num_tokens": 165678775.0, + "step": 6396 + }, + { + "epoch": 0.7025038436195915, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5002026557922363, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7055488228797913, + "num_tokens": 165704460.0, + "step": 6397 + }, + { + "epoch": 0.7026136613222052, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4914441108703613, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7123926877975464, + "num_tokens": 165731850.0, + "step": 6398 + }, + { + "epoch": 0.7027234790248188, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6211650371551514, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6947332620620728, + "num_tokens": 165757577.0, + "step": 6399 + }, + { + "epoch": 0.7028332967274324, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.412935972213745, + "learning_rate": 1e-06, + "loss": 1.1396, + "mean_token_accuracy": 0.672059953212738, + "num_tokens": 165787726.0, + "step": 6400 + }, + { + "epoch": 0.7029431144300461, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.7600648403167725, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7003085017204285, + "num_tokens": 165810931.0, + "step": 6401 + }, + { + "epoch": 0.7030529321326597, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5668463706970215, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6872256994247437, + "num_tokens": 165837238.0, + "step": 6402 + }, + { + "epoch": 0.7031627498352735, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.3057007789611816, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7031652331352234, + "num_tokens": 165867065.0, + "step": 6403 + }, + { + "epoch": 0.7032725675378871, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.511054277420044, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6901991367340088, + "num_tokens": 165895370.0, + "step": 6404 + }, + { + "epoch": 0.7033823852405008, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.54205584526062, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6888896822929382, + "num_tokens": 165928520.0, + "step": 6405 + }, + { + "epoch": 0.7034922029431144, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 3.073612689971924, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6868805885314941, + "num_tokens": 165952076.0, + "step": 6406 + }, + { + "epoch": 0.7036020206457281, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4993340969085693, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7016829252243042, + "num_tokens": 165978496.0, + "step": 6407 + }, + { + "epoch": 0.7037118383483417, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6727139949798584, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.6988101005554199, + "num_tokens": 166000883.0, + "step": 6408 + }, + { + "epoch": 0.7038216560509554, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4463040828704834, + "learning_rate": 1e-06, + "loss": 1.1215, + "mean_token_accuracy": 0.6769487261772156, + "num_tokens": 166029905.0, + "step": 6409 + }, + { + "epoch": 0.7039314737535691, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6439321041107178, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7009242177009583, + "num_tokens": 166052445.0, + "step": 6410 + }, + { + "epoch": 0.7040412914561828, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.4212188720703125, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6951174736022949, + "num_tokens": 166080139.0, + "step": 6411 + }, + { + "epoch": 0.7041511091587964, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.439142942428589, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7034258842468262, + "num_tokens": 166109391.0, + "step": 6412 + }, + { + "epoch": 0.7042609268614101, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5046844482421875, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6766790151596069, + "num_tokens": 166138729.0, + "step": 6413 + }, + { + "epoch": 0.7043707445640237, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6940293312072754, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6817358732223511, + "num_tokens": 166164711.0, + "step": 6414 + }, + { + "epoch": 0.7044805622666374, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 3.0766143798828125, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7149198055267334, + "num_tokens": 166182875.0, + "step": 6415 + }, + { + "epoch": 0.704590379969251, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.580000162124634, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6997224688529968, + "num_tokens": 166209426.0, + "step": 6416 + }, + { + "epoch": 0.7047001976718648, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.421905040740967, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6783868074417114, + "num_tokens": 166234077.0, + "step": 6417 + }, + { + "epoch": 0.7048100153744784, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.294743299484253, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6978074312210083, + "num_tokens": 166263591.0, + "step": 6418 + }, + { + "epoch": 0.704919833077092, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6690118312835693, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7062234878540039, + "num_tokens": 166287278.0, + "step": 6419 + }, + { + "epoch": 0.7050296507797057, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.5618932247161865, + "learning_rate": 1e-06, + "loss": 0.9171, + "mean_token_accuracy": 0.7283250093460083, + "num_tokens": 166309647.0, + "step": 6420 + }, + { + "epoch": 0.7051394684823193, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.513862371444702, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.681351900100708, + "num_tokens": 166336724.0, + "step": 6421 + }, + { + "epoch": 0.705249286184933, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.717104434967041, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7012119293212891, + "num_tokens": 166361211.0, + "step": 6422 + }, + { + "epoch": 0.7053591038875466, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.3679826259613037, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7243518829345703, + "num_tokens": 166385822.0, + "step": 6423 + }, + { + "epoch": 0.7054689215901603, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7291817665100098, + "learning_rate": 1e-06, + "loss": 0.8694, + "mean_token_accuracy": 0.7392474412918091, + "num_tokens": 166407055.0, + "step": 6424 + }, + { + "epoch": 0.705578739292774, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.674694776535034, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6928530931472778, + "num_tokens": 166430708.0, + "step": 6425 + }, + { + "epoch": 0.7056885569953877, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7305526733398438, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6995798349380493, + "num_tokens": 166456779.0, + "step": 6426 + }, + { + "epoch": 0.7057983746980013, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5126376152038574, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7065236568450928, + "num_tokens": 166481233.0, + "step": 6427 + }, + { + "epoch": 0.705908192400615, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7681567668914795, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7164131999015808, + "num_tokens": 166501202.0, + "step": 6428 + }, + { + "epoch": 0.7060180101032286, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.42867112159729, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.6990196704864502, + "num_tokens": 166531152.0, + "step": 6429 + }, + { + "epoch": 0.7061278278058423, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6949706077575684, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7131056785583496, + "num_tokens": 166553589.0, + "step": 6430 + }, + { + "epoch": 0.7062376455084559, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.9542577266693115, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.708562433719635, + "num_tokens": 166578512.0, + "step": 6431 + }, + { + "epoch": 0.7063474632110697, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7377216815948486, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7089142203330994, + "num_tokens": 166599150.0, + "step": 6432 + }, + { + "epoch": 0.7064572809136833, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.507035255432129, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6831396818161011, + "num_tokens": 166626449.0, + "step": 6433 + }, + { + "epoch": 0.706567098616297, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.599036931991577, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7195608615875244, + "num_tokens": 166648587.0, + "step": 6434 + }, + { + "epoch": 0.7066769163189106, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.741834878921509, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.7184686660766602, + "num_tokens": 166670674.0, + "step": 6435 + }, + { + "epoch": 0.7067867340215243, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.8710765838623047, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7037034630775452, + "num_tokens": 166695813.0, + "step": 6436 + }, + { + "epoch": 0.7068965517241379, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.8552448749542236, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7117006778717041, + "num_tokens": 166716119.0, + "step": 6437 + }, + { + "epoch": 0.7070063694267515, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.3966481685638428, + "learning_rate": 1e-06, + "loss": 1.1522, + "mean_token_accuracy": 0.6693872213363647, + "num_tokens": 166750666.0, + "step": 6438 + }, + { + "epoch": 0.7071161871293653, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7638368606567383, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7026252746582031, + "num_tokens": 166772775.0, + "step": 6439 + }, + { + "epoch": 0.707226004831979, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.9968926906585693, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7187741994857788, + "num_tokens": 166792522.0, + "step": 6440 + }, + { + "epoch": 0.7073358225345926, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5325000286102295, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6952841281890869, + "num_tokens": 166821634.0, + "step": 6441 + }, + { + "epoch": 0.7074456402372062, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5558438301086426, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.68914794921875, + "num_tokens": 166847977.0, + "step": 6442 + }, + { + "epoch": 0.7075554579398199, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.8146913051605225, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6958756446838379, + "num_tokens": 166871906.0, + "step": 6443 + }, + { + "epoch": 0.7076652756424335, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4463725090026855, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7004473209381104, + "num_tokens": 166903174.0, + "step": 6444 + }, + { + "epoch": 0.7077750933450472, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6029865741729736, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7222801446914673, + "num_tokens": 166927652.0, + "step": 6445 + }, + { + "epoch": 0.7078849110476609, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4871408939361572, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7039238214492798, + "num_tokens": 166951923.0, + "step": 6446 + }, + { + "epoch": 0.7079947287502746, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.888479471206665, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7059096097946167, + "num_tokens": 166973547.0, + "step": 6447 + }, + { + "epoch": 0.7081045464528882, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.792534828186035, + "learning_rate": 1e-06, + "loss": 0.9219, + "mean_token_accuracy": 0.7227564454078674, + "num_tokens": 166992928.0, + "step": 6448 + }, + { + "epoch": 0.7082143641555019, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7505011558532715, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7134820818901062, + "num_tokens": 167014443.0, + "step": 6449 + }, + { + "epoch": 0.7083241818581155, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5830135345458984, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.695685625076294, + "num_tokens": 167040164.0, + "step": 6450 + }, + { + "epoch": 0.7084339995607292, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4677815437316895, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6943814754486084, + "num_tokens": 167065933.0, + "step": 6451 + }, + { + "epoch": 0.7085438172633428, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.949406385421753, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7078925371170044, + "num_tokens": 167083429.0, + "step": 6452 + }, + { + "epoch": 0.7086536349659565, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.525850772857666, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7061159610748291, + "num_tokens": 167108697.0, + "step": 6453 + }, + { + "epoch": 0.7087634526685702, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.327244281768799, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7118588089942932, + "num_tokens": 167138333.0, + "step": 6454 + }, + { + "epoch": 0.7088732703711839, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.436917543411255, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.6982025504112244, + "num_tokens": 167166755.0, + "step": 6455 + }, + { + "epoch": 0.7089830880737975, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7230470180511475, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7120609283447266, + "num_tokens": 167189785.0, + "step": 6456 + }, + { + "epoch": 0.7090929057764112, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.836454153060913, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.71449875831604, + "num_tokens": 167211804.0, + "step": 6457 + }, + { + "epoch": 0.7092027234790248, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6962194442749023, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6917054653167725, + "num_tokens": 167233908.0, + "step": 6458 + }, + { + "epoch": 0.7093125411816384, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.758343458175659, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.7040156722068787, + "num_tokens": 167258245.0, + "step": 6459 + }, + { + "epoch": 0.7094223588842521, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.8196535110473633, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6947017312049866, + "num_tokens": 167284185.0, + "step": 6460 + }, + { + "epoch": 0.7095321765868658, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.491710662841797, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7037608623504639, + "num_tokens": 167309018.0, + "step": 6461 + }, + { + "epoch": 0.7096419942894795, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.363568067550659, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7094154357910156, + "num_tokens": 167337963.0, + "step": 6462 + }, + { + "epoch": 0.7097518119920931, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.2627131938934326, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6803911924362183, + "num_tokens": 167368817.0, + "step": 6463 + }, + { + "epoch": 0.7098616296947068, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4364709854125977, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7016819715499878, + "num_tokens": 167395048.0, + "step": 6464 + }, + { + "epoch": 0.7099714473973204, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.2832064628601074, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7064450979232788, + "num_tokens": 167424637.0, + "step": 6465 + }, + { + "epoch": 0.7100812650999341, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.2609567642211914, + "learning_rate": 1e-06, + "loss": 1.0984, + "mean_token_accuracy": 0.6797385215759277, + "num_tokens": 167459751.0, + "step": 6466 + }, + { + "epoch": 0.7101910828025477, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.559666633605957, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7136752009391785, + "num_tokens": 167484847.0, + "step": 6467 + }, + { + "epoch": 0.7103009005051615, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5488829612731934, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7088353037834167, + "num_tokens": 167507780.0, + "step": 6468 + }, + { + "epoch": 0.7104107182077751, + "ewc_loss": 7.62939453125e-06, + "grad_norm": 2.6555964946746826, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7164427042007446, + "num_tokens": 167530304.0, + "step": 6469 + }, + { + "epoch": 0.7105205359103888, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7865312099456787, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6911833882331848, + "num_tokens": 167551363.0, + "step": 6470 + }, + { + "epoch": 0.7106303536130024, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4716317653656006, + "learning_rate": 1e-06, + "loss": 1.0986, + "mean_token_accuracy": 0.6808887720108032, + "num_tokens": 167582669.0, + "step": 6471 + }, + { + "epoch": 0.7107401713156161, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.383599042892456, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.68513423204422, + "num_tokens": 167613850.0, + "step": 6472 + }, + { + "epoch": 0.7108499890182297, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.1947693824768066, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.700639545917511, + "num_tokens": 167646709.0, + "step": 6473 + }, + { + "epoch": 0.7109598067208434, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4600894451141357, + "learning_rate": 1e-06, + "loss": 1.0796, + "mean_token_accuracy": 0.7076269388198853, + "num_tokens": 167675359.0, + "step": 6474 + }, + { + "epoch": 0.7110696244234571, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6926159858703613, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.6960971355438232, + "num_tokens": 167698157.0, + "step": 6475 + }, + { + "epoch": 0.7111794421260708, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5890297889709473, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6923581957817078, + "num_tokens": 167725839.0, + "step": 6476 + }, + { + "epoch": 0.7112892598286844, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5980823040008545, + "learning_rate": 1e-06, + "loss": 0.9326, + "mean_token_accuracy": 0.7203357815742493, + "num_tokens": 167749767.0, + "step": 6477 + }, + { + "epoch": 0.711399077531298, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.457169532775879, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6900771260261536, + "num_tokens": 167777055.0, + "step": 6478 + }, + { + "epoch": 0.7115088952339117, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.737464427947998, + "learning_rate": 1e-06, + "loss": 0.9418, + "mean_token_accuracy": 0.7217708826065063, + "num_tokens": 167799343.0, + "step": 6479 + }, + { + "epoch": 0.7116187129365253, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.8093760013580322, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7176330089569092, + "num_tokens": 167822339.0, + "step": 6480 + }, + { + "epoch": 0.711728530639139, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5416440963745117, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.721733808517456, + "num_tokens": 167847874.0, + "step": 6481 + }, + { + "epoch": 0.7118383483417526, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6756343841552734, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6903983354568481, + "num_tokens": 167871087.0, + "step": 6482 + }, + { + "epoch": 0.7119481660443664, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5143067836761475, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.6854045391082764, + "num_tokens": 167898170.0, + "step": 6483 + }, + { + "epoch": 0.71205798374698, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.669783592224121, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7023801803588867, + "num_tokens": 167923656.0, + "step": 6484 + }, + { + "epoch": 0.7121678014495937, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4605889320373535, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6805336475372314, + "num_tokens": 167949266.0, + "step": 6485 + }, + { + "epoch": 0.7122776191522073, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.615013837814331, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.705548882484436, + "num_tokens": 167972994.0, + "step": 6486 + }, + { + "epoch": 0.712387436854821, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6906869411468506, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.698042631149292, + "num_tokens": 167997392.0, + "step": 6487 + }, + { + "epoch": 0.7124972545574346, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.7230491638183594, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6883929371833801, + "num_tokens": 168022751.0, + "step": 6488 + }, + { + "epoch": 0.7126070722600483, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.483710289001465, + "learning_rate": 1e-06, + "loss": 1.1252, + "mean_token_accuracy": 0.6814969182014465, + "num_tokens": 168052042.0, + "step": 6489 + }, + { + "epoch": 0.712716889962662, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 3.0141937732696533, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7006569504737854, + "num_tokens": 168071391.0, + "step": 6490 + }, + { + "epoch": 0.7128267076652757, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.411118507385254, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.69243323802948, + "num_tokens": 168102465.0, + "step": 6491 + }, + { + "epoch": 0.7129365253678893, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 3.0268874168395996, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.692676305770874, + "num_tokens": 168122278.0, + "step": 6492 + }, + { + "epoch": 0.713046343070503, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.4767067432403564, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.7035344839096069, + "num_tokens": 168151102.0, + "step": 6493 + }, + { + "epoch": 0.7131561607731166, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5533690452575684, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7290636301040649, + "num_tokens": 168175128.0, + "step": 6494 + }, + { + "epoch": 0.7132659784757303, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.3039724826812744, + "learning_rate": 1e-06, + "loss": 1.1173, + "mean_token_accuracy": 0.6715490818023682, + "num_tokens": 168207639.0, + "step": 6495 + }, + { + "epoch": 0.7133757961783439, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.6142845153808594, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7022006511688232, + "num_tokens": 168233243.0, + "step": 6496 + }, + { + "epoch": 0.7134856138809577, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5835063457489014, + "learning_rate": 1e-06, + "loss": 1.1381, + "mean_token_accuracy": 0.6750919818878174, + "num_tokens": 168260065.0, + "step": 6497 + }, + { + "epoch": 0.7135954315835713, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.615098476409912, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7175491452217102, + "num_tokens": 168284638.0, + "step": 6498 + }, + { + "epoch": 0.713705249286185, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.627628803253174, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.7011945247650146, + "num_tokens": 168310957.0, + "step": 6499 + }, + { + "epoch": 0.7138150669887986, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.730746030807495, + "learning_rate": 1e-06, + "loss": 1.1145, + "mean_token_accuracy": 0.6807606220245361, + "num_tokens": 168336000.0, + "step": 6500 + }, + { + "epoch": 0.7139248846914122, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.7571797370910645, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7057111263275146, + "num_tokens": 168358893.0, + "step": 6501 + }, + { + "epoch": 0.7140347023940259, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5739028453826904, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7218899726867676, + "num_tokens": 168382930.0, + "step": 6502 + }, + { + "epoch": 0.7141445200966395, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4968347549438477, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.685886561870575, + "num_tokens": 168410606.0, + "step": 6503 + }, + { + "epoch": 0.7142543377992533, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.35107421875, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7043391466140747, + "num_tokens": 168441184.0, + "step": 6504 + }, + { + "epoch": 0.7143641555018669, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.738698959350586, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7103044986724854, + "num_tokens": 168464355.0, + "step": 6505 + }, + { + "epoch": 0.7144739732044806, + "ewc_loss": 7.68899917602539e-06, + "grad_norm": 2.5820119380950928, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7034099102020264, + "num_tokens": 168491266.0, + "step": 6506 + }, + { + "epoch": 0.7145837909070942, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.83197021484375, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6894744038581848, + "num_tokens": 168516585.0, + "step": 6507 + }, + { + "epoch": 0.7146936086097079, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 3.0341970920562744, + "learning_rate": 1e-06, + "loss": 1.0847, + "mean_token_accuracy": 0.6853761076927185, + "num_tokens": 168535196.0, + "step": 6508 + }, + { + "epoch": 0.7148034263123215, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5279629230499268, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.7074611186981201, + "num_tokens": 168561111.0, + "step": 6509 + }, + { + "epoch": 0.7149132440149352, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 3.055546522140503, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7208709716796875, + "num_tokens": 168578428.0, + "step": 6510 + }, + { + "epoch": 0.7150230617175488, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.397566556930542, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.703065812587738, + "num_tokens": 168607648.0, + "step": 6511 + }, + { + "epoch": 0.7151328794201626, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.570554494857788, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6921653151512146, + "num_tokens": 168636005.0, + "step": 6512 + }, + { + "epoch": 0.7152426971227762, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3957295417785645, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.692030131816864, + "num_tokens": 168664419.0, + "step": 6513 + }, + { + "epoch": 0.7153525148253899, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.604888916015625, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7210371494293213, + "num_tokens": 168687639.0, + "step": 6514 + }, + { + "epoch": 0.7154623325280035, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.763855457305908, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.7020069360733032, + "num_tokens": 168712456.0, + "step": 6515 + }, + { + "epoch": 0.7155721502306172, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4635753631591797, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7164827585220337, + "num_tokens": 168740946.0, + "step": 6516 + }, + { + "epoch": 0.7156819679332308, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4332902431488037, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6826097965240479, + "num_tokens": 168769603.0, + "step": 6517 + }, + { + "epoch": 0.7157917856358444, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6394524574279785, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7215266227722168, + "num_tokens": 168795783.0, + "step": 6518 + }, + { + "epoch": 0.7159016033384582, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.192782163619995, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7243478894233704, + "num_tokens": 168813454.0, + "step": 6519 + }, + { + "epoch": 0.7160114210410718, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.846897840499878, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.715711236000061, + "num_tokens": 168834759.0, + "step": 6520 + }, + { + "epoch": 0.7161212387436855, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6876795291900635, + "learning_rate": 1e-06, + "loss": 1.1126, + "mean_token_accuracy": 0.6840198040008545, + "num_tokens": 168861563.0, + "step": 6521 + }, + { + "epoch": 0.7162310564462991, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.475022315979004, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7006876468658447, + "num_tokens": 168888884.0, + "step": 6522 + }, + { + "epoch": 0.7163408741489128, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4251837730407715, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7033888101577759, + "num_tokens": 168917774.0, + "step": 6523 + }, + { + "epoch": 0.7164506918515264, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7886526584625244, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7092359066009521, + "num_tokens": 168936506.0, + "step": 6524 + }, + { + "epoch": 0.7165605095541401, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.486250638961792, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6921546459197998, + "num_tokens": 168963308.0, + "step": 6525 + }, + { + "epoch": 0.7166703272567538, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.0158162117004395, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.726118266582489, + "num_tokens": 168985502.0, + "step": 6526 + }, + { + "epoch": 0.7167801449593675, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.596646547317505, + "learning_rate": 1e-06, + "loss": 0.8987, + "mean_token_accuracy": 0.7330793142318726, + "num_tokens": 169006486.0, + "step": 6527 + }, + { + "epoch": 0.7168899626619811, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5625579357147217, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.6971776485443115, + "num_tokens": 169030895.0, + "step": 6528 + }, + { + "epoch": 0.7169997803645948, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7645132541656494, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7006614208221436, + "num_tokens": 169054078.0, + "step": 6529 + }, + { + "epoch": 0.7171095980672084, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.922024726867676, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7147355079650879, + "num_tokens": 169072668.0, + "step": 6530 + }, + { + "epoch": 0.7172194157698221, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6220061779022217, + "learning_rate": 1e-06, + "loss": 1.0992, + "mean_token_accuracy": 0.6883829236030579, + "num_tokens": 169099329.0, + "step": 6531 + }, + { + "epoch": 0.7173292334724357, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.852778434753418, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7074190378189087, + "num_tokens": 169120903.0, + "step": 6532 + }, + { + "epoch": 0.7174390511750495, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4174180030822754, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6834529042243958, + "num_tokens": 169149640.0, + "step": 6533 + }, + { + "epoch": 0.7175488688776631, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.803001880645752, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7083821296691895, + "num_tokens": 169170780.0, + "step": 6534 + }, + { + "epoch": 0.7176586865802768, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4238686561584473, + "learning_rate": 1e-06, + "loss": 1.1066, + "mean_token_accuracy": 0.6881988644599915, + "num_tokens": 169200303.0, + "step": 6535 + }, + { + "epoch": 0.7177685042828904, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3281774520874023, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7210174202919006, + "num_tokens": 169229281.0, + "step": 6536 + }, + { + "epoch": 0.717878321985504, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6519041061401367, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7192650437355042, + "num_tokens": 169252296.0, + "step": 6537 + }, + { + "epoch": 0.7179881396881177, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.595874786376953, + "learning_rate": 1e-06, + "loss": 1.101, + "mean_token_accuracy": 0.6837137937545776, + "num_tokens": 169279224.0, + "step": 6538 + }, + { + "epoch": 0.7180979573907313, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.591029405593872, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6961514949798584, + "num_tokens": 169307045.0, + "step": 6539 + }, + { + "epoch": 0.7182077750933451, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7795217037200928, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7006306648254395, + "num_tokens": 169333309.0, + "step": 6540 + }, + { + "epoch": 0.7183175927959587, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6482319831848145, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6850814819335938, + "num_tokens": 169362714.0, + "step": 6541 + }, + { + "epoch": 0.7184274104985724, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7252960205078125, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6891430616378784, + "num_tokens": 169390450.0, + "step": 6542 + }, + { + "epoch": 0.718537228201186, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.466235876083374, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6914267539978027, + "num_tokens": 169419864.0, + "step": 6543 + }, + { + "epoch": 0.7186470459037997, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.726505756378174, + "learning_rate": 1e-06, + "loss": 0.907, + "mean_token_accuracy": 0.7291595339775085, + "num_tokens": 169441594.0, + "step": 6544 + }, + { + "epoch": 0.7187568636064133, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5970404148101807, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7000490427017212, + "num_tokens": 169467608.0, + "step": 6545 + }, + { + "epoch": 0.718866681309027, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5462539196014404, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.6956807971000671, + "num_tokens": 169494768.0, + "step": 6546 + }, + { + "epoch": 0.7189764990116406, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.769576072692871, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.6962252855300903, + "num_tokens": 169521616.0, + "step": 6547 + }, + { + "epoch": 0.7190863167142544, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.8867812156677246, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7174293994903564, + "num_tokens": 169542364.0, + "step": 6548 + }, + { + "epoch": 0.719196134416868, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.680644989013672, + "learning_rate": 1e-06, + "loss": 1.125, + "mean_token_accuracy": 0.6740937829017639, + "num_tokens": 169574336.0, + "step": 6549 + }, + { + "epoch": 0.7193059521194817, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.391037940979004, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.6986258029937744, + "num_tokens": 169602383.0, + "step": 6550 + }, + { + "epoch": 0.7194157698220953, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.606541633605957, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.6986051797866821, + "num_tokens": 169625680.0, + "step": 6551 + }, + { + "epoch": 0.719525587524709, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.58140230178833, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7084395289421082, + "num_tokens": 169649483.0, + "step": 6552 + }, + { + "epoch": 0.7196354052273226, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.3873744010925293, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7112633585929871, + "num_tokens": 169676176.0, + "step": 6553 + }, + { + "epoch": 0.7197452229299363, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.749831438064575, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7001769542694092, + "num_tokens": 169700230.0, + "step": 6554 + }, + { + "epoch": 0.71985504063255, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.575643301010132, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6950520277023315, + "num_tokens": 169724249.0, + "step": 6555 + }, + { + "epoch": 0.7199648583351637, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6485214233398438, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6959877014160156, + "num_tokens": 169750728.0, + "step": 6556 + }, + { + "epoch": 0.7200746760377773, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.343965530395508, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6966259479522705, + "num_tokens": 169782386.0, + "step": 6557 + }, + { + "epoch": 0.720184493740391, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.4572548866271973, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7152955532073975, + "num_tokens": 169809532.0, + "step": 6558 + }, + { + "epoch": 0.7202943114430046, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.425543785095215, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.6933755874633789, + "num_tokens": 169838593.0, + "step": 6559 + }, + { + "epoch": 0.7204041291456182, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.448899745941162, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7064540386199951, + "num_tokens": 169865464.0, + "step": 6560 + }, + { + "epoch": 0.7205139468482319, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.390068531036377, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7199345827102661, + "num_tokens": 169894499.0, + "step": 6561 + }, + { + "epoch": 0.7206237645508456, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.413001298904419, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6904267072677612, + "num_tokens": 169922616.0, + "step": 6562 + }, + { + "epoch": 0.7207335822534593, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.450401544570923, + "learning_rate": 1e-06, + "loss": 1.095, + "mean_token_accuracy": 0.6804085373878479, + "num_tokens": 169950672.0, + "step": 6563 + }, + { + "epoch": 0.7208433999560729, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.7062175273895264, + "learning_rate": 1e-06, + "loss": 1.0994, + "mean_token_accuracy": 0.6936231851577759, + "num_tokens": 169974535.0, + "step": 6564 + }, + { + "epoch": 0.7209532176586866, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.509315252304077, + "learning_rate": 1e-06, + "loss": 1.1441, + "mean_token_accuracy": 0.6764682531356812, + "num_tokens": 170002421.0, + "step": 6565 + }, + { + "epoch": 0.7210630353613002, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.285857677459717, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7143071889877319, + "num_tokens": 170032139.0, + "step": 6566 + }, + { + "epoch": 0.7211728530639139, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.737433910369873, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7001924514770508, + "num_tokens": 170051814.0, + "step": 6567 + }, + { + "epoch": 0.7212826707665275, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 5.26889705657959, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7104935646057129, + "num_tokens": 170080831.0, + "step": 6568 + }, + { + "epoch": 0.7213924884691413, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3920421600341797, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.6961748003959656, + "num_tokens": 170108149.0, + "step": 6569 + }, + { + "epoch": 0.7215023061717549, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5707876682281494, + "learning_rate": 1e-06, + "loss": 1.1133, + "mean_token_accuracy": 0.6803590059280396, + "num_tokens": 170137356.0, + "step": 6570 + }, + { + "epoch": 0.7216121238743686, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 3.111204147338867, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7032562494277954, + "num_tokens": 170161300.0, + "step": 6571 + }, + { + "epoch": 0.7217219415769822, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.647125720977783, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.7207435965538025, + "num_tokens": 170184667.0, + "step": 6572 + }, + { + "epoch": 0.7218317592795959, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.3538100719451904, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6883350610733032, + "num_tokens": 170215863.0, + "step": 6573 + }, + { + "epoch": 0.7219415769822095, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6251070499420166, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7079074382781982, + "num_tokens": 170239781.0, + "step": 6574 + }, + { + "epoch": 0.7220513946848232, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.597945213317871, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6901932954788208, + "num_tokens": 170266272.0, + "step": 6575 + }, + { + "epoch": 0.7221612123874368, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.586333751678467, + "learning_rate": 1e-06, + "loss": 1.137, + "mean_token_accuracy": 0.6689516305923462, + "num_tokens": 170293997.0, + "step": 6576 + }, + { + "epoch": 0.7222710300900506, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6256210803985596, + "learning_rate": 1e-06, + "loss": 0.8853, + "mean_token_accuracy": 0.7305681705474854, + "num_tokens": 170316894.0, + "step": 6577 + }, + { + "epoch": 0.7223808477926642, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.532167911529541, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7182464599609375, + "num_tokens": 170343035.0, + "step": 6578 + }, + { + "epoch": 0.7224906654952779, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.599188804626465, + "learning_rate": 1e-06, + "loss": 1.0977, + "mean_token_accuracy": 0.6926867365837097, + "num_tokens": 170368423.0, + "step": 6579 + }, + { + "epoch": 0.7226004831978915, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5645506381988525, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7201401591300964, + "num_tokens": 170393599.0, + "step": 6580 + }, + { + "epoch": 0.7227103009005051, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5086605548858643, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7214006185531616, + "num_tokens": 170419722.0, + "step": 6581 + }, + { + "epoch": 0.7228201186031188, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.583531379699707, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7072701454162598, + "num_tokens": 170443024.0, + "step": 6582 + }, + { + "epoch": 0.7229299363057324, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.259201765060425, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7043145298957825, + "num_tokens": 170471462.0, + "step": 6583 + }, + { + "epoch": 0.7230397540083462, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.3350698947906494, + "learning_rate": 1e-06, + "loss": 1.1221, + "mean_token_accuracy": 0.6789672374725342, + "num_tokens": 170502437.0, + "step": 6584 + }, + { + "epoch": 0.7231495717109598, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6660354137420654, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7147753238677979, + "num_tokens": 170525603.0, + "step": 6585 + }, + { + "epoch": 0.7232593894135735, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.3317198753356934, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6981861591339111, + "num_tokens": 170555958.0, + "step": 6586 + }, + { + "epoch": 0.7233692071161871, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6663405895233154, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.6975305080413818, + "num_tokens": 170579904.0, + "step": 6587 + }, + { + "epoch": 0.7234790248188008, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 3.0611062049865723, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7121129035949707, + "num_tokens": 170599073.0, + "step": 6588 + }, + { + "epoch": 0.7235888425214144, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.419095993041992, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7244453430175781, + "num_tokens": 170624812.0, + "step": 6589 + }, + { + "epoch": 0.7236986602240281, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.692405939102173, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6919969320297241, + "num_tokens": 170649253.0, + "step": 6590 + }, + { + "epoch": 0.7238084779266418, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6537277698516846, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6890096068382263, + "num_tokens": 170672443.0, + "step": 6591 + }, + { + "epoch": 0.7239182956292555, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.43833065032959, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6885823011398315, + "num_tokens": 170701394.0, + "step": 6592 + }, + { + "epoch": 0.7240281133318691, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.65830135345459, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7044251561164856, + "num_tokens": 170725227.0, + "step": 6593 + }, + { + "epoch": 0.7241379310344828, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7662482261657715, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7039386630058289, + "num_tokens": 170747080.0, + "step": 6594 + }, + { + "epoch": 0.7242477487370964, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.471539258956909, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7188311815261841, + "num_tokens": 170769740.0, + "step": 6595 + }, + { + "epoch": 0.72435756643971, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.56584095954895, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6871424913406372, + "num_tokens": 170795664.0, + "step": 6596 + }, + { + "epoch": 0.7244673841423237, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.429666042327881, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7100217342376709, + "num_tokens": 170824411.0, + "step": 6597 + }, + { + "epoch": 0.7245772018449375, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5340559482574463, + "learning_rate": 1e-06, + "loss": 1.0942, + "mean_token_accuracy": 0.6824049949645996, + "num_tokens": 170850123.0, + "step": 6598 + }, + { + "epoch": 0.7246870195475511, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7064623832702637, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.697292149066925, + "num_tokens": 170873265.0, + "step": 6599 + }, + { + "epoch": 0.7247968372501647, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5229127407073975, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6937578916549683, + "num_tokens": 170899152.0, + "step": 6600 + }, + { + "epoch": 0.7249066549527784, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4956154823303223, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6927740573883057, + "num_tokens": 170929558.0, + "step": 6601 + }, + { + "epoch": 0.725016472655392, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4794929027557373, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7131646871566772, + "num_tokens": 170955891.0, + "step": 6602 + }, + { + "epoch": 0.7251262903580057, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3420636653900146, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.7115316390991211, + "num_tokens": 170984853.0, + "step": 6603 + }, + { + "epoch": 0.7252361080606193, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.323370933532715, + "learning_rate": 1e-06, + "loss": 1.1523, + "mean_token_accuracy": 0.6730906963348389, + "num_tokens": 171017608.0, + "step": 6604 + }, + { + "epoch": 0.725345925763233, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6636264324188232, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.6944030523300171, + "num_tokens": 171043941.0, + "step": 6605 + }, + { + "epoch": 0.7254557434658467, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.689148187637329, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.690678060054779, + "num_tokens": 171067231.0, + "step": 6606 + }, + { + "epoch": 0.7255655611684604, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4252865314483643, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6941359043121338, + "num_tokens": 171094751.0, + "step": 6607 + }, + { + "epoch": 0.725675378871074, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5886311531066895, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6884641051292419, + "num_tokens": 171121793.0, + "step": 6608 + }, + { + "epoch": 0.7257851965736877, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 16.7182559967041, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6904290914535522, + "num_tokens": 171147151.0, + "step": 6609 + }, + { + "epoch": 0.7258950142763013, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7495172023773193, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6967991590499878, + "num_tokens": 171173165.0, + "step": 6610 + }, + { + "epoch": 0.726004831978915, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.806612730026245, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7184818387031555, + "num_tokens": 171195083.0, + "step": 6611 + }, + { + "epoch": 0.7261146496815286, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.1607344150543213, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6849554777145386, + "num_tokens": 171227814.0, + "step": 6612 + }, + { + "epoch": 0.7262244673841424, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.532938241958618, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6888954043388367, + "num_tokens": 171255157.0, + "step": 6613 + }, + { + "epoch": 0.726334285086756, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4698007106781006, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7041188478469849, + "num_tokens": 171279921.0, + "step": 6614 + }, + { + "epoch": 0.7264441027893697, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.549030065536499, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6861303448677063, + "num_tokens": 171306367.0, + "step": 6615 + }, + { + "epoch": 0.7265539204919833, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.44645094871521, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6843097805976868, + "num_tokens": 171335470.0, + "step": 6616 + }, + { + "epoch": 0.726663738194597, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.538450002670288, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6920282244682312, + "num_tokens": 171360291.0, + "step": 6617 + }, + { + "epoch": 0.7267735558972106, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.484180212020874, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.7014885544776917, + "num_tokens": 171386241.0, + "step": 6618 + }, + { + "epoch": 0.7268833735998242, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5686872005462646, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6843423843383789, + "num_tokens": 171412587.0, + "step": 6619 + }, + { + "epoch": 0.726993191302438, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.7231388092041016, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.700474202632904, + "num_tokens": 171434516.0, + "step": 6620 + }, + { + "epoch": 0.7271030090050516, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5093066692352295, + "learning_rate": 1e-06, + "loss": 1.0991, + "mean_token_accuracy": 0.6812794208526611, + "num_tokens": 171464916.0, + "step": 6621 + }, + { + "epoch": 0.7272128267076653, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6468348503112793, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6922046542167664, + "num_tokens": 171489467.0, + "step": 6622 + }, + { + "epoch": 0.7273226444102789, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.044962167739868, + "learning_rate": 1e-06, + "loss": 0.8986, + "mean_token_accuracy": 0.7347167730331421, + "num_tokens": 171506265.0, + "step": 6623 + }, + { + "epoch": 0.7274324621128926, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.648505687713623, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.724021315574646, + "num_tokens": 171527160.0, + "step": 6624 + }, + { + "epoch": 0.7275422798155062, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 16.696083068847656, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.707950234413147, + "num_tokens": 171552198.0, + "step": 6625 + }, + { + "epoch": 0.7276520975181199, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9495325088500977, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7023926377296448, + "num_tokens": 171573001.0, + "step": 6626 + }, + { + "epoch": 0.7277619152207336, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5493228435516357, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6861860752105713, + "num_tokens": 171602431.0, + "step": 6627 + }, + { + "epoch": 0.7278717329233473, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5387163162231445, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6970087289810181, + "num_tokens": 171631809.0, + "step": 6628 + }, + { + "epoch": 0.7279815506259609, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5656850337982178, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6974433660507202, + "num_tokens": 171655463.0, + "step": 6629 + }, + { + "epoch": 0.7280913683285746, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5394632816314697, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7161886692047119, + "num_tokens": 171679233.0, + "step": 6630 + }, + { + "epoch": 0.7282011860311882, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8364834785461426, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7087951898574829, + "num_tokens": 171704875.0, + "step": 6631 + }, + { + "epoch": 0.7283110037338019, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3691818714141846, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.7096745371818542, + "num_tokens": 171736957.0, + "step": 6632 + }, + { + "epoch": 0.7284208214364155, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.433781385421753, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6775819063186646, + "num_tokens": 171765231.0, + "step": 6633 + }, + { + "epoch": 0.7285306391390292, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8059473037719727, + "learning_rate": 1e-06, + "loss": 1.1119, + "mean_token_accuracy": 0.6822937726974487, + "num_tokens": 171790280.0, + "step": 6634 + }, + { + "epoch": 0.7286404568416429, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.424773693084717, + "learning_rate": 1e-06, + "loss": 1.1356, + "mean_token_accuracy": 0.6801608800888062, + "num_tokens": 171823396.0, + "step": 6635 + }, + { + "epoch": 0.7287502745442566, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.281989812850952, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7079729437828064, + "num_tokens": 171852932.0, + "step": 6636 + }, + { + "epoch": 0.7288600922468702, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.473813056945801, + "learning_rate": 1e-06, + "loss": 1.1372, + "mean_token_accuracy": 0.6711257696151733, + "num_tokens": 171882645.0, + "step": 6637 + }, + { + "epoch": 0.7289699099494839, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 3.276343822479248, + "learning_rate": 1e-06, + "loss": 1.1014, + "mean_token_accuracy": 0.6838535070419312, + "num_tokens": 171907999.0, + "step": 6638 + }, + { + "epoch": 0.7290797276520975, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.7951884269714355, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7136422991752625, + "num_tokens": 171929437.0, + "step": 6639 + }, + { + "epoch": 0.7291895453547111, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.538416862487793, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7172143459320068, + "num_tokens": 171952768.0, + "step": 6640 + }, + { + "epoch": 0.7292993630573248, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.4000473022460938, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6915056109428406, + "num_tokens": 171982817.0, + "step": 6641 + }, + { + "epoch": 0.7294091807599385, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.634577512741089, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.699893593788147, + "num_tokens": 172008264.0, + "step": 6642 + }, + { + "epoch": 0.7295189984625522, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.3860037326812744, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.695472776889801, + "num_tokens": 172037187.0, + "step": 6643 + }, + { + "epoch": 0.7296288161651658, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6070399284362793, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6948542594909668, + "num_tokens": 172061265.0, + "step": 6644 + }, + { + "epoch": 0.7297386338677795, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.51678204536438, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.707501232624054, + "num_tokens": 172088106.0, + "step": 6645 + }, + { + "epoch": 0.7298484515703931, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6030707359313965, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6919746398925781, + "num_tokens": 172111278.0, + "step": 6646 + }, + { + "epoch": 0.7299582692730068, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4526445865631104, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6915876269340515, + "num_tokens": 172139161.0, + "step": 6647 + }, + { + "epoch": 0.7300680869756204, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5824668407440186, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6808556914329529, + "num_tokens": 172164497.0, + "step": 6648 + }, + { + "epoch": 0.7301779046782342, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.3138034343719482, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6849149465560913, + "num_tokens": 172195776.0, + "step": 6649 + }, + { + "epoch": 0.7302877223808478, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.478858232498169, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7077677249908447, + "num_tokens": 172222583.0, + "step": 6650 + }, + { + "epoch": 0.7303975400834615, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6792173385620117, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6901102066040039, + "num_tokens": 172247462.0, + "step": 6651 + }, + { + "epoch": 0.7305073577860751, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.4110536575317383, + "learning_rate": 1e-06, + "loss": 1.0923, + "mean_token_accuracy": 0.6875016689300537, + "num_tokens": 172275377.0, + "step": 6652 + }, + { + "epoch": 0.7306171754886888, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.721200466156006, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7219040393829346, + "num_tokens": 172297610.0, + "step": 6653 + }, + { + "epoch": 0.7307269931913024, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6143102645874023, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.708003044128418, + "num_tokens": 172322826.0, + "step": 6654 + }, + { + "epoch": 0.7308368108939161, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.584700584411621, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.692298173904419, + "num_tokens": 172347222.0, + "step": 6655 + }, + { + "epoch": 0.7309466285965298, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.337313413619995, + "learning_rate": 1e-06, + "loss": 0.9478, + "mean_token_accuracy": 0.7222402095794678, + "num_tokens": 172373660.0, + "step": 6656 + }, + { + "epoch": 0.7310564462991435, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3549087047576904, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6907370090484619, + "num_tokens": 172402897.0, + "step": 6657 + }, + { + "epoch": 0.7311662640017571, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6812596321105957, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7102978825569153, + "num_tokens": 172424908.0, + "step": 6658 + }, + { + "epoch": 0.7312760817043708, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.4398865699768066, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6986204385757446, + "num_tokens": 172452343.0, + "step": 6659 + }, + { + "epoch": 0.7313858994069844, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5764377117156982, + "learning_rate": 1e-06, + "loss": 0.9091, + "mean_token_accuracy": 0.7275513410568237, + "num_tokens": 172474687.0, + "step": 6660 + }, + { + "epoch": 0.731495717109598, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.43385648727417, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7122761607170105, + "num_tokens": 172501083.0, + "step": 6661 + }, + { + "epoch": 0.7316055348122117, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4923739433288574, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6897560358047485, + "num_tokens": 172529837.0, + "step": 6662 + }, + { + "epoch": 0.7317153525148253, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3254246711730957, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6967300772666931, + "num_tokens": 172565242.0, + "step": 6663 + }, + { + "epoch": 0.7318251702174391, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4570393562316895, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7161328792572021, + "num_tokens": 172590053.0, + "step": 6664 + }, + { + "epoch": 0.7319349879200527, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3043270111083984, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7118948698043823, + "num_tokens": 172617737.0, + "step": 6665 + }, + { + "epoch": 0.7320448056226664, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.493565082550049, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7092064619064331, + "num_tokens": 172643719.0, + "step": 6666 + }, + { + "epoch": 0.73215462332528, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.4431257247924805, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7072430849075317, + "num_tokens": 172670802.0, + "step": 6667 + }, + { + "epoch": 0.7322644410278937, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9690794944763184, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6965515613555908, + "num_tokens": 172701329.0, + "step": 6668 + }, + { + "epoch": 0.7323742587305073, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6809775829315186, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6937544345855713, + "num_tokens": 172723687.0, + "step": 6669 + }, + { + "epoch": 0.732484076433121, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3530941009521484, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6973252296447754, + "num_tokens": 172752354.0, + "step": 6670 + }, + { + "epoch": 0.7325938941357347, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.5654656887054443, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6944012641906738, + "num_tokens": 172776323.0, + "step": 6671 + }, + { + "epoch": 0.7327037118383484, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5283708572387695, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7104372978210449, + "num_tokens": 172800904.0, + "step": 6672 + }, + { + "epoch": 0.732813529540962, + "ewc_loss": 7.748603820800781e-06, + "grad_norm": 2.6365253925323486, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7085298299789429, + "num_tokens": 172823370.0, + "step": 6673 + }, + { + "epoch": 0.7329233472435757, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6589174270629883, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7072461247444153, + "num_tokens": 172846258.0, + "step": 6674 + }, + { + "epoch": 0.7330331649461893, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.862128496170044, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.6986689567565918, + "num_tokens": 172867940.0, + "step": 6675 + }, + { + "epoch": 0.733142982648803, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.152509927749634, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.6970856785774231, + "num_tokens": 172886700.0, + "step": 6676 + }, + { + "epoch": 0.7332528003514166, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5675971508026123, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7002314329147339, + "num_tokens": 172914941.0, + "step": 6677 + }, + { + "epoch": 0.7333626180540304, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9649877548217773, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7126941680908203, + "num_tokens": 172939971.0, + "step": 6678 + }, + { + "epoch": 0.733472435756644, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.564087152481079, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.7059445381164551, + "num_tokens": 172968136.0, + "step": 6679 + }, + { + "epoch": 0.7335822534592576, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.545095920562744, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6993449330329895, + "num_tokens": 172993683.0, + "step": 6680 + }, + { + "epoch": 0.7336920711618713, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5626678466796875, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6952716112136841, + "num_tokens": 173018448.0, + "step": 6681 + }, + { + "epoch": 0.7338018888644849, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.479402780532837, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.69675612449646, + "num_tokens": 173044670.0, + "step": 6682 + }, + { + "epoch": 0.7339117065670986, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5184295177459717, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6899142861366272, + "num_tokens": 173072158.0, + "step": 6683 + }, + { + "epoch": 0.7340215242697122, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.443479299545288, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7053515315055847, + "num_tokens": 173101918.0, + "step": 6684 + }, + { + "epoch": 0.734131341972326, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9328439235687256, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6953884363174438, + "num_tokens": 173127160.0, + "step": 6685 + }, + { + "epoch": 0.7342411596749396, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4161605834960938, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.6880121827125549, + "num_tokens": 173153811.0, + "step": 6686 + }, + { + "epoch": 0.7343509773775533, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6772067546844482, + "learning_rate": 1e-06, + "loss": 1.0928, + "mean_token_accuracy": 0.682500958442688, + "num_tokens": 173178814.0, + "step": 6687 + }, + { + "epoch": 0.7344607950801669, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7140824794769287, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.687724232673645, + "num_tokens": 173200290.0, + "step": 6688 + }, + { + "epoch": 0.7345706127827806, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.502786874771118, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6889687776565552, + "num_tokens": 173226366.0, + "step": 6689 + }, + { + "epoch": 0.7346804304853942, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.771644353866577, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.692901611328125, + "num_tokens": 173253655.0, + "step": 6690 + }, + { + "epoch": 0.7347902481880079, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.784984588623047, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7149075865745544, + "num_tokens": 173275181.0, + "step": 6691 + }, + { + "epoch": 0.7349000658906216, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3957817554473877, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7074201107025146, + "num_tokens": 173302924.0, + "step": 6692 + }, + { + "epoch": 0.7350098835932353, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.67546010017395, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6941248178482056, + "num_tokens": 173327534.0, + "step": 6693 + }, + { + "epoch": 0.7351197012958489, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6990015506744385, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7146502733230591, + "num_tokens": 173352614.0, + "step": 6694 + }, + { + "epoch": 0.7352295189984626, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8780274391174316, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.70171719789505, + "num_tokens": 173374986.0, + "step": 6695 + }, + { + "epoch": 0.7353393367010762, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.034402370452881, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6992604732513428, + "num_tokens": 173394416.0, + "step": 6696 + }, + { + "epoch": 0.7354491544036899, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.609372615814209, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7114881277084351, + "num_tokens": 173417035.0, + "step": 6697 + }, + { + "epoch": 0.7355589721063035, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3450331687927246, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7111304998397827, + "num_tokens": 173447210.0, + "step": 6698 + }, + { + "epoch": 0.7356687898089171, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5174310207366943, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.6979259848594666, + "num_tokens": 173473313.0, + "step": 6699 + }, + { + "epoch": 0.7357786075115309, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.56485652923584, + "learning_rate": 1e-06, + "loss": 0.8955, + "mean_token_accuracy": 0.7335994243621826, + "num_tokens": 173495252.0, + "step": 6700 + }, + { + "epoch": 0.7358884252141445, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.596118450164795, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7227895259857178, + "num_tokens": 173521442.0, + "step": 6701 + }, + { + "epoch": 0.7359982429167582, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4298977851867676, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7187550067901611, + "num_tokens": 173547300.0, + "step": 6702 + }, + { + "epoch": 0.7361080606193718, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4887595176696777, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7109419703483582, + "num_tokens": 173576514.0, + "step": 6703 + }, + { + "epoch": 0.7362178783219855, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.672123432159424, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6995202898979187, + "num_tokens": 173602159.0, + "step": 6704 + }, + { + "epoch": 0.7363276960245991, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5420384407043457, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7042297720909119, + "num_tokens": 173627353.0, + "step": 6705 + }, + { + "epoch": 0.7364375137272128, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.352681875228882, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7195162773132324, + "num_tokens": 173655914.0, + "step": 6706 + }, + { + "epoch": 0.7365473314298265, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.559889793395996, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6982969045639038, + "num_tokens": 173682089.0, + "step": 6707 + }, + { + "epoch": 0.7366571491324402, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.463792324066162, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7161145210266113, + "num_tokens": 173710899.0, + "step": 6708 + }, + { + "epoch": 0.7367669668350538, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.834429979324341, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6950907707214355, + "num_tokens": 173734020.0, + "step": 6709 + }, + { + "epoch": 0.7368767845376675, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6074726581573486, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7123264670372009, + "num_tokens": 173763730.0, + "step": 6710 + }, + { + "epoch": 0.7369866022402811, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.666840076446533, + "learning_rate": 1e-06, + "loss": 1.1203, + "mean_token_accuracy": 0.670343816280365, + "num_tokens": 173789322.0, + "step": 6711 + }, + { + "epoch": 0.7370964199428948, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3118896484375, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.685530424118042, + "num_tokens": 173821838.0, + "step": 6712 + }, + { + "epoch": 0.7372062376455084, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.51570725440979, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7160719633102417, + "num_tokens": 173848132.0, + "step": 6713 + }, + { + "epoch": 0.7373160553481222, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.804872751235962, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7284987568855286, + "num_tokens": 173872793.0, + "step": 6714 + }, + { + "epoch": 0.7374258730507358, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.524763584136963, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7124587893486023, + "num_tokens": 173897476.0, + "step": 6715 + }, + { + "epoch": 0.7375356907533495, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6136155128479004, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7009437084197998, + "num_tokens": 173921544.0, + "step": 6716 + }, + { + "epoch": 0.7376455084559631, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.645583152770996, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.7152388095855713, + "num_tokens": 173942604.0, + "step": 6717 + }, + { + "epoch": 0.7377553261585768, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4465813636779785, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7019438147544861, + "num_tokens": 173967385.0, + "step": 6718 + }, + { + "epoch": 0.7378651438611904, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.713406562805176, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.7100023031234741, + "num_tokens": 173989018.0, + "step": 6719 + }, + { + "epoch": 0.737974961563804, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.576463222503662, + "learning_rate": 1e-06, + "loss": 1.1215, + "mean_token_accuracy": 0.6850392818450928, + "num_tokens": 174014756.0, + "step": 6720 + }, + { + "epoch": 0.7380847792664178, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8310277462005615, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.691535472869873, + "num_tokens": 174036322.0, + "step": 6721 + }, + { + "epoch": 0.7381945969690314, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5257608890533447, + "learning_rate": 1e-06, + "loss": 0.9301, + "mean_token_accuracy": 0.7253142595291138, + "num_tokens": 174061966.0, + "step": 6722 + }, + { + "epoch": 0.7383044146716451, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.18471360206604, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6927914619445801, + "num_tokens": 174082133.0, + "step": 6723 + }, + { + "epoch": 0.7384142323742587, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3174538612365723, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.6977497339248657, + "num_tokens": 174114587.0, + "step": 6724 + }, + { + "epoch": 0.7385240500768724, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.321537971496582, + "learning_rate": 1e-06, + "loss": 1.1055, + "mean_token_accuracy": 0.6810479760169983, + "num_tokens": 174144910.0, + "step": 6725 + }, + { + "epoch": 0.738633867779486, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5724337100982666, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7012447714805603, + "num_tokens": 174170044.0, + "step": 6726 + }, + { + "epoch": 0.7387436854820997, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.4476168155670166, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7261029481887817, + "num_tokens": 174194128.0, + "step": 6727 + }, + { + "epoch": 0.7388535031847133, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7986342906951904, + "learning_rate": 1e-06, + "loss": 0.916, + "mean_token_accuracy": 0.725865364074707, + "num_tokens": 174213613.0, + "step": 6728 + }, + { + "epoch": 0.7389633208873271, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5296390056610107, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7158300280570984, + "num_tokens": 174239072.0, + "step": 6729 + }, + { + "epoch": 0.7390731385899407, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4926276206970215, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6892104744911194, + "num_tokens": 174266710.0, + "step": 6730 + }, + { + "epoch": 0.7391829562925544, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.622997522354126, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7043795585632324, + "num_tokens": 174289578.0, + "step": 6731 + }, + { + "epoch": 0.739292773995168, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.608597993850708, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7048224806785583, + "num_tokens": 174312357.0, + "step": 6732 + }, + { + "epoch": 0.7394025916977817, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7850534915924072, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7068291902542114, + "num_tokens": 174333509.0, + "step": 6733 + }, + { + "epoch": 0.7395124094003953, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.744398832321167, + "learning_rate": 1e-06, + "loss": 0.9175, + "mean_token_accuracy": 0.7191633582115173, + "num_tokens": 174354178.0, + "step": 6734 + }, + { + "epoch": 0.739622227103009, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7776501178741455, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6994016170501709, + "num_tokens": 174377254.0, + "step": 6735 + }, + { + "epoch": 0.7397320448056227, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6538054943084717, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7090891599655151, + "num_tokens": 174400581.0, + "step": 6736 + }, + { + "epoch": 0.7398418625082364, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.303675889968872, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6942130327224731, + "num_tokens": 174431436.0, + "step": 6737 + }, + { + "epoch": 0.73995168021085, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.756131172180176, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.7022420167922974, + "num_tokens": 174456476.0, + "step": 6738 + }, + { + "epoch": 0.7400614979134637, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6797966957092285, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.6829867959022522, + "num_tokens": 174479634.0, + "step": 6739 + }, + { + "epoch": 0.7401713156160773, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6762406826019287, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.6993425488471985, + "num_tokens": 174504223.0, + "step": 6740 + }, + { + "epoch": 0.7402811333186909, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4328384399414062, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7203691005706787, + "num_tokens": 174532251.0, + "step": 6741 + }, + { + "epoch": 0.7403909510213046, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.787155866622925, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.716208815574646, + "num_tokens": 174551363.0, + "step": 6742 + }, + { + "epoch": 0.7405007687239183, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.137686252593994, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7085720896720886, + "num_tokens": 174585552.0, + "step": 6743 + }, + { + "epoch": 0.740610586426532, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.545433521270752, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6884846687316895, + "num_tokens": 174612006.0, + "step": 6744 + }, + { + "epoch": 0.7407204041291456, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.812035322189331, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.6934103965759277, + "num_tokens": 174635438.0, + "step": 6745 + }, + { + "epoch": 0.7408302218317593, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.517971992492676, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6956883072853088, + "num_tokens": 174663764.0, + "step": 6746 + }, + { + "epoch": 0.7409400395343729, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.725743055343628, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7051823735237122, + "num_tokens": 174686855.0, + "step": 6747 + }, + { + "epoch": 0.7410498572369866, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.574044942855835, + "learning_rate": 1e-06, + "loss": 1.0745, + "mean_token_accuracy": 0.6907795071601868, + "num_tokens": 174712408.0, + "step": 6748 + }, + { + "epoch": 0.7411596749396002, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.310762882232666, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6950769424438477, + "num_tokens": 174746301.0, + "step": 6749 + }, + { + "epoch": 0.741269492642214, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.506310224533081, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7025099396705627, + "num_tokens": 174772707.0, + "step": 6750 + }, + { + "epoch": 0.7413793103448276, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.492824077606201, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6982861161231995, + "num_tokens": 174800887.0, + "step": 6751 + }, + { + "epoch": 0.7414891280474413, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.2871930599212646, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7154186367988586, + "num_tokens": 174825484.0, + "step": 6752 + }, + { + "epoch": 0.7415989457500549, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4778635501861572, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6923351287841797, + "num_tokens": 174856788.0, + "step": 6753 + }, + { + "epoch": 0.7417087634526686, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5085079669952393, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7121500372886658, + "num_tokens": 174883870.0, + "step": 6754 + }, + { + "epoch": 0.7418185811552822, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.499082565307617, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7098495960235596, + "num_tokens": 174911277.0, + "step": 6755 + }, + { + "epoch": 0.7419283988578959, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4208807945251465, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7126421928405762, + "num_tokens": 174937395.0, + "step": 6756 + }, + { + "epoch": 0.7420382165605095, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7142457962036133, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6820467710494995, + "num_tokens": 174962830.0, + "step": 6757 + }, + { + "epoch": 0.7421480342631233, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4217488765716553, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.6976191401481628, + "num_tokens": 174989345.0, + "step": 6758 + }, + { + "epoch": 0.7422578519657369, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6641085147857666, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.71550452709198, + "num_tokens": 175012986.0, + "step": 6759 + }, + { + "epoch": 0.7423676696683505, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.93416690826416, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7045000791549683, + "num_tokens": 175032096.0, + "step": 6760 + }, + { + "epoch": 0.7424774873709642, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.889090061187744, + "learning_rate": 1e-06, + "loss": 0.933, + "mean_token_accuracy": 0.7305910587310791, + "num_tokens": 175051955.0, + "step": 6761 + }, + { + "epoch": 0.7425873050735778, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6358044147491455, + "learning_rate": 1e-06, + "loss": 1.135, + "mean_token_accuracy": 0.6764606237411499, + "num_tokens": 175078938.0, + "step": 6762 + }, + { + "epoch": 0.7426971227761915, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.488455057144165, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.71628737449646, + "num_tokens": 175104952.0, + "step": 6763 + }, + { + "epoch": 0.7428069404788051, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.2900829315185547, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6928464770317078, + "num_tokens": 175129499.0, + "step": 6764 + }, + { + "epoch": 0.7429167581814189, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.594592332839966, + "learning_rate": 1e-06, + "loss": 1.1339, + "mean_token_accuracy": 0.6727662086486816, + "num_tokens": 175155643.0, + "step": 6765 + }, + { + "epoch": 0.7430265758840325, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.626307249069214, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6923472285270691, + "num_tokens": 175179215.0, + "step": 6766 + }, + { + "epoch": 0.7431363935866462, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3818931579589844, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6937698125839233, + "num_tokens": 175208654.0, + "step": 6767 + }, + { + "epoch": 0.7432462112892598, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.59306263923645, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6969077587127686, + "num_tokens": 175235570.0, + "step": 6768 + }, + { + "epoch": 0.7433560289918735, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.564624547958374, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6927503943443298, + "num_tokens": 175260521.0, + "step": 6769 + }, + { + "epoch": 0.7434658466944871, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3693323135375977, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7154794335365295, + "num_tokens": 175287083.0, + "step": 6770 + }, + { + "epoch": 0.7435756643971008, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.791186809539795, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7118273973464966, + "num_tokens": 175309073.0, + "step": 6771 + }, + { + "epoch": 0.7436854820997145, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.950683832168579, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7051018476486206, + "num_tokens": 175327506.0, + "step": 6772 + }, + { + "epoch": 0.7437952998023282, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.9964799880981445, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7049646377563477, + "num_tokens": 175350031.0, + "step": 6773 + }, + { + "epoch": 0.7439051175049418, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7030093669891357, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7054451704025269, + "num_tokens": 175371750.0, + "step": 6774 + }, + { + "epoch": 0.7440149352075555, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.423180103302002, + "learning_rate": 1e-06, + "loss": 1.1139, + "mean_token_accuracy": 0.6746318936347961, + "num_tokens": 175398988.0, + "step": 6775 + }, + { + "epoch": 0.7441247529101691, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.403531312942505, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7057201862335205, + "num_tokens": 175427102.0, + "step": 6776 + }, + { + "epoch": 0.7442345706127828, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4942336082458496, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7009563446044922, + "num_tokens": 175454098.0, + "step": 6777 + }, + { + "epoch": 0.7443443883153964, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.471608877182007, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6860933303833008, + "num_tokens": 175481485.0, + "step": 6778 + }, + { + "epoch": 0.7444542060180102, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6556332111358643, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7009320259094238, + "num_tokens": 175505092.0, + "step": 6779 + }, + { + "epoch": 0.7445640237206238, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5298945903778076, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6932390928268433, + "num_tokens": 175529187.0, + "step": 6780 + }, + { + "epoch": 0.7446738414232374, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6512250900268555, + "learning_rate": 1e-06, + "loss": 1.1138, + "mean_token_accuracy": 0.6856590509414673, + "num_tokens": 175554714.0, + "step": 6781 + }, + { + "epoch": 0.7447836591258511, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.512356996536255, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6949194073677063, + "num_tokens": 175582796.0, + "step": 6782 + }, + { + "epoch": 0.7448934768284647, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7531356811523438, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6970479488372803, + "num_tokens": 175606081.0, + "step": 6783 + }, + { + "epoch": 0.7450032945310784, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6235272884368896, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7162989974021912, + "num_tokens": 175632000.0, + "step": 6784 + }, + { + "epoch": 0.745113112233692, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.478245258331299, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7105184197425842, + "num_tokens": 175658364.0, + "step": 6785 + }, + { + "epoch": 0.7452229299363057, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6999034881591797, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7060812711715698, + "num_tokens": 175683050.0, + "step": 6786 + }, + { + "epoch": 0.7453327476389194, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6770434379577637, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7119252681732178, + "num_tokens": 175706060.0, + "step": 6787 + }, + { + "epoch": 0.7454425653415331, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3847296237945557, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7114990949630737, + "num_tokens": 175734301.0, + "step": 6788 + }, + { + "epoch": 0.7455523830441467, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.650512218475342, + "learning_rate": 1e-06, + "loss": 0.939, + "mean_token_accuracy": 0.7178425192832947, + "num_tokens": 175757178.0, + "step": 6789 + }, + { + "epoch": 0.7456622007467604, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9482967853546143, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.6834378242492676, + "num_tokens": 175778863.0, + "step": 6790 + }, + { + "epoch": 0.745772018449374, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.493496894836426, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7064350247383118, + "num_tokens": 175806023.0, + "step": 6791 + }, + { + "epoch": 0.7458818361519877, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6944336891174316, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.6991747617721558, + "num_tokens": 175828049.0, + "step": 6792 + }, + { + "epoch": 0.7459916538546013, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.864231586456299, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7048521637916565, + "num_tokens": 175848135.0, + "step": 6793 + }, + { + "epoch": 0.7461014715572151, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3810112476348877, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.697253942489624, + "num_tokens": 175876498.0, + "step": 6794 + }, + { + "epoch": 0.7462112892598287, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4826200008392334, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.6945856809616089, + "num_tokens": 175902733.0, + "step": 6795 + }, + { + "epoch": 0.7463211069624424, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.93438720703125, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7045292854309082, + "num_tokens": 175925566.0, + "step": 6796 + }, + { + "epoch": 0.746430924665056, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4871768951416016, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6827301383018494, + "num_tokens": 175955080.0, + "step": 6797 + }, + { + "epoch": 0.7465407423676697, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.741783857345581, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7138687372207642, + "num_tokens": 175977085.0, + "step": 6798 + }, + { + "epoch": 0.7466505600702833, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.628051996231079, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.702904224395752, + "num_tokens": 176003377.0, + "step": 6799 + }, + { + "epoch": 0.7467603777728969, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.324852228164673, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7033833861351013, + "num_tokens": 176030427.0, + "step": 6800 + }, + { + "epoch": 0.7468701954755107, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.630965232849121, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6825817823410034, + "num_tokens": 176058425.0, + "step": 6801 + }, + { + "epoch": 0.7469800131781243, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5916504859924316, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.728345513343811, + "num_tokens": 176081969.0, + "step": 6802 + }, + { + "epoch": 0.747089830880738, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.472036600112915, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7132216095924377, + "num_tokens": 176106404.0, + "step": 6803 + }, + { + "epoch": 0.7471996485833516, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.441166639328003, + "learning_rate": 1e-06, + "loss": 1.1285, + "mean_token_accuracy": 0.6666237115859985, + "num_tokens": 176137354.0, + "step": 6804 + }, + { + "epoch": 0.7473094662859653, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5947299003601074, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6994645595550537, + "num_tokens": 176162288.0, + "step": 6805 + }, + { + "epoch": 0.7474192839885789, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9564316272735596, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6958709955215454, + "num_tokens": 176189523.0, + "step": 6806 + }, + { + "epoch": 0.7475291016911926, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7508561611175537, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.7000777721405029, + "num_tokens": 176213616.0, + "step": 6807 + }, + { + "epoch": 0.7476389193938063, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5345752239227295, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7183864116668701, + "num_tokens": 176236767.0, + "step": 6808 + }, + { + "epoch": 0.74774873709642, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6007492542266846, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7175787687301636, + "num_tokens": 176263344.0, + "step": 6809 + }, + { + "epoch": 0.7478585547990336, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 4.306359767913818, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7005614638328552, + "num_tokens": 176288047.0, + "step": 6810 + }, + { + "epoch": 0.7479683725016473, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.399552583694458, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.6958432197570801, + "num_tokens": 176317781.0, + "step": 6811 + }, + { + "epoch": 0.7480781902042609, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5051090717315674, + "learning_rate": 1e-06, + "loss": 1.1164, + "mean_token_accuracy": 0.6786552667617798, + "num_tokens": 176347624.0, + "step": 6812 + }, + { + "epoch": 0.7481880079068746, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7594940662384033, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7117759585380554, + "num_tokens": 176367772.0, + "step": 6813 + }, + { + "epoch": 0.7482978256094882, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4840469360351562, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7064892649650574, + "num_tokens": 176400767.0, + "step": 6814 + }, + { + "epoch": 0.7484076433121019, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4409701824188232, + "learning_rate": 1e-06, + "loss": 1.0991, + "mean_token_accuracy": 0.6810364723205566, + "num_tokens": 176425357.0, + "step": 6815 + }, + { + "epoch": 0.7485174610147156, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.793971300125122, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6922354698181152, + "num_tokens": 176445214.0, + "step": 6816 + }, + { + "epoch": 0.7486272787173293, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5666399002075195, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.701127290725708, + "num_tokens": 176472967.0, + "step": 6817 + }, + { + "epoch": 0.7487370964199429, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.398672342300415, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7140028476715088, + "num_tokens": 176502370.0, + "step": 6818 + }, + { + "epoch": 0.7488469141225566, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7163634300231934, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.6999978423118591, + "num_tokens": 176525646.0, + "step": 6819 + }, + { + "epoch": 0.7489567318251702, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 7.912642955780029, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.692216157913208, + "num_tokens": 176552922.0, + "step": 6820 + }, + { + "epoch": 0.7490665495277838, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5709681510925293, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7052081227302551, + "num_tokens": 176580090.0, + "step": 6821 + }, + { + "epoch": 0.7491763672303975, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7097651958465576, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.70256507396698, + "num_tokens": 176607392.0, + "step": 6822 + }, + { + "epoch": 0.7492861849330112, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8044071197509766, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7155988216400146, + "num_tokens": 176631020.0, + "step": 6823 + }, + { + "epoch": 0.7493960026356249, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3465216159820557, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7210138440132141, + "num_tokens": 176659765.0, + "step": 6824 + }, + { + "epoch": 0.7495058203382385, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5085549354553223, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7227644324302673, + "num_tokens": 176688642.0, + "step": 6825 + }, + { + "epoch": 0.7496156380408522, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7822465896606445, + "learning_rate": 1e-06, + "loss": 0.9456, + "mean_token_accuracy": 0.7258874773979187, + "num_tokens": 176712306.0, + "step": 6826 + }, + { + "epoch": 0.7497254557434658, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7071304321289062, + "learning_rate": 1e-06, + "loss": 0.9321, + "mean_token_accuracy": 0.7268017530441284, + "num_tokens": 176731830.0, + "step": 6827 + }, + { + "epoch": 0.7498352734460795, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3435075283050537, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6948248147964478, + "num_tokens": 176762912.0, + "step": 6828 + }, + { + "epoch": 0.7499450911486931, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.58420467376709, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6976489424705505, + "num_tokens": 176787808.0, + "step": 6829 + }, + { + "epoch": 0.7500549088513069, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.8761181831359863, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7063311338424683, + "num_tokens": 176809620.0, + "step": 6830 + }, + { + "epoch": 0.7501647265539205, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 8.049546241760254, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6958936452865601, + "num_tokens": 176838777.0, + "step": 6831 + }, + { + "epoch": 0.7502745442565342, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.0913655757904053, + "learning_rate": 1e-06, + "loss": 1.1334, + "mean_token_accuracy": 0.6785176992416382, + "num_tokens": 176863801.0, + "step": 6832 + }, + { + "epoch": 0.7503843619591478, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.530925989151001, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.7038352489471436, + "num_tokens": 176892098.0, + "step": 6833 + }, + { + "epoch": 0.7504941796617615, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3536040782928467, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7158252596855164, + "num_tokens": 176919806.0, + "step": 6834 + }, + { + "epoch": 0.7506039973643751, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6580278873443604, + "learning_rate": 1e-06, + "loss": 1.0883, + "mean_token_accuracy": 0.6796025037765503, + "num_tokens": 176944799.0, + "step": 6835 + }, + { + "epoch": 0.7507138150669888, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8992671966552734, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6923754215240479, + "num_tokens": 176968499.0, + "step": 6836 + }, + { + "epoch": 0.7508236327696025, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6568171977996826, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7084367871284485, + "num_tokens": 176992760.0, + "step": 6837 + }, + { + "epoch": 0.7509334504722162, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4271297454833984, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6901202201843262, + "num_tokens": 177019679.0, + "step": 6838 + }, + { + "epoch": 0.7510432681748298, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9765090942382812, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6906217336654663, + "num_tokens": 177048441.0, + "step": 6839 + }, + { + "epoch": 0.7511530858774434, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4585864543914795, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.69993656873703, + "num_tokens": 177076460.0, + "step": 6840 + }, + { + "epoch": 0.7512629035800571, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.690638780593872, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6981201171875, + "num_tokens": 177099730.0, + "step": 6841 + }, + { + "epoch": 0.7513727212826707, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.658362865447998, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.696945309638977, + "num_tokens": 177122154.0, + "step": 6842 + }, + { + "epoch": 0.7514825389852844, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.790471315383911, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7092123627662659, + "num_tokens": 177142945.0, + "step": 6843 + }, + { + "epoch": 0.7515923566878981, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5389909744262695, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6982992887496948, + "num_tokens": 177170514.0, + "step": 6844 + }, + { + "epoch": 0.7517021743905118, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5247950553894043, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.6946676969528198, + "num_tokens": 177196408.0, + "step": 6845 + }, + { + "epoch": 0.7518119920931254, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6202239990234375, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.718529462814331, + "num_tokens": 177220300.0, + "step": 6846 + }, + { + "epoch": 0.7519218097957391, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3273162841796875, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.6922721266746521, + "num_tokens": 177251674.0, + "step": 6847 + }, + { + "epoch": 0.7520316274983527, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3159234523773193, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6939581632614136, + "num_tokens": 177284386.0, + "step": 6848 + }, + { + "epoch": 0.7521414452009664, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.340914726257324, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6934198141098022, + "num_tokens": 177314575.0, + "step": 6849 + }, + { + "epoch": 0.75225126290358, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.659087896347046, + "learning_rate": 1e-06, + "loss": 1.0844, + "mean_token_accuracy": 0.6876214742660522, + "num_tokens": 177340530.0, + "step": 6850 + }, + { + "epoch": 0.7523610806061937, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7963125705718994, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.698936402797699, + "num_tokens": 177363114.0, + "step": 6851 + }, + { + "epoch": 0.7524708983088074, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.364452362060547, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7009220123291016, + "num_tokens": 177392813.0, + "step": 6852 + }, + { + "epoch": 0.7525807160114211, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.592219829559326, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7032344341278076, + "num_tokens": 177416349.0, + "step": 6853 + }, + { + "epoch": 0.7526905337140347, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.581570625305176, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6819217205047607, + "num_tokens": 177441313.0, + "step": 6854 + }, + { + "epoch": 0.7528003514166484, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7266783714294434, + "learning_rate": 1e-06, + "loss": 0.852, + "mean_token_accuracy": 0.7419652938842773, + "num_tokens": 177462066.0, + "step": 6855 + }, + { + "epoch": 0.752910169119262, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6520237922668457, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.703421950340271, + "num_tokens": 177488623.0, + "step": 6856 + }, + { + "epoch": 0.7530199868218757, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5958259105682373, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.70368492603302, + "num_tokens": 177517433.0, + "step": 6857 + }, + { + "epoch": 0.7531298045244893, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5293166637420654, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6955747008323669, + "num_tokens": 177546971.0, + "step": 6858 + }, + { + "epoch": 0.7532396222271031, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6964054107666016, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.6838673949241638, + "num_tokens": 177568348.0, + "step": 6859 + }, + { + "epoch": 0.7533494399297167, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.573976516723633, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7048416137695312, + "num_tokens": 177595015.0, + "step": 6860 + }, + { + "epoch": 0.7534592576323303, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.692456007003784, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7060861587524414, + "num_tokens": 177621428.0, + "step": 6861 + }, + { + "epoch": 0.753569075334944, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4801738262176514, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7066301107406616, + "num_tokens": 177648253.0, + "step": 6862 + }, + { + "epoch": 0.7536788930375576, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.0392770767211914, + "learning_rate": 1e-06, + "loss": 0.9166, + "mean_token_accuracy": 0.7288268208503723, + "num_tokens": 177666539.0, + "step": 6863 + }, + { + "epoch": 0.7537887107401713, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6987974643707275, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7131768465042114, + "num_tokens": 177691319.0, + "step": 6864 + }, + { + "epoch": 0.7538985284427849, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4650776386260986, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6968367099761963, + "num_tokens": 177721356.0, + "step": 6865 + }, + { + "epoch": 0.7540083461453987, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3915302753448486, + "learning_rate": 1e-06, + "loss": 1.113, + "mean_token_accuracy": 0.6807376146316528, + "num_tokens": 177753114.0, + "step": 6866 + }, + { + "epoch": 0.7541181638480123, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5274341106414795, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.7190468311309814, + "num_tokens": 177777685.0, + "step": 6867 + }, + { + "epoch": 0.754227981550626, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.504417896270752, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7048587799072266, + "num_tokens": 177805844.0, + "step": 6868 + }, + { + "epoch": 0.7543377992532396, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.620009660720825, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6916782259941101, + "num_tokens": 177835153.0, + "step": 6869 + }, + { + "epoch": 0.7544476169558533, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5507588386535645, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7073808908462524, + "num_tokens": 177861316.0, + "step": 6870 + }, + { + "epoch": 0.7545574346584669, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4593496322631836, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7064303755760193, + "num_tokens": 177888119.0, + "step": 6871 + }, + { + "epoch": 0.7546672523610806, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.484234571456909, + "learning_rate": 1e-06, + "loss": 1.1046, + "mean_token_accuracy": 0.6779784560203552, + "num_tokens": 177920001.0, + "step": 6872 + }, + { + "epoch": 0.7547770700636943, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.0694150924682617, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7064728140830994, + "num_tokens": 177938096.0, + "step": 6873 + }, + { + "epoch": 0.754886887766308, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6207687854766846, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6946598291397095, + "num_tokens": 177961331.0, + "step": 6874 + }, + { + "epoch": 0.7549967054689216, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.595379590988159, + "learning_rate": 1e-06, + "loss": 0.9158, + "mean_token_accuracy": 0.7239288091659546, + "num_tokens": 177985902.0, + "step": 6875 + }, + { + "epoch": 0.7551065231715353, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.875119686126709, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6835429668426514, + "num_tokens": 178011077.0, + "step": 6876 + }, + { + "epoch": 0.7552163408741489, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.834474563598633, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7125977873802185, + "num_tokens": 178033075.0, + "step": 6877 + }, + { + "epoch": 0.7553261585767626, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5410289764404297, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.69820636510849, + "num_tokens": 178060752.0, + "step": 6878 + }, + { + "epoch": 0.7554359762793762, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3776912689208984, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7202122211456299, + "num_tokens": 178089622.0, + "step": 6879 + }, + { + "epoch": 0.7555457939819898, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.424586296081543, + "learning_rate": 1e-06, + "loss": 1.1175, + "mean_token_accuracy": 0.6724720001220703, + "num_tokens": 178120948.0, + "step": 6880 + }, + { + "epoch": 0.7556556116846036, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.000802516937256, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6854361295700073, + "num_tokens": 178143058.0, + "step": 6881 + }, + { + "epoch": 0.7557654293872172, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6142971515655518, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6984882354736328, + "num_tokens": 178166367.0, + "step": 6882 + }, + { + "epoch": 0.7558752470898309, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.956156015396118, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7109761834144592, + "num_tokens": 178185632.0, + "step": 6883 + }, + { + "epoch": 0.7559850647924445, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5842392444610596, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6902629137039185, + "num_tokens": 178210509.0, + "step": 6884 + }, + { + "epoch": 0.7560948824950582, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.435425043106079, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6899730563163757, + "num_tokens": 178246098.0, + "step": 6885 + }, + { + "epoch": 0.7562047001976718, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5951859951019287, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.699921727180481, + "num_tokens": 178269653.0, + "step": 6886 + }, + { + "epoch": 0.7563145179002855, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.504302978515625, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6999878883361816, + "num_tokens": 178297507.0, + "step": 6887 + }, + { + "epoch": 0.7564243356028992, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4641547203063965, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7083436250686646, + "num_tokens": 178322444.0, + "step": 6888 + }, + { + "epoch": 0.7565341533055129, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8498964309692383, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6856081485748291, + "num_tokens": 178345073.0, + "step": 6889 + }, + { + "epoch": 0.7566439710081265, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6022279262542725, + "learning_rate": 1e-06, + "loss": 1.0855, + "mean_token_accuracy": 0.6888031959533691, + "num_tokens": 178373921.0, + "step": 6890 + }, + { + "epoch": 0.7567537887107402, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3403842449188232, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.7010492086410522, + "num_tokens": 178401628.0, + "step": 6891 + }, + { + "epoch": 0.7568636064133538, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.556938648223877, + "learning_rate": 1e-06, + "loss": 0.9162, + "mean_token_accuracy": 0.7267862558364868, + "num_tokens": 178424561.0, + "step": 6892 + }, + { + "epoch": 0.7569734241159675, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.52717924118042, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7251927852630615, + "num_tokens": 178447980.0, + "step": 6893 + }, + { + "epoch": 0.7570832418185811, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5374696254730225, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.711732029914856, + "num_tokens": 178471718.0, + "step": 6894 + }, + { + "epoch": 0.7571930595211949, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.6077077388763428, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.6989078521728516, + "num_tokens": 178496847.0, + "step": 6895 + }, + { + "epoch": 0.7573028772238085, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6290299892425537, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6839131116867065, + "num_tokens": 178527000.0, + "step": 6896 + }, + { + "epoch": 0.7574126949264222, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7561895847320557, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6981487274169922, + "num_tokens": 178550133.0, + "step": 6897 + }, + { + "epoch": 0.7575225126290358, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.592867374420166, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6970826387405396, + "num_tokens": 178577050.0, + "step": 6898 + }, + { + "epoch": 0.7576323303316495, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4795100688934326, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.720233142375946, + "num_tokens": 178602725.0, + "step": 6899 + }, + { + "epoch": 0.7577421480342631, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3359785079956055, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7088984251022339, + "num_tokens": 178633785.0, + "step": 6900 + }, + { + "epoch": 0.7578519657368767, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.811039447784424, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7029231786727905, + "num_tokens": 178654144.0, + "step": 6901 + }, + { + "epoch": 0.7579617834394905, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5298311710357666, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6971141695976257, + "num_tokens": 178680025.0, + "step": 6902 + }, + { + "epoch": 0.7580716011421041, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.015693187713623, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6970032453536987, + "num_tokens": 178702119.0, + "step": 6903 + }, + { + "epoch": 0.7581814188447178, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.424287796020508, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.69395512342453, + "num_tokens": 178731672.0, + "step": 6904 + }, + { + "epoch": 0.7582912365473314, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3998425006866455, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7221399545669556, + "num_tokens": 178758582.0, + "step": 6905 + }, + { + "epoch": 0.7584010542499451, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3600661754608154, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.684913158416748, + "num_tokens": 178787672.0, + "step": 6906 + }, + { + "epoch": 0.7585108719525587, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4940245151519775, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6826188564300537, + "num_tokens": 178814095.0, + "step": 6907 + }, + { + "epoch": 0.7586206896551724, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.398507595062256, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6887482404708862, + "num_tokens": 178843461.0, + "step": 6908 + }, + { + "epoch": 0.758730507357786, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4865875244140625, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6970468163490295, + "num_tokens": 178870250.0, + "step": 6909 + }, + { + "epoch": 0.7588403250603998, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.6661477088928223, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7144270539283752, + "num_tokens": 178893417.0, + "step": 6910 + }, + { + "epoch": 0.7589501427630134, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7195072174072266, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.6984753608703613, + "num_tokens": 178915244.0, + "step": 6911 + }, + { + "epoch": 0.7590599604656271, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.654430389404297, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6968701481819153, + "num_tokens": 178938120.0, + "step": 6912 + }, + { + "epoch": 0.7591697781682407, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4526402950286865, + "learning_rate": 1e-06, + "loss": 1.1213, + "mean_token_accuracy": 0.678901731967926, + "num_tokens": 178965558.0, + "step": 6913 + }, + { + "epoch": 0.7592795958708544, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3381407260894775, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7114983797073364, + "num_tokens": 178994111.0, + "step": 6914 + }, + { + "epoch": 0.759389413573468, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.712890863418579, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7001693844795227, + "num_tokens": 179017532.0, + "step": 6915 + }, + { + "epoch": 0.7594992312760817, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9266130924224854, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.6972252130508423, + "num_tokens": 179036660.0, + "step": 6916 + }, + { + "epoch": 0.7596090489786954, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.1936943531036377, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.7006328105926514, + "num_tokens": 179064753.0, + "step": 6917 + }, + { + "epoch": 0.7597188666813091, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.609272003173828, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7131510972976685, + "num_tokens": 179087979.0, + "step": 6918 + }, + { + "epoch": 0.7598286843839227, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 20.557762145996094, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.696406364440918, + "num_tokens": 179113525.0, + "step": 6919 + }, + { + "epoch": 0.7599385020865363, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5956015586853027, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7125359773635864, + "num_tokens": 179140003.0, + "step": 6920 + }, + { + "epoch": 0.76004831978915, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5939571857452393, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7183048129081726, + "num_tokens": 179164191.0, + "step": 6921 + }, + { + "epoch": 0.7601581374917636, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5304501056671143, + "learning_rate": 1e-06, + "loss": 1.1056, + "mean_token_accuracy": 0.680127739906311, + "num_tokens": 179194456.0, + "step": 6922 + }, + { + "epoch": 0.7602679551943773, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6798696517944336, + "learning_rate": 1e-06, + "loss": 0.9197, + "mean_token_accuracy": 0.7226271033287048, + "num_tokens": 179220806.0, + "step": 6923 + }, + { + "epoch": 0.760377772896991, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7654614448547363, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6977918148040771, + "num_tokens": 179243326.0, + "step": 6924 + }, + { + "epoch": 0.7604875905996047, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.44992733001709, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.7039582133293152, + "num_tokens": 179270466.0, + "step": 6925 + }, + { + "epoch": 0.7605974083022183, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.2821571826934814, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7149922251701355, + "num_tokens": 179300590.0, + "step": 6926 + }, + { + "epoch": 0.760707226004832, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5704667568206787, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6827288866043091, + "num_tokens": 179324837.0, + "step": 6927 + }, + { + "epoch": 0.7608170437074456, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5052080154418945, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6945207118988037, + "num_tokens": 179356085.0, + "step": 6928 + }, + { + "epoch": 0.7609268614100593, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.676210403442383, + "learning_rate": 1e-06, + "loss": 1.1299, + "mean_token_accuracy": 0.6758967638015747, + "num_tokens": 179380943.0, + "step": 6929 + }, + { + "epoch": 0.7610366791126729, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.0857045650482178, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6948287487030029, + "num_tokens": 179406636.0, + "step": 6930 + }, + { + "epoch": 0.7611464968152867, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4610342979431152, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6927435398101807, + "num_tokens": 179432128.0, + "step": 6931 + }, + { + "epoch": 0.7612563145179003, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5880374908447266, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6972894668579102, + "num_tokens": 179457591.0, + "step": 6932 + }, + { + "epoch": 0.761366132220514, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5391294956207275, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7074206471443176, + "num_tokens": 179483646.0, + "step": 6933 + }, + { + "epoch": 0.7614759499231276, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3897831439971924, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7057522535324097, + "num_tokens": 179511204.0, + "step": 6934 + }, + { + "epoch": 0.7615857676257413, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.542119264602661, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7014543414115906, + "num_tokens": 179536067.0, + "step": 6935 + }, + { + "epoch": 0.7616955853283549, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.3955070972442627, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7307983040809631, + "num_tokens": 179563707.0, + "step": 6936 + }, + { + "epoch": 0.7618054030309686, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.51882004737854, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.718777596950531, + "num_tokens": 179587191.0, + "step": 6937 + }, + { + "epoch": 0.7619152207335822, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5142359733581543, + "learning_rate": 1e-06, + "loss": 0.8891, + "mean_token_accuracy": 0.7300387620925903, + "num_tokens": 179610040.0, + "step": 6938 + }, + { + "epoch": 0.762025038436196, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.3692784309387207, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6977174878120422, + "num_tokens": 179635594.0, + "step": 6939 + }, + { + "epoch": 0.7621348561388096, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.236182928085327, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7084137797355652, + "num_tokens": 179657892.0, + "step": 6940 + }, + { + "epoch": 0.7622446738414232, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4967586994171143, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.7234745621681213, + "num_tokens": 179683394.0, + "step": 6941 + }, + { + "epoch": 0.7623544915440369, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.4314074516296387, + "learning_rate": 1e-06, + "loss": 1.1116, + "mean_token_accuracy": 0.6809301376342773, + "num_tokens": 179711403.0, + "step": 6942 + }, + { + "epoch": 0.7624643092466505, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.8954601287841797, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.71535325050354, + "num_tokens": 179730406.0, + "step": 6943 + }, + { + "epoch": 0.7625741269492642, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.341869354248047, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6882115602493286, + "num_tokens": 179761665.0, + "step": 6944 + }, + { + "epoch": 0.7626839446518778, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.5630006790161133, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7061024904251099, + "num_tokens": 179784254.0, + "step": 6945 + }, + { + "epoch": 0.7627937623544916, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.7607123851776123, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6988123655319214, + "num_tokens": 179807077.0, + "step": 6946 + }, + { + "epoch": 0.7629035800571052, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 2.249213457107544, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7123554944992065, + "num_tokens": 179838235.0, + "step": 6947 + }, + { + "epoch": 0.7630133977597189, + "ewc_loss": 7.808208465576172e-06, + "grad_norm": 3.137380361557007, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7013320922851562, + "num_tokens": 179866103.0, + "step": 6948 + }, + { + "epoch": 0.7631232154623325, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.616629123687744, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7161424160003662, + "num_tokens": 179888382.0, + "step": 6949 + }, + { + "epoch": 0.7632330331649462, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.718446731567383, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.707503080368042, + "num_tokens": 179908263.0, + "step": 6950 + }, + { + "epoch": 0.7633428508675598, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5610902309417725, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7104914784431458, + "num_tokens": 179935433.0, + "step": 6951 + }, + { + "epoch": 0.7634526685701735, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4671738147735596, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6958710551261902, + "num_tokens": 179961446.0, + "step": 6952 + }, + { + "epoch": 0.7635624862727872, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4619827270507812, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7144430875778198, + "num_tokens": 179987906.0, + "step": 6953 + }, + { + "epoch": 0.7636723039754009, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7835640907287598, + "learning_rate": 1e-06, + "loss": 1.122, + "mean_token_accuracy": 0.6699331998825073, + "num_tokens": 180011551.0, + "step": 6954 + }, + { + "epoch": 0.7637821216780145, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6298022270202637, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7120821475982666, + "num_tokens": 180037809.0, + "step": 6955 + }, + { + "epoch": 0.7638919393806282, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4590988159179688, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7090530395507812, + "num_tokens": 180063827.0, + "step": 6956 + }, + { + "epoch": 0.7640017570832418, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.506690740585327, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6886998414993286, + "num_tokens": 180089028.0, + "step": 6957 + }, + { + "epoch": 0.7641115747858555, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6175692081451416, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7063766717910767, + "num_tokens": 180113074.0, + "step": 6958 + }, + { + "epoch": 0.7642213924884691, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6635611057281494, + "learning_rate": 1e-06, + "loss": 1.0745, + "mean_token_accuracy": 0.6899547576904297, + "num_tokens": 180140288.0, + "step": 6959 + }, + { + "epoch": 0.7643312101910829, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.494992971420288, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6877613663673401, + "num_tokens": 180165848.0, + "step": 6960 + }, + { + "epoch": 0.7644410278936965, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4550960063934326, + "learning_rate": 1e-06, + "loss": 1.1275, + "mean_token_accuracy": 0.6768757700920105, + "num_tokens": 180193944.0, + "step": 6961 + }, + { + "epoch": 0.7645508455963101, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6646695137023926, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.699401319026947, + "num_tokens": 180217722.0, + "step": 6962 + }, + { + "epoch": 0.7646606632989238, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5138931274414062, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7182457447052002, + "num_tokens": 180242650.0, + "step": 6963 + }, + { + "epoch": 0.7647704810015374, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.230083465576172, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7281737327575684, + "num_tokens": 180272453.0, + "step": 6964 + }, + { + "epoch": 0.7648802987041511, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.333907127380371, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7136791944503784, + "num_tokens": 180300422.0, + "step": 6965 + }, + { + "epoch": 0.7649901164067647, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6796436309814453, + "learning_rate": 1e-06, + "loss": 0.9475, + "mean_token_accuracy": 0.7237852811813354, + "num_tokens": 180322956.0, + "step": 6966 + }, + { + "epoch": 0.7650999341093784, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.474979877471924, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.6980841159820557, + "num_tokens": 180348589.0, + "step": 6967 + }, + { + "epoch": 0.7652097518119921, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8013837337493896, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7006757259368896, + "num_tokens": 180369574.0, + "step": 6968 + }, + { + "epoch": 0.7653195695146058, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.683027982711792, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6992524862289429, + "num_tokens": 180391533.0, + "step": 6969 + }, + { + "epoch": 0.7654293872172194, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.598578691482544, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6821516156196594, + "num_tokens": 180419677.0, + "step": 6970 + }, + { + "epoch": 0.7655392049198331, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.480625629425049, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6830692887306213, + "num_tokens": 180449792.0, + "step": 6971 + }, + { + "epoch": 0.7656490226224467, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.899704694747925, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7026606798171997, + "num_tokens": 180470960.0, + "step": 6972 + }, + { + "epoch": 0.7657588403250604, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3602285385131836, + "learning_rate": 1e-06, + "loss": 1.1066, + "mean_token_accuracy": 0.6758769154548645, + "num_tokens": 180501704.0, + "step": 6973 + }, + { + "epoch": 0.765868658027674, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7693755626678467, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.6987624168395996, + "num_tokens": 180522239.0, + "step": 6974 + }, + { + "epoch": 0.7659784757302878, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.582197666168213, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.7010515332221985, + "num_tokens": 180545659.0, + "step": 6975 + }, + { + "epoch": 0.7660882934329014, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5060131549835205, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.719919741153717, + "num_tokens": 180570244.0, + "step": 6976 + }, + { + "epoch": 0.7661981111355151, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.520862579345703, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7164362668991089, + "num_tokens": 180597361.0, + "step": 6977 + }, + { + "epoch": 0.7663079288381287, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.544489860534668, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.695290744304657, + "num_tokens": 180622400.0, + "step": 6978 + }, + { + "epoch": 0.7664177465407424, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6055476665496826, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7032851576805115, + "num_tokens": 180649632.0, + "step": 6979 + }, + { + "epoch": 0.766527564243356, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3560001850128174, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7119026780128479, + "num_tokens": 180678695.0, + "step": 6980 + }, + { + "epoch": 0.7666373819459696, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.803957462310791, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.716594934463501, + "num_tokens": 180700718.0, + "step": 6981 + }, + { + "epoch": 0.7667471996485834, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.728097438812256, + "learning_rate": 1e-06, + "loss": 0.9186, + "mean_token_accuracy": 0.7257276177406311, + "num_tokens": 180723553.0, + "step": 6982 + }, + { + "epoch": 0.766857017351197, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.598996639251709, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7094196081161499, + "num_tokens": 180748214.0, + "step": 6983 + }, + { + "epoch": 0.7669668350538107, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.523667097091675, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6998263001441956, + "num_tokens": 180773817.0, + "step": 6984 + }, + { + "epoch": 0.7670766527564243, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9295225143432617, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7046700716018677, + "num_tokens": 180794784.0, + "step": 6985 + }, + { + "epoch": 0.767186470459038, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7716455459594727, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6983440518379211, + "num_tokens": 180815988.0, + "step": 6986 + }, + { + "epoch": 0.7672962881616516, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.42169451713562, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6928544640541077, + "num_tokens": 180842777.0, + "step": 6987 + }, + { + "epoch": 0.7674061058642653, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.643552780151367, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7082681059837341, + "num_tokens": 180868011.0, + "step": 6988 + }, + { + "epoch": 0.767515923566879, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3048481941223145, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6869779229164124, + "num_tokens": 180899163.0, + "step": 6989 + }, + { + "epoch": 0.7676257412694927, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4901225566864014, + "learning_rate": 1e-06, + "loss": 1.1491, + "mean_token_accuracy": 0.6682901978492737, + "num_tokens": 180926392.0, + "step": 6990 + }, + { + "epoch": 0.7677355589721063, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.576371192932129, + "learning_rate": 1e-06, + "loss": 1.0799, + "mean_token_accuracy": 0.6923480033874512, + "num_tokens": 180951585.0, + "step": 6991 + }, + { + "epoch": 0.76784537667472, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6313109397888184, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7027250528335571, + "num_tokens": 180975363.0, + "step": 6992 + }, + { + "epoch": 0.7679551943773336, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.63845157623291, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6872862577438354, + "num_tokens": 181003116.0, + "step": 6993 + }, + { + "epoch": 0.7680650120799473, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4728736877441406, + "learning_rate": 1e-06, + "loss": 1.1039, + "mean_token_accuracy": 0.6862276792526245, + "num_tokens": 181029285.0, + "step": 6994 + }, + { + "epoch": 0.7681748297825609, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.495959997177124, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7087253332138062, + "num_tokens": 181053653.0, + "step": 6995 + }, + { + "epoch": 0.7682846474851747, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5402276515960693, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.702469527721405, + "num_tokens": 181079899.0, + "step": 6996 + }, + { + "epoch": 0.7683944651877883, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.554898738861084, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6887383460998535, + "num_tokens": 181105803.0, + "step": 6997 + }, + { + "epoch": 0.768504282890402, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.502068281173706, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.7019908428192139, + "num_tokens": 181132456.0, + "step": 6998 + }, + { + "epoch": 0.7686141005930156, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7900145053863525, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6897433996200562, + "num_tokens": 181155100.0, + "step": 6999 + }, + { + "epoch": 0.7687239182956292, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.756957769393921, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7029455304145813, + "num_tokens": 181176842.0, + "step": 7000 + }, + { + "epoch": 0.7688337359982429, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.621137857437134, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.6961244344711304, + "num_tokens": 181200229.0, + "step": 7001 + }, + { + "epoch": 0.7689435537008565, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4275403022766113, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6991506814956665, + "num_tokens": 181229968.0, + "step": 7002 + }, + { + "epoch": 0.7690533714034702, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.635066032409668, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.692255973815918, + "num_tokens": 181256302.0, + "step": 7003 + }, + { + "epoch": 0.7691631891060839, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.687544345855713, + "learning_rate": 1e-06, + "loss": 0.9332, + "mean_token_accuracy": 0.7268086671829224, + "num_tokens": 181286191.0, + "step": 7004 + }, + { + "epoch": 0.7692730068086976, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7829766273498535, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6989948749542236, + "num_tokens": 181307029.0, + "step": 7005 + }, + { + "epoch": 0.7693828245113112, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6787807941436768, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.700464129447937, + "num_tokens": 181332716.0, + "step": 7006 + }, + { + "epoch": 0.7694926422139249, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5012946128845215, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.702334463596344, + "num_tokens": 181361598.0, + "step": 7007 + }, + { + "epoch": 0.7696024599165385, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.682798147201538, + "learning_rate": 1e-06, + "loss": 1.0754, + "mean_token_accuracy": 0.6867367029190063, + "num_tokens": 181385434.0, + "step": 7008 + }, + { + "epoch": 0.7697122776191522, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5467309951782227, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.692571759223938, + "num_tokens": 181411830.0, + "step": 7009 + }, + { + "epoch": 0.7698220953217658, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3946049213409424, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.6965713500976562, + "num_tokens": 181442415.0, + "step": 7010 + }, + { + "epoch": 0.7699319130243796, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5299534797668457, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.6975850462913513, + "num_tokens": 181466967.0, + "step": 7011 + }, + { + "epoch": 0.7700417307269932, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9808387756347656, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7134305834770203, + "num_tokens": 181487298.0, + "step": 7012 + }, + { + "epoch": 0.7701515484296069, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8628036975860596, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.7329508066177368, + "num_tokens": 181507506.0, + "step": 7013 + }, + { + "epoch": 0.7702613661322205, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5005886554718018, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.7005756497383118, + "num_tokens": 181536470.0, + "step": 7014 + }, + { + "epoch": 0.7703711838348342, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.43953013420105, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.7029115557670593, + "num_tokens": 181562596.0, + "step": 7015 + }, + { + "epoch": 0.7704810015374478, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5084218978881836, + "learning_rate": 1e-06, + "loss": 1.0883, + "mean_token_accuracy": 0.6861944198608398, + "num_tokens": 181591441.0, + "step": 7016 + }, + { + "epoch": 0.7705908192400615, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.67854905128479, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.6940985918045044, + "num_tokens": 181617441.0, + "step": 7017 + }, + { + "epoch": 0.7707006369426752, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4763214588165283, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.7076637744903564, + "num_tokens": 181644105.0, + "step": 7018 + }, + { + "epoch": 0.7708104546452889, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5337672233581543, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7079067826271057, + "num_tokens": 181669368.0, + "step": 7019 + }, + { + "epoch": 0.7709202723479025, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7133970260620117, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.700096607208252, + "num_tokens": 181690185.0, + "step": 7020 + }, + { + "epoch": 0.7710300900505161, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.399082899093628, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.712445855140686, + "num_tokens": 181717010.0, + "step": 7021 + }, + { + "epoch": 0.7711399077531298, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.542274236679077, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6865096092224121, + "num_tokens": 181742561.0, + "step": 7022 + }, + { + "epoch": 0.7712497254557434, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6544361114501953, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6858683824539185, + "num_tokens": 181766667.0, + "step": 7023 + }, + { + "epoch": 0.7713595431583571, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5739972591400146, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7255235314369202, + "num_tokens": 181792346.0, + "step": 7024 + }, + { + "epoch": 0.7714693608609708, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.760072946548462, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6783949136734009, + "num_tokens": 181817478.0, + "step": 7025 + }, + { + "epoch": 0.7715791785635845, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.666809558868408, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.6887502074241638, + "num_tokens": 181841823.0, + "step": 7026 + }, + { + "epoch": 0.7716889962661981, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7047834396362305, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.6958752870559692, + "num_tokens": 181863086.0, + "step": 7027 + }, + { + "epoch": 0.7717988139688118, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7875711917877197, + "learning_rate": 1e-06, + "loss": 1.1167, + "mean_token_accuracy": 0.6799672842025757, + "num_tokens": 181888119.0, + "step": 7028 + }, + { + "epoch": 0.7719086316714254, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4562368392944336, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.6884435415267944, + "num_tokens": 181914649.0, + "step": 7029 + }, + { + "epoch": 0.7720184493740391, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6280364990234375, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.6941003799438477, + "num_tokens": 181940137.0, + "step": 7030 + }, + { + "epoch": 0.7721282670766527, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.668614387512207, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7210642695426941, + "num_tokens": 181963552.0, + "step": 7031 + }, + { + "epoch": 0.7722380847792664, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8536605834960938, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7325469255447388, + "num_tokens": 181983514.0, + "step": 7032 + }, + { + "epoch": 0.7723479024818801, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.355947494506836, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7060092091560364, + "num_tokens": 182012525.0, + "step": 7033 + }, + { + "epoch": 0.7724577201844938, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.621703863143921, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6859422922134399, + "num_tokens": 182036340.0, + "step": 7034 + }, + { + "epoch": 0.7725675378871074, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9216041564941406, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7111033201217651, + "num_tokens": 182055692.0, + "step": 7035 + }, + { + "epoch": 0.7726773555897211, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3886847496032715, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7036786675453186, + "num_tokens": 182084359.0, + "step": 7036 + }, + { + "epoch": 0.7727871732923347, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.820298910140991, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6840322017669678, + "num_tokens": 182108290.0, + "step": 7037 + }, + { + "epoch": 0.7728969909949484, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.360252857208252, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.6843596696853638, + "num_tokens": 182137242.0, + "step": 7038 + }, + { + "epoch": 0.773006808697562, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5092737674713135, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.6867722272872925, + "num_tokens": 182166149.0, + "step": 7039 + }, + { + "epoch": 0.7731166264001758, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5854010581970215, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7092865705490112, + "num_tokens": 182189087.0, + "step": 7040 + }, + { + "epoch": 0.7732264441027894, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.701544761657715, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6881794929504395, + "num_tokens": 182213522.0, + "step": 7041 + }, + { + "epoch": 0.773336261805403, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7704100608825684, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6794146299362183, + "num_tokens": 182236627.0, + "step": 7042 + }, + { + "epoch": 0.7734460795080167, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7659804821014404, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7248932123184204, + "num_tokens": 182259748.0, + "step": 7043 + }, + { + "epoch": 0.7735558972106303, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.471081256866455, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6994752883911133, + "num_tokens": 182288135.0, + "step": 7044 + }, + { + "epoch": 0.773665714913244, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.575791597366333, + "learning_rate": 1e-06, + "loss": 1.0679, + "mean_token_accuracy": 0.688499391078949, + "num_tokens": 182313404.0, + "step": 7045 + }, + { + "epoch": 0.7737755326158576, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.390800952911377, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7069925665855408, + "num_tokens": 182342822.0, + "step": 7046 + }, + { + "epoch": 0.7738853503184714, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.932811975479126, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.720973551273346, + "num_tokens": 182360758.0, + "step": 7047 + }, + { + "epoch": 0.773995168021085, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3935086727142334, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6908023357391357, + "num_tokens": 182393280.0, + "step": 7048 + }, + { + "epoch": 0.7741049857236987, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.569469690322876, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6986560225486755, + "num_tokens": 182417912.0, + "step": 7049 + }, + { + "epoch": 0.7742148034263123, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5918784141540527, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6744333505630493, + "num_tokens": 182442234.0, + "step": 7050 + }, + { + "epoch": 0.774324621128926, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6074326038360596, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7114425897598267, + "num_tokens": 182466737.0, + "step": 7051 + }, + { + "epoch": 0.7744344388315396, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.573004961013794, + "learning_rate": 1e-06, + "loss": 1.1021, + "mean_token_accuracy": 0.6844088435173035, + "num_tokens": 182494786.0, + "step": 7052 + }, + { + "epoch": 0.7745442565341533, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5768892765045166, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7094988822937012, + "num_tokens": 182517728.0, + "step": 7053 + }, + { + "epoch": 0.774654074236767, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8501038551330566, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7168223261833191, + "num_tokens": 182539508.0, + "step": 7054 + }, + { + "epoch": 0.7747638919393807, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5685296058654785, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.688282310962677, + "num_tokens": 182565504.0, + "step": 7055 + }, + { + "epoch": 0.7748737096419943, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4364864826202393, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6901128888130188, + "num_tokens": 182593815.0, + "step": 7056 + }, + { + "epoch": 0.774983527344608, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.43564510345459, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6960365772247314, + "num_tokens": 182621249.0, + "step": 7057 + }, + { + "epoch": 0.7750933450472216, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6107401847839355, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7169322967529297, + "num_tokens": 182645287.0, + "step": 7058 + }, + { + "epoch": 0.7752031627498353, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5578627586364746, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6930853128433228, + "num_tokens": 182670630.0, + "step": 7059 + }, + { + "epoch": 0.7753129804524489, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 16.712234497070312, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7229315042495728, + "num_tokens": 182693031.0, + "step": 7060 + }, + { + "epoch": 0.7754227981550625, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 20.54537582397461, + "learning_rate": 1e-06, + "loss": 1.1341, + "mean_token_accuracy": 0.6748978495597839, + "num_tokens": 182720336.0, + "step": 7061 + }, + { + "epoch": 0.7755326158576763, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8721864223480225, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6991443634033203, + "num_tokens": 182744926.0, + "step": 7062 + }, + { + "epoch": 0.77564243356029, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 3.1552326679229736, + "learning_rate": 1e-06, + "loss": 1.1245, + "mean_token_accuracy": 0.677260160446167, + "num_tokens": 182771478.0, + "step": 7063 + }, + { + "epoch": 0.7757522512629036, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9048714637756348, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6971919536590576, + "num_tokens": 182795271.0, + "step": 7064 + }, + { + "epoch": 0.7758620689655172, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.486180067062378, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6903082728385925, + "num_tokens": 182820752.0, + "step": 7065 + }, + { + "epoch": 0.7759718866681309, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3751778602600098, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6857352256774902, + "num_tokens": 182850623.0, + "step": 7066 + }, + { + "epoch": 0.7760817043707445, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4584121704101562, + "learning_rate": 1e-06, + "loss": 1.1195, + "mean_token_accuracy": 0.680019199848175, + "num_tokens": 182876812.0, + "step": 7067 + }, + { + "epoch": 0.7761915220733582, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4755868911743164, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7207561135292053, + "num_tokens": 182901447.0, + "step": 7068 + }, + { + "epoch": 0.7763013397759719, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8241965770721436, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6919285655021667, + "num_tokens": 182922621.0, + "step": 7069 + }, + { + "epoch": 0.7764111574785856, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.561305522918701, + "learning_rate": 1e-06, + "loss": 0.9552, + "mean_token_accuracy": 0.7148823738098145, + "num_tokens": 182945907.0, + "step": 7070 + }, + { + "epoch": 0.7765209751811992, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7672126293182373, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.7030298709869385, + "num_tokens": 182967961.0, + "step": 7071 + }, + { + "epoch": 0.7766307928838129, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.496333122253418, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6978826522827148, + "num_tokens": 182997028.0, + "step": 7072 + }, + { + "epoch": 0.7767406105864265, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9986681938171387, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7144224047660828, + "num_tokens": 183020275.0, + "step": 7073 + }, + { + "epoch": 0.7768504282890402, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.714948892593384, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.703540027141571, + "num_tokens": 183042996.0, + "step": 7074 + }, + { + "epoch": 0.7769602459916538, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.651102304458618, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6982351541519165, + "num_tokens": 183069469.0, + "step": 7075 + }, + { + "epoch": 0.7770700636942676, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.9350337982177734, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6961055994033813, + "num_tokens": 183094259.0, + "step": 7076 + }, + { + "epoch": 0.7771798813968812, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.585390567779541, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7303484082221985, + "num_tokens": 183114980.0, + "step": 7077 + }, + { + "epoch": 0.7772896990994949, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.648481607437134, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.7021500468254089, + "num_tokens": 183140267.0, + "step": 7078 + }, + { + "epoch": 0.7773995168021085, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.632256507873535, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7155870199203491, + "num_tokens": 183165152.0, + "step": 7079 + }, + { + "epoch": 0.7775093345047221, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6294238567352295, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7075308561325073, + "num_tokens": 183192657.0, + "step": 7080 + }, + { + "epoch": 0.7776191522073358, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.304198980331421, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.7332215905189514, + "num_tokens": 183221786.0, + "step": 7081 + }, + { + "epoch": 0.7777289699099494, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 16.66400909423828, + "learning_rate": 1e-06, + "loss": 0.9403, + "mean_token_accuracy": 0.7232668399810791, + "num_tokens": 183248432.0, + "step": 7082 + }, + { + "epoch": 0.7778387876125632, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.501600980758667, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7080479860305786, + "num_tokens": 183276389.0, + "step": 7083 + }, + { + "epoch": 0.7779486053151768, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5291712284088135, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.7075915336608887, + "num_tokens": 183302609.0, + "step": 7084 + }, + { + "epoch": 0.7780584230177905, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.317110061645508, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6910156011581421, + "num_tokens": 183336102.0, + "step": 7085 + }, + { + "epoch": 0.7781682407204041, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.688293695449829, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6927303671836853, + "num_tokens": 183365553.0, + "step": 7086 + }, + { + "epoch": 0.7782780584230178, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5720267295837402, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.692190408706665, + "num_tokens": 183391577.0, + "step": 7087 + }, + { + "epoch": 0.7783878761256314, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5049972534179688, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6893072128295898, + "num_tokens": 183416801.0, + "step": 7088 + }, + { + "epoch": 0.7784976938282451, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.957679033279419, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7300738096237183, + "num_tokens": 183435573.0, + "step": 7089 + }, + { + "epoch": 0.7786075115308587, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.338109254837036, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6852765083312988, + "num_tokens": 183467300.0, + "step": 7090 + }, + { + "epoch": 0.7787173292334725, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.677915096282959, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7019489407539368, + "num_tokens": 183490195.0, + "step": 7091 + }, + { + "epoch": 0.7788271469360861, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.65368914604187, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.6982558965682983, + "num_tokens": 183514431.0, + "step": 7092 + }, + { + "epoch": 0.7789369646386998, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.458235740661621, + "learning_rate": 1e-06, + "loss": 0.9099, + "mean_token_accuracy": 0.7375066876411438, + "num_tokens": 183539682.0, + "step": 7093 + }, + { + "epoch": 0.7790467823413134, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.876134157180786, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7120928764343262, + "num_tokens": 183562472.0, + "step": 7094 + }, + { + "epoch": 0.7791566000439271, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.573326349258423, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7022544145584106, + "num_tokens": 183589502.0, + "step": 7095 + }, + { + "epoch": 0.7792664177465407, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8943378925323486, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6962625980377197, + "num_tokens": 183615986.0, + "step": 7096 + }, + { + "epoch": 0.7793762354491544, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.3446078300476074, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6885379552841187, + "num_tokens": 183642905.0, + "step": 7097 + }, + { + "epoch": 0.7794860531517681, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.4447383880615234, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7106665372848511, + "num_tokens": 183664664.0, + "step": 7098 + }, + { + "epoch": 0.7795958708543818, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5821897983551025, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.7243573069572449, + "num_tokens": 183688895.0, + "step": 7099 + }, + { + "epoch": 0.7797056885569954, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4183945655822754, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6928304433822632, + "num_tokens": 183716026.0, + "step": 7100 + }, + { + "epoch": 0.779815506259609, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.336925506591797, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7026584148406982, + "num_tokens": 183744988.0, + "step": 7101 + }, + { + "epoch": 0.7799253239622227, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.598930597305298, + "learning_rate": 1e-06, + "loss": 1.1257, + "mean_token_accuracy": 0.6747381091117859, + "num_tokens": 183771237.0, + "step": 7102 + }, + { + "epoch": 0.7800351416648363, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4929380416870117, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6956976652145386, + "num_tokens": 183798604.0, + "step": 7103 + }, + { + "epoch": 0.78014495936745, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7832608222961426, + "learning_rate": 1e-06, + "loss": 1.1143, + "mean_token_accuracy": 0.6765766143798828, + "num_tokens": 183824031.0, + "step": 7104 + }, + { + "epoch": 0.7802547770700637, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6682915687561035, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7044956684112549, + "num_tokens": 183847258.0, + "step": 7105 + }, + { + "epoch": 0.7803645947726774, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5000789165496826, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7131374478340149, + "num_tokens": 183872247.0, + "step": 7106 + }, + { + "epoch": 0.780474412475291, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.432145118713379, + "learning_rate": 1e-06, + "loss": 1.1297, + "mean_token_accuracy": 0.6821500062942505, + "num_tokens": 183899771.0, + "step": 7107 + }, + { + "epoch": 0.7805842301779047, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.959789276123047, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.6993507146835327, + "num_tokens": 183927184.0, + "step": 7108 + }, + { + "epoch": 0.7806940478805183, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2814178466796875, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.7003250122070312, + "num_tokens": 183957129.0, + "step": 7109 + }, + { + "epoch": 0.780803865583132, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.450363874435425, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6988709568977356, + "num_tokens": 183985840.0, + "step": 7110 + }, + { + "epoch": 0.7809136832857456, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.457385540008545, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6879408359527588, + "num_tokens": 184012366.0, + "step": 7111 + }, + { + "epoch": 0.7810235009883594, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.469583511352539, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6849528551101685, + "num_tokens": 184040951.0, + "step": 7112 + }, + { + "epoch": 0.781133318690973, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.470472574234009, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7081462144851685, + "num_tokens": 184067588.0, + "step": 7113 + }, + { + "epoch": 0.7812431363935867, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7227542400360107, + "learning_rate": 1e-06, + "loss": 0.9282, + "mean_token_accuracy": 0.729305624961853, + "num_tokens": 184089857.0, + "step": 7114 + }, + { + "epoch": 0.7813529540962003, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.615692377090454, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7184982299804688, + "num_tokens": 184113498.0, + "step": 7115 + }, + { + "epoch": 0.781462771798814, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.897894859313965, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.687492311000824, + "num_tokens": 184137404.0, + "step": 7116 + }, + { + "epoch": 0.7815725895014276, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.819748878479004, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6954255104064941, + "num_tokens": 184159063.0, + "step": 7117 + }, + { + "epoch": 0.7816824072040413, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.661424398422241, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7070648670196533, + "num_tokens": 184181374.0, + "step": 7118 + }, + { + "epoch": 0.7817922249066549, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6040709018707275, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7140364050865173, + "num_tokens": 184204543.0, + "step": 7119 + }, + { + "epoch": 0.7819020426092687, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.7497010231018066, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7081328630447388, + "num_tokens": 184227436.0, + "step": 7120 + }, + { + "epoch": 0.7820118603118823, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6721580028533936, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7108169794082642, + "num_tokens": 184252789.0, + "step": 7121 + }, + { + "epoch": 0.782121678014496, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.428212881088257, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7109324932098389, + "num_tokens": 184279574.0, + "step": 7122 + }, + { + "epoch": 0.7822314957171096, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4314160346984863, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6928120851516724, + "num_tokens": 184311071.0, + "step": 7123 + }, + { + "epoch": 0.7823413134197232, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4153106212615967, + "learning_rate": 1e-06, + "loss": 1.1, + "mean_token_accuracy": 0.6797776818275452, + "num_tokens": 184339486.0, + "step": 7124 + }, + { + "epoch": 0.7824511311223369, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.643648147583008, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.720451831817627, + "num_tokens": 184362691.0, + "step": 7125 + }, + { + "epoch": 0.7825609488249505, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 16.6658878326416, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.6991594433784485, + "num_tokens": 184391019.0, + "step": 7126 + }, + { + "epoch": 0.7826707665275643, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7870500087738037, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7144348621368408, + "num_tokens": 184412939.0, + "step": 7127 + }, + { + "epoch": 0.7827805842301779, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.482635259628296, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.70817631483078, + "num_tokens": 184441180.0, + "step": 7128 + }, + { + "epoch": 0.7828904019327916, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.33870530128479, + "learning_rate": 1e-06, + "loss": 0.9088, + "mean_token_accuracy": 0.7246808409690857, + "num_tokens": 184471774.0, + "step": 7129 + }, + { + "epoch": 0.7830002196354052, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.679490327835083, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7051381468772888, + "num_tokens": 184494838.0, + "step": 7130 + }, + { + "epoch": 0.7831100373380189, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.427382469177246, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7177112698554993, + "num_tokens": 184520574.0, + "step": 7131 + }, + { + "epoch": 0.7832198550406325, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7402775287628174, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7067519426345825, + "num_tokens": 184542020.0, + "step": 7132 + }, + { + "epoch": 0.7833296727432462, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9300084114074707, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6843552589416504, + "num_tokens": 184567106.0, + "step": 7133 + }, + { + "epoch": 0.7834394904458599, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4582183361053467, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7087797522544861, + "num_tokens": 184592293.0, + "step": 7134 + }, + { + "epoch": 0.7835493081484736, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.810208320617676, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6750541925430298, + "num_tokens": 184614675.0, + "step": 7135 + }, + { + "epoch": 0.7836591258510872, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.418119430541992, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6869441866874695, + "num_tokens": 184643311.0, + "step": 7136 + }, + { + "epoch": 0.7837689435537009, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.501258611679077, + "learning_rate": 1e-06, + "loss": 1.1139, + "mean_token_accuracy": 0.6821702718734741, + "num_tokens": 184671028.0, + "step": 7137 + }, + { + "epoch": 0.7838787612563145, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.441493272781372, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6973767876625061, + "num_tokens": 184697887.0, + "step": 7138 + }, + { + "epoch": 0.7839885789589282, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3798656463623047, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7068483233451843, + "num_tokens": 184728433.0, + "step": 7139 + }, + { + "epoch": 0.7840983966615418, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5278286933898926, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7027162313461304, + "num_tokens": 184754313.0, + "step": 7140 + }, + { + "epoch": 0.7842082143641556, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3930447101593018, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6912258863449097, + "num_tokens": 184785843.0, + "step": 7141 + }, + { + "epoch": 0.7843180320667692, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.525228500366211, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7123105525970459, + "num_tokens": 184811996.0, + "step": 7142 + }, + { + "epoch": 0.7844278497693828, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7637569904327393, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6871428489685059, + "num_tokens": 184834675.0, + "step": 7143 + }, + { + "epoch": 0.7845376674719965, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 6.593572616577148, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7046813368797302, + "num_tokens": 184863266.0, + "step": 7144 + }, + { + "epoch": 0.7846474851746101, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.771219253540039, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7189142107963562, + "num_tokens": 184886024.0, + "step": 7145 + }, + { + "epoch": 0.7847573028772238, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.810408115386963, + "learning_rate": 1e-06, + "loss": 1.0623, + "mean_token_accuracy": 0.6935126781463623, + "num_tokens": 184909965.0, + "step": 7146 + }, + { + "epoch": 0.7848671205798374, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.715872287750244, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7144944071769714, + "num_tokens": 184931186.0, + "step": 7147 + }, + { + "epoch": 0.7849769382824512, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5071756839752197, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.685706377029419, + "num_tokens": 184959886.0, + "step": 7148 + }, + { + "epoch": 0.7850867559850648, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.489309549331665, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7280482649803162, + "num_tokens": 184983109.0, + "step": 7149 + }, + { + "epoch": 0.7851965736876785, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.479973793029785, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6824072599411011, + "num_tokens": 185012198.0, + "step": 7150 + }, + { + "epoch": 0.7853063913902921, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6436877250671387, + "learning_rate": 1e-06, + "loss": 0.8824, + "mean_token_accuracy": 0.7328448295593262, + "num_tokens": 185033034.0, + "step": 7151 + }, + { + "epoch": 0.7854162090929058, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4561550617218018, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6927692890167236, + "num_tokens": 185059299.0, + "step": 7152 + }, + { + "epoch": 0.7855260267955194, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3718512058258057, + "learning_rate": 1e-06, + "loss": 1.1465, + "mean_token_accuracy": 0.6612296104431152, + "num_tokens": 185087271.0, + "step": 7153 + }, + { + "epoch": 0.7856358444981331, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5025475025177, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6949702501296997, + "num_tokens": 185113305.0, + "step": 7154 + }, + { + "epoch": 0.7857456622007467, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9313576221466064, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.711636483669281, + "num_tokens": 185134830.0, + "step": 7155 + }, + { + "epoch": 0.7858554799033605, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7818217277526855, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.7213922739028931, + "num_tokens": 185154646.0, + "step": 7156 + }, + { + "epoch": 0.7859652976059741, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4386074542999268, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6932126879692078, + "num_tokens": 185183577.0, + "step": 7157 + }, + { + "epoch": 0.7860751153085878, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.907806873321533, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7145050168037415, + "num_tokens": 185203013.0, + "step": 7158 + }, + { + "epoch": 0.7861849330112014, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.480809211730957, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6821699738502502, + "num_tokens": 185229869.0, + "step": 7159 + }, + { + "epoch": 0.786294750713815, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5422449111938477, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6937898993492126, + "num_tokens": 185255496.0, + "step": 7160 + }, + { + "epoch": 0.7864045684164287, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.410061836242676, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6881210207939148, + "num_tokens": 185284771.0, + "step": 7161 + }, + { + "epoch": 0.7865143861190423, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.714040517807007, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7206457257270813, + "num_tokens": 185306814.0, + "step": 7162 + }, + { + "epoch": 0.7866242038216561, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6036670207977295, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6848218441009521, + "num_tokens": 185331860.0, + "step": 7163 + }, + { + "epoch": 0.7867340215242697, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4183638095855713, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6975820660591125, + "num_tokens": 185359863.0, + "step": 7164 + }, + { + "epoch": 0.7868438392268834, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.557013988494873, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6934127807617188, + "num_tokens": 185387084.0, + "step": 7165 + }, + { + "epoch": 0.786953656929497, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4203057289123535, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7221288681030273, + "num_tokens": 185413133.0, + "step": 7166 + }, + { + "epoch": 0.7870634746321107, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.503047227859497, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6901077032089233, + "num_tokens": 185437925.0, + "step": 7167 + }, + { + "epoch": 0.7871732923347243, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.521496295928955, + "learning_rate": 1e-06, + "loss": 1.1164, + "mean_token_accuracy": 0.6777642369270325, + "num_tokens": 185466086.0, + "step": 7168 + }, + { + "epoch": 0.787283110037338, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.541771650314331, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7059777975082397, + "num_tokens": 185493799.0, + "step": 7169 + }, + { + "epoch": 0.7873929277399517, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3982925415039062, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6966577768325806, + "num_tokens": 185522535.0, + "step": 7170 + }, + { + "epoch": 0.7875027454425654, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.355631113052368, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6790528297424316, + "num_tokens": 185556122.0, + "step": 7171 + }, + { + "epoch": 0.787612563145179, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6177186965942383, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7126481533050537, + "num_tokens": 185582159.0, + "step": 7172 + }, + { + "epoch": 0.7877223808477927, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 3.361710548400879, + "learning_rate": 1e-06, + "loss": 1.1084, + "mean_token_accuracy": 0.6758294105529785, + "num_tokens": 185607770.0, + "step": 7173 + }, + { + "epoch": 0.7878321985504063, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.590071201324463, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7121708989143372, + "num_tokens": 185633153.0, + "step": 7174 + }, + { + "epoch": 0.78794201625302, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.534543752670288, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7078258991241455, + "num_tokens": 185662012.0, + "step": 7175 + }, + { + "epoch": 0.7880518339556336, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.448852777481079, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.6837136745452881, + "num_tokens": 185691755.0, + "step": 7176 + }, + { + "epoch": 0.7881616516582474, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.652672529220581, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.6986010074615479, + "num_tokens": 185715651.0, + "step": 7177 + }, + { + "epoch": 0.788271469360861, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5139191150665283, + "learning_rate": 1e-06, + "loss": 1.128, + "mean_token_accuracy": 0.6741921901702881, + "num_tokens": 185741311.0, + "step": 7178 + }, + { + "epoch": 0.7883812870634747, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3688297271728516, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7184156179428101, + "num_tokens": 185768557.0, + "step": 7179 + }, + { + "epoch": 0.7884911047660883, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8075828552246094, + "learning_rate": 1e-06, + "loss": 1.0857, + "mean_token_accuracy": 0.6872023344039917, + "num_tokens": 185795015.0, + "step": 7180 + }, + { + "epoch": 0.788600922468702, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5924386978149414, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6879276633262634, + "num_tokens": 185819116.0, + "step": 7181 + }, + { + "epoch": 0.7887107401713156, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.745314836502075, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.70313560962677, + "num_tokens": 185841426.0, + "step": 7182 + }, + { + "epoch": 0.7888205578739292, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6585962772369385, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6974025964736938, + "num_tokens": 185866540.0, + "step": 7183 + }, + { + "epoch": 0.7889303755765429, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.526564121246338, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6975998878479004, + "num_tokens": 185892827.0, + "step": 7184 + }, + { + "epoch": 0.7890401932791566, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.8635058403015137, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.724738359451294, + "num_tokens": 185913702.0, + "step": 7185 + }, + { + "epoch": 0.7891500109817703, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.518834114074707, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.709905207157135, + "num_tokens": 185939475.0, + "step": 7186 + }, + { + "epoch": 0.7892598286843839, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2779030799865723, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6835500597953796, + "num_tokens": 185972425.0, + "step": 7187 + }, + { + "epoch": 0.7893696463869976, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.2609963417053223, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.6927030086517334, + "num_tokens": 186007050.0, + "step": 7188 + }, + { + "epoch": 0.7894794640896112, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5760765075683594, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7101282477378845, + "num_tokens": 186030479.0, + "step": 7189 + }, + { + "epoch": 0.7895892817922249, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.4198474884033203, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6964111328125, + "num_tokens": 186061037.0, + "step": 7190 + }, + { + "epoch": 0.7896990994948385, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.538719415664673, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7009460926055908, + "num_tokens": 186086817.0, + "step": 7191 + }, + { + "epoch": 0.7898089171974523, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.6065244674682617, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7071117162704468, + "num_tokens": 186114481.0, + "step": 7192 + }, + { + "epoch": 0.7899187349000659, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.726280689239502, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6843609809875488, + "num_tokens": 186142585.0, + "step": 7193 + }, + { + "epoch": 0.7900285526026796, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6263017654418945, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6974208354949951, + "num_tokens": 186170466.0, + "step": 7194 + }, + { + "epoch": 0.7901383703052932, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.469007968902588, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7027089595794678, + "num_tokens": 186197787.0, + "step": 7195 + }, + { + "epoch": 0.7902481880079069, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.0343117713928223, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7129037976264954, + "num_tokens": 186220283.0, + "step": 7196 + }, + { + "epoch": 0.7903580057105205, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.352635145187378, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.6983261108398438, + "num_tokens": 186250044.0, + "step": 7197 + }, + { + "epoch": 0.7904678234131342, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.2225072383880615, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6850270628929138, + "num_tokens": 186279918.0, + "step": 7198 + }, + { + "epoch": 0.7905776411157479, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5507822036743164, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6929739713668823, + "num_tokens": 186306253.0, + "step": 7199 + }, + { + "epoch": 0.7906874588183616, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.42044734954834, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7056560516357422, + "num_tokens": 186334988.0, + "step": 7200 + }, + { + "epoch": 0.7907972765209752, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4576520919799805, + "learning_rate": 1e-06, + "loss": 1.1264, + "mean_token_accuracy": 0.6724697947502136, + "num_tokens": 186362393.0, + "step": 7201 + }, + { + "epoch": 0.7909070942235888, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2762465476989746, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7222984433174133, + "num_tokens": 186391206.0, + "step": 7202 + }, + { + "epoch": 0.7910169119262025, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4001991748809814, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7003370523452759, + "num_tokens": 186420976.0, + "step": 7203 + }, + { + "epoch": 0.7911267296288161, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5215327739715576, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.7061774730682373, + "num_tokens": 186447364.0, + "step": 7204 + }, + { + "epoch": 0.7912365473314298, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.756723165512085, + "learning_rate": 1e-06, + "loss": 1.1323, + "mean_token_accuracy": 0.6889216899871826, + "num_tokens": 186470996.0, + "step": 7205 + }, + { + "epoch": 0.7913463650340435, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.5121939182281494, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7230316400527954, + "num_tokens": 186496877.0, + "step": 7206 + }, + { + "epoch": 0.7914561827366572, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 2.739313840866089, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7159944176673889, + "num_tokens": 186519265.0, + "step": 7207 + }, + { + "epoch": 0.7915660004392708, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.53389310836792, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.6968474388122559, + "num_tokens": 186548132.0, + "step": 7208 + }, + { + "epoch": 0.7916758181418845, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4879724979400635, + "learning_rate": 1e-06, + "loss": 1.1794, + "mean_token_accuracy": 0.6607807278633118, + "num_tokens": 186577678.0, + "step": 7209 + }, + { + "epoch": 0.7917856358444981, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2938709259033203, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7058656215667725, + "num_tokens": 186608047.0, + "step": 7210 + }, + { + "epoch": 0.7918954535471118, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.1358718872070312, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6963474154472351, + "num_tokens": 186642725.0, + "step": 7211 + }, + { + "epoch": 0.7920052712497254, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2346434593200684, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.7027536630630493, + "num_tokens": 186673867.0, + "step": 7212 + }, + { + "epoch": 0.7921150889523391, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3053972721099854, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6865593194961548, + "num_tokens": 186704477.0, + "step": 7213 + }, + { + "epoch": 0.7922249066549528, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5910866260528564, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6927433013916016, + "num_tokens": 186730133.0, + "step": 7214 + }, + { + "epoch": 0.7923347243575665, + "ewc_loss": 7.867813110351562e-06, + "grad_norm": 3.0153493881225586, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.7047025561332703, + "num_tokens": 186750994.0, + "step": 7215 + }, + { + "epoch": 0.7924445420601801, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.057318925857544, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7093621492385864, + "num_tokens": 186770534.0, + "step": 7216 + }, + { + "epoch": 0.7925543597627938, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.531975269317627, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.690320611000061, + "num_tokens": 186798520.0, + "step": 7217 + }, + { + "epoch": 0.7926641774654074, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3827321529388428, + "learning_rate": 1e-06, + "loss": 1.1477, + "mean_token_accuracy": 0.6677790284156799, + "num_tokens": 186828948.0, + "step": 7218 + }, + { + "epoch": 0.792773995168021, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5761516094207764, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6943816542625427, + "num_tokens": 186857082.0, + "step": 7219 + }, + { + "epoch": 0.7928838128706347, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 16.733531951904297, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6856414675712585, + "num_tokens": 186880348.0, + "step": 7220 + }, + { + "epoch": 0.7929936305732485, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.014770746231079, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7021388411521912, + "num_tokens": 186899871.0, + "step": 7221 + }, + { + "epoch": 0.7931034482758621, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5040032863616943, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.6864637732505798, + "num_tokens": 186930246.0, + "step": 7222 + }, + { + "epoch": 0.7932132659784757, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4225006103515625, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7083712220191956, + "num_tokens": 186957223.0, + "step": 7223 + }, + { + "epoch": 0.7933230836810894, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7528393268585205, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7024368643760681, + "num_tokens": 186980450.0, + "step": 7224 + }, + { + "epoch": 0.793432901383703, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.541222095489502, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7081999182701111, + "num_tokens": 187004492.0, + "step": 7225 + }, + { + "epoch": 0.7935427190863167, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.480330228805542, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7106349468231201, + "num_tokens": 187030339.0, + "step": 7226 + }, + { + "epoch": 0.7936525367889303, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6211349964141846, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6835021376609802, + "num_tokens": 187057485.0, + "step": 7227 + }, + { + "epoch": 0.7937623544915441, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7369697093963623, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.704480767250061, + "num_tokens": 187077206.0, + "step": 7228 + }, + { + "epoch": 0.7938721721941577, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.462306022644043, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7067289352416992, + "num_tokens": 187104941.0, + "step": 7229 + }, + { + "epoch": 0.7939819898967714, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.120039224624634, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7196336388587952, + "num_tokens": 187127950.0, + "step": 7230 + }, + { + "epoch": 0.794091807599385, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7395334243774414, + "learning_rate": 1e-06, + "loss": 1.1021, + "mean_token_accuracy": 0.686305582523346, + "num_tokens": 187153998.0, + "step": 7231 + }, + { + "epoch": 0.7942016253019987, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6330020427703857, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.7203010320663452, + "num_tokens": 187174998.0, + "step": 7232 + }, + { + "epoch": 0.7943114430046123, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.472637176513672, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.704651951789856, + "num_tokens": 187201294.0, + "step": 7233 + }, + { + "epoch": 0.794421260707226, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3861265182495117, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.6937935948371887, + "num_tokens": 187229083.0, + "step": 7234 + }, + { + "epoch": 0.7945310784098397, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.69661021232605, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7080790996551514, + "num_tokens": 187250489.0, + "step": 7235 + }, + { + "epoch": 0.7946408961124534, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5295653343200684, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.714332103729248, + "num_tokens": 187277302.0, + "step": 7236 + }, + { + "epoch": 0.794750713815067, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.663616895675659, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7056938409805298, + "num_tokens": 187300062.0, + "step": 7237 + }, + { + "epoch": 0.7948605315176807, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8512279987335205, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7148769497871399, + "num_tokens": 187319118.0, + "step": 7238 + }, + { + "epoch": 0.7949703492202943, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.470930576324463, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7008013725280762, + "num_tokens": 187345374.0, + "step": 7239 + }, + { + "epoch": 0.795080166922908, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.468219757080078, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7230352163314819, + "num_tokens": 187370379.0, + "step": 7240 + }, + { + "epoch": 0.7951899846255216, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4711992740631104, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7155346274375916, + "num_tokens": 187395194.0, + "step": 7241 + }, + { + "epoch": 0.7952998023281352, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.585907459259033, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6984038352966309, + "num_tokens": 187418656.0, + "step": 7242 + }, + { + "epoch": 0.795409620030749, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6687796115875244, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6965962648391724, + "num_tokens": 187442710.0, + "step": 7243 + }, + { + "epoch": 0.7955194377333626, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.522158145904541, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7067602872848511, + "num_tokens": 187470929.0, + "step": 7244 + }, + { + "epoch": 0.7956292554359763, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3626065254211426, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7137584090232849, + "num_tokens": 187499503.0, + "step": 7245 + }, + { + "epoch": 0.7957390731385899, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5388941764831543, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7181977033615112, + "num_tokens": 187523173.0, + "step": 7246 + }, + { + "epoch": 0.7958488908412036, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.369784355163574, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7049408555030823, + "num_tokens": 187551392.0, + "step": 7247 + }, + { + "epoch": 0.7959587085438172, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2844650745391846, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6826710104942322, + "num_tokens": 187583865.0, + "step": 7248 + }, + { + "epoch": 0.7960685262464309, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6831421852111816, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7089953422546387, + "num_tokens": 187605803.0, + "step": 7249 + }, + { + "epoch": 0.7961783439490446, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7865660190582275, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6974745988845825, + "num_tokens": 187628523.0, + "step": 7250 + }, + { + "epoch": 0.7962881616516583, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8981974124908447, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.69809889793396, + "num_tokens": 187648844.0, + "step": 7251 + }, + { + "epoch": 0.7963979793542719, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.0173444747924805, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.6936357021331787, + "num_tokens": 187668170.0, + "step": 7252 + }, + { + "epoch": 0.7965077970568856, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4791979789733887, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7043535113334656, + "num_tokens": 187695231.0, + "step": 7253 + }, + { + "epoch": 0.7966176147594992, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7794368267059326, + "learning_rate": 1e-06, + "loss": 0.8952, + "mean_token_accuracy": 0.7434231042861938, + "num_tokens": 187716433.0, + "step": 7254 + }, + { + "epoch": 0.7967274324621129, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.70155930519104, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7320947647094727, + "num_tokens": 187736921.0, + "step": 7255 + }, + { + "epoch": 0.7968372501647265, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.54508900642395, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6940909028053284, + "num_tokens": 187761659.0, + "step": 7256 + }, + { + "epoch": 0.7969470678673403, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.373670816421509, + "learning_rate": 1e-06, + "loss": 1.1333, + "mean_token_accuracy": 0.6721992492675781, + "num_tokens": 187794860.0, + "step": 7257 + }, + { + "epoch": 0.7970568855699539, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.514848470687866, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.69892418384552, + "num_tokens": 187822637.0, + "step": 7258 + }, + { + "epoch": 0.7971667032725676, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.314157009124756, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.6978772282600403, + "num_tokens": 187853797.0, + "step": 7259 + }, + { + "epoch": 0.7972765209751812, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8050312995910645, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7138522863388062, + "num_tokens": 187874537.0, + "step": 7260 + }, + { + "epoch": 0.7973863386777948, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.383751153945923, + "learning_rate": 1e-06, + "loss": 1.147, + "mean_token_accuracy": 0.6671825051307678, + "num_tokens": 187907296.0, + "step": 7261 + }, + { + "epoch": 0.7974961563804085, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.533245086669922, + "learning_rate": 1e-06, + "loss": 1.1767, + "mean_token_accuracy": 0.6602976322174072, + "num_tokens": 187935828.0, + "step": 7262 + }, + { + "epoch": 0.7976059740830221, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.573377847671509, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7194325923919678, + "num_tokens": 187958585.0, + "step": 7263 + }, + { + "epoch": 0.7977157917856359, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.853299379348755, + "learning_rate": 1e-06, + "loss": 0.9508, + "mean_token_accuracy": 0.7133786678314209, + "num_tokens": 187979118.0, + "step": 7264 + }, + { + "epoch": 0.7978256094882495, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6561710834503174, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7033653855323792, + "num_tokens": 188005438.0, + "step": 7265 + }, + { + "epoch": 0.7979354271908632, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6898279190063477, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7044883966445923, + "num_tokens": 188029090.0, + "step": 7266 + }, + { + "epoch": 0.7980452448934768, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6550066471099854, + "learning_rate": 1e-06, + "loss": 1.1267, + "mean_token_accuracy": 0.6749796867370605, + "num_tokens": 188053893.0, + "step": 7267 + }, + { + "epoch": 0.7981550625960905, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9985620975494385, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6936242580413818, + "num_tokens": 188078874.0, + "step": 7268 + }, + { + "epoch": 0.7982648802987041, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6371536254882812, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.695268452167511, + "num_tokens": 188105436.0, + "step": 7269 + }, + { + "epoch": 0.7983746980013178, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3875694274902344, + "learning_rate": 1e-06, + "loss": 1.1316, + "mean_token_accuracy": 0.6834020614624023, + "num_tokens": 188139579.0, + "step": 7270 + }, + { + "epoch": 0.7984845157039314, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4713802337646484, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7135164141654968, + "num_tokens": 188165248.0, + "step": 7271 + }, + { + "epoch": 0.7985943334065452, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4491915702819824, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6891725659370422, + "num_tokens": 188192813.0, + "step": 7272 + }, + { + "epoch": 0.7987041511091588, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.0301570892333984, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7098703980445862, + "num_tokens": 188212555.0, + "step": 7273 + }, + { + "epoch": 0.7988139688117725, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.527606248855591, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.6987501978874207, + "num_tokens": 188240737.0, + "step": 7274 + }, + { + "epoch": 0.7989237865143861, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.388066291809082, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7009773850440979, + "num_tokens": 188268722.0, + "step": 7275 + }, + { + "epoch": 0.7990336042169998, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6537883281707764, + "learning_rate": 1e-06, + "loss": 1.1079, + "mean_token_accuracy": 0.6801185607910156, + "num_tokens": 188295061.0, + "step": 7276 + }, + { + "epoch": 0.7991434219196134, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5091421604156494, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7124594449996948, + "num_tokens": 188322008.0, + "step": 7277 + }, + { + "epoch": 0.799253239622227, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5216071605682373, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6837567687034607, + "num_tokens": 188351369.0, + "step": 7278 + }, + { + "epoch": 0.7993630573248408, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6060307025909424, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7178230285644531, + "num_tokens": 188374342.0, + "step": 7279 + }, + { + "epoch": 0.7994728750274545, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.469820499420166, + "learning_rate": 1e-06, + "loss": 1.1369, + "mean_token_accuracy": 0.6761728525161743, + "num_tokens": 188403080.0, + "step": 7280 + }, + { + "epoch": 0.7995826927300681, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.278773069381714, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6827394962310791, + "num_tokens": 188434808.0, + "step": 7281 + }, + { + "epoch": 0.7996925104326817, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3778834342956543, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.7089447975158691, + "num_tokens": 188465375.0, + "step": 7282 + }, + { + "epoch": 0.7998023281352954, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.384701728820801, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7032574415206909, + "num_tokens": 188495748.0, + "step": 7283 + }, + { + "epoch": 0.799912145837909, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 16.736360549926758, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7034420371055603, + "num_tokens": 188517345.0, + "step": 7284 + }, + { + "epoch": 0.8000219635405227, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.95246958732605, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.704859733581543, + "num_tokens": 188543347.0, + "step": 7285 + }, + { + "epoch": 0.8001317812431364, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3307087421417236, + "learning_rate": 1e-06, + "loss": 1.1498, + "mean_token_accuracy": 0.6749135255813599, + "num_tokens": 188577493.0, + "step": 7286 + }, + { + "epoch": 0.8002415989457501, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.495195150375366, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7093685865402222, + "num_tokens": 188602905.0, + "step": 7287 + }, + { + "epoch": 0.8003514166483637, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5619256496429443, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7025368213653564, + "num_tokens": 188629470.0, + "step": 7288 + }, + { + "epoch": 0.8004612343509774, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3762035369873047, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6913439631462097, + "num_tokens": 188657305.0, + "step": 7289 + }, + { + "epoch": 0.800571052053591, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3468618392944336, + "learning_rate": 1e-06, + "loss": 1.1104, + "mean_token_accuracy": 0.6769712567329407, + "num_tokens": 188685231.0, + "step": 7290 + }, + { + "epoch": 0.8006808697562047, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.644059181213379, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6800057291984558, + "num_tokens": 188719881.0, + "step": 7291 + }, + { + "epoch": 0.8007906874588183, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7913429737091064, + "learning_rate": 1e-06, + "loss": 0.9026, + "mean_token_accuracy": 0.7288468480110168, + "num_tokens": 188740271.0, + "step": 7292 + }, + { + "epoch": 0.8009005051614321, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.2546164989471436, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7122693061828613, + "num_tokens": 188771459.0, + "step": 7293 + }, + { + "epoch": 0.8010103228640457, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.570979118347168, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7156618237495422, + "num_tokens": 188796163.0, + "step": 7294 + }, + { + "epoch": 0.8011201405666594, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7526206970214844, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.704150915145874, + "num_tokens": 188818117.0, + "step": 7295 + }, + { + "epoch": 0.801229958269273, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7497129440307617, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7044490575790405, + "num_tokens": 188839835.0, + "step": 7296 + }, + { + "epoch": 0.8013397759718867, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6540064811706543, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6995376348495483, + "num_tokens": 188864548.0, + "step": 7297 + }, + { + "epoch": 0.8014495936745003, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4335522651672363, + "learning_rate": 1e-06, + "loss": 1.0796, + "mean_token_accuracy": 0.6912201642990112, + "num_tokens": 188894610.0, + "step": 7298 + }, + { + "epoch": 0.801559411377114, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.753293514251709, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7175549268722534, + "num_tokens": 188915161.0, + "step": 7299 + }, + { + "epoch": 0.8016692290797276, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.451272487640381, + "learning_rate": 1e-06, + "loss": 1.0917, + "mean_token_accuracy": 0.6853024959564209, + "num_tokens": 188944775.0, + "step": 7300 + }, + { + "epoch": 0.8017790467823414, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4352967739105225, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7066686153411865, + "num_tokens": 188973964.0, + "step": 7301 + }, + { + "epoch": 0.801888864484955, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.008460760116577, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7078194618225098, + "num_tokens": 189000178.0, + "step": 7302 + }, + { + "epoch": 0.8019986821875686, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6717257499694824, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6983809471130371, + "num_tokens": 189030193.0, + "step": 7303 + }, + { + "epoch": 0.8021084998901823, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.694767713546753, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6835348606109619, + "num_tokens": 189053274.0, + "step": 7304 + }, + { + "epoch": 0.8022183175927959, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.400087356567383, + "learning_rate": 1e-06, + "loss": 1.0845, + "mean_token_accuracy": 0.685854434967041, + "num_tokens": 189082363.0, + "step": 7305 + }, + { + "epoch": 0.8023281352954096, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4119362831115723, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6914495229721069, + "num_tokens": 189109955.0, + "step": 7306 + }, + { + "epoch": 0.8024379529980232, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4004437923431396, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7200812697410583, + "num_tokens": 189134395.0, + "step": 7307 + }, + { + "epoch": 0.802547770700637, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5485451221466064, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6998157501220703, + "num_tokens": 189162058.0, + "step": 7308 + }, + { + "epoch": 0.8026575884032506, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.516470432281494, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.693970799446106, + "num_tokens": 189187227.0, + "step": 7309 + }, + { + "epoch": 0.8027674061058643, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.484651803970337, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7120904922485352, + "num_tokens": 189212907.0, + "step": 7310 + }, + { + "epoch": 0.8028772238084779, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4710323810577393, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7069154977798462, + "num_tokens": 189241018.0, + "step": 7311 + }, + { + "epoch": 0.8029870415110916, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6348018646240234, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6996915936470032, + "num_tokens": 189263735.0, + "step": 7312 + }, + { + "epoch": 0.8030968592137052, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.403933525085449, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7205278277397156, + "num_tokens": 189293717.0, + "step": 7313 + }, + { + "epoch": 0.8032066769163189, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6504857540130615, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6951149702072144, + "num_tokens": 189316183.0, + "step": 7314 + }, + { + "epoch": 0.8033164946189326, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4105982780456543, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.7015889883041382, + "num_tokens": 189343617.0, + "step": 7315 + }, + { + "epoch": 0.8034263123215463, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5882771015167236, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.692909836769104, + "num_tokens": 189371355.0, + "step": 7316 + }, + { + "epoch": 0.8035361300241599, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9629082679748535, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6983568668365479, + "num_tokens": 189391749.0, + "step": 7317 + }, + { + "epoch": 0.8036459477267736, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5160632133483887, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6865935325622559, + "num_tokens": 189421161.0, + "step": 7318 + }, + { + "epoch": 0.8037557654293872, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.811842679977417, + "learning_rate": 1e-06, + "loss": 0.9308, + "mean_token_accuracy": 0.7166119813919067, + "num_tokens": 189440703.0, + "step": 7319 + }, + { + "epoch": 0.8038655831320008, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.883939027786255, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7072484493255615, + "num_tokens": 189460939.0, + "step": 7320 + }, + { + "epoch": 0.8039754008346145, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.538919687271118, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.689761757850647, + "num_tokens": 189492736.0, + "step": 7321 + }, + { + "epoch": 0.8040852185372283, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8470137119293213, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6959325075149536, + "num_tokens": 189517571.0, + "step": 7322 + }, + { + "epoch": 0.8041950362398419, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4950919151306152, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7089830040931702, + "num_tokens": 189544763.0, + "step": 7323 + }, + { + "epoch": 0.8043048539424555, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4729936122894287, + "learning_rate": 1e-06, + "loss": 1.1426, + "mean_token_accuracy": 0.6795070171356201, + "num_tokens": 189571734.0, + "step": 7324 + }, + { + "epoch": 0.8044146716450692, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3855154514312744, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.712924063205719, + "num_tokens": 189599544.0, + "step": 7325 + }, + { + "epoch": 0.8045244893476828, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8181111812591553, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6884801387786865, + "num_tokens": 189620801.0, + "step": 7326 + }, + { + "epoch": 0.8046343070502965, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7481558322906494, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6898412108421326, + "num_tokens": 189642798.0, + "step": 7327 + }, + { + "epoch": 0.8047441247529101, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.833235740661621, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7087064385414124, + "num_tokens": 189667045.0, + "step": 7328 + }, + { + "epoch": 0.8048539424555239, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.698232650756836, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7101086378097534, + "num_tokens": 189693170.0, + "step": 7329 + }, + { + "epoch": 0.8049637601581375, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7296340465545654, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6910514831542969, + "num_tokens": 189717385.0, + "step": 7330 + }, + { + "epoch": 0.8050735778607512, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.487656354904175, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7020766139030457, + "num_tokens": 189743773.0, + "step": 7331 + }, + { + "epoch": 0.8051833955633648, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.543832778930664, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6780984401702881, + "num_tokens": 189775152.0, + "step": 7332 + }, + { + "epoch": 0.8052932132659785, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4502599239349365, + "learning_rate": 1e-06, + "loss": 0.9211, + "mean_token_accuracy": 0.7294246554374695, + "num_tokens": 189801396.0, + "step": 7333 + }, + { + "epoch": 0.8054030309685921, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6369478702545166, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6849784851074219, + "num_tokens": 189824889.0, + "step": 7334 + }, + { + "epoch": 0.8055128486712058, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.923025131225586, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7090845704078674, + "num_tokens": 189845178.0, + "step": 7335 + }, + { + "epoch": 0.8056226663738194, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.520395040512085, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7031542062759399, + "num_tokens": 189869934.0, + "step": 7336 + }, + { + "epoch": 0.8057324840764332, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4267122745513916, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6857936382293701, + "num_tokens": 189897957.0, + "step": 7337 + }, + { + "epoch": 0.8058423017790468, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5003252029418945, + "learning_rate": 1e-06, + "loss": 1.1188, + "mean_token_accuracy": 0.6778400540351868, + "num_tokens": 189926967.0, + "step": 7338 + }, + { + "epoch": 0.8059521194816605, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6869428157806396, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7164196968078613, + "num_tokens": 189948959.0, + "step": 7339 + }, + { + "epoch": 0.8060619371842741, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3871676921844482, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.6966801881790161, + "num_tokens": 189979548.0, + "step": 7340 + }, + { + "epoch": 0.8061717548868877, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8031442165374756, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6840961575508118, + "num_tokens": 190003615.0, + "step": 7341 + }, + { + "epoch": 0.8062815725895014, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.882211685180664, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7263275384902954, + "num_tokens": 190027255.0, + "step": 7342 + }, + { + "epoch": 0.806391390292115, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4953081607818604, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.701905369758606, + "num_tokens": 190054949.0, + "step": 7343 + }, + { + "epoch": 0.8065012079947288, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6971182823181152, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7267354130744934, + "num_tokens": 190077434.0, + "step": 7344 + }, + { + "epoch": 0.8066110256973424, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.52077054977417, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7049844264984131, + "num_tokens": 190105048.0, + "step": 7345 + }, + { + "epoch": 0.8067208433999561, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3934741020202637, + "learning_rate": 1e-06, + "loss": 1.193, + "mean_token_accuracy": 0.6626566648483276, + "num_tokens": 190135929.0, + "step": 7346 + }, + { + "epoch": 0.8068306611025697, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.357301950454712, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6889008283615112, + "num_tokens": 190164509.0, + "step": 7347 + }, + { + "epoch": 0.8069404788051834, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.613068103790283, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7191757559776306, + "num_tokens": 190186916.0, + "step": 7348 + }, + { + "epoch": 0.807050296507797, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7425622940063477, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6921353340148926, + "num_tokens": 190215352.0, + "step": 7349 + }, + { + "epoch": 0.8071601142104107, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4937620162963867, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7110965251922607, + "num_tokens": 190240627.0, + "step": 7350 + }, + { + "epoch": 0.8072699319130244, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5700266361236572, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6970080137252808, + "num_tokens": 190266976.0, + "step": 7351 + }, + { + "epoch": 0.8073797496156381, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.722182035446167, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7217077016830444, + "num_tokens": 190290414.0, + "step": 7352 + }, + { + "epoch": 0.8074895673182517, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.906458616256714, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6853015422821045, + "num_tokens": 190314442.0, + "step": 7353 + }, + { + "epoch": 0.8075993850208654, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5235049724578857, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7219851016998291, + "num_tokens": 190341818.0, + "step": 7354 + }, + { + "epoch": 0.807709202723479, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.409724473953247, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7229531407356262, + "num_tokens": 190368152.0, + "step": 7355 + }, + { + "epoch": 0.8078190204260927, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8344385623931885, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7020093202590942, + "num_tokens": 190387620.0, + "step": 7356 + }, + { + "epoch": 0.8079288381287063, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6145517826080322, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7018520832061768, + "num_tokens": 190411877.0, + "step": 7357 + }, + { + "epoch": 0.8080386558313201, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.653047800064087, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7092914581298828, + "num_tokens": 190436207.0, + "step": 7358 + }, + { + "epoch": 0.8081484735339337, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3505592346191406, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7006601095199585, + "num_tokens": 190469819.0, + "step": 7359 + }, + { + "epoch": 0.8082582912365474, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4824626445770264, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7087366580963135, + "num_tokens": 190497336.0, + "step": 7360 + }, + { + "epoch": 0.808368108939161, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.36523175239563, + "learning_rate": 1e-06, + "loss": 1.1063, + "mean_token_accuracy": 0.6812103986740112, + "num_tokens": 190526504.0, + "step": 7361 + }, + { + "epoch": 0.8084779266417746, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3336358070373535, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6900151968002319, + "num_tokens": 190559033.0, + "step": 7362 + }, + { + "epoch": 0.8085877443443883, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.9120490550994873, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7218643426895142, + "num_tokens": 190576453.0, + "step": 7363 + }, + { + "epoch": 0.8086975620470019, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7488064765930176, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6916435360908508, + "num_tokens": 190601617.0, + "step": 7364 + }, + { + "epoch": 0.8088073797496156, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.896489143371582, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6907380819320679, + "num_tokens": 190622769.0, + "step": 7365 + }, + { + "epoch": 0.8089171974522293, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3939590454101562, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6995702981948853, + "num_tokens": 190654941.0, + "step": 7366 + }, + { + "epoch": 0.809027015154843, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5492029190063477, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6873482465744019, + "num_tokens": 190679722.0, + "step": 7367 + }, + { + "epoch": 0.8091368328574566, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.438251256942749, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6953659057617188, + "num_tokens": 190708192.0, + "step": 7368 + }, + { + "epoch": 0.8092466505600703, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.674264907836914, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6919889450073242, + "num_tokens": 190731374.0, + "step": 7369 + }, + { + "epoch": 0.8093564682626839, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5228915214538574, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6940486431121826, + "num_tokens": 190759008.0, + "step": 7370 + }, + { + "epoch": 0.8094662859652976, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.511915445327759, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7013428211212158, + "num_tokens": 190784306.0, + "step": 7371 + }, + { + "epoch": 0.8095761036679112, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6471099853515625, + "learning_rate": 1e-06, + "loss": 1.1321, + "mean_token_accuracy": 0.6780049800872803, + "num_tokens": 190812971.0, + "step": 7372 + }, + { + "epoch": 0.809685921370525, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.69387149810791, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7107323408126831, + "num_tokens": 190835592.0, + "step": 7373 + }, + { + "epoch": 0.8097957390731386, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.666264295578003, + "learning_rate": 1e-06, + "loss": 1.1284, + "mean_token_accuracy": 0.678575873374939, + "num_tokens": 190864363.0, + "step": 7374 + }, + { + "epoch": 0.8099055567757523, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 3.0845448970794678, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.7042766809463501, + "num_tokens": 190882458.0, + "step": 7375 + }, + { + "epoch": 0.8100153744783659, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6055963039398193, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6915839910507202, + "num_tokens": 190909073.0, + "step": 7376 + }, + { + "epoch": 0.8101251921809796, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5735130310058594, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7143152952194214, + "num_tokens": 190933418.0, + "step": 7377 + }, + { + "epoch": 0.8102350098835932, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6381475925445557, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6866596341133118, + "num_tokens": 190957448.0, + "step": 7378 + }, + { + "epoch": 0.8103448275862069, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.571582794189453, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7157704830169678, + "num_tokens": 190984754.0, + "step": 7379 + }, + { + "epoch": 0.8104546452888206, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.788560152053833, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6932320594787598, + "num_tokens": 191007254.0, + "step": 7380 + }, + { + "epoch": 0.8105644629914343, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5978660583496094, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7010205984115601, + "num_tokens": 191030203.0, + "step": 7381 + }, + { + "epoch": 0.8106742806940479, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.685230016708374, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6971384286880493, + "num_tokens": 191053705.0, + "step": 7382 + }, + { + "epoch": 0.8107840983966615, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.532050848007202, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.705065131187439, + "num_tokens": 191081547.0, + "step": 7383 + }, + { + "epoch": 0.8108939160992752, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 3.0534510612487793, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7193700075149536, + "num_tokens": 191100151.0, + "step": 7384 + }, + { + "epoch": 0.8110037338018888, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.525630235671997, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7134509682655334, + "num_tokens": 191128014.0, + "step": 7385 + }, + { + "epoch": 0.8111135515045025, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.450437545776367, + "learning_rate": 1e-06, + "loss": 1.1274, + "mean_token_accuracy": 0.6780149340629578, + "num_tokens": 191154703.0, + "step": 7386 + }, + { + "epoch": 0.8112233692071162, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8727047443389893, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7002505660057068, + "num_tokens": 191180285.0, + "step": 7387 + }, + { + "epoch": 0.8113331869097299, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.799797296524048, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6945737600326538, + "num_tokens": 191201760.0, + "step": 7388 + }, + { + "epoch": 0.8114430046123435, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3414499759674072, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6923125982284546, + "num_tokens": 191231226.0, + "step": 7389 + }, + { + "epoch": 0.8115528223149572, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.039034128189087, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6802939772605896, + "num_tokens": 191254797.0, + "step": 7390 + }, + { + "epoch": 0.8116626400175708, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7553625106811523, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7199920415878296, + "num_tokens": 191278029.0, + "step": 7391 + }, + { + "epoch": 0.8117724577201845, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 3.0381462574005127, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6987347602844238, + "num_tokens": 191306205.0, + "step": 7392 + }, + { + "epoch": 0.8118822754227981, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5167760848999023, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6860764026641846, + "num_tokens": 191331158.0, + "step": 7393 + }, + { + "epoch": 0.8119920931254118, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3447885513305664, + "learning_rate": 1e-06, + "loss": 1.1047, + "mean_token_accuracy": 0.6893993616104126, + "num_tokens": 191362487.0, + "step": 7394 + }, + { + "epoch": 0.8121019108280255, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6072885990142822, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6971043348312378, + "num_tokens": 191389257.0, + "step": 7395 + }, + { + "epoch": 0.8122117285306392, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6330370903015137, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6974207162857056, + "num_tokens": 191414774.0, + "step": 7396 + }, + { + "epoch": 0.8123215462332528, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.701953887939453, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7091897130012512, + "num_tokens": 191438055.0, + "step": 7397 + }, + { + "epoch": 0.8124313639358665, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.568619966506958, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6956916451454163, + "num_tokens": 191461414.0, + "step": 7398 + }, + { + "epoch": 0.8125411816384801, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.617849111557007, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7142941355705261, + "num_tokens": 191485426.0, + "step": 7399 + }, + { + "epoch": 0.8126509993410937, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.480419158935547, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.6984533071517944, + "num_tokens": 191513727.0, + "step": 7400 + }, + { + "epoch": 0.8127608170437074, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4584133625030518, + "learning_rate": 1e-06, + "loss": 1.0954, + "mean_token_accuracy": 0.6814063191413879, + "num_tokens": 191541044.0, + "step": 7401 + }, + { + "epoch": 0.8128706347463212, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.162165641784668, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7296634912490845, + "num_tokens": 191575243.0, + "step": 7402 + }, + { + "epoch": 0.8129804524489348, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.59623646736145, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.709568440914154, + "num_tokens": 191599817.0, + "step": 7403 + }, + { + "epoch": 0.8130902701515484, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6922829151153564, + "learning_rate": 1e-06, + "loss": 1.1419, + "mean_token_accuracy": 0.6682311296463013, + "num_tokens": 191633627.0, + "step": 7404 + }, + { + "epoch": 0.8132000878541621, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.481477975845337, + "learning_rate": 1e-06, + "loss": 1.0999, + "mean_token_accuracy": 0.6845430731773376, + "num_tokens": 191661966.0, + "step": 7405 + }, + { + "epoch": 0.8133099055567757, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4970333576202393, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6962100863456726, + "num_tokens": 191689951.0, + "step": 7406 + }, + { + "epoch": 0.8134197232593894, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5371267795562744, + "learning_rate": 1e-06, + "loss": 1.1251, + "mean_token_accuracy": 0.6773718595504761, + "num_tokens": 191716482.0, + "step": 7407 + }, + { + "epoch": 0.813529540962003, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8192191123962402, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6975644826889038, + "num_tokens": 191741044.0, + "step": 7408 + }, + { + "epoch": 0.8136393586646168, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.384371280670166, + "learning_rate": 1e-06, + "loss": 1.1751, + "mean_token_accuracy": 0.6675494313240051, + "num_tokens": 191772667.0, + "step": 7409 + }, + { + "epoch": 0.8137491763672304, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8835551738739014, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7065430879592896, + "num_tokens": 191801028.0, + "step": 7410 + }, + { + "epoch": 0.8138589940698441, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.532824754714966, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.6997961401939392, + "num_tokens": 191825693.0, + "step": 7411 + }, + { + "epoch": 0.8139688117724577, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5114715099334717, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7139673829078674, + "num_tokens": 191850044.0, + "step": 7412 + }, + { + "epoch": 0.8140786294750714, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.567638635635376, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7050598859786987, + "num_tokens": 191873183.0, + "step": 7413 + }, + { + "epoch": 0.814188447177685, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3467226028442383, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7034080028533936, + "num_tokens": 191899944.0, + "step": 7414 + }, + { + "epoch": 0.8142982648802987, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.442680597305298, + "learning_rate": 1e-06, + "loss": 1.1235, + "mean_token_accuracy": 0.6756547689437866, + "num_tokens": 191927847.0, + "step": 7415 + }, + { + "epoch": 0.8144080825829124, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.514720916748047, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7001692056655884, + "num_tokens": 191953831.0, + "step": 7416 + }, + { + "epoch": 0.8145179002855261, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.530897855758667, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.694354772567749, + "num_tokens": 191979703.0, + "step": 7417 + }, + { + "epoch": 0.8146277179881397, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4211952686309814, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7182519435882568, + "num_tokens": 192007014.0, + "step": 7418 + }, + { + "epoch": 0.8147375356907534, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5286026000976562, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6825430989265442, + "num_tokens": 192034396.0, + "step": 7419 + }, + { + "epoch": 0.814847353393367, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.460726022720337, + "learning_rate": 1e-06, + "loss": 1.082, + "mean_token_accuracy": 0.6845695972442627, + "num_tokens": 192065478.0, + "step": 7420 + }, + { + "epoch": 0.8149571710959806, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4959311485290527, + "learning_rate": 1e-06, + "loss": 1.1149, + "mean_token_accuracy": 0.6870704889297485, + "num_tokens": 192091541.0, + "step": 7421 + }, + { + "epoch": 0.8150669887985943, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7401580810546875, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7134323120117188, + "num_tokens": 192113111.0, + "step": 7422 + }, + { + "epoch": 0.8151768065012079, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.700749635696411, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6885637044906616, + "num_tokens": 192139874.0, + "step": 7423 + }, + { + "epoch": 0.8152866242038217, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.274563789367676, + "learning_rate": 1e-06, + "loss": 1.123, + "mean_token_accuracy": 0.6798638105392456, + "num_tokens": 192173021.0, + "step": 7424 + }, + { + "epoch": 0.8153964419064353, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4367311000823975, + "learning_rate": 1e-06, + "loss": 0.9418, + "mean_token_accuracy": 0.7212905883789062, + "num_tokens": 192200042.0, + "step": 7425 + }, + { + "epoch": 0.815506259609049, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.320373058319092, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.6984258890151978, + "num_tokens": 192233652.0, + "step": 7426 + }, + { + "epoch": 0.8156160773116626, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.594128131866455, + "learning_rate": 1e-06, + "loss": 1.0736, + "mean_token_accuracy": 0.6887596845626831, + "num_tokens": 192259177.0, + "step": 7427 + }, + { + "epoch": 0.8157258950142763, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3951973915100098, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.6870643496513367, + "num_tokens": 192289215.0, + "step": 7428 + }, + { + "epoch": 0.8158357127168899, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.555201292037964, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7205895185470581, + "num_tokens": 192311361.0, + "step": 7429 + }, + { + "epoch": 0.8159455304195036, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5146355628967285, + "learning_rate": 1e-06, + "loss": 0.9761, + "mean_token_accuracy": 0.7142056226730347, + "num_tokens": 192336211.0, + "step": 7430 + }, + { + "epoch": 0.8160553481221173, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.554086685180664, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7109717130661011, + "num_tokens": 192359588.0, + "step": 7431 + }, + { + "epoch": 0.816165165824731, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.864992141723633, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6875765323638916, + "num_tokens": 192387109.0, + "step": 7432 + }, + { + "epoch": 0.8162749835273446, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6482813358306885, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6997068524360657, + "num_tokens": 192414021.0, + "step": 7433 + }, + { + "epoch": 0.8163848012299583, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4284324645996094, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6888097524642944, + "num_tokens": 192444406.0, + "step": 7434 + }, + { + "epoch": 0.8164946189325719, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6112818717956543, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6831506490707397, + "num_tokens": 192472462.0, + "step": 7435 + }, + { + "epoch": 0.8166044366351856, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.435347318649292, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.6958500742912292, + "num_tokens": 192501003.0, + "step": 7436 + }, + { + "epoch": 0.8167142543377992, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6923375129699707, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.715314507484436, + "num_tokens": 192521744.0, + "step": 7437 + }, + { + "epoch": 0.816824072040413, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3600382804870605, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7162436246871948, + "num_tokens": 192551757.0, + "step": 7438 + }, + { + "epoch": 0.8169338897430266, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.3537144660949707, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6872761845588684, + "num_tokens": 192585787.0, + "step": 7439 + }, + { + "epoch": 0.8170437074456403, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.847857713699341, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7116273641586304, + "num_tokens": 192605935.0, + "step": 7440 + }, + { + "epoch": 0.8171535251482539, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.621190071105957, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7085654735565186, + "num_tokens": 192629921.0, + "step": 7441 + }, + { + "epoch": 0.8172633428508675, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.338611125946045, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7126673460006714, + "num_tokens": 192660071.0, + "step": 7442 + }, + { + "epoch": 0.8173731605534812, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6095786094665527, + "learning_rate": 1e-06, + "loss": 1.1187, + "mean_token_accuracy": 0.6738131046295166, + "num_tokens": 192686540.0, + "step": 7443 + }, + { + "epoch": 0.8174829782560948, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.362663507461548, + "learning_rate": 1e-06, + "loss": 1.1121, + "mean_token_accuracy": 0.6846005916595459, + "num_tokens": 192718722.0, + "step": 7444 + }, + { + "epoch": 0.8175927959587086, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6708693504333496, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.695383608341217, + "num_tokens": 192743632.0, + "step": 7445 + }, + { + "epoch": 0.8177026136613222, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6578738689422607, + "learning_rate": 1e-06, + "loss": 0.9049, + "mean_token_accuracy": 0.7256734371185303, + "num_tokens": 192768444.0, + "step": 7446 + }, + { + "epoch": 0.8178124313639359, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5787272453308105, + "learning_rate": 1e-06, + "loss": 1.1308, + "mean_token_accuracy": 0.6808507442474365, + "num_tokens": 192795814.0, + "step": 7447 + }, + { + "epoch": 0.8179222490665495, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5362892150878906, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7010054588317871, + "num_tokens": 192821027.0, + "step": 7448 + }, + { + "epoch": 0.8180320667691632, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6525025367736816, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6857572793960571, + "num_tokens": 192846075.0, + "step": 7449 + }, + { + "epoch": 0.8181418844717768, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.281806468963623, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7131630182266235, + "num_tokens": 192875195.0, + "step": 7450 + }, + { + "epoch": 0.8182517021743905, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.603367567062378, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6958536505699158, + "num_tokens": 192899409.0, + "step": 7451 + }, + { + "epoch": 0.8183615198770041, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.455723762512207, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7021309733390808, + "num_tokens": 192924224.0, + "step": 7452 + }, + { + "epoch": 0.8184713375796179, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.843202590942383, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.6994900703430176, + "num_tokens": 192944988.0, + "step": 7453 + }, + { + "epoch": 0.8185811552822315, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.740844249725342, + "learning_rate": 1e-06, + "loss": 0.938, + "mean_token_accuracy": 0.724074125289917, + "num_tokens": 192967089.0, + "step": 7454 + }, + { + "epoch": 0.8186909729848452, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5208401679992676, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.69929039478302, + "num_tokens": 192993389.0, + "step": 7455 + }, + { + "epoch": 0.8188007906874588, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.742033004760742, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7055585384368896, + "num_tokens": 193016217.0, + "step": 7456 + }, + { + "epoch": 0.8189106083900725, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.540781021118164, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6909341216087341, + "num_tokens": 193043656.0, + "step": 7457 + }, + { + "epoch": 0.8190204260926861, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.748796224594116, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.7033425569534302, + "num_tokens": 193066359.0, + "step": 7458 + }, + { + "epoch": 0.8191302437952998, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.34765625, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6850709915161133, + "num_tokens": 193096540.0, + "step": 7459 + }, + { + "epoch": 0.8192400614979135, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.734152317047119, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.698143482208252, + "num_tokens": 193119171.0, + "step": 7460 + }, + { + "epoch": 0.8193498792005272, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5202577114105225, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.695097804069519, + "num_tokens": 193149032.0, + "step": 7461 + }, + { + "epoch": 0.8194596969031408, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.514171600341797, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.702668309211731, + "num_tokens": 193174665.0, + "step": 7462 + }, + { + "epoch": 0.8195695146057544, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7043216228485107, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7046988010406494, + "num_tokens": 193198821.0, + "step": 7463 + }, + { + "epoch": 0.8196793323083681, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.683786153793335, + "learning_rate": 1e-06, + "loss": 0.8462, + "mean_token_accuracy": 0.746364176273346, + "num_tokens": 193219550.0, + "step": 7464 + }, + { + "epoch": 0.8197891500109817, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5575363636016846, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7134832739830017, + "num_tokens": 193245055.0, + "step": 7465 + }, + { + "epoch": 0.8198989677135954, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.550175905227661, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7109532356262207, + "num_tokens": 193274167.0, + "step": 7466 + }, + { + "epoch": 0.8200087854162091, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6576056480407715, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7174994945526123, + "num_tokens": 193297555.0, + "step": 7467 + }, + { + "epoch": 0.8201186031188228, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6220240592956543, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7026705145835876, + "num_tokens": 193324070.0, + "step": 7468 + }, + { + "epoch": 0.8202284208214364, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.858910322189331, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6954132318496704, + "num_tokens": 193347653.0, + "step": 7469 + }, + { + "epoch": 0.8203382385240501, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.510514497756958, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7056787610054016, + "num_tokens": 193372170.0, + "step": 7470 + }, + { + "epoch": 0.8204480562266637, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.473717212677002, + "learning_rate": 1e-06, + "loss": 1.1587, + "mean_token_accuracy": 0.6648059487342834, + "num_tokens": 193404706.0, + "step": 7471 + }, + { + "epoch": 0.8205578739292774, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5055835247039795, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.711728036403656, + "num_tokens": 193431115.0, + "step": 7472 + }, + { + "epoch": 0.820667691631891, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5872244834899902, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.690085232257843, + "num_tokens": 193458507.0, + "step": 7473 + }, + { + "epoch": 0.8207775093345048, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.523031711578369, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6959730386734009, + "num_tokens": 193488381.0, + "step": 7474 + }, + { + "epoch": 0.8208873270371184, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4217209815979004, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6895283460617065, + "num_tokens": 193519142.0, + "step": 7475 + }, + { + "epoch": 0.8209971447397321, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5506420135498047, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7113293409347534, + "num_tokens": 193543492.0, + "step": 7476 + }, + { + "epoch": 0.8211069624423457, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5825161933898926, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6936179399490356, + "num_tokens": 193567255.0, + "step": 7477 + }, + { + "epoch": 0.8212167801449594, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.903043508529663, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7056292295455933, + "num_tokens": 193588674.0, + "step": 7478 + }, + { + "epoch": 0.821326597847573, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5144667625427246, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.689952552318573, + "num_tokens": 193617385.0, + "step": 7479 + }, + { + "epoch": 0.8214364155501866, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5379421710968018, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7269570231437683, + "num_tokens": 193641889.0, + "step": 7480 + }, + { + "epoch": 0.8215462332528004, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8520658016204834, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6846886873245239, + "num_tokens": 193663586.0, + "step": 7481 + }, + { + "epoch": 0.821656050955414, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8099136352539062, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7021284103393555, + "num_tokens": 193685218.0, + "step": 7482 + }, + { + "epoch": 0.8217658686580277, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7021307945251465, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7126708030700684, + "num_tokens": 193708146.0, + "step": 7483 + }, + { + "epoch": 0.8218756863606413, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8373496532440186, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7075046300888062, + "num_tokens": 193727761.0, + "step": 7484 + }, + { + "epoch": 0.821985504063255, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.766291856765747, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6936913728713989, + "num_tokens": 193751345.0, + "step": 7485 + }, + { + "epoch": 0.8220953217658686, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6065869331359863, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.725210428237915, + "num_tokens": 193781026.0, + "step": 7486 + }, + { + "epoch": 0.8222051394684823, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.549457550048828, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7078523635864258, + "num_tokens": 193807025.0, + "step": 7487 + }, + { + "epoch": 0.8223149571710959, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.412123680114746, + "learning_rate": 1e-06, + "loss": 0.9316, + "mean_token_accuracy": 0.7277195453643799, + "num_tokens": 193831146.0, + "step": 7488 + }, + { + "epoch": 0.8224247748737097, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 3.015861988067627, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7236212491989136, + "num_tokens": 193856325.0, + "step": 7489 + }, + { + "epoch": 0.8225345925763233, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.617638111114502, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.692720890045166, + "num_tokens": 193881662.0, + "step": 7490 + }, + { + "epoch": 0.822644410278937, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5376603603363037, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6973844766616821, + "num_tokens": 193909299.0, + "step": 7491 + }, + { + "epoch": 0.8227542279815506, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3382344245910645, + "learning_rate": 1e-06, + "loss": 1.171, + "mean_token_accuracy": 0.6637200713157654, + "num_tokens": 193945491.0, + "step": 7492 + }, + { + "epoch": 0.8228640456841643, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4032845497131348, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.6934273838996887, + "num_tokens": 193972892.0, + "step": 7493 + }, + { + "epoch": 0.8229738633867779, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7870190143585205, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6916593313217163, + "num_tokens": 193994630.0, + "step": 7494 + }, + { + "epoch": 0.8230836810893916, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.744868278503418, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.720012903213501, + "num_tokens": 194016765.0, + "step": 7495 + }, + { + "epoch": 0.8231934987920053, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.531055450439453, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7069995999336243, + "num_tokens": 194041630.0, + "step": 7496 + }, + { + "epoch": 0.823303316494619, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.9954023361206055, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.6965159177780151, + "num_tokens": 194059924.0, + "step": 7497 + }, + { + "epoch": 0.8234131341972326, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.480908155441284, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7013596296310425, + "num_tokens": 194087610.0, + "step": 7498 + }, + { + "epoch": 0.8235229518998463, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5672216415405273, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7147809863090515, + "num_tokens": 194112974.0, + "step": 7499 + }, + { + "epoch": 0.8236327696024599, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5254733562469482, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7216728925704956, + "num_tokens": 194145084.0, + "step": 7500 + }, + { + "epoch": 0.8237425873050735, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4946160316467285, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7028627395629883, + "num_tokens": 194170949.0, + "step": 7501 + }, + { + "epoch": 0.8238524050076872, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.713178873062134, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7057730555534363, + "num_tokens": 194194584.0, + "step": 7502 + }, + { + "epoch": 0.823962222710301, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6035635471343994, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6970340013504028, + "num_tokens": 194219238.0, + "step": 7503 + }, + { + "epoch": 0.8240720404129146, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.463254451751709, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.713924765586853, + "num_tokens": 194245490.0, + "step": 7504 + }, + { + "epoch": 0.8241818581155282, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4603517055511475, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6917884349822998, + "num_tokens": 194277489.0, + "step": 7505 + }, + { + "epoch": 0.8242916758181419, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.444333791732788, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7214541435241699, + "num_tokens": 194304550.0, + "step": 7506 + }, + { + "epoch": 0.8244014935207555, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6349990367889404, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7160636186599731, + "num_tokens": 194328241.0, + "step": 7507 + }, + { + "epoch": 0.8245113112233692, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6788992881774902, + "learning_rate": 1e-06, + "loss": 1.1557, + "mean_token_accuracy": 0.683986246585846, + "num_tokens": 194353809.0, + "step": 7508 + }, + { + "epoch": 0.8246211289259828, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7954633235931396, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.7353741526603699, + "num_tokens": 194374441.0, + "step": 7509 + }, + { + "epoch": 0.8247309466285966, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.4925966262817383, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7101352214813232, + "num_tokens": 194401876.0, + "step": 7510 + }, + { + "epoch": 0.8248407643312102, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6469154357910156, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7221066951751709, + "num_tokens": 194423064.0, + "step": 7511 + }, + { + "epoch": 0.8249505820338239, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.629354953765869, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7153609991073608, + "num_tokens": 194449636.0, + "step": 7512 + }, + { + "epoch": 0.8250603997364375, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.702155590057373, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.71125328540802, + "num_tokens": 194476248.0, + "step": 7513 + }, + { + "epoch": 0.8251702174390512, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3882346153259277, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6952641010284424, + "num_tokens": 194504609.0, + "step": 7514 + }, + { + "epoch": 0.8252800351416648, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6503617763519287, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.703610360622406, + "num_tokens": 194526147.0, + "step": 7515 + }, + { + "epoch": 0.8253898528442785, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4860386848449707, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6963460445404053, + "num_tokens": 194552878.0, + "step": 7516 + }, + { + "epoch": 0.8254996705468921, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.423618793487549, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6935280561447144, + "num_tokens": 194581847.0, + "step": 7517 + }, + { + "epoch": 0.8256094882495059, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3735036849975586, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7011265754699707, + "num_tokens": 194610977.0, + "step": 7518 + }, + { + "epoch": 0.8257193059521195, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.681637763977051, + "learning_rate": 1e-06, + "loss": 0.8527, + "mean_token_accuracy": 0.7480819821357727, + "num_tokens": 194633139.0, + "step": 7519 + }, + { + "epoch": 0.8258291236547332, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.532802104949951, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6923723220825195, + "num_tokens": 194658417.0, + "step": 7520 + }, + { + "epoch": 0.8259389413573468, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7535815238952637, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6999082565307617, + "num_tokens": 194682723.0, + "step": 7521 + }, + { + "epoch": 0.8260487590599604, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.612271308898926, + "learning_rate": 1e-06, + "loss": 1.1514, + "mean_token_accuracy": 0.6679619550704956, + "num_tokens": 194709719.0, + "step": 7522 + }, + { + "epoch": 0.8261585767625741, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.582524538040161, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7150643467903137, + "num_tokens": 194735545.0, + "step": 7523 + }, + { + "epoch": 0.8262683944651877, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7137842178344727, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.704910397529602, + "num_tokens": 194764463.0, + "step": 7524 + }, + { + "epoch": 0.8263782121678015, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5752828121185303, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6992732882499695, + "num_tokens": 194788477.0, + "step": 7525 + }, + { + "epoch": 0.8264880298704151, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4531424045562744, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7069175243377686, + "num_tokens": 194818075.0, + "step": 7526 + }, + { + "epoch": 0.8265978475730288, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4878101348876953, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7042787075042725, + "num_tokens": 194844525.0, + "step": 7527 + }, + { + "epoch": 0.8267076652756424, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.2504448890686035, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6866653561592102, + "num_tokens": 194875526.0, + "step": 7528 + }, + { + "epoch": 0.8268174829782561, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7321720123291016, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7016184329986572, + "num_tokens": 194902455.0, + "step": 7529 + }, + { + "epoch": 0.8269273006808697, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7067408561706543, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7235653400421143, + "num_tokens": 194924955.0, + "step": 7530 + }, + { + "epoch": 0.8270371183834834, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3846335411071777, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.711200475692749, + "num_tokens": 194956923.0, + "step": 7531 + }, + { + "epoch": 0.8271469360860971, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5819313526153564, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7043859958648682, + "num_tokens": 194982918.0, + "step": 7532 + }, + { + "epoch": 0.8272567537887108, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6614277362823486, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6930674314498901, + "num_tokens": 195006390.0, + "step": 7533 + }, + { + "epoch": 0.8273665714913244, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.670125722885132, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7167065143585205, + "num_tokens": 195030450.0, + "step": 7534 + }, + { + "epoch": 0.8274763891939381, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.528331995010376, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.680106520652771, + "num_tokens": 195057157.0, + "step": 7535 + }, + { + "epoch": 0.8275862068965517, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5960581302642822, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6840393543243408, + "num_tokens": 195086748.0, + "step": 7536 + }, + { + "epoch": 0.8276960245991654, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.433323383331299, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6977382302284241, + "num_tokens": 195116699.0, + "step": 7537 + }, + { + "epoch": 0.827805842301779, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4594287872314453, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7039399743080139, + "num_tokens": 195144108.0, + "step": 7538 + }, + { + "epoch": 0.8279156600043928, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6338367462158203, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7085679769515991, + "num_tokens": 195167984.0, + "step": 7539 + }, + { + "epoch": 0.8280254777070064, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8490819931030273, + "learning_rate": 1e-06, + "loss": 0.9043, + "mean_token_accuracy": 0.7348509430885315, + "num_tokens": 195186092.0, + "step": 7540 + }, + { + "epoch": 0.82813529540962, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4731364250183105, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.6996064186096191, + "num_tokens": 195213490.0, + "step": 7541 + }, + { + "epoch": 0.8282451131122337, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7673826217651367, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7017232775688171, + "num_tokens": 195235215.0, + "step": 7542 + }, + { + "epoch": 0.8283549308148473, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3379664421081543, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.684846818447113, + "num_tokens": 195268719.0, + "step": 7543 + }, + { + "epoch": 0.828464748517461, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.306703567504883, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6814860701560974, + "num_tokens": 195302496.0, + "step": 7544 + }, + { + "epoch": 0.8285745662200746, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6636550426483154, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6988651752471924, + "num_tokens": 195328694.0, + "step": 7545 + }, + { + "epoch": 0.8286843839226883, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6679294109344482, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7113039493560791, + "num_tokens": 195352880.0, + "step": 7546 + }, + { + "epoch": 0.828794201625302, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.2627434730529785, + "learning_rate": 1e-06, + "loss": 0.8958, + "mean_token_accuracy": 0.7297804951667786, + "num_tokens": 195381758.0, + "step": 7547 + }, + { + "epoch": 0.8289040193279157, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.583639144897461, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6882157325744629, + "num_tokens": 195406281.0, + "step": 7548 + }, + { + "epoch": 0.8290138370305293, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.720184564590454, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6897695064544678, + "num_tokens": 195431651.0, + "step": 7549 + }, + { + "epoch": 0.829123654733143, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3613831996917725, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7072103023529053, + "num_tokens": 195460148.0, + "step": 7550 + }, + { + "epoch": 0.8292334724357566, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.650317668914795, + "learning_rate": 1e-06, + "loss": 0.8749, + "mean_token_accuracy": 0.7413041591644287, + "num_tokens": 195484865.0, + "step": 7551 + }, + { + "epoch": 0.8293432901383703, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6125171184539795, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7075009942054749, + "num_tokens": 195510871.0, + "step": 7552 + }, + { + "epoch": 0.8294531078409839, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6797282695770264, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6916310787200928, + "num_tokens": 195540329.0, + "step": 7553 + }, + { + "epoch": 0.8295629255435977, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.578026533126831, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7229345440864563, + "num_tokens": 195563569.0, + "step": 7554 + }, + { + "epoch": 0.8296727432462113, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.82421875, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7041857242584229, + "num_tokens": 195584615.0, + "step": 7555 + }, + { + "epoch": 0.829782560948825, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6323320865631104, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6788734793663025, + "num_tokens": 195612512.0, + "step": 7556 + }, + { + "epoch": 0.8298923786514386, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.9072258472442627, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7118914723396301, + "num_tokens": 195632874.0, + "step": 7557 + }, + { + "epoch": 0.8300021963540523, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7639291286468506, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6894477009773254, + "num_tokens": 195655467.0, + "step": 7558 + }, + { + "epoch": 0.8301120140566659, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3674864768981934, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.7019672393798828, + "num_tokens": 195687886.0, + "step": 7559 + }, + { + "epoch": 0.8302218317592795, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5653066635131836, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6999028325080872, + "num_tokens": 195711794.0, + "step": 7560 + }, + { + "epoch": 0.8303316494618933, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5521488189697266, + "learning_rate": 1e-06, + "loss": 1.1343, + "mean_token_accuracy": 0.673820436000824, + "num_tokens": 195742931.0, + "step": 7561 + }, + { + "epoch": 0.830441467164507, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.796034097671509, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7111146450042725, + "num_tokens": 195762619.0, + "step": 7562 + }, + { + "epoch": 0.8305512848671206, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6020421981811523, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7213898301124573, + "num_tokens": 195784610.0, + "step": 7563 + }, + { + "epoch": 0.8306611025697342, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4794294834136963, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6992204189300537, + "num_tokens": 195810897.0, + "step": 7564 + }, + { + "epoch": 0.8307709202723479, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.2288434505462646, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6819637417793274, + "num_tokens": 195846421.0, + "step": 7565 + }, + { + "epoch": 0.8308807379749615, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7259743213653564, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6956368088722229, + "num_tokens": 195868398.0, + "step": 7566 + }, + { + "epoch": 0.8309905556775752, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7309820652008057, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.6948297023773193, + "num_tokens": 195889234.0, + "step": 7567 + }, + { + "epoch": 0.8311003733801889, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.844083786010742, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6994469165802002, + "num_tokens": 195908887.0, + "step": 7568 + }, + { + "epoch": 0.8312101910828026, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.73435640335083, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6854212284088135, + "num_tokens": 195932480.0, + "step": 7569 + }, + { + "epoch": 0.8313200087854162, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.514662504196167, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6978304386138916, + "num_tokens": 195959682.0, + "step": 7570 + }, + { + "epoch": 0.8314298264880299, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5207161903381348, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6832795143127441, + "num_tokens": 195992913.0, + "step": 7571 + }, + { + "epoch": 0.8315396441906435, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.911553382873535, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7057598829269409, + "num_tokens": 196013906.0, + "step": 7572 + }, + { + "epoch": 0.8316494618932572, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7981605529785156, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6797875165939331, + "num_tokens": 196039076.0, + "step": 7573 + }, + { + "epoch": 0.8317592795958708, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.380674362182617, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.6973565220832825, + "num_tokens": 196070521.0, + "step": 7574 + }, + { + "epoch": 0.8318690972984845, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.44832706451416, + "learning_rate": 1e-06, + "loss": 1.1309, + "mean_token_accuracy": 0.6729180812835693, + "num_tokens": 196097730.0, + "step": 7575 + }, + { + "epoch": 0.8319789150010982, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7864797115325928, + "learning_rate": 1e-06, + "loss": 0.8824, + "mean_token_accuracy": 0.7357010841369629, + "num_tokens": 196117974.0, + "step": 7576 + }, + { + "epoch": 0.8320887327037119, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4708471298217773, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.698691725730896, + "num_tokens": 196146404.0, + "step": 7577 + }, + { + "epoch": 0.8321985504063255, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5429656505584717, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7098303437232971, + "num_tokens": 196171219.0, + "step": 7578 + }, + { + "epoch": 0.8323083681089392, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8018107414245605, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6786074638366699, + "num_tokens": 196194458.0, + "step": 7579 + }, + { + "epoch": 0.8324181858115528, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.499143362045288, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7193555235862732, + "num_tokens": 196219471.0, + "step": 7580 + }, + { + "epoch": 0.8325280035141664, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6842057704925537, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6893472075462341, + "num_tokens": 196243947.0, + "step": 7581 + }, + { + "epoch": 0.8326378212167801, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4665443897247314, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.6949819922447205, + "num_tokens": 196270203.0, + "step": 7582 + }, + { + "epoch": 0.8327476389193939, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6520822048187256, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7077393531799316, + "num_tokens": 196297066.0, + "step": 7583 + }, + { + "epoch": 0.8328574566220075, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7264339923858643, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7079591751098633, + "num_tokens": 196318379.0, + "step": 7584 + }, + { + "epoch": 0.8329672743246211, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8107826709747314, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7131353616714478, + "num_tokens": 196338137.0, + "step": 7585 + }, + { + "epoch": 0.8330770920272348, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.4794862270355225, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.692079484462738, + "num_tokens": 196367424.0, + "step": 7586 + }, + { + "epoch": 0.8331869097298484, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.5583696365356445, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7113518714904785, + "num_tokens": 196391314.0, + "step": 7587 + }, + { + "epoch": 0.8332967274324621, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.7140650749206543, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.6949665546417236, + "num_tokens": 196416374.0, + "step": 7588 + }, + { + "epoch": 0.8334065451350757, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4436020851135254, + "learning_rate": 1e-06, + "loss": 0.9186, + "mean_token_accuracy": 0.7258202433586121, + "num_tokens": 196441639.0, + "step": 7589 + }, + { + "epoch": 0.8335163628376895, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.262420415878296, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7092334032058716, + "num_tokens": 196471450.0, + "step": 7590 + }, + { + "epoch": 0.8336261805403031, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.432633638381958, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.6857525110244751, + "num_tokens": 196500994.0, + "step": 7591 + }, + { + "epoch": 0.8337359982429168, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4865481853485107, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7070808410644531, + "num_tokens": 196526947.0, + "step": 7592 + }, + { + "epoch": 0.8338458159455304, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7365763187408447, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7009513974189758, + "num_tokens": 196550249.0, + "step": 7593 + }, + { + "epoch": 0.8339556336481441, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6869678497314453, + "learning_rate": 1e-06, + "loss": 1.1349, + "mean_token_accuracy": 0.6808210611343384, + "num_tokens": 196577281.0, + "step": 7594 + }, + { + "epoch": 0.8340654513507577, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5908772945404053, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7034900188446045, + "num_tokens": 196602095.0, + "step": 7595 + }, + { + "epoch": 0.8341752690533714, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.5659754276275635, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6903752684593201, + "num_tokens": 196629629.0, + "step": 7596 + }, + { + "epoch": 0.8342850867559851, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8493454456329346, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7010093927383423, + "num_tokens": 196652009.0, + "step": 7597 + }, + { + "epoch": 0.8343949044585988, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6003947257995605, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7175739407539368, + "num_tokens": 196677268.0, + "step": 7598 + }, + { + "epoch": 0.8345047221612124, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6162712574005127, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6909070014953613, + "num_tokens": 196703110.0, + "step": 7599 + }, + { + "epoch": 0.834614539863826, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.2817909717559814, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.6954537630081177, + "num_tokens": 196735845.0, + "step": 7600 + }, + { + "epoch": 0.8347243575664397, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6338796615600586, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.6983458995819092, + "num_tokens": 196764382.0, + "step": 7601 + }, + { + "epoch": 0.8348341752690533, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.8043665885925293, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7170277833938599, + "num_tokens": 196785955.0, + "step": 7602 + }, + { + "epoch": 0.834943992971667, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.500408411026001, + "learning_rate": 1e-06, + "loss": 1.1539, + "mean_token_accuracy": 0.6694648265838623, + "num_tokens": 196815121.0, + "step": 7603 + }, + { + "epoch": 0.8350538106742806, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3059778213500977, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7085064053535461, + "num_tokens": 196844968.0, + "step": 7604 + }, + { + "epoch": 0.8351636283768944, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.498866081237793, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6969321966171265, + "num_tokens": 196872347.0, + "step": 7605 + }, + { + "epoch": 0.835273446079508, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.62595796585083, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6925830841064453, + "num_tokens": 196896899.0, + "step": 7606 + }, + { + "epoch": 0.8353832637821217, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.602736234664917, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6981855630874634, + "num_tokens": 196925898.0, + "step": 7607 + }, + { + "epoch": 0.8354930814847353, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6696834564208984, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6934030055999756, + "num_tokens": 196949512.0, + "step": 7608 + }, + { + "epoch": 0.835602899187349, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.888420343399048, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7201399803161621, + "num_tokens": 196969131.0, + "step": 7609 + }, + { + "epoch": 0.8357127168899626, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.705899953842163, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6881306171417236, + "num_tokens": 196996637.0, + "step": 7610 + }, + { + "epoch": 0.8358225345925763, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.793205976486206, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.7030684351921082, + "num_tokens": 197017944.0, + "step": 7611 + }, + { + "epoch": 0.83593235229519, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.791999340057373, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7123825550079346, + "num_tokens": 197039417.0, + "step": 7612 + }, + { + "epoch": 0.8360421699978037, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.8228342533111572, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7067244648933411, + "num_tokens": 197063264.0, + "step": 7613 + }, + { + "epoch": 0.8361519877004173, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6575162410736084, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7243895530700684, + "num_tokens": 197085752.0, + "step": 7614 + }, + { + "epoch": 0.836261805403031, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.3916471004486084, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6878846883773804, + "num_tokens": 197114138.0, + "step": 7615 + }, + { + "epoch": 0.8363716231056446, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.6752054691314697, + "learning_rate": 1e-06, + "loss": 1.1055, + "mean_token_accuracy": 0.6787605881690979, + "num_tokens": 197136693.0, + "step": 7616 + }, + { + "epoch": 0.8364814408082583, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4739339351654053, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7172647714614868, + "num_tokens": 197161869.0, + "step": 7617 + }, + { + "epoch": 0.8365912585108719, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5748980045318604, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7108585834503174, + "num_tokens": 197192424.0, + "step": 7618 + }, + { + "epoch": 0.8367010762134857, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.3239474296569824, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7048007249832153, + "num_tokens": 197221800.0, + "step": 7619 + }, + { + "epoch": 0.8368108939160993, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.497877359390259, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7028254270553589, + "num_tokens": 197249437.0, + "step": 7620 + }, + { + "epoch": 0.836920711618713, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.7698426246643066, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.711388111114502, + "num_tokens": 197271744.0, + "step": 7621 + }, + { + "epoch": 0.8370305293213266, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.68569278717041, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7044545412063599, + "num_tokens": 197295207.0, + "step": 7622 + }, + { + "epoch": 0.8371403470239402, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.798051357269287, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.703513503074646, + "num_tokens": 197319632.0, + "step": 7623 + }, + { + "epoch": 0.8372501647265539, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5408246517181396, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6794347167015076, + "num_tokens": 197350363.0, + "step": 7624 + }, + { + "epoch": 0.8373599824291675, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.512035369873047, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7174171209335327, + "num_tokens": 197375405.0, + "step": 7625 + }, + { + "epoch": 0.8374698001317813, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6135916709899902, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6882331371307373, + "num_tokens": 197401800.0, + "step": 7626 + }, + { + "epoch": 0.8375796178343949, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6285150051116943, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6902967691421509, + "num_tokens": 197425361.0, + "step": 7627 + }, + { + "epoch": 0.8376894355370086, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.440068006515503, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.7066551446914673, + "num_tokens": 197451855.0, + "step": 7628 + }, + { + "epoch": 0.8377992532396222, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.593456983566284, + "learning_rate": 1e-06, + "loss": 1.091, + "mean_token_accuracy": 0.6864117383956909, + "num_tokens": 197475433.0, + "step": 7629 + }, + { + "epoch": 0.8379090709422359, + "ewc_loss": 7.927417755126953e-06, + "grad_norm": 2.841139316558838, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7085601091384888, + "num_tokens": 197495708.0, + "step": 7630 + }, + { + "epoch": 0.8380188886448495, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.34376859664917, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6906985640525818, + "num_tokens": 197529731.0, + "step": 7631 + }, + { + "epoch": 0.8381287063474632, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5354652404785156, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.68550705909729, + "num_tokens": 197555343.0, + "step": 7632 + }, + { + "epoch": 0.8382385240500769, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.308946132659912, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7119461894035339, + "num_tokens": 197585097.0, + "step": 7633 + }, + { + "epoch": 0.8383483417526906, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.6697356700897217, + "learning_rate": 1e-06, + "loss": 0.8907, + "mean_token_accuracy": 0.7348807454109192, + "num_tokens": 197610696.0, + "step": 7634 + }, + { + "epoch": 0.8384581594553042, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4644534587860107, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6932960748672485, + "num_tokens": 197638076.0, + "step": 7635 + }, + { + "epoch": 0.8385679771579179, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8925485610961914, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7063673734664917, + "num_tokens": 197659044.0, + "step": 7636 + }, + { + "epoch": 0.8386777948605315, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6500070095062256, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7018798589706421, + "num_tokens": 197684359.0, + "step": 7637 + }, + { + "epoch": 0.8387876125631452, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8354313373565674, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7018906474113464, + "num_tokens": 197706907.0, + "step": 7638 + }, + { + "epoch": 0.8388974302657588, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.399345636367798, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7017886638641357, + "num_tokens": 197734448.0, + "step": 7639 + }, + { + "epoch": 0.8390072479683724, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.497506618499756, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7010531425476074, + "num_tokens": 197760172.0, + "step": 7640 + }, + { + "epoch": 0.8391170656709862, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.467101812362671, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.6989219784736633, + "num_tokens": 197786884.0, + "step": 7641 + }, + { + "epoch": 0.8392268833735999, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5832669734954834, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.704648494720459, + "num_tokens": 197809095.0, + "step": 7642 + }, + { + "epoch": 0.8393367010762135, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5045132637023926, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7070890069007874, + "num_tokens": 197835389.0, + "step": 7643 + }, + { + "epoch": 0.8394465187788271, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.766418933868408, + "learning_rate": 1e-06, + "loss": 0.9022, + "mean_token_accuracy": 0.7272559404373169, + "num_tokens": 197857238.0, + "step": 7644 + }, + { + "epoch": 0.8395563364814408, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.2538633346557617, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6909735202789307, + "num_tokens": 197891332.0, + "step": 7645 + }, + { + "epoch": 0.8396661541840544, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.712233543395996, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6863323450088501, + "num_tokens": 197917516.0, + "step": 7646 + }, + { + "epoch": 0.8397759718866681, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.763026714324951, + "learning_rate": 1e-06, + "loss": 0.9341, + "mean_token_accuracy": 0.7162560224533081, + "num_tokens": 197937808.0, + "step": 7647 + }, + { + "epoch": 0.8398857895892818, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5533769130706787, + "learning_rate": 1e-06, + "loss": 1.149, + "mean_token_accuracy": 0.667086124420166, + "num_tokens": 197970058.0, + "step": 7648 + }, + { + "epoch": 0.8399956072918955, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5343477725982666, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6866833567619324, + "num_tokens": 198000588.0, + "step": 7649 + }, + { + "epoch": 0.8401054249945091, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.286360502243042, + "learning_rate": 1e-06, + "loss": 1.0982, + "mean_token_accuracy": 0.6909617185592651, + "num_tokens": 198030599.0, + "step": 7650 + }, + { + "epoch": 0.8402152426971228, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4883768558502197, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.705103874206543, + "num_tokens": 198055412.0, + "step": 7651 + }, + { + "epoch": 0.8403250603997364, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4832491874694824, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7169821858406067, + "num_tokens": 198079265.0, + "step": 7652 + }, + { + "epoch": 0.8404348781023501, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.933971881866455, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.706061840057373, + "num_tokens": 198099381.0, + "step": 7653 + }, + { + "epoch": 0.8405446958049637, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.646223306655884, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6968151330947876, + "num_tokens": 198122592.0, + "step": 7654 + }, + { + "epoch": 0.8406545135075775, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6320180892944336, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.7040222883224487, + "num_tokens": 198146233.0, + "step": 7655 + }, + { + "epoch": 0.8407643312101911, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.62361741065979, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7065004110336304, + "num_tokens": 198170128.0, + "step": 7656 + }, + { + "epoch": 0.8408741489128048, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.868300437927246, + "learning_rate": 1e-06, + "loss": 0.9237, + "mean_token_accuracy": 0.7222634553909302, + "num_tokens": 198195310.0, + "step": 7657 + }, + { + "epoch": 0.8409839666154184, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.532855749130249, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6970070600509644, + "num_tokens": 198221303.0, + "step": 7658 + }, + { + "epoch": 0.8410937843180321, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.713963270187378, + "learning_rate": 1e-06, + "loss": 1.1307, + "mean_token_accuracy": 0.6827064752578735, + "num_tokens": 198247562.0, + "step": 7659 + }, + { + "epoch": 0.8412036020206457, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5006041526794434, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6849634647369385, + "num_tokens": 198276906.0, + "step": 7660 + }, + { + "epoch": 0.8413134197232593, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.9890835285186768, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7055881023406982, + "num_tokens": 198295104.0, + "step": 7661 + }, + { + "epoch": 0.8414232374258731, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4005308151245117, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6964821815490723, + "num_tokens": 198323331.0, + "step": 7662 + }, + { + "epoch": 0.8415330551284868, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.474144697189331, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7000769376754761, + "num_tokens": 198350668.0, + "step": 7663 + }, + { + "epoch": 0.8416428728311004, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.517448902130127, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7102069854736328, + "num_tokens": 198381292.0, + "step": 7664 + }, + { + "epoch": 0.841752690533714, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.543637275695801, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7067031860351562, + "num_tokens": 198406274.0, + "step": 7665 + }, + { + "epoch": 0.8418625082363277, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.520970344543457, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6921101808547974, + "num_tokens": 198430130.0, + "step": 7666 + }, + { + "epoch": 0.8419723259389413, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.2580883502960205, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7231990098953247, + "num_tokens": 198459777.0, + "step": 7667 + }, + { + "epoch": 0.842082143641555, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.627995014190674, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6949748992919922, + "num_tokens": 198485192.0, + "step": 7668 + }, + { + "epoch": 0.8421919613441686, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8710782527923584, + "learning_rate": 1e-06, + "loss": 0.9144, + "mean_token_accuracy": 0.7221161127090454, + "num_tokens": 198503231.0, + "step": 7669 + }, + { + "epoch": 0.8423017790467824, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6491429805755615, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7139805555343628, + "num_tokens": 198528863.0, + "step": 7670 + }, + { + "epoch": 0.842411596749396, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5543389320373535, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7112724781036377, + "num_tokens": 198553857.0, + "step": 7671 + }, + { + "epoch": 0.8425214144520097, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6329379081726074, + "learning_rate": 1e-06, + "loss": 1.1102, + "mean_token_accuracy": 0.6948934197425842, + "num_tokens": 198583558.0, + "step": 7672 + }, + { + "epoch": 0.8426312321546233, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6474568843841553, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7221270799636841, + "num_tokens": 198609291.0, + "step": 7673 + }, + { + "epoch": 0.842741049857237, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.465731620788574, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7064921259880066, + "num_tokens": 198636900.0, + "step": 7674 + }, + { + "epoch": 0.8428508675598506, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.3786611557006836, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6864822506904602, + "num_tokens": 198668433.0, + "step": 7675 + }, + { + "epoch": 0.8429606852624643, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6482326984405518, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7081082463264465, + "num_tokens": 198694562.0, + "step": 7676 + }, + { + "epoch": 0.843070502965078, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6783409118652344, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.7003454566001892, + "num_tokens": 198717369.0, + "step": 7677 + }, + { + "epoch": 0.8431803206676917, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5481810569763184, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7051075100898743, + "num_tokens": 198744856.0, + "step": 7678 + }, + { + "epoch": 0.8432901383703053, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5012261867523193, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6834094524383545, + "num_tokens": 198771538.0, + "step": 7679 + }, + { + "epoch": 0.843399956072919, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6132121086120605, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.6843176484107971, + "num_tokens": 198796885.0, + "step": 7680 + }, + { + "epoch": 0.8435097737755326, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.752469062805176, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.711341381072998, + "num_tokens": 198818940.0, + "step": 7681 + }, + { + "epoch": 0.8436195914781462, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 3.1045877933502197, + "learning_rate": 1e-06, + "loss": 1.116, + "mean_token_accuracy": 0.6702189445495605, + "num_tokens": 198845612.0, + "step": 7682 + }, + { + "epoch": 0.8437294091807599, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5818581581115723, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.704744815826416, + "num_tokens": 198874429.0, + "step": 7683 + }, + { + "epoch": 0.8438392268833736, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.766913890838623, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7060973644256592, + "num_tokens": 198899257.0, + "step": 7684 + }, + { + "epoch": 0.8439490445859873, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6515610218048096, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.708510160446167, + "num_tokens": 198923303.0, + "step": 7685 + }, + { + "epoch": 0.8440588622886009, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5222842693328857, + "learning_rate": 1e-06, + "loss": 1.1783, + "mean_token_accuracy": 0.657034158706665, + "num_tokens": 198953795.0, + "step": 7686 + }, + { + "epoch": 0.8441686799912146, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5300278663635254, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.6973620057106018, + "num_tokens": 198981725.0, + "step": 7687 + }, + { + "epoch": 0.8442784976938282, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.603724241256714, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.707838773727417, + "num_tokens": 199005153.0, + "step": 7688 + }, + { + "epoch": 0.8443883153964419, + "ewc_loss": 7.987022399902344e-06, + "grad_norm": 2.805786371231079, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7216131687164307, + "num_tokens": 199027042.0, + "step": 7689 + }, + { + "epoch": 0.8444981330990555, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.713881492614746, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7070742249488831, + "num_tokens": 199050146.0, + "step": 7690 + }, + { + "epoch": 0.8446079508016693, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.709575891494751, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6923760175704956, + "num_tokens": 199073404.0, + "step": 7691 + }, + { + "epoch": 0.8447177685042829, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.531423568725586, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6966606378555298, + "num_tokens": 199101008.0, + "step": 7692 + }, + { + "epoch": 0.8448275862068966, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6621320247650146, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.6897307634353638, + "num_tokens": 199126165.0, + "step": 7693 + }, + { + "epoch": 0.8449374039095102, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.633840560913086, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7001530528068542, + "num_tokens": 199153533.0, + "step": 7694 + }, + { + "epoch": 0.8450472216121239, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.438962459564209, + "learning_rate": 1e-06, + "loss": 0.8972, + "mean_token_accuracy": 0.7321383953094482, + "num_tokens": 199179424.0, + "step": 7695 + }, + { + "epoch": 0.8451570393147375, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.721179962158203, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6963902711868286, + "num_tokens": 199202764.0, + "step": 7696 + }, + { + "epoch": 0.8452668570173512, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.526827096939087, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7106120586395264, + "num_tokens": 199227940.0, + "step": 7697 + }, + { + "epoch": 0.8453766747199648, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6420209407806396, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.704809308052063, + "num_tokens": 199252255.0, + "step": 7698 + }, + { + "epoch": 0.8454864924225786, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.563837766647339, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7071776390075684, + "num_tokens": 199278817.0, + "step": 7699 + }, + { + "epoch": 0.8455963101251922, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6011204719543457, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7047383785247803, + "num_tokens": 199304937.0, + "step": 7700 + }, + { + "epoch": 0.8457061278278059, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.2295901775360107, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7257944345474243, + "num_tokens": 199337317.0, + "step": 7701 + }, + { + "epoch": 0.8458159455304195, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.559978723526001, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7032864093780518, + "num_tokens": 199363824.0, + "step": 7702 + }, + { + "epoch": 0.8459257632330331, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.661229133605957, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6995335221290588, + "num_tokens": 199388320.0, + "step": 7703 + }, + { + "epoch": 0.8460355809356468, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.521697759628296, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.6938324570655823, + "num_tokens": 199415523.0, + "step": 7704 + }, + { + "epoch": 0.8461453986382604, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6336829662323, + "learning_rate": 1e-06, + "loss": 1.1504, + "mean_token_accuracy": 0.6828229427337646, + "num_tokens": 199441645.0, + "step": 7705 + }, + { + "epoch": 0.8462552163408742, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7155826091766357, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7140535116195679, + "num_tokens": 199462863.0, + "step": 7706 + }, + { + "epoch": 0.8463650340434878, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7391445636749268, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.7236614227294922, + "num_tokens": 199484387.0, + "step": 7707 + }, + { + "epoch": 0.8464748517461015, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.568882942199707, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.6977373957633972, + "num_tokens": 199513679.0, + "step": 7708 + }, + { + "epoch": 0.8465846694487151, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.575147867202759, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.6845710277557373, + "num_tokens": 199542808.0, + "step": 7709 + }, + { + "epoch": 0.8466944871513288, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7072830200195312, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6852067112922668, + "num_tokens": 199566968.0, + "step": 7710 + }, + { + "epoch": 0.8468043048539424, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.507636308670044, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7132889628410339, + "num_tokens": 199593692.0, + "step": 7711 + }, + { + "epoch": 0.8469141225565561, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4356558322906494, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.6995173096656799, + "num_tokens": 199622402.0, + "step": 7712 + }, + { + "epoch": 0.8470239402591698, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 3.1038620471954346, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7135370969772339, + "num_tokens": 199641522.0, + "step": 7713 + }, + { + "epoch": 0.8471337579617835, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5476653575897217, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7039300203323364, + "num_tokens": 199667927.0, + "step": 7714 + }, + { + "epoch": 0.8472435756643971, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6766858100891113, + "learning_rate": 1e-06, + "loss": 1.1076, + "mean_token_accuracy": 0.6846613883972168, + "num_tokens": 199693552.0, + "step": 7715 + }, + { + "epoch": 0.8473533933670108, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6580538749694824, + "learning_rate": 1e-06, + "loss": 1.1218, + "mean_token_accuracy": 0.676958441734314, + "num_tokens": 199716886.0, + "step": 7716 + }, + { + "epoch": 0.8474632110696244, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7953310012817383, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7063977718353271, + "num_tokens": 199737928.0, + "step": 7717 + }, + { + "epoch": 0.8475730287722381, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.572554588317871, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7055703401565552, + "num_tokens": 199766914.0, + "step": 7718 + }, + { + "epoch": 0.8476828464748517, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4467949867248535, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7019981145858765, + "num_tokens": 199797223.0, + "step": 7719 + }, + { + "epoch": 0.8477926641774655, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.594059944152832, + "learning_rate": 1e-06, + "loss": 0.9005, + "mean_token_accuracy": 0.7316868901252747, + "num_tokens": 199819060.0, + "step": 7720 + }, + { + "epoch": 0.8479024818800791, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.63383412361145, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7091323733329773, + "num_tokens": 199845241.0, + "step": 7721 + }, + { + "epoch": 0.8480122995826928, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.486556053161621, + "learning_rate": 1e-06, + "loss": 0.9405, + "mean_token_accuracy": 0.7203918099403381, + "num_tokens": 199868686.0, + "step": 7722 + }, + { + "epoch": 0.8481221172853064, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.782301664352417, + "learning_rate": 1e-06, + "loss": 1.1322, + "mean_token_accuracy": 0.6777105331420898, + "num_tokens": 199892312.0, + "step": 7723 + }, + { + "epoch": 0.84823193498792, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5645506381988525, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.706390380859375, + "num_tokens": 199917331.0, + "step": 7724 + }, + { + "epoch": 0.8483417526905337, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.498414993286133, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.703220009803772, + "num_tokens": 199942069.0, + "step": 7725 + }, + { + "epoch": 0.8484515703931473, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.461815357208252, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6871762275695801, + "num_tokens": 199967921.0, + "step": 7726 + }, + { + "epoch": 0.848561388095761, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6638331413269043, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7144359350204468, + "num_tokens": 199993258.0, + "step": 7727 + }, + { + "epoch": 0.8486712057983747, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.443998098373413, + "learning_rate": 1e-06, + "loss": 1.1275, + "mean_token_accuracy": 0.6708998084068298, + "num_tokens": 200025395.0, + "step": 7728 + }, + { + "epoch": 0.8487810235009884, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8495066165924072, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6889198422431946, + "num_tokens": 200047088.0, + "step": 7729 + }, + { + "epoch": 0.848890841203602, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5175747871398926, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.711414098739624, + "num_tokens": 200073044.0, + "step": 7730 + }, + { + "epoch": 0.8490006589062157, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.3856253623962402, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.698102593421936, + "num_tokens": 200103693.0, + "step": 7731 + }, + { + "epoch": 0.8491104766088293, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4633986949920654, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7105329036712646, + "num_tokens": 200129605.0, + "step": 7732 + }, + { + "epoch": 0.849220294311443, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.442441701889038, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6945861577987671, + "num_tokens": 200159419.0, + "step": 7733 + }, + { + "epoch": 0.8493301120140566, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6514334678649902, + "learning_rate": 1e-06, + "loss": 1.0997, + "mean_token_accuracy": 0.6806303858757019, + "num_tokens": 200183858.0, + "step": 7734 + }, + { + "epoch": 0.8494399297166704, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.418437957763672, + "learning_rate": 1e-06, + "loss": 1.0909, + "mean_token_accuracy": 0.6810750961303711, + "num_tokens": 200215080.0, + "step": 7735 + }, + { + "epoch": 0.849549747419284, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6955976486206055, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6862948536872864, + "num_tokens": 200236944.0, + "step": 7736 + }, + { + "epoch": 0.8496595651218977, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4855663776397705, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6974133849143982, + "num_tokens": 200265185.0, + "step": 7737 + }, + { + "epoch": 0.8497693828245113, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.2806358337402344, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7016178369522095, + "num_tokens": 200293716.0, + "step": 7738 + }, + { + "epoch": 0.849879200527125, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.685612201690674, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.6980810165405273, + "num_tokens": 200318403.0, + "step": 7739 + }, + { + "epoch": 0.8499890182297386, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.460782527923584, + "learning_rate": 1e-06, + "loss": 0.9201, + "mean_token_accuracy": 0.7277446389198303, + "num_tokens": 200343584.0, + "step": 7740 + }, + { + "epoch": 0.8500988359323522, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.2840213775634766, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7033133506774902, + "num_tokens": 200375251.0, + "step": 7741 + }, + { + "epoch": 0.850208653634966, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4016854763031006, + "learning_rate": 1e-06, + "loss": 1.1049, + "mean_token_accuracy": 0.6888015270233154, + "num_tokens": 200406054.0, + "step": 7742 + }, + { + "epoch": 0.8503184713375797, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.930964946746826, + "learning_rate": 1e-06, + "loss": 1.0956, + "mean_token_accuracy": 0.7023193836212158, + "num_tokens": 200427875.0, + "step": 7743 + }, + { + "epoch": 0.8504282890401933, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.541227102279663, + "learning_rate": 1e-06, + "loss": 0.9318, + "mean_token_accuracy": 0.7223106622695923, + "num_tokens": 200454987.0, + "step": 7744 + }, + { + "epoch": 0.8505381067428069, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5473453998565674, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7257658839225769, + "num_tokens": 200478848.0, + "step": 7745 + }, + { + "epoch": 0.8506479244454206, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.459627389907837, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7107933163642883, + "num_tokens": 200502938.0, + "step": 7746 + }, + { + "epoch": 0.8507577421480342, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.523592233657837, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7245330810546875, + "num_tokens": 200527962.0, + "step": 7747 + }, + { + "epoch": 0.8508675598506479, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.471940517425537, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6998940110206604, + "num_tokens": 200554384.0, + "step": 7748 + }, + { + "epoch": 0.8509773775532616, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.498340129852295, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.6938768625259399, + "num_tokens": 200580256.0, + "step": 7749 + }, + { + "epoch": 0.8510871952558753, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7472121715545654, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6985175609588623, + "num_tokens": 200602913.0, + "step": 7750 + }, + { + "epoch": 0.8511970129584889, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.325263261795044, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6918260455131531, + "num_tokens": 200637997.0, + "step": 7751 + }, + { + "epoch": 0.8513068306611026, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6583383083343506, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7147073745727539, + "num_tokens": 200663680.0, + "step": 7752 + }, + { + "epoch": 0.8514166483637162, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.766357421875, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7184482216835022, + "num_tokens": 200687698.0, + "step": 7753 + }, + { + "epoch": 0.8515264660663299, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6376748085021973, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.6888010501861572, + "num_tokens": 200718240.0, + "step": 7754 + }, + { + "epoch": 0.8516362837689435, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.593703508377075, + "learning_rate": 1e-06, + "loss": 1.0975, + "mean_token_accuracy": 0.6774673461914062, + "num_tokens": 200743003.0, + "step": 7755 + }, + { + "epoch": 0.8517461014715572, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.431981325149536, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6810882091522217, + "num_tokens": 200772187.0, + "step": 7756 + }, + { + "epoch": 0.8518559191741709, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5698022842407227, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7187808156013489, + "num_tokens": 200800494.0, + "step": 7757 + }, + { + "epoch": 0.8519657368767846, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8490402698516846, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7160501480102539, + "num_tokens": 200820127.0, + "step": 7758 + }, + { + "epoch": 0.8520755545793982, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 3.1456892490386963, + "learning_rate": 1e-06, + "loss": 0.9031, + "mean_token_accuracy": 0.7280089855194092, + "num_tokens": 200838496.0, + "step": 7759 + }, + { + "epoch": 0.8521853722820119, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.701946258544922, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.699893593788147, + "num_tokens": 200863720.0, + "step": 7760 + }, + { + "epoch": 0.8522951899846255, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.565335273742676, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.7031404972076416, + "num_tokens": 200890496.0, + "step": 7761 + }, + { + "epoch": 0.8524050076872391, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.449045181274414, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7078338861465454, + "num_tokens": 200920864.0, + "step": 7762 + }, + { + "epoch": 0.8525148253898528, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.757519245147705, + "learning_rate": 1e-06, + "loss": 1.1129, + "mean_token_accuracy": 0.6774375438690186, + "num_tokens": 200945141.0, + "step": 7763 + }, + { + "epoch": 0.8526246430924665, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.523534059524536, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.702878475189209, + "num_tokens": 200973178.0, + "step": 7764 + }, + { + "epoch": 0.8527344607950802, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7093117237091064, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7145445942878723, + "num_tokens": 200996047.0, + "step": 7765 + }, + { + "epoch": 0.8528442784976938, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.587181329727173, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6970641613006592, + "num_tokens": 201021146.0, + "step": 7766 + }, + { + "epoch": 0.8529540962003075, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5712814331054688, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7162384390830994, + "num_tokens": 201046614.0, + "step": 7767 + }, + { + "epoch": 0.8530639139029211, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6830265522003174, + "learning_rate": 1e-06, + "loss": 0.9442, + "mean_token_accuracy": 0.7291595935821533, + "num_tokens": 201068484.0, + "step": 7768 + }, + { + "epoch": 0.8531737316055348, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.717940092086792, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7084318399429321, + "num_tokens": 201089170.0, + "step": 7769 + }, + { + "epoch": 0.8532835493081484, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9232733249664307, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6860420107841492, + "num_tokens": 201112542.0, + "step": 7770 + }, + { + "epoch": 0.8533933670107622, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5227439403533936, + "learning_rate": 1e-06, + "loss": 0.9036, + "mean_token_accuracy": 0.7317303419113159, + "num_tokens": 201136180.0, + "step": 7771 + }, + { + "epoch": 0.8535031847133758, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.806551933288574, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.701840877532959, + "num_tokens": 201159136.0, + "step": 7772 + }, + { + "epoch": 0.8536130024159895, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4168899059295654, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.689307689666748, + "num_tokens": 201188466.0, + "step": 7773 + }, + { + "epoch": 0.8537228201186031, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4300830364227295, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6837204098701477, + "num_tokens": 201214870.0, + "step": 7774 + }, + { + "epoch": 0.8538326378212168, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6458232402801514, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.7104212045669556, + "num_tokens": 201238516.0, + "step": 7775 + }, + { + "epoch": 0.8539424555238304, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4316940307617188, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7054705619812012, + "num_tokens": 201264492.0, + "step": 7776 + }, + { + "epoch": 0.8540522732264441, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.2397329807281494, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6819934248924255, + "num_tokens": 201300458.0, + "step": 7777 + }, + { + "epoch": 0.8541620909290578, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.173158884048462, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7088811993598938, + "num_tokens": 201337032.0, + "step": 7778 + }, + { + "epoch": 0.8542719086316715, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7400848865509033, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7053540945053101, + "num_tokens": 201360124.0, + "step": 7779 + }, + { + "epoch": 0.8543817263342851, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8961434364318848, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7012261152267456, + "num_tokens": 201379881.0, + "step": 7780 + }, + { + "epoch": 0.8544915440368988, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.689476728439331, + "learning_rate": 1e-06, + "loss": 1.1477, + "mean_token_accuracy": 0.6658515930175781, + "num_tokens": 201407020.0, + "step": 7781 + }, + { + "epoch": 0.8546013617395124, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7540197372436523, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.711989164352417, + "num_tokens": 201430308.0, + "step": 7782 + }, + { + "epoch": 0.854711179442126, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5956625938415527, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6864191293716431, + "num_tokens": 201456423.0, + "step": 7783 + }, + { + "epoch": 0.8548209971447397, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6471450328826904, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7116726636886597, + "num_tokens": 201479965.0, + "step": 7784 + }, + { + "epoch": 0.8549308148473534, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6054725646972656, + "learning_rate": 1e-06, + "loss": 1.1215, + "mean_token_accuracy": 0.679868221282959, + "num_tokens": 201506398.0, + "step": 7785 + }, + { + "epoch": 0.8550406325499671, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4546594619750977, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7005612850189209, + "num_tokens": 201533902.0, + "step": 7786 + }, + { + "epoch": 0.8551504502525807, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5928266048431396, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7142240405082703, + "num_tokens": 201555508.0, + "step": 7787 + }, + { + "epoch": 0.8552602679551944, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5637459754943848, + "learning_rate": 1e-06, + "loss": 1.1453, + "mean_token_accuracy": 0.6713786125183105, + "num_tokens": 201583580.0, + "step": 7788 + }, + { + "epoch": 0.855370085657808, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4841220378875732, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6954928636550903, + "num_tokens": 201614497.0, + "step": 7789 + }, + { + "epoch": 0.8554799033604217, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6092236042022705, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7064918279647827, + "num_tokens": 201637724.0, + "step": 7790 + }, + { + "epoch": 0.8555897210630353, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.416550636291504, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7151591181755066, + "num_tokens": 201663407.0, + "step": 7791 + }, + { + "epoch": 0.855699538765649, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9039089679718018, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6961588859558105, + "num_tokens": 201683632.0, + "step": 7792 + }, + { + "epoch": 0.8558093564682627, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7097747325897217, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7143514752388, + "num_tokens": 201704739.0, + "step": 7793 + }, + { + "epoch": 0.8559191741708764, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6418516635894775, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7030630707740784, + "num_tokens": 201727432.0, + "step": 7794 + }, + { + "epoch": 0.85602899187349, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6847784519195557, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.6984533071517944, + "num_tokens": 201751819.0, + "step": 7795 + }, + { + "epoch": 0.8561388095761037, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.3471412658691406, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.707840621471405, + "num_tokens": 201783478.0, + "step": 7796 + }, + { + "epoch": 0.8562486272787173, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.444077491760254, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7037532925605774, + "num_tokens": 201808626.0, + "step": 7797 + }, + { + "epoch": 0.856358444981331, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7464303970336914, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7101284265518188, + "num_tokens": 201830947.0, + "step": 7798 + }, + { + "epoch": 0.8564682626839446, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6608259677886963, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6904008388519287, + "num_tokens": 201855299.0, + "step": 7799 + }, + { + "epoch": 0.8565780803865584, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5587663650512695, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7095844745635986, + "num_tokens": 201881901.0, + "step": 7800 + }, + { + "epoch": 0.856687898089172, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 7.872049808502197, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6951530575752258, + "num_tokens": 201914176.0, + "step": 7801 + }, + { + "epoch": 0.8567977157917857, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.483560562133789, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6962838172912598, + "num_tokens": 201941746.0, + "step": 7802 + }, + { + "epoch": 0.8569075334943993, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.56795597076416, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7182559967041016, + "num_tokens": 201968023.0, + "step": 7803 + }, + { + "epoch": 0.8570173511970129, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.491556167602539, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7067131996154785, + "num_tokens": 201992326.0, + "step": 7804 + }, + { + "epoch": 0.8571271688996266, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.140705108642578, + "learning_rate": 1e-06, + "loss": 1.1447, + "mean_token_accuracy": 0.6667683124542236, + "num_tokens": 202030782.0, + "step": 7805 + }, + { + "epoch": 0.8572369866022402, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4721782207489014, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7008655667304993, + "num_tokens": 202058757.0, + "step": 7806 + }, + { + "epoch": 0.857346804304854, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.617650032043457, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6985499262809753, + "num_tokens": 202083103.0, + "step": 7807 + }, + { + "epoch": 0.8574566220074676, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4773340225219727, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6900194883346558, + "num_tokens": 202112823.0, + "step": 7808 + }, + { + "epoch": 0.8575664397100813, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.3294918537139893, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7208164930343628, + "num_tokens": 202143862.0, + "step": 7809 + }, + { + "epoch": 0.8576762574126949, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5718777179718018, + "learning_rate": 1e-06, + "loss": 1.1, + "mean_token_accuracy": 0.6793705224990845, + "num_tokens": 202176427.0, + "step": 7810 + }, + { + "epoch": 0.8577860751153086, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.469642162322998, + "learning_rate": 1e-06, + "loss": 0.9565, + "mean_token_accuracy": 0.7180883884429932, + "num_tokens": 202205462.0, + "step": 7811 + }, + { + "epoch": 0.8578958928179222, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.732971429824829, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.7051317691802979, + "num_tokens": 202228124.0, + "step": 7812 + }, + { + "epoch": 0.8580057105205359, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 7.907739639282227, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7067201137542725, + "num_tokens": 202253879.0, + "step": 7813 + }, + { + "epoch": 0.8581155282231496, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5872886180877686, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6951799392700195, + "num_tokens": 202283863.0, + "step": 7814 + }, + { + "epoch": 0.8582253459257633, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.611982583999634, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7055752873420715, + "num_tokens": 202308816.0, + "step": 7815 + }, + { + "epoch": 0.8583351636283769, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 3.0172677040100098, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7170121669769287, + "num_tokens": 202328376.0, + "step": 7816 + }, + { + "epoch": 0.8584449813309906, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4611244201660156, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7053190469741821, + "num_tokens": 202355544.0, + "step": 7817 + }, + { + "epoch": 0.8585547990336042, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.360445022583008, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7236345410346985, + "num_tokens": 202381441.0, + "step": 7818 + }, + { + "epoch": 0.8586646167362179, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5590500831604004, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7256097793579102, + "num_tokens": 202408014.0, + "step": 7819 + }, + { + "epoch": 0.8587744344388315, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.399610757827759, + "learning_rate": 1e-06, + "loss": 0.9069, + "mean_token_accuracy": 0.7391159534454346, + "num_tokens": 202435919.0, + "step": 7820 + }, + { + "epoch": 0.8588842521414451, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.542847156524658, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7142791748046875, + "num_tokens": 202458627.0, + "step": 7821 + }, + { + "epoch": 0.8589940698440589, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4912147521972656, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7041885256767273, + "num_tokens": 202484299.0, + "step": 7822 + }, + { + "epoch": 0.8591038875466726, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.467360258102417, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7025171518325806, + "num_tokens": 202510414.0, + "step": 7823 + }, + { + "epoch": 0.8592137052492862, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6503536701202393, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7100054025650024, + "num_tokens": 202533513.0, + "step": 7824 + }, + { + "epoch": 0.8593235229518998, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5892443656921387, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6933872699737549, + "num_tokens": 202559780.0, + "step": 7825 + }, + { + "epoch": 0.8594333406545135, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.372797966003418, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.6974657773971558, + "num_tokens": 202589175.0, + "step": 7826 + }, + { + "epoch": 0.8595431583571271, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.748244285583496, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7098515033721924, + "num_tokens": 202611568.0, + "step": 7827 + }, + { + "epoch": 0.8596529760597408, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.409236192703247, + "learning_rate": 1e-06, + "loss": 1.1125, + "mean_token_accuracy": 0.6731530427932739, + "num_tokens": 202640644.0, + "step": 7828 + }, + { + "epoch": 0.8597627937623545, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.1928651332855225, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6973708271980286, + "num_tokens": 202674639.0, + "step": 7829 + }, + { + "epoch": 0.8598726114649682, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4923017024993896, + "learning_rate": 1e-06, + "loss": 1.1026, + "mean_token_accuracy": 0.6872209310531616, + "num_tokens": 202700287.0, + "step": 7830 + }, + { + "epoch": 0.8599824291675818, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.152585983276367, + "learning_rate": 1e-06, + "loss": 1.1106, + "mean_token_accuracy": 0.6768378019332886, + "num_tokens": 202736629.0, + "step": 7831 + }, + { + "epoch": 0.8600922468701955, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6784629821777344, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7122811079025269, + "num_tokens": 202757926.0, + "step": 7832 + }, + { + "epoch": 0.8602020645728091, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.55318021774292, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.6938673257827759, + "num_tokens": 202784249.0, + "step": 7833 + }, + { + "epoch": 0.8603118822754228, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4864256381988525, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.693785548210144, + "num_tokens": 202809701.0, + "step": 7834 + }, + { + "epoch": 0.8604216999780364, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4508254528045654, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.696043074131012, + "num_tokens": 202837782.0, + "step": 7835 + }, + { + "epoch": 0.8605315176806502, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.511413097381592, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7214498519897461, + "num_tokens": 202864615.0, + "step": 7836 + }, + { + "epoch": 0.8606413353832638, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5631957054138184, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7031733989715576, + "num_tokens": 202891672.0, + "step": 7837 + }, + { + "epoch": 0.8607511530858775, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.513205051422119, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.6927598714828491, + "num_tokens": 202922336.0, + "step": 7838 + }, + { + "epoch": 0.8608609707884911, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5173027515411377, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6936330795288086, + "num_tokens": 202949143.0, + "step": 7839 + }, + { + "epoch": 0.8609707884911048, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.883012533187866, + "learning_rate": 1e-06, + "loss": 0.8697, + "mean_token_accuracy": 0.7349717020988464, + "num_tokens": 202967791.0, + "step": 7840 + }, + { + "epoch": 0.8610806061937184, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6675870418548584, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6872320175170898, + "num_tokens": 202990516.0, + "step": 7841 + }, + { + "epoch": 0.861190423896332, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5523252487182617, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6958465576171875, + "num_tokens": 203014639.0, + "step": 7842 + }, + { + "epoch": 0.8613002415989458, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.490915298461914, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7072510719299316, + "num_tokens": 203042275.0, + "step": 7843 + }, + { + "epoch": 0.8614100593015594, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6855106353759766, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6936813592910767, + "num_tokens": 203066848.0, + "step": 7844 + }, + { + "epoch": 0.8615198770041731, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.939934492111206, + "learning_rate": 1e-06, + "loss": 1.1033, + "mean_token_accuracy": 0.6785605549812317, + "num_tokens": 203097751.0, + "step": 7845 + }, + { + "epoch": 0.8616296947067867, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.528146266937256, + "learning_rate": 1e-06, + "loss": 1.091, + "mean_token_accuracy": 0.685479462146759, + "num_tokens": 203124755.0, + "step": 7846 + }, + { + "epoch": 0.8617395124094004, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7566895484924316, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6999335289001465, + "num_tokens": 203146497.0, + "step": 7847 + }, + { + "epoch": 0.861849330112014, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7798428535461426, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.7084715366363525, + "num_tokens": 203167468.0, + "step": 7848 + }, + { + "epoch": 0.8619591478146277, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.476957321166992, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7141045331954956, + "num_tokens": 203198814.0, + "step": 7849 + }, + { + "epoch": 0.8620689655172413, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.462850332260132, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.6999896764755249, + "num_tokens": 203224152.0, + "step": 7850 + }, + { + "epoch": 0.8621787832198551, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.437000274658203, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7275099754333496, + "num_tokens": 203248679.0, + "step": 7851 + }, + { + "epoch": 0.8622886009224687, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.288623332977295, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6816562414169312, + "num_tokens": 203279320.0, + "step": 7852 + }, + { + "epoch": 0.8623984186250824, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5302202701568604, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7120120525360107, + "num_tokens": 203303420.0, + "step": 7853 + }, + { + "epoch": 0.862508236327696, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 4.509467124938965, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7088066339492798, + "num_tokens": 203327137.0, + "step": 7854 + }, + { + "epoch": 0.8626180540303097, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5736470222473145, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7093826532363892, + "num_tokens": 203353839.0, + "step": 7855 + }, + { + "epoch": 0.8627278717329233, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.398561477661133, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6841796040534973, + "num_tokens": 203383795.0, + "step": 7856 + }, + { + "epoch": 0.862837689435537, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7377817630767822, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7089289426803589, + "num_tokens": 203407162.0, + "step": 7857 + }, + { + "epoch": 0.8629475071381507, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.599616050720215, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.6962337493896484, + "num_tokens": 203434099.0, + "step": 7858 + }, + { + "epoch": 0.8630573248407644, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8136839866638184, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.7126728892326355, + "num_tokens": 203456548.0, + "step": 7859 + }, + { + "epoch": 0.863167142543378, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.779393434524536, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7128370404243469, + "num_tokens": 203484169.0, + "step": 7860 + }, + { + "epoch": 0.8632769602459917, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6886913776397705, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6899397373199463, + "num_tokens": 203509125.0, + "step": 7861 + }, + { + "epoch": 0.8633867779486053, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5481152534484863, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7099733352661133, + "num_tokens": 203534755.0, + "step": 7862 + }, + { + "epoch": 0.863496595651219, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.583801746368408, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6889483332633972, + "num_tokens": 203558568.0, + "step": 7863 + }, + { + "epoch": 0.8636064133538326, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4862794876098633, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7044290900230408, + "num_tokens": 203582837.0, + "step": 7864 + }, + { + "epoch": 0.8637162310564463, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4992945194244385, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7121646404266357, + "num_tokens": 203611467.0, + "step": 7865 + }, + { + "epoch": 0.86382604875906, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5940706729888916, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.705722987651825, + "num_tokens": 203636023.0, + "step": 7866 + }, + { + "epoch": 0.8639358664616736, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5912399291992188, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6955200433731079, + "num_tokens": 203660568.0, + "step": 7867 + }, + { + "epoch": 0.8640456841642873, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.509920358657837, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6884002089500427, + "num_tokens": 203688505.0, + "step": 7868 + }, + { + "epoch": 0.8641555018669009, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8776328563690186, + "learning_rate": 1e-06, + "loss": 1.1325, + "mean_token_accuracy": 0.6774119138717651, + "num_tokens": 203718239.0, + "step": 7869 + }, + { + "epoch": 0.8642653195695146, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6247398853302, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7101683616638184, + "num_tokens": 203743844.0, + "step": 7870 + }, + { + "epoch": 0.8643751372721282, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.596083641052246, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7058152556419373, + "num_tokens": 203767153.0, + "step": 7871 + }, + { + "epoch": 0.864484954974742, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5121703147888184, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7000374794006348, + "num_tokens": 203794917.0, + "step": 7872 + }, + { + "epoch": 0.8645947726773556, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.834625720977783, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7010890245437622, + "num_tokens": 203817801.0, + "step": 7873 + }, + { + "epoch": 0.8647045903799693, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6327507495880127, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6952575445175171, + "num_tokens": 203841656.0, + "step": 7874 + }, + { + "epoch": 0.8648144080825829, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.743826150894165, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7149500250816345, + "num_tokens": 203864128.0, + "step": 7875 + }, + { + "epoch": 0.8649242257851966, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7669436931610107, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7193938493728638, + "num_tokens": 203886915.0, + "step": 7876 + }, + { + "epoch": 0.8650340434878102, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.6524789333343506, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7118479013442993, + "num_tokens": 203912550.0, + "step": 7877 + }, + { + "epoch": 0.8651438611904239, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7107601165771484, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.705815315246582, + "num_tokens": 203935089.0, + "step": 7878 + }, + { + "epoch": 0.8652536788930375, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.3957085609436035, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7307657599449158, + "num_tokens": 203963748.0, + "step": 7879 + }, + { + "epoch": 0.8653634965956513, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8874921798706055, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7171562314033508, + "num_tokens": 203984878.0, + "step": 7880 + }, + { + "epoch": 0.8654733142982649, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.306830883026123, + "learning_rate": 1e-06, + "loss": 1.1455, + "mean_token_accuracy": 0.6760058403015137, + "num_tokens": 204014695.0, + "step": 7881 + }, + { + "epoch": 0.8655831320008786, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.821843147277832, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7142739295959473, + "num_tokens": 204035281.0, + "step": 7882 + }, + { + "epoch": 0.8656929497034922, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6563684940338135, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7109683752059937, + "num_tokens": 204059679.0, + "step": 7883 + }, + { + "epoch": 0.8658027674061058, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9129514694213867, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7084336280822754, + "num_tokens": 204081454.0, + "step": 7884 + }, + { + "epoch": 0.8659125851087195, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8836560249328613, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.717900812625885, + "num_tokens": 204103325.0, + "step": 7885 + }, + { + "epoch": 0.8660224028113331, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.867624282836914, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.7005641460418701, + "num_tokens": 204124511.0, + "step": 7886 + }, + { + "epoch": 0.8661322205139469, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5532028675079346, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7010135054588318, + "num_tokens": 204151632.0, + "step": 7887 + }, + { + "epoch": 0.8662420382165605, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4433820247650146, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6811341643333435, + "num_tokens": 204177986.0, + "step": 7888 + }, + { + "epoch": 0.8663518559191742, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4899415969848633, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6926224231719971, + "num_tokens": 204205111.0, + "step": 7889 + }, + { + "epoch": 0.8664616736217878, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6779630184173584, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.6990025043487549, + "num_tokens": 204227601.0, + "step": 7890 + }, + { + "epoch": 0.8665714913244015, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.660702705383301, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7176287174224854, + "num_tokens": 204250245.0, + "step": 7891 + }, + { + "epoch": 0.8666813090270151, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.821620225906372, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7038266062736511, + "num_tokens": 204274328.0, + "step": 7892 + }, + { + "epoch": 0.8667911267296288, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.911411762237549, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7155759930610657, + "num_tokens": 204297123.0, + "step": 7893 + }, + { + "epoch": 0.8669009444322425, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.3694205284118652, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7050669193267822, + "num_tokens": 204327408.0, + "step": 7894 + }, + { + "epoch": 0.8670107621348562, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9982218742370605, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.712500274181366, + "num_tokens": 204349246.0, + "step": 7895 + }, + { + "epoch": 0.8671205798374698, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9118783473968506, + "learning_rate": 1e-06, + "loss": 0.9093, + "mean_token_accuracy": 0.7256333827972412, + "num_tokens": 204367333.0, + "step": 7896 + }, + { + "epoch": 0.8672303975400835, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.2582104206085205, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.72219318151474, + "num_tokens": 204400348.0, + "step": 7897 + }, + { + "epoch": 0.8673402152426971, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5941808223724365, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6820840835571289, + "num_tokens": 204425388.0, + "step": 7898 + }, + { + "epoch": 0.8674500329453108, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.648357391357422, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6902764439582825, + "num_tokens": 204449858.0, + "step": 7899 + }, + { + "epoch": 0.8675598506479244, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4980950355529785, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6817188262939453, + "num_tokens": 204476824.0, + "step": 7900 + }, + { + "epoch": 0.8676696683505382, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.3562934398651123, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6807494163513184, + "num_tokens": 204505730.0, + "step": 7901 + }, + { + "epoch": 0.8677794860531518, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9478957653045654, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.6998104453086853, + "num_tokens": 204526003.0, + "step": 7902 + }, + { + "epoch": 0.8678893037557655, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.894205331802368, + "learning_rate": 1e-06, + "loss": 1.1071, + "mean_token_accuracy": 0.6936488151550293, + "num_tokens": 204548218.0, + "step": 7903 + }, + { + "epoch": 0.8679991214583791, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4582624435424805, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6890463829040527, + "num_tokens": 204576206.0, + "step": 7904 + }, + { + "epoch": 0.8681089391609927, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5970511436462402, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6937528252601624, + "num_tokens": 204608378.0, + "step": 7905 + }, + { + "epoch": 0.8682187568636064, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.9049344062805176, + "learning_rate": 1e-06, + "loss": 1.0923, + "mean_token_accuracy": 0.6834646463394165, + "num_tokens": 204628158.0, + "step": 7906 + }, + { + "epoch": 0.86832857456622, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.729382038116455, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7103111743927002, + "num_tokens": 204650319.0, + "step": 7907 + }, + { + "epoch": 0.8684383922688337, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 3.092845916748047, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7164003849029541, + "num_tokens": 204668863.0, + "step": 7908 + }, + { + "epoch": 0.8685482099714474, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.594163179397583, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7157782316207886, + "num_tokens": 204692670.0, + "step": 7909 + }, + { + "epoch": 0.8686580276740611, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.776531457901001, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.684583306312561, + "num_tokens": 204715673.0, + "step": 7910 + }, + { + "epoch": 0.8687678453766747, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7904980182647705, + "learning_rate": 1e-06, + "loss": 0.919, + "mean_token_accuracy": 0.728480875492096, + "num_tokens": 204738523.0, + "step": 7911 + }, + { + "epoch": 0.8688776630792884, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4192421436309814, + "learning_rate": 1e-06, + "loss": 1.1363, + "mean_token_accuracy": 0.6688681840896606, + "num_tokens": 204767872.0, + "step": 7912 + }, + { + "epoch": 0.868987480781902, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4877188205718994, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6922903060913086, + "num_tokens": 204795170.0, + "step": 7913 + }, + { + "epoch": 0.8690972984845157, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.865422010421753, + "learning_rate": 1e-06, + "loss": 1.1, + "mean_token_accuracy": 0.6793423295021057, + "num_tokens": 204825372.0, + "step": 7914 + }, + { + "epoch": 0.8692071161871293, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5448405742645264, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6944403648376465, + "num_tokens": 204853010.0, + "step": 7915 + }, + { + "epoch": 0.8693169338897431, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.438631534576416, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6989661455154419, + "num_tokens": 204883722.0, + "step": 7916 + }, + { + "epoch": 0.8694267515923567, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5299153327941895, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.70284104347229, + "num_tokens": 204908704.0, + "step": 7917 + }, + { + "epoch": 0.8695365692949704, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.620093822479248, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7077723741531372, + "num_tokens": 204931118.0, + "step": 7918 + }, + { + "epoch": 0.869646386997584, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4977898597717285, + "learning_rate": 1e-06, + "loss": 1.11, + "mean_token_accuracy": 0.6750672459602356, + "num_tokens": 204959245.0, + "step": 7919 + }, + { + "epoch": 0.8697562047001977, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8596434593200684, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7002630233764648, + "num_tokens": 204980960.0, + "step": 7920 + }, + { + "epoch": 0.8698660224028113, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.324714183807373, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7084505558013916, + "num_tokens": 205012385.0, + "step": 7921 + }, + { + "epoch": 0.869975840105425, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.4465525150299072, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.701104998588562, + "num_tokens": 205042316.0, + "step": 7922 + }, + { + "epoch": 0.8700856578080387, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.52199649810791, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6984745860099792, + "num_tokens": 205069593.0, + "step": 7923 + }, + { + "epoch": 0.8701954755106523, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6834869384765625, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7061182260513306, + "num_tokens": 205095869.0, + "step": 7924 + }, + { + "epoch": 0.870305293213266, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.708838701248169, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6843444108963013, + "num_tokens": 205120650.0, + "step": 7925 + }, + { + "epoch": 0.8704151109158796, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.7266085147857666, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.692008376121521, + "num_tokens": 205145191.0, + "step": 7926 + }, + { + "epoch": 0.8705249286184933, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.721348524093628, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7111638784408569, + "num_tokens": 205171048.0, + "step": 7927 + }, + { + "epoch": 0.8706347463211069, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3598179817199707, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7119853496551514, + "num_tokens": 205197124.0, + "step": 7928 + }, + { + "epoch": 0.8707445640237206, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.239173650741577, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.7020403146743774, + "num_tokens": 205228334.0, + "step": 7929 + }, + { + "epoch": 0.8708543817263343, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5907442569732666, + "learning_rate": 1e-06, + "loss": 0.942, + "mean_token_accuracy": 0.7170019149780273, + "num_tokens": 205252781.0, + "step": 7930 + }, + { + "epoch": 0.870964199428948, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5656931400299072, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6949223279953003, + "num_tokens": 205279568.0, + "step": 7931 + }, + { + "epoch": 0.8710740171315616, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.932223081588745, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7134431004524231, + "num_tokens": 205298739.0, + "step": 7932 + }, + { + "epoch": 0.8711838348341753, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5992236137390137, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.7234230041503906, + "num_tokens": 205323281.0, + "step": 7933 + }, + { + "epoch": 0.8712936525367889, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.965794324874878, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.7001910209655762, + "num_tokens": 205346345.0, + "step": 7934 + }, + { + "epoch": 0.8714034702394026, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7408597469329834, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7093899250030518, + "num_tokens": 205367389.0, + "step": 7935 + }, + { + "epoch": 0.8715132879420162, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.659327268600464, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6952337026596069, + "num_tokens": 205390695.0, + "step": 7936 + }, + { + "epoch": 0.87162310564463, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.645634889602661, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6876183152198792, + "num_tokens": 205415769.0, + "step": 7937 + }, + { + "epoch": 0.8717329233472436, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6517257690429688, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.715194046497345, + "num_tokens": 205438970.0, + "step": 7938 + }, + { + "epoch": 0.8718427410498573, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.56479549407959, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7001967430114746, + "num_tokens": 205463899.0, + "step": 7939 + }, + { + "epoch": 0.8719525587524709, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.447723627090454, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6953606605529785, + "num_tokens": 205493986.0, + "step": 7940 + }, + { + "epoch": 0.8720623764550846, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.8254950046539307, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7005233764648438, + "num_tokens": 205516520.0, + "step": 7941 + }, + { + "epoch": 0.8721721941576982, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 3.8265438079833984, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6890172958374023, + "num_tokens": 205544956.0, + "step": 7942 + }, + { + "epoch": 0.8722820118603118, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9917845726013184, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6997146010398865, + "num_tokens": 205565308.0, + "step": 7943 + }, + { + "epoch": 0.8723918295629255, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.414106845855713, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6941131353378296, + "num_tokens": 205592370.0, + "step": 7944 + }, + { + "epoch": 0.8725016472655392, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4207160472869873, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6917440891265869, + "num_tokens": 205622412.0, + "step": 7945 + }, + { + "epoch": 0.8726114649681529, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6163036823272705, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7005476951599121, + "num_tokens": 205646197.0, + "step": 7946 + }, + { + "epoch": 0.8727212826707665, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5871479511260986, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7043138742446899, + "num_tokens": 205669733.0, + "step": 7947 + }, + { + "epoch": 0.8728311003733802, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.3875086307525635, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.700799822807312, + "num_tokens": 205697984.0, + "step": 7948 + }, + { + "epoch": 0.8729409180759938, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.699749708175659, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6836386919021606, + "num_tokens": 205722254.0, + "step": 7949 + }, + { + "epoch": 0.8730507357786075, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7193219661712646, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.6991168260574341, + "num_tokens": 205748538.0, + "step": 7950 + }, + { + "epoch": 0.8731605534812211, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6252453327178955, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7298205494880676, + "num_tokens": 205772601.0, + "step": 7951 + }, + { + "epoch": 0.8732703711838349, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7995269298553467, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.687478244304657, + "num_tokens": 205796423.0, + "step": 7952 + }, + { + "epoch": 0.8733801888864485, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5135271549224854, + "learning_rate": 1e-06, + "loss": 0.931, + "mean_token_accuracy": 0.7219125628471375, + "num_tokens": 205826257.0, + "step": 7953 + }, + { + "epoch": 0.8734900065890622, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.0090994834899902, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.689822793006897, + "num_tokens": 205850383.0, + "step": 7954 + }, + { + "epoch": 0.8735998242916758, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.908068895339966, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7004225254058838, + "num_tokens": 205869212.0, + "step": 7955 + }, + { + "epoch": 0.8737096419942895, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7415452003479004, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7077636122703552, + "num_tokens": 205891962.0, + "step": 7956 + }, + { + "epoch": 0.8738194596969031, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6120476722717285, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7106571197509766, + "num_tokens": 205916887.0, + "step": 7957 + }, + { + "epoch": 0.8739292773995168, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4807169437408447, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7045316100120544, + "num_tokens": 205943337.0, + "step": 7958 + }, + { + "epoch": 0.8740390951021305, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4263455867767334, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.696060299873352, + "num_tokens": 205969059.0, + "step": 7959 + }, + { + "epoch": 0.8741489128047442, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.2814133167266846, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7166025042533875, + "num_tokens": 205998041.0, + "step": 7960 + }, + { + "epoch": 0.8742587305073578, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.5741677284240723, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7112245559692383, + "num_tokens": 206022857.0, + "step": 7961 + }, + { + "epoch": 0.8743685482099715, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.976303815841675, + "learning_rate": 1e-06, + "loss": 1.1361, + "mean_token_accuracy": 0.6817623972892761, + "num_tokens": 206051159.0, + "step": 7962 + }, + { + "epoch": 0.8744783659125851, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4332706928253174, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7118488550186157, + "num_tokens": 206079039.0, + "step": 7963 + }, + { + "epoch": 0.8745881836151987, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.572916269302368, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.6991896629333496, + "num_tokens": 206103595.0, + "step": 7964 + }, + { + "epoch": 0.8746980013178124, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5407309532165527, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6955218315124512, + "num_tokens": 206128831.0, + "step": 7965 + }, + { + "epoch": 0.8748078190204261, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.621818780899048, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6938660144805908, + "num_tokens": 206153310.0, + "step": 7966 + }, + { + "epoch": 0.8749176367230398, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.317483901977539, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.7216793894767761, + "num_tokens": 206179791.0, + "step": 7967 + }, + { + "epoch": 0.8750274544256534, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8953633308410645, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7098028063774109, + "num_tokens": 206198392.0, + "step": 7968 + }, + { + "epoch": 0.8751372721282671, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4777567386627197, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.7016377449035645, + "num_tokens": 206226572.0, + "step": 7969 + }, + { + "epoch": 0.8752470898308807, + "ewc_loss": 8.046627044677734e-06, + "grad_norm": 2.770775079727173, + "learning_rate": 1e-06, + "loss": 0.908, + "mean_token_accuracy": 0.7313308715820312, + "num_tokens": 206247127.0, + "step": 7970 + }, + { + "epoch": 0.8753569075334944, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5501651763916016, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7079628109931946, + "num_tokens": 206271724.0, + "step": 7971 + }, + { + "epoch": 0.875466725236108, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.270400285720825, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6936473250389099, + "num_tokens": 206302578.0, + "step": 7972 + }, + { + "epoch": 0.8755765429387217, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.549855947494507, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7089004516601562, + "num_tokens": 206328020.0, + "step": 7973 + }, + { + "epoch": 0.8756863606413354, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.491520643234253, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7150136232376099, + "num_tokens": 206353497.0, + "step": 7974 + }, + { + "epoch": 0.8757961783439491, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4938738346099854, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7037479281425476, + "num_tokens": 206377663.0, + "step": 7975 + }, + { + "epoch": 0.8759059960465627, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.529257297515869, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.7007368803024292, + "num_tokens": 206404111.0, + "step": 7976 + }, + { + "epoch": 0.8760158137491764, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.63798189163208, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.6996475458145142, + "num_tokens": 206425714.0, + "step": 7977 + }, + { + "epoch": 0.87612563145179, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4475581645965576, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.695514440536499, + "num_tokens": 206454801.0, + "step": 7978 + }, + { + "epoch": 0.8762354491544037, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5289127826690674, + "learning_rate": 1e-06, + "loss": 1.13, + "mean_token_accuracy": 0.6820271015167236, + "num_tokens": 206480724.0, + "step": 7979 + }, + { + "epoch": 0.8763452668570173, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4411892890930176, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7105443477630615, + "num_tokens": 206508803.0, + "step": 7980 + }, + { + "epoch": 0.8764550845596311, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5930070877075195, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6996380686759949, + "num_tokens": 206532791.0, + "step": 7981 + }, + { + "epoch": 0.8765649022622447, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 3.056523561477661, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7018823027610779, + "num_tokens": 206550452.0, + "step": 7982 + }, + { + "epoch": 0.8766747199648584, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4988770484924316, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.6984871029853821, + "num_tokens": 206575311.0, + "step": 7983 + }, + { + "epoch": 0.876784537667472, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5857419967651367, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6872771978378296, + "num_tokens": 206606387.0, + "step": 7984 + }, + { + "epoch": 0.8768943553700856, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.446601152420044, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7184614539146423, + "num_tokens": 206631078.0, + "step": 7985 + }, + { + "epoch": 0.8770041730726993, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7484562397003174, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.6981219053268433, + "num_tokens": 206656326.0, + "step": 7986 + }, + { + "epoch": 0.8771139907753129, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3239970207214355, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6961886882781982, + "num_tokens": 206687052.0, + "step": 7987 + }, + { + "epoch": 0.8772238084779267, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.961061716079712, + "learning_rate": 1e-06, + "loss": 1.1309, + "mean_token_accuracy": 0.6710128784179688, + "num_tokens": 206708479.0, + "step": 7988 + }, + { + "epoch": 0.8773336261805403, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6607916355133057, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.68537437915802, + "num_tokens": 206734578.0, + "step": 7989 + }, + { + "epoch": 0.877443443883154, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4274888038635254, + "learning_rate": 1e-06, + "loss": 0.9785, + "mean_token_accuracy": 0.7170984148979187, + "num_tokens": 206761012.0, + "step": 7990 + }, + { + "epoch": 0.8775532615857676, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6637518405914307, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.720802903175354, + "num_tokens": 206782739.0, + "step": 7991 + }, + { + "epoch": 0.8776630792883813, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.284512519836426, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.6741486191749573, + "num_tokens": 206816363.0, + "step": 7992 + }, + { + "epoch": 0.8777728969909949, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.686229944229126, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7140157222747803, + "num_tokens": 206840364.0, + "step": 7993 + }, + { + "epoch": 0.8778827146936086, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5713536739349365, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6841384172439575, + "num_tokens": 206868088.0, + "step": 7994 + }, + { + "epoch": 0.8779925323962223, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.250610113143921, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.6993309855461121, + "num_tokens": 206889482.0, + "step": 7995 + }, + { + "epoch": 0.878102350098836, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8441543579101562, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7029460668563843, + "num_tokens": 206912349.0, + "step": 7996 + }, + { + "epoch": 0.8782121678014496, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.477175235748291, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7148484587669373, + "num_tokens": 206937852.0, + "step": 7997 + }, + { + "epoch": 0.8783219855040633, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5929322242736816, + "learning_rate": 1e-06, + "loss": 0.9089, + "mean_token_accuracy": 0.7245357036590576, + "num_tokens": 206961008.0, + "step": 7998 + }, + { + "epoch": 0.8784318032066769, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.625471830368042, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6928302049636841, + "num_tokens": 206983322.0, + "step": 7999 + }, + { + "epoch": 0.8785416209092906, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6580450534820557, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.6982017755508423, + "num_tokens": 207007351.0, + "step": 8000 + }, + { + "epoch": 0.8786514386119042, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.369684934616089, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7063434720039368, + "num_tokens": 207038395.0, + "step": 8001 + }, + { + "epoch": 0.8787612563145178, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.719576120376587, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.704770028591156, + "num_tokens": 207062632.0, + "step": 8002 + }, + { + "epoch": 0.8788710740171316, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.42679762840271, + "learning_rate": 1e-06, + "loss": 1.094, + "mean_token_accuracy": 0.6821209788322449, + "num_tokens": 207092243.0, + "step": 8003 + }, + { + "epoch": 0.8789808917197452, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.347580909729004, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6944386959075928, + "num_tokens": 207122819.0, + "step": 8004 + }, + { + "epoch": 0.8790907094223589, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.683255910873413, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.7090078592300415, + "num_tokens": 207147033.0, + "step": 8005 + }, + { + "epoch": 0.8792005271249725, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.489320993423462, + "learning_rate": 1e-06, + "loss": 1.1096, + "mean_token_accuracy": 0.6836493015289307, + "num_tokens": 207172566.0, + "step": 8006 + }, + { + "epoch": 0.8793103448275862, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.350898265838623, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6984531879425049, + "num_tokens": 207201224.0, + "step": 8007 + }, + { + "epoch": 0.8794201625301998, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4863767623901367, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7165981531143188, + "num_tokens": 207228077.0, + "step": 8008 + }, + { + "epoch": 0.8795299802328135, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.410423994064331, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6771700382232666, + "num_tokens": 207256588.0, + "step": 8009 + }, + { + "epoch": 0.8796397979354272, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.644228219985962, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7137348651885986, + "num_tokens": 207278042.0, + "step": 8010 + }, + { + "epoch": 0.8797496156380409, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.556713342666626, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.6830673217773438, + "num_tokens": 207305518.0, + "step": 8011 + }, + { + "epoch": 0.8798594333406545, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6685590744018555, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7044764757156372, + "num_tokens": 207335128.0, + "step": 8012 + }, + { + "epoch": 0.8799692510432682, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8170979022979736, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7118414044380188, + "num_tokens": 207357774.0, + "step": 8013 + }, + { + "epoch": 0.8800790687458818, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5911917686462402, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7080672979354858, + "num_tokens": 207383683.0, + "step": 8014 + }, + { + "epoch": 0.8801888864484955, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6912424564361572, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7091405391693115, + "num_tokens": 207413913.0, + "step": 8015 + }, + { + "epoch": 0.8802987041511091, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.2256109714508057, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7023941874504089, + "num_tokens": 207449426.0, + "step": 8016 + }, + { + "epoch": 0.8804085218537229, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.507319211959839, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.703902542591095, + "num_tokens": 207476016.0, + "step": 8017 + }, + { + "epoch": 0.8805183395563365, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6940228939056396, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.710219144821167, + "num_tokens": 207497955.0, + "step": 8018 + }, + { + "epoch": 0.8806281572589502, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8366336822509766, + "learning_rate": 1e-06, + "loss": 0.8999, + "mean_token_accuracy": 0.7309305667877197, + "num_tokens": 207516932.0, + "step": 8019 + }, + { + "epoch": 0.8807379749615638, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.594881772994995, + "learning_rate": 1e-06, + "loss": 1.1077, + "mean_token_accuracy": 0.6826387047767639, + "num_tokens": 207545331.0, + "step": 8020 + }, + { + "epoch": 0.8808477926641775, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.654723882675171, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.705986738204956, + "num_tokens": 207568324.0, + "step": 8021 + }, + { + "epoch": 0.8809576103667911, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.6517622470855713, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7166433334350586, + "num_tokens": 207590390.0, + "step": 8022 + }, + { + "epoch": 0.8810674280694047, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.616623640060425, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7225707173347473, + "num_tokens": 207614861.0, + "step": 8023 + }, + { + "epoch": 0.8811772457720185, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3852689266204834, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6957539319992065, + "num_tokens": 207643294.0, + "step": 8024 + }, + { + "epoch": 0.8812870634746321, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.8298771381378174, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.6980269551277161, + "num_tokens": 207663473.0, + "step": 8025 + }, + { + "epoch": 0.8813968811772458, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4383270740509033, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7076664566993713, + "num_tokens": 207693923.0, + "step": 8026 + }, + { + "epoch": 0.8815066988798594, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4879932403564453, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7253633141517639, + "num_tokens": 207722358.0, + "step": 8027 + }, + { + "epoch": 0.8816165165824731, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7421276569366455, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.7057015895843506, + "num_tokens": 207748579.0, + "step": 8028 + }, + { + "epoch": 0.8817263342850867, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.779801607131958, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7175595760345459, + "num_tokens": 207771070.0, + "step": 8029 + }, + { + "epoch": 0.8818361519877004, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9528775215148926, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7125876545906067, + "num_tokens": 207789767.0, + "step": 8030 + }, + { + "epoch": 0.881945969690314, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7281527519226074, + "learning_rate": 1e-06, + "loss": 0.9552, + "mean_token_accuracy": 0.7155625820159912, + "num_tokens": 207810612.0, + "step": 8031 + }, + { + "epoch": 0.8820557873929278, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 17.07219696044922, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6919674277305603, + "num_tokens": 207836096.0, + "step": 8032 + }, + { + "epoch": 0.8821656050955414, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3743298053741455, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6985092163085938, + "num_tokens": 207868083.0, + "step": 8033 + }, + { + "epoch": 0.8822754227981551, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6932520866394043, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.715348482131958, + "num_tokens": 207892881.0, + "step": 8034 + }, + { + "epoch": 0.8823852405007687, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.460172176361084, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.689715564250946, + "num_tokens": 207924786.0, + "step": 8035 + }, + { + "epoch": 0.8824950582033824, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8784849643707275, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7185970544815063, + "num_tokens": 207946705.0, + "step": 8036 + }, + { + "epoch": 0.882604875905996, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.706709384918213, + "learning_rate": 1e-06, + "loss": 0.9259, + "mean_token_accuracy": 0.7274994850158691, + "num_tokens": 207968344.0, + "step": 8037 + }, + { + "epoch": 0.8827146936086097, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6843438148498535, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7059042453765869, + "num_tokens": 207991588.0, + "step": 8038 + }, + { + "epoch": 0.8828245113112234, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.392331600189209, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6903179883956909, + "num_tokens": 208020156.0, + "step": 8039 + }, + { + "epoch": 0.8829343290138371, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.963195323944092, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7074595093727112, + "num_tokens": 208039776.0, + "step": 8040 + }, + { + "epoch": 0.8830441467164507, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2036702632904053, + "learning_rate": 1e-06, + "loss": 1.1344, + "mean_token_accuracy": 0.678120493888855, + "num_tokens": 208078399.0, + "step": 8041 + }, + { + "epoch": 0.8831539644190644, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.757727861404419, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.70887690782547, + "num_tokens": 208098993.0, + "step": 8042 + }, + { + "epoch": 0.883263782121678, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.337207555770874, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7037694454193115, + "num_tokens": 208132064.0, + "step": 8043 + }, + { + "epoch": 0.8833735998242916, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4323112964630127, + "learning_rate": 1e-06, + "loss": 0.9364, + "mean_token_accuracy": 0.7286460995674133, + "num_tokens": 208158784.0, + "step": 8044 + }, + { + "epoch": 0.8834834175269053, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4703168869018555, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.698051929473877, + "num_tokens": 208184449.0, + "step": 8045 + }, + { + "epoch": 0.883593235229519, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.806932210922241, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7181345224380493, + "num_tokens": 208204624.0, + "step": 8046 + }, + { + "epoch": 0.8837030529321327, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.071591377258301, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6941725611686707, + "num_tokens": 208227598.0, + "step": 8047 + }, + { + "epoch": 0.8838128706347463, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.622147560119629, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7207474708557129, + "num_tokens": 208251223.0, + "step": 8048 + }, + { + "epoch": 0.88392268833736, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.7638814449310303, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.6971073150634766, + "num_tokens": 208271542.0, + "step": 8049 + }, + { + "epoch": 0.8840325060399736, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4718995094299316, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.6958657503128052, + "num_tokens": 208300464.0, + "step": 8050 + }, + { + "epoch": 0.8841423237425873, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.810939073562622, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7091443538665771, + "num_tokens": 208320189.0, + "step": 8051 + }, + { + "epoch": 0.8842521414452009, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.348358392715454, + "learning_rate": 1e-06, + "loss": 1.1339, + "mean_token_accuracy": 0.6870076060295105, + "num_tokens": 208355152.0, + "step": 8052 + }, + { + "epoch": 0.8843619591478147, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.458078384399414, + "learning_rate": 1e-06, + "loss": 1.0874, + "mean_token_accuracy": 0.680440366268158, + "num_tokens": 208383984.0, + "step": 8053 + }, + { + "epoch": 0.8844717768504283, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.423818349838257, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.711158812046051, + "num_tokens": 208410915.0, + "step": 8054 + }, + { + "epoch": 0.884581594553042, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.505678176879883, + "learning_rate": 1e-06, + "loss": 0.898, + "mean_token_accuracy": 0.7389045357704163, + "num_tokens": 208436148.0, + "step": 8055 + }, + { + "epoch": 0.8846914122556556, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.488147020339966, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.7058793306350708, + "num_tokens": 208465224.0, + "step": 8056 + }, + { + "epoch": 0.8848012299582693, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3043229579925537, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7280287742614746, + "num_tokens": 208494111.0, + "step": 8057 + }, + { + "epoch": 0.8849110476608829, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.805508852005005, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7189720869064331, + "num_tokens": 208514390.0, + "step": 8058 + }, + { + "epoch": 0.8850208653634966, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4132773876190186, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.693037748336792, + "num_tokens": 208543421.0, + "step": 8059 + }, + { + "epoch": 0.8851306830661102, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.533785581588745, + "learning_rate": 1e-06, + "loss": 1.1224, + "mean_token_accuracy": 0.6824009418487549, + "num_tokens": 208570744.0, + "step": 8060 + }, + { + "epoch": 0.885240500768724, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5118792057037354, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7147698998451233, + "num_tokens": 208596170.0, + "step": 8061 + }, + { + "epoch": 0.8853503184713376, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8090152740478516, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6923247575759888, + "num_tokens": 208625206.0, + "step": 8062 + }, + { + "epoch": 0.8854601361739513, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.818627119064331, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7187271118164062, + "num_tokens": 208645910.0, + "step": 8063 + }, + { + "epoch": 0.8855699538765649, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4155471324920654, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7022720575332642, + "num_tokens": 208672805.0, + "step": 8064 + }, + { + "epoch": 0.8856797715791785, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.377274751663208, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.7017645835876465, + "num_tokens": 208703429.0, + "step": 8065 + }, + { + "epoch": 0.8857895892817922, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.608293056488037, + "learning_rate": 1e-06, + "loss": 1.1296, + "mean_token_accuracy": 0.6779832243919373, + "num_tokens": 208732360.0, + "step": 8066 + }, + { + "epoch": 0.8858994069844058, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.535658836364746, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6960721611976624, + "num_tokens": 208759178.0, + "step": 8067 + }, + { + "epoch": 0.8860092246870196, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8495378494262695, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7079023122787476, + "num_tokens": 208778917.0, + "step": 8068 + }, + { + "epoch": 0.8861190423896332, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.40373158454895, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7053681015968323, + "num_tokens": 208808384.0, + "step": 8069 + }, + { + "epoch": 0.8862288600922469, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7786073684692383, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7077158689498901, + "num_tokens": 208829418.0, + "step": 8070 + }, + { + "epoch": 0.8863386777948605, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.379044771194458, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6984964609146118, + "num_tokens": 208857521.0, + "step": 8071 + }, + { + "epoch": 0.8864484954974742, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6193244457244873, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6914017796516418, + "num_tokens": 208881906.0, + "step": 8072 + }, + { + "epoch": 0.8865583132000878, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.375778913497925, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.7276136875152588, + "num_tokens": 208914194.0, + "step": 8073 + }, + { + "epoch": 0.8866681309027015, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5672173500061035, + "learning_rate": 1e-06, + "loss": 1.1151, + "mean_token_accuracy": 0.6813076734542847, + "num_tokens": 208939704.0, + "step": 8074 + }, + { + "epoch": 0.8867779486053152, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5511960983276367, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7239958643913269, + "num_tokens": 208963714.0, + "step": 8075 + }, + { + "epoch": 0.8868877663079289, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5224947929382324, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6848984360694885, + "num_tokens": 208992718.0, + "step": 8076 + }, + { + "epoch": 0.8869975840105425, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.472529888153076, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7105097770690918, + "num_tokens": 209021300.0, + "step": 8077 + }, + { + "epoch": 0.8871074017131562, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6753909587860107, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6947569847106934, + "num_tokens": 209048287.0, + "step": 8078 + }, + { + "epoch": 0.8872172194157698, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.877497673034668, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.6998164653778076, + "num_tokens": 209069545.0, + "step": 8079 + }, + { + "epoch": 0.8873270371183835, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.494924783706665, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.6957777738571167, + "num_tokens": 209095028.0, + "step": 8080 + }, + { + "epoch": 0.8874368548209971, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.669111490249634, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7193485498428345, + "num_tokens": 209118193.0, + "step": 8081 + }, + { + "epoch": 0.8875466725236109, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.467019557952881, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7022440433502197, + "num_tokens": 209146980.0, + "step": 8082 + }, + { + "epoch": 0.8876564902262245, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.508010149002075, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6979007124900818, + "num_tokens": 209174812.0, + "step": 8083 + }, + { + "epoch": 0.8877663079288381, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.467625379562378, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7166538238525391, + "num_tokens": 209200809.0, + "step": 8084 + }, + { + "epoch": 0.8878761256314518, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.590848207473755, + "learning_rate": 1e-06, + "loss": 0.8953, + "mean_token_accuracy": 0.7261219620704651, + "num_tokens": 209225031.0, + "step": 8085 + }, + { + "epoch": 0.8879859433340654, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5337166786193848, + "learning_rate": 1e-06, + "loss": 1.1254, + "mean_token_accuracy": 0.6741964817047119, + "num_tokens": 209253581.0, + "step": 8086 + }, + { + "epoch": 0.8880957610366791, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.461634874343872, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6774032115936279, + "num_tokens": 209283539.0, + "step": 8087 + }, + { + "epoch": 0.8882055787392927, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.402974843978882, + "learning_rate": 1e-06, + "loss": 1.0915, + "mean_token_accuracy": 0.6839942336082458, + "num_tokens": 209313896.0, + "step": 8088 + }, + { + "epoch": 0.8883153964419065, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4307451248168945, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7164381742477417, + "num_tokens": 209338742.0, + "step": 8089 + }, + { + "epoch": 0.8884252141445201, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.77282977104187, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7070232629776001, + "num_tokens": 209358816.0, + "step": 8090 + }, + { + "epoch": 0.8885350318471338, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.675060987472534, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7048807144165039, + "num_tokens": 209383834.0, + "step": 8091 + }, + { + "epoch": 0.8886448495497474, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7361881732940674, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6895463466644287, + "num_tokens": 209410554.0, + "step": 8092 + }, + { + "epoch": 0.8887546672523611, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9878852367401123, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7139865159988403, + "num_tokens": 209430014.0, + "step": 8093 + }, + { + "epoch": 0.8888644849549747, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.9005274772644043, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7021682262420654, + "num_tokens": 209450901.0, + "step": 8094 + }, + { + "epoch": 0.8889743026575884, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5347883701324463, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.692814290523529, + "num_tokens": 209480423.0, + "step": 8095 + }, + { + "epoch": 0.889084120360202, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.416504144668579, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.71352219581604, + "num_tokens": 209509147.0, + "step": 8096 + }, + { + "epoch": 0.8891939380628158, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3808538913726807, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7096427083015442, + "num_tokens": 209539620.0, + "step": 8097 + }, + { + "epoch": 0.8893037557654294, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8714051246643066, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7236038446426392, + "num_tokens": 209560177.0, + "step": 8098 + }, + { + "epoch": 0.8894135734680431, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4641335010528564, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7090334892272949, + "num_tokens": 209585869.0, + "step": 8099 + }, + { + "epoch": 0.8895233911706567, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.977778673171997, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6930596232414246, + "num_tokens": 209607506.0, + "step": 8100 + }, + { + "epoch": 0.8896332088732704, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.550029754638672, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.6974831819534302, + "num_tokens": 209634975.0, + "step": 8101 + }, + { + "epoch": 0.889743026575884, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6235620975494385, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7027068138122559, + "num_tokens": 209661090.0, + "step": 8102 + }, + { + "epoch": 0.8898528442784976, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6967740058898926, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.6972335577011108, + "num_tokens": 209687591.0, + "step": 8103 + }, + { + "epoch": 0.8899626619811114, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.757119655609131, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7105714678764343, + "num_tokens": 209709367.0, + "step": 8104 + }, + { + "epoch": 0.890072479683725, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.447687864303589, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.7028972506523132, + "num_tokens": 209740280.0, + "step": 8105 + }, + { + "epoch": 0.8901822973863387, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4719033241271973, + "learning_rate": 1e-06, + "loss": 0.9323, + "mean_token_accuracy": 0.7216970324516296, + "num_tokens": 209766365.0, + "step": 8106 + }, + { + "epoch": 0.8902921150889523, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6218650341033936, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.7010130286216736, + "num_tokens": 209792007.0, + "step": 8107 + }, + { + "epoch": 0.890401932791566, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4952428340911865, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7021464109420776, + "num_tokens": 209820382.0, + "step": 8108 + }, + { + "epoch": 0.8905117504941796, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6261773109436035, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.6947622895240784, + "num_tokens": 209844582.0, + "step": 8109 + }, + { + "epoch": 0.8906215681967933, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4167723655700684, + "learning_rate": 1e-06, + "loss": 1.1588, + "mean_token_accuracy": 0.6656374931335449, + "num_tokens": 209876586.0, + "step": 8110 + }, + { + "epoch": 0.890731385899407, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.1486129760742188, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7152971029281616, + "num_tokens": 209901208.0, + "step": 8111 + }, + { + "epoch": 0.8908412036020207, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.682265520095825, + "learning_rate": 1e-06, + "loss": 0.8757, + "mean_token_accuracy": 0.734700083732605, + "num_tokens": 209922889.0, + "step": 8112 + }, + { + "epoch": 0.8909510213046343, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.48075795173645, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.7020701169967651, + "num_tokens": 209948901.0, + "step": 8113 + }, + { + "epoch": 0.891060839007248, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6066932678222656, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7047137022018433, + "num_tokens": 209972783.0, + "step": 8114 + }, + { + "epoch": 0.8911706567098616, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.280378818511963, + "learning_rate": 1e-06, + "loss": 1.1211, + "mean_token_accuracy": 0.6767025589942932, + "num_tokens": 210003463.0, + "step": 8115 + }, + { + "epoch": 0.8912804744124753, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7020108699798584, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7047449946403503, + "num_tokens": 210026671.0, + "step": 8116 + }, + { + "epoch": 0.8913902921150889, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.536789655685425, + "learning_rate": 1e-06, + "loss": 0.9262, + "mean_token_accuracy": 0.7212815284729004, + "num_tokens": 210049798.0, + "step": 8117 + }, + { + "epoch": 0.8915001098177027, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5412046909332275, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.6934226155281067, + "num_tokens": 210076020.0, + "step": 8118 + }, + { + "epoch": 0.8916099275203163, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3504066467285156, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6851611733436584, + "num_tokens": 210106761.0, + "step": 8119 + }, + { + "epoch": 0.89171974522293, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6468262672424316, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7154051065444946, + "num_tokens": 210132344.0, + "step": 8120 + }, + { + "epoch": 0.8918295629255436, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4645299911499023, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7008435726165771, + "num_tokens": 210159175.0, + "step": 8121 + }, + { + "epoch": 0.8919393806281573, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4393470287323, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.686429500579834, + "num_tokens": 210189422.0, + "step": 8122 + }, + { + "epoch": 0.8920491983307709, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 3.0233817100524902, + "learning_rate": 1e-06, + "loss": 1.1091, + "mean_token_accuracy": 0.6840361952781677, + "num_tokens": 210220355.0, + "step": 8123 + }, + { + "epoch": 0.8921590160333845, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.729667901992798, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7085236310958862, + "num_tokens": 210243366.0, + "step": 8124 + }, + { + "epoch": 0.8922688337359982, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.459146499633789, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6958355903625488, + "num_tokens": 210273857.0, + "step": 8125 + }, + { + "epoch": 0.892378651438612, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7355706691741943, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6876441836357117, + "num_tokens": 210298244.0, + "step": 8126 + }, + { + "epoch": 0.8924884691412256, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3837904930114746, + "learning_rate": 1e-06, + "loss": 1.1117, + "mean_token_accuracy": 0.6813998222351074, + "num_tokens": 210329923.0, + "step": 8127 + }, + { + "epoch": 0.8925982868438392, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5197348594665527, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6973259449005127, + "num_tokens": 210358538.0, + "step": 8128 + }, + { + "epoch": 0.8927081045464529, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3817834854125977, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7043616771697998, + "num_tokens": 210386966.0, + "step": 8129 + }, + { + "epoch": 0.8928179222490665, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.46359920501709, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6889763474464417, + "num_tokens": 210414246.0, + "step": 8130 + }, + { + "epoch": 0.8929277399516802, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.658024549484253, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6888515949249268, + "num_tokens": 210437784.0, + "step": 8131 + }, + { + "epoch": 0.8930375576542938, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.5538883209228516, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6960130333900452, + "num_tokens": 210463775.0, + "step": 8132 + }, + { + "epoch": 0.8931473753569076, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5245423316955566, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6930290460586548, + "num_tokens": 210492247.0, + "step": 8133 + }, + { + "epoch": 0.8932571930595212, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6570167541503906, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6969050168991089, + "num_tokens": 210516236.0, + "step": 8134 + }, + { + "epoch": 0.8933670107621349, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5755412578582764, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.7177208065986633, + "num_tokens": 210539167.0, + "step": 8135 + }, + { + "epoch": 0.8934768284647485, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6001152992248535, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7097569704055786, + "num_tokens": 210562684.0, + "step": 8136 + }, + { + "epoch": 0.8935866461673622, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.2265450954437256, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7181896567344666, + "num_tokens": 210596516.0, + "step": 8137 + }, + { + "epoch": 0.8936964638699758, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4241743087768555, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7188771963119507, + "num_tokens": 210623295.0, + "step": 8138 + }, + { + "epoch": 0.8938062815725895, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.8543806076049805, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7214902639389038, + "num_tokens": 210643221.0, + "step": 8139 + }, + { + "epoch": 0.8939160992752032, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.537829637527466, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6928200721740723, + "num_tokens": 210670898.0, + "step": 8140 + }, + { + "epoch": 0.8940259169778169, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7613189220428467, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.7017937898635864, + "num_tokens": 210694162.0, + "step": 8141 + }, + { + "epoch": 0.8941357346804305, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.552365303039551, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6915774941444397, + "num_tokens": 210719930.0, + "step": 8142 + }, + { + "epoch": 0.8942455523830442, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5226991176605225, + "learning_rate": 1e-06, + "loss": 1.1037, + "mean_token_accuracy": 0.6812785267829895, + "num_tokens": 210747367.0, + "step": 8143 + }, + { + "epoch": 0.8943553700856578, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.778909683227539, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.7036638259887695, + "num_tokens": 210773069.0, + "step": 8144 + }, + { + "epoch": 0.8944651877882714, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.493135690689087, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7086619138717651, + "num_tokens": 210801344.0, + "step": 8145 + }, + { + "epoch": 0.8945750054908851, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.2651631832122803, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7111043930053711, + "num_tokens": 210835548.0, + "step": 8146 + }, + { + "epoch": 0.8946848231934988, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.669625997543335, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7206628918647766, + "num_tokens": 210857281.0, + "step": 8147 + }, + { + "epoch": 0.8947946408961125, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.418649196624756, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.7028360366821289, + "num_tokens": 210884292.0, + "step": 8148 + }, + { + "epoch": 0.8949044585987261, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.622973918914795, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7038096189498901, + "num_tokens": 210908448.0, + "step": 8149 + }, + { + "epoch": 0.8950142763013398, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.8458874225616455, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6863013505935669, + "num_tokens": 210932876.0, + "step": 8150 + }, + { + "epoch": 0.8951240940039534, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.306079387664795, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7099571824073792, + "num_tokens": 210963703.0, + "step": 8151 + }, + { + "epoch": 0.8952339117065671, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.542940855026245, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7047399282455444, + "num_tokens": 210990458.0, + "step": 8152 + }, + { + "epoch": 0.8953437294091807, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.3510332107543945, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.6893740892410278, + "num_tokens": 211021700.0, + "step": 8153 + }, + { + "epoch": 0.8954535471117944, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.630131721496582, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7011464834213257, + "num_tokens": 211044523.0, + "step": 8154 + }, + { + "epoch": 0.8955633648144081, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7693471908569336, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7073929905891418, + "num_tokens": 211067809.0, + "step": 8155 + }, + { + "epoch": 0.8956731825170218, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6489014625549316, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6994861364364624, + "num_tokens": 211093559.0, + "step": 8156 + }, + { + "epoch": 0.8957830002196354, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4800708293914795, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7149041891098022, + "num_tokens": 211120767.0, + "step": 8157 + }, + { + "epoch": 0.8958928179222491, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.577558755874634, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7115429639816284, + "num_tokens": 211144996.0, + "step": 8158 + }, + { + "epoch": 0.8960026356248627, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3885531425476074, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7147314548492432, + "num_tokens": 211175424.0, + "step": 8159 + }, + { + "epoch": 0.8961124533274764, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3327622413635254, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7028764486312866, + "num_tokens": 211204392.0, + "step": 8160 + }, + { + "epoch": 0.89622227103009, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.547588586807251, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6966419219970703, + "num_tokens": 211229550.0, + "step": 8161 + }, + { + "epoch": 0.8963320887327038, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5515189170837402, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.716506838798523, + "num_tokens": 211255545.0, + "step": 8162 + }, + { + "epoch": 0.8964419064353174, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4625165462493896, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.694242000579834, + "num_tokens": 211283602.0, + "step": 8163 + }, + { + "epoch": 0.896551724137931, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.8857691287994385, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7018694877624512, + "num_tokens": 211302935.0, + "step": 8164 + }, + { + "epoch": 0.8966615418405447, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.0382254123687744, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7031782865524292, + "num_tokens": 211321835.0, + "step": 8165 + }, + { + "epoch": 0.8967713595431583, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4958202838897705, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6949930787086487, + "num_tokens": 211351891.0, + "step": 8166 + }, + { + "epoch": 0.896881177245772, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7536768913269043, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6929417848587036, + "num_tokens": 211372892.0, + "step": 8167 + }, + { + "epoch": 0.8969909949483856, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7919585704803467, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7132179737091064, + "num_tokens": 211394671.0, + "step": 8168 + }, + { + "epoch": 0.8971008126509994, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4318201541900635, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7090514302253723, + "num_tokens": 211422879.0, + "step": 8169 + }, + { + "epoch": 0.897210630353613, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4183483123779297, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6964157223701477, + "num_tokens": 211450565.0, + "step": 8170 + }, + { + "epoch": 0.8973204480562267, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7231266498565674, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.6954624056816101, + "num_tokens": 211471829.0, + "step": 8171 + }, + { + "epoch": 0.8974302657588403, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8399322032928467, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7182486057281494, + "num_tokens": 211495522.0, + "step": 8172 + }, + { + "epoch": 0.897540083461454, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.45096492767334, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7070677280426025, + "num_tokens": 211522860.0, + "step": 8173 + }, + { + "epoch": 0.8976499011640676, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.554245710372925, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6913427710533142, + "num_tokens": 211548583.0, + "step": 8174 + }, + { + "epoch": 0.8977597188666813, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6564173698425293, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6996902227401733, + "num_tokens": 211572444.0, + "step": 8175 + }, + { + "epoch": 0.897869536569295, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6460251808166504, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7206918001174927, + "num_tokens": 211594694.0, + "step": 8176 + }, + { + "epoch": 0.8979793542719087, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9455840587615967, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7015806436538696, + "num_tokens": 211614529.0, + "step": 8177 + }, + { + "epoch": 0.8980891719745223, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.569582939147949, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.708114743232727, + "num_tokens": 211640705.0, + "step": 8178 + }, + { + "epoch": 0.898198989677136, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.65990948677063, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7090168595314026, + "num_tokens": 211664393.0, + "step": 8179 + }, + { + "epoch": 0.8983088073797496, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.526221513748169, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7131417393684387, + "num_tokens": 211689982.0, + "step": 8180 + }, + { + "epoch": 0.8984186250823633, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8377177715301514, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7082632780075073, + "num_tokens": 211708693.0, + "step": 8181 + }, + { + "epoch": 0.8985284427849769, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8440418243408203, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6905648708343506, + "num_tokens": 211732231.0, + "step": 8182 + }, + { + "epoch": 0.8986382604875905, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2956056594848633, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7121831178665161, + "num_tokens": 211760191.0, + "step": 8183 + }, + { + "epoch": 0.8987480781902043, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.554060935974121, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6921209096908569, + "num_tokens": 211786463.0, + "step": 8184 + }, + { + "epoch": 0.898857895892818, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5817861557006836, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7043681740760803, + "num_tokens": 211812926.0, + "step": 8185 + }, + { + "epoch": 0.8989677135954316, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.518927574157715, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7244869470596313, + "num_tokens": 211834368.0, + "step": 8186 + }, + { + "epoch": 0.8990775312980452, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.994459390640259, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7103887796401978, + "num_tokens": 211853863.0, + "step": 8187 + }, + { + "epoch": 0.8991873490006589, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.638885974884033, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6948206424713135, + "num_tokens": 211877620.0, + "step": 8188 + }, + { + "epoch": 0.8992971667032725, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.400315523147583, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.687791645526886, + "num_tokens": 211908189.0, + "step": 8189 + }, + { + "epoch": 0.8994069844058862, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.584364652633667, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6925626993179321, + "num_tokens": 211935489.0, + "step": 8190 + }, + { + "epoch": 0.8995168021084999, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7243754863739014, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.706912636756897, + "num_tokens": 211958611.0, + "step": 8191 + }, + { + "epoch": 0.8996266198111136, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5180842876434326, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7043739557266235, + "num_tokens": 211983462.0, + "step": 8192 + }, + { + "epoch": 0.8997364375137272, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4015350341796875, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7185540199279785, + "num_tokens": 212011557.0, + "step": 8193 + }, + { + "epoch": 0.8998462552163409, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.269036054611206, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.7225043773651123, + "num_tokens": 212042302.0, + "step": 8194 + }, + { + "epoch": 0.8999560729189545, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6669812202453613, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6985478401184082, + "num_tokens": 212065357.0, + "step": 8195 + }, + { + "epoch": 0.9000658906215682, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5317046642303467, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7015402913093567, + "num_tokens": 212092573.0, + "step": 8196 + }, + { + "epoch": 0.9001757083241818, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.473799467086792, + "learning_rate": 1e-06, + "loss": 1.1243, + "mean_token_accuracy": 0.6755180358886719, + "num_tokens": 212120452.0, + "step": 8197 + }, + { + "epoch": 0.9002855260267956, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5453453063964844, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7149986624717712, + "num_tokens": 212146205.0, + "step": 8198 + }, + { + "epoch": 0.9003953437294092, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.583019256591797, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7154773473739624, + "num_tokens": 212170022.0, + "step": 8199 + }, + { + "epoch": 0.9005051614320229, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3844549655914307, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7007390260696411, + "num_tokens": 212197763.0, + "step": 8200 + }, + { + "epoch": 0.9006149791346365, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5467145442962646, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6935524940490723, + "num_tokens": 212223065.0, + "step": 8201 + }, + { + "epoch": 0.9007247968372502, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.82236385345459, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7014292478561401, + "num_tokens": 212245488.0, + "step": 8202 + }, + { + "epoch": 0.9008346145398638, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.450221538543701, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6887184381484985, + "num_tokens": 212274724.0, + "step": 8203 + }, + { + "epoch": 0.9009444322424774, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5512633323669434, + "learning_rate": 1e-06, + "loss": 0.9125, + "mean_token_accuracy": 0.7296367883682251, + "num_tokens": 212296850.0, + "step": 8204 + }, + { + "epoch": 0.9010542499450912, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.49422025680542, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6799021363258362, + "num_tokens": 212324861.0, + "step": 8205 + }, + { + "epoch": 0.9011640676477048, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.578144073486328, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6907731294631958, + "num_tokens": 212350599.0, + "step": 8206 + }, + { + "epoch": 0.9012738853503185, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8072171211242676, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7072514295578003, + "num_tokens": 212374498.0, + "step": 8207 + }, + { + "epoch": 0.9013837030529321, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.60611891746521, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6859729290008545, + "num_tokens": 212400734.0, + "step": 8208 + }, + { + "epoch": 0.9014935207555458, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7549421787261963, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7226758003234863, + "num_tokens": 212428540.0, + "step": 8209 + }, + { + "epoch": 0.9016033384581594, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7203593254089355, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7179655432701111, + "num_tokens": 212450911.0, + "step": 8210 + }, + { + "epoch": 0.9017131561607731, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6837713718414307, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7287344336509705, + "num_tokens": 212473053.0, + "step": 8211 + }, + { + "epoch": 0.9018229738633867, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6241559982299805, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6939438581466675, + "num_tokens": 212498830.0, + "step": 8212 + }, + { + "epoch": 0.9019327915660005, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3284358978271484, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6918683648109436, + "num_tokens": 212530401.0, + "step": 8213 + }, + { + "epoch": 0.9020426092686141, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.392922878265381, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6837484836578369, + "num_tokens": 212558520.0, + "step": 8214 + }, + { + "epoch": 0.9021524269712278, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.544747829437256, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6885578632354736, + "num_tokens": 212586735.0, + "step": 8215 + }, + { + "epoch": 0.9022622446738414, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5789828300476074, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6925352811813354, + "num_tokens": 212613134.0, + "step": 8216 + }, + { + "epoch": 0.9023720623764551, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.541224956512451, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6955249309539795, + "num_tokens": 212638391.0, + "step": 8217 + }, + { + "epoch": 0.9024818800790687, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5531463623046875, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7142577767372131, + "num_tokens": 212663725.0, + "step": 8218 + }, + { + "epoch": 0.9025916977816824, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5149290561676025, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7046576738357544, + "num_tokens": 212693559.0, + "step": 8219 + }, + { + "epoch": 0.9027015154842961, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5890414714813232, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7036252021789551, + "num_tokens": 212716262.0, + "step": 8220 + }, + { + "epoch": 0.9028113331869098, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4193105697631836, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.711622416973114, + "num_tokens": 212747179.0, + "step": 8221 + }, + { + "epoch": 0.9029211508895234, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.475909471511841, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7049059867858887, + "num_tokens": 212774828.0, + "step": 8222 + }, + { + "epoch": 0.903030968592137, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.46102237701416, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.6883687973022461, + "num_tokens": 212808125.0, + "step": 8223 + }, + { + "epoch": 0.9031407862947507, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.578467607498169, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6821398138999939, + "num_tokens": 212834200.0, + "step": 8224 + }, + { + "epoch": 0.9032506039973643, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5404281616210938, + "learning_rate": 1e-06, + "loss": 1.1215, + "mean_token_accuracy": 0.678185224533081, + "num_tokens": 212861783.0, + "step": 8225 + }, + { + "epoch": 0.903360421699978, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.357595682144165, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6864607334136963, + "num_tokens": 212893688.0, + "step": 8226 + }, + { + "epoch": 0.9034702394025917, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5052900314331055, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7017489075660706, + "num_tokens": 212923304.0, + "step": 8227 + }, + { + "epoch": 0.9035800571052054, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.385610580444336, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6823911070823669, + "num_tokens": 212955307.0, + "step": 8228 + }, + { + "epoch": 0.903689874807819, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.736823320388794, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7036259174346924, + "num_tokens": 212976112.0, + "step": 8229 + }, + { + "epoch": 0.9037996925104327, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.838771343231201, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7099428176879883, + "num_tokens": 212995197.0, + "step": 8230 + }, + { + "epoch": 0.9039095102130463, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5168581008911133, + "learning_rate": 1e-06, + "loss": 1.1604, + "mean_token_accuracy": 0.6715384721755981, + "num_tokens": 213024273.0, + "step": 8231 + }, + { + "epoch": 0.90401932791566, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3269612789154053, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7112264037132263, + "num_tokens": 213053233.0, + "step": 8232 + }, + { + "epoch": 0.9041291456182736, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.8519818782806396, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7226787805557251, + "num_tokens": 213074531.0, + "step": 8233 + }, + { + "epoch": 0.9042389633208874, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3652327060699463, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7001878023147583, + "num_tokens": 213103128.0, + "step": 8234 + }, + { + "epoch": 0.904348781023501, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.741084337234497, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7094463109970093, + "num_tokens": 213124284.0, + "step": 8235 + }, + { + "epoch": 0.9044585987261147, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4488258361816406, + "learning_rate": 1e-06, + "loss": 0.8973, + "mean_token_accuracy": 0.7342125177383423, + "num_tokens": 213153427.0, + "step": 8236 + }, + { + "epoch": 0.9045684164287283, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.537578821182251, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6894515752792358, + "num_tokens": 213180649.0, + "step": 8237 + }, + { + "epoch": 0.904678234131342, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4494802951812744, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.6880106329917908, + "num_tokens": 213212215.0, + "step": 8238 + }, + { + "epoch": 0.9047880518339556, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4759628772735596, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6859322190284729, + "num_tokens": 213242563.0, + "step": 8239 + }, + { + "epoch": 0.9048978695365693, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.532806158065796, + "learning_rate": 1e-06, + "loss": 1.0943, + "mean_token_accuracy": 0.6827760934829712, + "num_tokens": 213271046.0, + "step": 8240 + }, + { + "epoch": 0.9050076872391829, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3343451023101807, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7148066163063049, + "num_tokens": 213301255.0, + "step": 8241 + }, + { + "epoch": 0.9051175049417967, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.621643543243408, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.719745397567749, + "num_tokens": 213324833.0, + "step": 8242 + }, + { + "epoch": 0.9052273226444103, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.4840049743652344, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.6881319880485535, + "num_tokens": 213351024.0, + "step": 8243 + }, + { + "epoch": 0.905337140347024, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.481807231903076, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6839302182197571, + "num_tokens": 213379797.0, + "step": 8244 + }, + { + "epoch": 0.9054469580496376, + "ewc_loss": 8.106231689453125e-06, + "grad_norm": 2.397834062576294, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6824756860733032, + "num_tokens": 213409556.0, + "step": 8245 + }, + { + "epoch": 0.9055567757522512, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5830178260803223, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7033466100692749, + "num_tokens": 213433519.0, + "step": 8246 + }, + { + "epoch": 0.9056665934548649, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5350215435028076, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.6883190870285034, + "num_tokens": 213462154.0, + "step": 8247 + }, + { + "epoch": 0.9057764111574785, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.776862621307373, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.707804799079895, + "num_tokens": 213485521.0, + "step": 8248 + }, + { + "epoch": 0.9058862288600923, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4339993000030518, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.6971416473388672, + "num_tokens": 213514699.0, + "step": 8249 + }, + { + "epoch": 0.9059960465627059, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.626864433288574, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6921123266220093, + "num_tokens": 213539877.0, + "step": 8250 + }, + { + "epoch": 0.9061058642653196, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4806771278381348, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6931052207946777, + "num_tokens": 213567267.0, + "step": 8251 + }, + { + "epoch": 0.9062156819679332, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6960103511810303, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7126045823097229, + "num_tokens": 213593174.0, + "step": 8252 + }, + { + "epoch": 0.9063254996705469, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6042394638061523, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.697332501411438, + "num_tokens": 213619039.0, + "step": 8253 + }, + { + "epoch": 0.9064353173731605, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5741093158721924, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7095812559127808, + "num_tokens": 213644412.0, + "step": 8254 + }, + { + "epoch": 0.9065451350757742, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6112279891967773, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7058151960372925, + "num_tokens": 213668556.0, + "step": 8255 + }, + { + "epoch": 0.9066549527783879, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.281358480453491, + "learning_rate": 1e-06, + "loss": 0.8729, + "mean_token_accuracy": 0.7390309572219849, + "num_tokens": 213694420.0, + "step": 8256 + }, + { + "epoch": 0.9067647704810016, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.8190555572509766, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6847761869430542, + "num_tokens": 213721246.0, + "step": 8257 + }, + { + "epoch": 0.9068745881836152, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5394532680511475, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7089066505432129, + "num_tokens": 213747046.0, + "step": 8258 + }, + { + "epoch": 0.9069844058862289, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7153735160827637, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6965012550354004, + "num_tokens": 213771105.0, + "step": 8259 + }, + { + "epoch": 0.9070942235888425, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6088998317718506, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7190366983413696, + "num_tokens": 213795878.0, + "step": 8260 + }, + { + "epoch": 0.9072040412914562, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.442290782928467, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.689167857170105, + "num_tokens": 213824083.0, + "step": 8261 + }, + { + "epoch": 0.9073138589940698, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5362493991851807, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7131211161613464, + "num_tokens": 213851743.0, + "step": 8262 + }, + { + "epoch": 0.9074236766966836, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.754711627960205, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7023999691009521, + "num_tokens": 213872726.0, + "step": 8263 + }, + { + "epoch": 0.9075334943992972, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.3194401264190674, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6955960988998413, + "num_tokens": 213904422.0, + "step": 8264 + }, + { + "epoch": 0.9076433121019108, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4935202598571777, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7046991586685181, + "num_tokens": 213929673.0, + "step": 8265 + }, + { + "epoch": 0.9077531298045245, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6547443866729736, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7101929187774658, + "num_tokens": 213952516.0, + "step": 8266 + }, + { + "epoch": 0.9078629475071381, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.2797486782073975, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6883553266525269, + "num_tokens": 213982810.0, + "step": 8267 + }, + { + "epoch": 0.9079727652097518, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6466381549835205, + "learning_rate": 1e-06, + "loss": 1.1181, + "mean_token_accuracy": 0.6816738247871399, + "num_tokens": 214010916.0, + "step": 8268 + }, + { + "epoch": 0.9080825829123654, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.546374797821045, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6987868547439575, + "num_tokens": 214035488.0, + "step": 8269 + }, + { + "epoch": 0.9081924006149792, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.319300651550293, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7180435657501221, + "num_tokens": 214064658.0, + "step": 8270 + }, + { + "epoch": 0.9083022183175928, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.575861930847168, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.6981304883956909, + "num_tokens": 214090493.0, + "step": 8271 + }, + { + "epoch": 0.9084120360202065, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.8849141597747803, + "learning_rate": 1e-06, + "loss": 0.9419, + "mean_token_accuracy": 0.7165868282318115, + "num_tokens": 214109319.0, + "step": 8272 + }, + { + "epoch": 0.9085218537228201, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4437315464019775, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.7019128203392029, + "num_tokens": 214135566.0, + "step": 8273 + }, + { + "epoch": 0.9086316714254338, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6155176162719727, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7146427631378174, + "num_tokens": 214158638.0, + "step": 8274 + }, + { + "epoch": 0.9087414891280474, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.298060894012451, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6922409534454346, + "num_tokens": 214186129.0, + "step": 8275 + }, + { + "epoch": 0.9088513068306611, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.146932601928711, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6919394731521606, + "num_tokens": 214216102.0, + "step": 8276 + }, + { + "epoch": 0.9089611245332747, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4304990768432617, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6992282867431641, + "num_tokens": 214244267.0, + "step": 8277 + }, + { + "epoch": 0.9090709422358885, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3119359016418457, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.689570426940918, + "num_tokens": 214274438.0, + "step": 8278 + }, + { + "epoch": 0.9091807599385021, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.249903440475464, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6867209076881409, + "num_tokens": 214306660.0, + "step": 8279 + }, + { + "epoch": 0.9092905776411158, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.519472122192383, + "learning_rate": 1e-06, + "loss": 0.8944, + "mean_token_accuracy": 0.7389172315597534, + "num_tokens": 214331101.0, + "step": 8280 + }, + { + "epoch": 0.9094003953437294, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4429383277893066, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7174919843673706, + "num_tokens": 214358743.0, + "step": 8281 + }, + { + "epoch": 0.909510213046343, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.7154364585876465, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6939911842346191, + "num_tokens": 214386075.0, + "step": 8282 + }, + { + "epoch": 0.9096200307489567, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.0353705883026123, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7039977312088013, + "num_tokens": 214405293.0, + "step": 8283 + }, + { + "epoch": 0.9097298484515703, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4272782802581787, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7020120620727539, + "num_tokens": 214432685.0, + "step": 8284 + }, + { + "epoch": 0.9098396661541841, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4971344470977783, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6967815160751343, + "num_tokens": 214458723.0, + "step": 8285 + }, + { + "epoch": 0.9099494838567977, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6435365676879883, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7102149128913879, + "num_tokens": 214486020.0, + "step": 8286 + }, + { + "epoch": 0.9100593015594114, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6457176208496094, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6994101405143738, + "num_tokens": 214509254.0, + "step": 8287 + }, + { + "epoch": 0.910169119262025, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4101061820983887, + "learning_rate": 1e-06, + "loss": 1.1112, + "mean_token_accuracy": 0.6837297081947327, + "num_tokens": 214540063.0, + "step": 8288 + }, + { + "epoch": 0.9102789369646387, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.617272138595581, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6909698843955994, + "num_tokens": 214565092.0, + "step": 8289 + }, + { + "epoch": 0.9103887546672523, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.999157190322876, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7000098824501038, + "num_tokens": 214598113.0, + "step": 8290 + }, + { + "epoch": 0.910498572369866, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 7.961052894592285, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7054187655448914, + "num_tokens": 214621759.0, + "step": 8291 + }, + { + "epoch": 0.9106083900724797, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.645218849182129, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7146161794662476, + "num_tokens": 214646977.0, + "step": 8292 + }, + { + "epoch": 0.9107182077750934, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7923507690429688, + "learning_rate": 1e-06, + "loss": 0.8975, + "mean_token_accuracy": 0.7295801043510437, + "num_tokens": 214667009.0, + "step": 8293 + }, + { + "epoch": 0.910828025477707, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6633706092834473, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6874865889549255, + "num_tokens": 214690203.0, + "step": 8294 + }, + { + "epoch": 0.9109378431803207, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.74538254737854, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7217735052108765, + "num_tokens": 214711257.0, + "step": 8295 + }, + { + "epoch": 0.9110476608829343, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2368361949920654, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7008222341537476, + "num_tokens": 214740590.0, + "step": 8296 + }, + { + "epoch": 0.911157478585548, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2225887775421143, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7089083194732666, + "num_tokens": 214772673.0, + "step": 8297 + }, + { + "epoch": 0.9112672962881616, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.694298267364502, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6920256614685059, + "num_tokens": 214796315.0, + "step": 8298 + }, + { + "epoch": 0.9113771139907754, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.542846202850342, + "learning_rate": 1e-06, + "loss": 1.112, + "mean_token_accuracy": 0.6810131669044495, + "num_tokens": 214822372.0, + "step": 8299 + }, + { + "epoch": 0.911486931693389, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.249741792678833, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6878989934921265, + "num_tokens": 214855213.0, + "step": 8300 + }, + { + "epoch": 0.9115967493960027, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.406902551651001, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.6985538005828857, + "num_tokens": 214890209.0, + "step": 8301 + }, + { + "epoch": 0.9117065670986163, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.2680890560150146, + "learning_rate": 1e-06, + "loss": 1.1222, + "mean_token_accuracy": 0.6809657216072083, + "num_tokens": 214924274.0, + "step": 8302 + }, + { + "epoch": 0.91181638480123, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.675342082977295, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6944923400878906, + "num_tokens": 214946732.0, + "step": 8303 + }, + { + "epoch": 0.9119262025038436, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.9483766555786133, + "learning_rate": 1e-06, + "loss": 0.9195, + "mean_token_accuracy": 0.7249599695205688, + "num_tokens": 214965979.0, + "step": 8304 + }, + { + "epoch": 0.9120360202064572, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.9729745388031006, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6953268051147461, + "num_tokens": 214989147.0, + "step": 8305 + }, + { + "epoch": 0.9121458379090709, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.700076103210449, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7097458839416504, + "num_tokens": 215012035.0, + "step": 8306 + }, + { + "epoch": 0.9122556556116846, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6897284984588623, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7183573842048645, + "num_tokens": 215036677.0, + "step": 8307 + }, + { + "epoch": 0.9123654733142983, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6936452388763428, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7228928804397583, + "num_tokens": 215057133.0, + "step": 8308 + }, + { + "epoch": 0.9124752910169119, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 7.87802267074585, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6913923025131226, + "num_tokens": 215082831.0, + "step": 8309 + }, + { + "epoch": 0.9125851087195256, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6849820613861084, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6991236209869385, + "num_tokens": 215106983.0, + "step": 8310 + }, + { + "epoch": 0.9126949264221392, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.498948335647583, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7158550024032593, + "num_tokens": 215131950.0, + "step": 8311 + }, + { + "epoch": 0.9128047441247529, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.707014322280884, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7038785219192505, + "num_tokens": 215153394.0, + "step": 8312 + }, + { + "epoch": 0.9129145618273665, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3557114601135254, + "learning_rate": 1e-06, + "loss": 1.1141, + "mean_token_accuracy": 0.6728184819221497, + "num_tokens": 215184587.0, + "step": 8313 + }, + { + "epoch": 0.9130243795299803, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.444551467895508, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6940498352050781, + "num_tokens": 215212894.0, + "step": 8314 + }, + { + "epoch": 0.9131341972325939, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.540933609008789, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6910244226455688, + "num_tokens": 215240789.0, + "step": 8315 + }, + { + "epoch": 0.9132440149352076, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6312506198883057, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6819170117378235, + "num_tokens": 215265364.0, + "step": 8316 + }, + { + "epoch": 0.9133538326378212, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.70813250541687, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7038671970367432, + "num_tokens": 215285525.0, + "step": 8317 + }, + { + "epoch": 0.9134636503404349, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6577186584472656, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7125723958015442, + "num_tokens": 215308355.0, + "step": 8318 + }, + { + "epoch": 0.9135734680430485, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8212890625, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6917127370834351, + "num_tokens": 215336239.0, + "step": 8319 + }, + { + "epoch": 0.9136832857456622, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.734971046447754, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7010928988456726, + "num_tokens": 215360536.0, + "step": 8320 + }, + { + "epoch": 0.9137931034482759, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5216009616851807, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7031937837600708, + "num_tokens": 215386149.0, + "step": 8321 + }, + { + "epoch": 0.9139029211508896, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.85727596282959, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7164193987846375, + "num_tokens": 215407934.0, + "step": 8322 + }, + { + "epoch": 0.9140127388535032, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.4285426139831543, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7229171991348267, + "num_tokens": 215435377.0, + "step": 8323 + }, + { + "epoch": 0.9141225565561168, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6656856536865234, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.693702220916748, + "num_tokens": 215461261.0, + "step": 8324 + }, + { + "epoch": 0.9142323742587305, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.872384548187256, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7169744968414307, + "num_tokens": 215483834.0, + "step": 8325 + }, + { + "epoch": 0.9143421919613441, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5734167098999023, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6905076503753662, + "num_tokens": 215508844.0, + "step": 8326 + }, + { + "epoch": 0.9144520096639578, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4349019527435303, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7230041027069092, + "num_tokens": 215536103.0, + "step": 8327 + }, + { + "epoch": 0.9145618273665715, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.906182289123535, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7221312522888184, + "num_tokens": 215555919.0, + "step": 8328 + }, + { + "epoch": 0.9146716450691852, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5426740646362305, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.7019799947738647, + "num_tokens": 215580581.0, + "step": 8329 + }, + { + "epoch": 0.9147814627717988, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.584479331970215, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7180284261703491, + "num_tokens": 215605680.0, + "step": 8330 + }, + { + "epoch": 0.9148912804744125, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.451124668121338, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6876056790351868, + "num_tokens": 215634982.0, + "step": 8331 + }, + { + "epoch": 0.9150010981770261, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.341641426086426, + "learning_rate": 1e-06, + "loss": 1.0874, + "mean_token_accuracy": 0.6856659650802612, + "num_tokens": 215667249.0, + "step": 8332 + }, + { + "epoch": 0.9151109158796398, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.722954034805298, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7164344787597656, + "num_tokens": 215690494.0, + "step": 8333 + }, + { + "epoch": 0.9152207335822534, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5111966133117676, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6901670098304749, + "num_tokens": 215716541.0, + "step": 8334 + }, + { + "epoch": 0.9153305512848671, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4814913272857666, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7012508511543274, + "num_tokens": 215741018.0, + "step": 8335 + }, + { + "epoch": 0.9154403689874808, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4925615787506104, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7114572525024414, + "num_tokens": 215765566.0, + "step": 8336 + }, + { + "epoch": 0.9155501866900945, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.8698859214782715, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7134971022605896, + "num_tokens": 215789434.0, + "step": 8337 + }, + { + "epoch": 0.9156600043927081, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7416021823883057, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.711140513420105, + "num_tokens": 215816457.0, + "step": 8338 + }, + { + "epoch": 0.9157698220953218, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7464053630828857, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7165133953094482, + "num_tokens": 215839046.0, + "step": 8339 + }, + { + "epoch": 0.9158796397979354, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.704432487487793, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7196660041809082, + "num_tokens": 215861389.0, + "step": 8340 + }, + { + "epoch": 0.915989457500549, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.945462226867676, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7118606567382812, + "num_tokens": 215881662.0, + "step": 8341 + }, + { + "epoch": 0.9160992752031627, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.727350950241089, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7192652821540833, + "num_tokens": 215901617.0, + "step": 8342 + }, + { + "epoch": 0.9162090929057765, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7103307247161865, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7084710001945496, + "num_tokens": 215927452.0, + "step": 8343 + }, + { + "epoch": 0.9163189106083901, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5400524139404297, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7096366882324219, + "num_tokens": 215953475.0, + "step": 8344 + }, + { + "epoch": 0.9164287283110037, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.675649404525757, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6976062059402466, + "num_tokens": 215978069.0, + "step": 8345 + }, + { + "epoch": 0.9165385460136174, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3900458812713623, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.712897777557373, + "num_tokens": 216005963.0, + "step": 8346 + }, + { + "epoch": 0.916648363716231, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5488357543945312, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7083496451377869, + "num_tokens": 216032217.0, + "step": 8347 + }, + { + "epoch": 0.9167581814188447, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4687142372131348, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6765410900115967, + "num_tokens": 216061479.0, + "step": 8348 + }, + { + "epoch": 0.9168679991214583, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.437742233276367, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7023227214813232, + "num_tokens": 216090079.0, + "step": 8349 + }, + { + "epoch": 0.9169778168240721, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.2440261840820312, + "learning_rate": 1e-06, + "loss": 1.1, + "mean_token_accuracy": 0.6816058158874512, + "num_tokens": 216117217.0, + "step": 8350 + }, + { + "epoch": 0.9170876345266857, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.465473175048828, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7212931513786316, + "num_tokens": 216144407.0, + "step": 8351 + }, + { + "epoch": 0.9171974522292994, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7074787616729736, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7193605899810791, + "num_tokens": 216168373.0, + "step": 8352 + }, + { + "epoch": 0.917307269931913, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6297683715820312, + "learning_rate": 1e-06, + "loss": 1.1411, + "mean_token_accuracy": 0.6652956008911133, + "num_tokens": 216193251.0, + "step": 8353 + }, + { + "epoch": 0.9174170876345267, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8045814037323, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7188129425048828, + "num_tokens": 216213242.0, + "step": 8354 + }, + { + "epoch": 0.9175269053371403, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 8.802074432373047, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.711249589920044, + "num_tokens": 216240426.0, + "step": 8355 + }, + { + "epoch": 0.917636723039754, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.42974853515625, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.693101167678833, + "num_tokens": 216270015.0, + "step": 8356 + }, + { + "epoch": 0.9177465407423677, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3914802074432373, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.69671630859375, + "num_tokens": 216300090.0, + "step": 8357 + }, + { + "epoch": 0.9178563584449814, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4965879917144775, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7059885859489441, + "num_tokens": 216330103.0, + "step": 8358 + }, + { + "epoch": 0.917966176147595, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.367176055908203, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7143311500549316, + "num_tokens": 216356619.0, + "step": 8359 + }, + { + "epoch": 0.9180759938502087, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.609138250350952, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7013437747955322, + "num_tokens": 216379171.0, + "step": 8360 + }, + { + "epoch": 0.9181858115528223, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.426873207092285, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6908524632453918, + "num_tokens": 216409387.0, + "step": 8361 + }, + { + "epoch": 0.918295629255436, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5455148220062256, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7104994058609009, + "num_tokens": 216433635.0, + "step": 8362 + }, + { + "epoch": 0.9184054469580496, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3046016693115234, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6932034492492676, + "num_tokens": 216466485.0, + "step": 8363 + }, + { + "epoch": 0.9185152646606632, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6223044395446777, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7097615599632263, + "num_tokens": 216488907.0, + "step": 8364 + }, + { + "epoch": 0.918625082363277, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2313072681427, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.694786548614502, + "num_tokens": 216521259.0, + "step": 8365 + }, + { + "epoch": 0.9187349000658906, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.9888839721679688, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7212991714477539, + "num_tokens": 216538864.0, + "step": 8366 + }, + { + "epoch": 0.9188447177685043, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3969712257385254, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7010617256164551, + "num_tokens": 216568362.0, + "step": 8367 + }, + { + "epoch": 0.9189545354711179, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4440760612487793, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.6982911825180054, + "num_tokens": 216598722.0, + "step": 8368 + }, + { + "epoch": 0.9190643531737316, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.765228271484375, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6891575455665588, + "num_tokens": 216623241.0, + "step": 8369 + }, + { + "epoch": 0.9191741708763452, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.440757989883423, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7169272303581238, + "num_tokens": 216650065.0, + "step": 8370 + }, + { + "epoch": 0.9192839885789589, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.555734395980835, + "learning_rate": 1e-06, + "loss": 1.1125, + "mean_token_accuracy": 0.6834851503372192, + "num_tokens": 216675984.0, + "step": 8371 + }, + { + "epoch": 0.9193938062815726, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3108043670654297, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6994592547416687, + "num_tokens": 216706335.0, + "step": 8372 + }, + { + "epoch": 0.9195036239841863, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.314635753631592, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.680180549621582, + "num_tokens": 216738678.0, + "step": 8373 + }, + { + "epoch": 0.9196134416867999, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5593812465667725, + "learning_rate": 1e-06, + "loss": 0.825, + "mean_token_accuracy": 0.7573158740997314, + "num_tokens": 216764265.0, + "step": 8374 + }, + { + "epoch": 0.9197232593894136, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7687270641326904, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7246017456054688, + "num_tokens": 216787019.0, + "step": 8375 + }, + { + "epoch": 0.9198330770920272, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6501896381378174, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7009981870651245, + "num_tokens": 216810780.0, + "step": 8376 + }, + { + "epoch": 0.9199428947946409, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8128693103790283, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6860435605049133, + "num_tokens": 216833727.0, + "step": 8377 + }, + { + "epoch": 0.9200527124972545, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.581833839416504, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6827337145805359, + "num_tokens": 216860498.0, + "step": 8378 + }, + { + "epoch": 0.9201625301998683, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.780165910720825, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7172539234161377, + "num_tokens": 216880677.0, + "step": 8379 + }, + { + "epoch": 0.9202723479024819, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6135449409484863, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7086780071258545, + "num_tokens": 216903748.0, + "step": 8380 + }, + { + "epoch": 0.9203821656050956, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.665621519088745, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6843538880348206, + "num_tokens": 216935680.0, + "step": 8381 + }, + { + "epoch": 0.9204919833077092, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.60868763923645, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7011440992355347, + "num_tokens": 216960966.0, + "step": 8382 + }, + { + "epoch": 0.9206018010103229, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4943151473999023, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7100156545639038, + "num_tokens": 216984655.0, + "step": 8383 + }, + { + "epoch": 0.9207116187129365, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4558095932006836, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.721878170967102, + "num_tokens": 217013859.0, + "step": 8384 + }, + { + "epoch": 0.9208214364155501, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.794144630432129, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6858035922050476, + "num_tokens": 217039191.0, + "step": 8385 + }, + { + "epoch": 0.9209312541181639, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6890320777893066, + "learning_rate": 1e-06, + "loss": 1.0466, + "mean_token_accuracy": 0.6975817680358887, + "num_tokens": 217065249.0, + "step": 8386 + }, + { + "epoch": 0.9210410718207775, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.576279878616333, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6924070119857788, + "num_tokens": 217092553.0, + "step": 8387 + }, + { + "epoch": 0.9211508895233912, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7138619422912598, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7059409618377686, + "num_tokens": 217117661.0, + "step": 8388 + }, + { + "epoch": 0.9212607072260048, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.738463878631592, + "learning_rate": 1e-06, + "loss": 1.1049, + "mean_token_accuracy": 0.6789010167121887, + "num_tokens": 217146021.0, + "step": 8389 + }, + { + "epoch": 0.9213705249286185, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5622706413269043, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6915055513381958, + "num_tokens": 217171003.0, + "step": 8390 + }, + { + "epoch": 0.9214803426312321, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.635289192199707, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6955803632736206, + "num_tokens": 217194716.0, + "step": 8391 + }, + { + "epoch": 0.9215901603338458, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5712192058563232, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.688646137714386, + "num_tokens": 217220352.0, + "step": 8392 + }, + { + "epoch": 0.9216999780364594, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6796817779541016, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7230517864227295, + "num_tokens": 217241491.0, + "step": 8393 + }, + { + "epoch": 0.9218097957390732, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7503199577331543, + "learning_rate": 1e-06, + "loss": 0.9604, + "mean_token_accuracy": 0.7178632020950317, + "num_tokens": 217261562.0, + "step": 8394 + }, + { + "epoch": 0.9219196134416868, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2707834243774414, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7050141096115112, + "num_tokens": 217294792.0, + "step": 8395 + }, + { + "epoch": 0.9220294311443005, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3738365173339844, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7067137956619263, + "num_tokens": 217322645.0, + "step": 8396 + }, + { + "epoch": 0.9221392488469141, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2264292240142822, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6969417333602905, + "num_tokens": 217357087.0, + "step": 8397 + }, + { + "epoch": 0.9222490665495278, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6178414821624756, + "learning_rate": 1e-06, + "loss": 1.1699, + "mean_token_accuracy": 0.6669713258743286, + "num_tokens": 217383181.0, + "step": 8398 + }, + { + "epoch": 0.9223588842521414, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.894033193588257, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.708095133304596, + "num_tokens": 217401323.0, + "step": 8399 + }, + { + "epoch": 0.922468701954755, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.303769826889038, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6951212882995605, + "num_tokens": 217430911.0, + "step": 8400 + }, + { + "epoch": 0.9225785196573688, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8540496826171875, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6868973970413208, + "num_tokens": 217460849.0, + "step": 8401 + }, + { + "epoch": 0.9226883373599825, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.735598087310791, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7120349407196045, + "num_tokens": 217487254.0, + "step": 8402 + }, + { + "epoch": 0.9227981550625961, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5748400688171387, + "learning_rate": 1e-06, + "loss": 1.1401, + "mean_token_accuracy": 0.6705107092857361, + "num_tokens": 217515029.0, + "step": 8403 + }, + { + "epoch": 0.9229079727652097, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6257164478302, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.69406658411026, + "num_tokens": 217539472.0, + "step": 8404 + }, + { + "epoch": 0.9230177904678234, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5961859226226807, + "learning_rate": 1e-06, + "loss": 1.1501, + "mean_token_accuracy": 0.6927950382232666, + "num_tokens": 217565277.0, + "step": 8405 + }, + { + "epoch": 0.923127608170437, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.5605695247650146, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6958258748054504, + "num_tokens": 217591244.0, + "step": 8406 + }, + { + "epoch": 0.9232374258730507, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7219040393829346, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.699225902557373, + "num_tokens": 217619176.0, + "step": 8407 + }, + { + "epoch": 0.9233472435756644, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4217588901519775, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7181697487831116, + "num_tokens": 217646909.0, + "step": 8408 + }, + { + "epoch": 0.9234570612782781, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8530337810516357, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7033262848854065, + "num_tokens": 217667989.0, + "step": 8409 + }, + { + "epoch": 0.9235668789808917, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3941903114318848, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7170822620391846, + "num_tokens": 217694095.0, + "step": 8410 + }, + { + "epoch": 0.9236766966835054, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9657182693481445, + "learning_rate": 1e-06, + "loss": 1.1034, + "mean_token_accuracy": 0.6796211004257202, + "num_tokens": 217723548.0, + "step": 8411 + }, + { + "epoch": 0.923786514386119, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.874964952468872, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7025778293609619, + "num_tokens": 217747359.0, + "step": 8412 + }, + { + "epoch": 0.9238963320887327, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7246718406677246, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6934095621109009, + "num_tokens": 217772721.0, + "step": 8413 + }, + { + "epoch": 0.9240061497913463, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.798408031463623, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6860045790672302, + "num_tokens": 217797517.0, + "step": 8414 + }, + { + "epoch": 0.9241159674939601, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.609771728515625, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6859403848648071, + "num_tokens": 217826190.0, + "step": 8415 + }, + { + "epoch": 0.9242257851965737, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.816819429397583, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6819415092468262, + "num_tokens": 217852227.0, + "step": 8416 + }, + { + "epoch": 0.9243356028991874, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.402134418487549, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6959244012832642, + "num_tokens": 217880720.0, + "step": 8417 + }, + { + "epoch": 0.924445420601801, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.858081102371216, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7010819315910339, + "num_tokens": 217901861.0, + "step": 8418 + }, + { + "epoch": 0.9245552383044147, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4662885665893555, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6895495057106018, + "num_tokens": 217931798.0, + "step": 8419 + }, + { + "epoch": 0.9246650560070283, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.447968006134033, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6858851313591003, + "num_tokens": 217959279.0, + "step": 8420 + }, + { + "epoch": 0.924774873709642, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4981839656829834, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.6989675760269165, + "num_tokens": 217984120.0, + "step": 8421 + }, + { + "epoch": 0.9248846914122557, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.486665725708008, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7255455255508423, + "num_tokens": 218010085.0, + "step": 8422 + }, + { + "epoch": 0.9249945091148694, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2285168170928955, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.6981106996536255, + "num_tokens": 218042423.0, + "step": 8423 + }, + { + "epoch": 0.925104326817483, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.628828287124634, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.695398211479187, + "num_tokens": 218067588.0, + "step": 8424 + }, + { + "epoch": 0.9252141445200966, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.372720718383789, + "learning_rate": 1e-06, + "loss": 1.2009, + "mean_token_accuracy": 0.6573957204818726, + "num_tokens": 218101759.0, + "step": 8425 + }, + { + "epoch": 0.9253239622227103, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.468252420425415, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6964648962020874, + "num_tokens": 218129303.0, + "step": 8426 + }, + { + "epoch": 0.9254337799253239, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.626283884048462, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6951307654380798, + "num_tokens": 218154835.0, + "step": 8427 + }, + { + "epoch": 0.9255435976279376, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3981034755706787, + "learning_rate": 1e-06, + "loss": 1.1054, + "mean_token_accuracy": 0.6794754266738892, + "num_tokens": 218183636.0, + "step": 8428 + }, + { + "epoch": 0.9256534153305512, + "ewc_loss": 8.165836334228516e-06, + "grad_norm": 2.6859374046325684, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7058185935020447, + "num_tokens": 218207019.0, + "step": 8429 + }, + { + "epoch": 0.925763233033165, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8895246982574463, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7122153043746948, + "num_tokens": 218229583.0, + "step": 8430 + }, + { + "epoch": 0.9258730507357786, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.567089557647705, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6976344585418701, + "num_tokens": 218256194.0, + "step": 8431 + }, + { + "epoch": 0.9259828684383923, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.651535749435425, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6796720623970032, + "num_tokens": 218280046.0, + "step": 8432 + }, + { + "epoch": 0.9260926861410059, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.002847194671631, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7127792835235596, + "num_tokens": 218299251.0, + "step": 8433 + }, + { + "epoch": 0.9262025038436196, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.598452091217041, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7074599862098694, + "num_tokens": 218323170.0, + "step": 8434 + }, + { + "epoch": 0.9263123215462332, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6818251609802246, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.7055565118789673, + "num_tokens": 218346012.0, + "step": 8435 + }, + { + "epoch": 0.9264221392488469, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8009824752807617, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.714104413986206, + "num_tokens": 218368155.0, + "step": 8436 + }, + { + "epoch": 0.9265319569514606, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9382314682006836, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6929774284362793, + "num_tokens": 218394595.0, + "step": 8437 + }, + { + "epoch": 0.9266417746540743, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6707162857055664, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.6992412209510803, + "num_tokens": 218419929.0, + "step": 8438 + }, + { + "epoch": 0.9267515923566879, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.792344570159912, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7304868698120117, + "num_tokens": 218441566.0, + "step": 8439 + }, + { + "epoch": 0.9268614100593016, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.786750078201294, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7119436264038086, + "num_tokens": 218464451.0, + "step": 8440 + }, + { + "epoch": 0.9269712277619152, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.614912986755371, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6957516074180603, + "num_tokens": 218492548.0, + "step": 8441 + }, + { + "epoch": 0.9270810454645289, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5884478092193604, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7045740485191345, + "num_tokens": 218519220.0, + "step": 8442 + }, + { + "epoch": 0.9271908631671425, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8334569931030273, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.6930361986160278, + "num_tokens": 218541685.0, + "step": 8443 + }, + { + "epoch": 0.9273006808697563, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6515629291534424, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7193251848220825, + "num_tokens": 218565855.0, + "step": 8444 + }, + { + "epoch": 0.9274104985723699, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.457467794418335, + "learning_rate": 1e-06, + "loss": 1.1105, + "mean_token_accuracy": 0.685685396194458, + "num_tokens": 218592633.0, + "step": 8445 + }, + { + "epoch": 0.9275203162749835, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.488213300704956, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.7014051675796509, + "num_tokens": 218618417.0, + "step": 8446 + }, + { + "epoch": 0.9276301339775972, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.621084213256836, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7159209251403809, + "num_tokens": 218641999.0, + "step": 8447 + }, + { + "epoch": 0.9277399516802108, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4282965660095215, + "learning_rate": 1e-06, + "loss": 1.1224, + "mean_token_accuracy": 0.6778872609138489, + "num_tokens": 218670529.0, + "step": 8448 + }, + { + "epoch": 0.9278497693828245, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7454254627227783, + "learning_rate": 1e-06, + "loss": 1.1428, + "mean_token_accuracy": 0.6775109767913818, + "num_tokens": 218695484.0, + "step": 8449 + }, + { + "epoch": 0.9279595870854381, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4216468334198, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.7012404799461365, + "num_tokens": 218727725.0, + "step": 8450 + }, + { + "epoch": 0.9280694047880519, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.438131809234619, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7141488194465637, + "num_tokens": 218754505.0, + "step": 8451 + }, + { + "epoch": 0.9281792224906655, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6189823150634766, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7040371298789978, + "num_tokens": 218778201.0, + "step": 8452 + }, + { + "epoch": 0.9282890401932792, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7085886001586914, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7114671468734741, + "num_tokens": 218801071.0, + "step": 8453 + }, + { + "epoch": 0.9283988578958928, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7798101902008057, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6963251233100891, + "num_tokens": 218821290.0, + "step": 8454 + }, + { + "epoch": 0.9285086755985065, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6396894454956055, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7155904769897461, + "num_tokens": 218847790.0, + "step": 8455 + }, + { + "epoch": 0.9286184933011201, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.509071111679077, + "learning_rate": 1e-06, + "loss": 1.1055, + "mean_token_accuracy": 0.6757663488388062, + "num_tokens": 218872798.0, + "step": 8456 + }, + { + "epoch": 0.9287283110037338, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7599239349365234, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6916439533233643, + "num_tokens": 218896411.0, + "step": 8457 + }, + { + "epoch": 0.9288381287063474, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6230809688568115, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6969226002693176, + "num_tokens": 218919199.0, + "step": 8458 + }, + { + "epoch": 0.9289479464089612, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5788934230804443, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.6993051767349243, + "num_tokens": 218947282.0, + "step": 8459 + }, + { + "epoch": 0.9290577641115748, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6962170600891113, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7010260820388794, + "num_tokens": 218970788.0, + "step": 8460 + }, + { + "epoch": 0.9291675818141885, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.610572099685669, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7133569717407227, + "num_tokens": 219001070.0, + "step": 8461 + }, + { + "epoch": 0.9292773995168021, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5298972129821777, + "learning_rate": 1e-06, + "loss": 1.1147, + "mean_token_accuracy": 0.6722570657730103, + "num_tokens": 219029358.0, + "step": 8462 + }, + { + "epoch": 0.9293872172194158, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5762083530426025, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7237845659255981, + "num_tokens": 219055155.0, + "step": 8463 + }, + { + "epoch": 0.9294970349220294, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2020480632781982, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6833149194717407, + "num_tokens": 219088784.0, + "step": 8464 + }, + { + "epoch": 0.929606852624643, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.277657985687256, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7156897783279419, + "num_tokens": 219119790.0, + "step": 8465 + }, + { + "epoch": 0.9297166703272568, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.511427402496338, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.705741286277771, + "num_tokens": 219142801.0, + "step": 8466 + }, + { + "epoch": 0.9298264880298704, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5359911918640137, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7244980335235596, + "num_tokens": 219168134.0, + "step": 8467 + }, + { + "epoch": 0.9299363057324841, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.719050407409668, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7254132628440857, + "num_tokens": 219191748.0, + "step": 8468 + }, + { + "epoch": 0.9300461234350977, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.952064037322998, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6929463744163513, + "num_tokens": 219214834.0, + "step": 8469 + }, + { + "epoch": 0.9301559411377114, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.522571563720703, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7070685625076294, + "num_tokens": 219243265.0, + "step": 8470 + }, + { + "epoch": 0.930265758840325, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2988665103912354, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7122877836227417, + "num_tokens": 219275706.0, + "step": 8471 + }, + { + "epoch": 0.9303755765429387, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.606205940246582, + "learning_rate": 1e-06, + "loss": 1.1324, + "mean_token_accuracy": 0.6669725179672241, + "num_tokens": 219305391.0, + "step": 8472 + }, + { + "epoch": 0.9304853942455524, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7450437545776367, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6905266642570496, + "num_tokens": 219331399.0, + "step": 8473 + }, + { + "epoch": 0.9305952119481661, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.457136631011963, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7241817712783813, + "num_tokens": 219358967.0, + "step": 8474 + }, + { + "epoch": 0.9307050296507797, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2967498302459717, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6956263780593872, + "num_tokens": 219394839.0, + "step": 8475 + }, + { + "epoch": 0.9308148473533934, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8938546180725098, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7124396562576294, + "num_tokens": 219413954.0, + "step": 8476 + }, + { + "epoch": 0.930924665056007, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4461324214935303, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6995932459831238, + "num_tokens": 219440917.0, + "step": 8477 + }, + { + "epoch": 0.9310344827586207, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.420316457748413, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6819722652435303, + "num_tokens": 219469897.0, + "step": 8478 + }, + { + "epoch": 0.9311443004612343, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.825507164001465, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7047838568687439, + "num_tokens": 219491251.0, + "step": 8479 + }, + { + "epoch": 0.9312541181638481, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.79695725440979, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7023339867591858, + "num_tokens": 219512466.0, + "step": 8480 + }, + { + "epoch": 0.9313639358664617, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5576295852661133, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7015841603279114, + "num_tokens": 219539230.0, + "step": 8481 + }, + { + "epoch": 0.9314737535690754, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5086143016815186, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7083566188812256, + "num_tokens": 219567470.0, + "step": 8482 + }, + { + "epoch": 0.931583571271689, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.552906036376953, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7068608999252319, + "num_tokens": 219595741.0, + "step": 8483 + }, + { + "epoch": 0.9316933889743026, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6856837272644043, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7187758088111877, + "num_tokens": 219620741.0, + "step": 8484 + }, + { + "epoch": 0.9318032066769163, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.52668833732605, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.699691116809845, + "num_tokens": 219648414.0, + "step": 8485 + }, + { + "epoch": 0.9319130243795299, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4078710079193115, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6906434297561646, + "num_tokens": 219675933.0, + "step": 8486 + }, + { + "epoch": 0.9320228420821436, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.452162504196167, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6928725242614746, + "num_tokens": 219702289.0, + "step": 8487 + }, + { + "epoch": 0.9321326597847573, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.658806085586548, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.6915879249572754, + "num_tokens": 219724969.0, + "step": 8488 + }, + { + "epoch": 0.932242477487371, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9072463512420654, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7195611596107483, + "num_tokens": 219745247.0, + "step": 8489 + }, + { + "epoch": 0.9323522951899846, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8506832122802734, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7148112058639526, + "num_tokens": 219767121.0, + "step": 8490 + }, + { + "epoch": 0.9324621128925983, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4317522048950195, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.6998374462127686, + "num_tokens": 219793110.0, + "step": 8491 + }, + { + "epoch": 0.9325719305952119, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5841166973114014, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7137107849121094, + "num_tokens": 219817431.0, + "step": 8492 + }, + { + "epoch": 0.9326817482978256, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.640627861022949, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7077648043632507, + "num_tokens": 219840958.0, + "step": 8493 + }, + { + "epoch": 0.9327915660004392, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8416519165039062, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.6975878477096558, + "num_tokens": 219864220.0, + "step": 8494 + }, + { + "epoch": 0.932901383703053, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7802741527557373, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7199088335037231, + "num_tokens": 219885350.0, + "step": 8495 + }, + { + "epoch": 0.9330112014056666, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5654990673065186, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.7000881433486938, + "num_tokens": 219910183.0, + "step": 8496 + }, + { + "epoch": 0.9331210191082803, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6710357666015625, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7080682516098022, + "num_tokens": 219935143.0, + "step": 8497 + }, + { + "epoch": 0.9332308368108939, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.616194725036621, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6817485690116882, + "num_tokens": 219962271.0, + "step": 8498 + }, + { + "epoch": 0.9333406545135076, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.275181531906128, + "learning_rate": 1e-06, + "loss": 1.1182, + "mean_token_accuracy": 0.6805523037910461, + "num_tokens": 219996977.0, + "step": 8499 + }, + { + "epoch": 0.9334504722161212, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.819453001022339, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7046164274215698, + "num_tokens": 220019017.0, + "step": 8500 + }, + { + "epoch": 0.9335602899187349, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.426196336746216, + "learning_rate": 1e-06, + "loss": 0.9565, + "mean_token_accuracy": 0.7157206535339355, + "num_tokens": 220044241.0, + "step": 8501 + }, + { + "epoch": 0.9336701076213486, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7799692153930664, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6923063397407532, + "num_tokens": 220066675.0, + "step": 8502 + }, + { + "epoch": 0.9337799253239623, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.872419595718384, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7160636186599731, + "num_tokens": 220087579.0, + "step": 8503 + }, + { + "epoch": 0.9338897430265759, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.4878857135772705, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.706896185874939, + "num_tokens": 220105762.0, + "step": 8504 + }, + { + "epoch": 0.9339995607291895, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8485734462738037, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7055059671401978, + "num_tokens": 220127377.0, + "step": 8505 + }, + { + "epoch": 0.9341093784318032, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4290618896484375, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7141386866569519, + "num_tokens": 220156035.0, + "step": 8506 + }, + { + "epoch": 0.9342191961344168, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.401895046234131, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7059891819953918, + "num_tokens": 220186364.0, + "step": 8507 + }, + { + "epoch": 0.9343290138370305, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.46067476272583, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.6849519610404968, + "num_tokens": 220216923.0, + "step": 8508 + }, + { + "epoch": 0.9344388315396442, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.465510606765747, + "learning_rate": 1e-06, + "loss": 1.1035, + "mean_token_accuracy": 0.6800148487091064, + "num_tokens": 220243427.0, + "step": 8509 + }, + { + "epoch": 0.9345486492422579, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6378026008605957, + "learning_rate": 1e-06, + "loss": 1.1257, + "mean_token_accuracy": 0.6760111451148987, + "num_tokens": 220270601.0, + "step": 8510 + }, + { + "epoch": 0.9346584669448715, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.508063316345215, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6949849128723145, + "num_tokens": 220299537.0, + "step": 8511 + }, + { + "epoch": 0.9347682846474852, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.667257308959961, + "learning_rate": 1e-06, + "loss": 1.1324, + "mean_token_accuracy": 0.6725380420684814, + "num_tokens": 220325252.0, + "step": 8512 + }, + { + "epoch": 0.9348781023500988, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4965503215789795, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7061384916305542, + "num_tokens": 220351982.0, + "step": 8513 + }, + { + "epoch": 0.9349879200527125, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.375288486480713, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7014588117599487, + "num_tokens": 220381152.0, + "step": 8514 + }, + { + "epoch": 0.9350977377553261, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.88496470451355, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7249356508255005, + "num_tokens": 220399949.0, + "step": 8515 + }, + { + "epoch": 0.9352075554579398, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2419703006744385, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7060712575912476, + "num_tokens": 220432231.0, + "step": 8516 + }, + { + "epoch": 0.9353173731605535, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8641581535339355, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7124376893043518, + "num_tokens": 220455363.0, + "step": 8517 + }, + { + "epoch": 0.9354271908631672, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4811031818389893, + "learning_rate": 1e-06, + "loss": 1.1146, + "mean_token_accuracy": 0.6914874315261841, + "num_tokens": 220485388.0, + "step": 8518 + }, + { + "epoch": 0.9355370085657808, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.585550546646118, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.6941121816635132, + "num_tokens": 220510686.0, + "step": 8519 + }, + { + "epoch": 0.9356468262683945, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5191988945007324, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.6950356960296631, + "num_tokens": 220538470.0, + "step": 8520 + }, + { + "epoch": 0.9357566439710081, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.660022497177124, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7221746444702148, + "num_tokens": 220559701.0, + "step": 8521 + }, + { + "epoch": 0.9358664616736218, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4879074096679688, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7209868431091309, + "num_tokens": 220587287.0, + "step": 8522 + }, + { + "epoch": 0.9359762793762354, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5611798763275146, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.697934627532959, + "num_tokens": 220612885.0, + "step": 8523 + }, + { + "epoch": 0.9360860970788492, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4033474922180176, + "learning_rate": 1e-06, + "loss": 1.1071, + "mean_token_accuracy": 0.6812924146652222, + "num_tokens": 220642379.0, + "step": 8524 + }, + { + "epoch": 0.9361959147814628, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4151740074157715, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6935520172119141, + "num_tokens": 220670345.0, + "step": 8525 + }, + { + "epoch": 0.9363057324840764, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5836181640625, + "learning_rate": 1e-06, + "loss": 1.0841, + "mean_token_accuracy": 0.6812079548835754, + "num_tokens": 220695748.0, + "step": 8526 + }, + { + "epoch": 0.9364155501866901, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2876803874969482, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6928381323814392, + "num_tokens": 220725066.0, + "step": 8527 + }, + { + "epoch": 0.9365253678893037, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7232065200805664, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6873841285705566, + "num_tokens": 220750281.0, + "step": 8528 + }, + { + "epoch": 0.9366351855919174, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.493140697479248, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7029788494110107, + "num_tokens": 220778300.0, + "step": 8529 + }, + { + "epoch": 0.936745003294531, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6151576042175293, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.701063334941864, + "num_tokens": 220803943.0, + "step": 8530 + }, + { + "epoch": 0.9368548209971448, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5277304649353027, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7162556052207947, + "num_tokens": 220827576.0, + "step": 8531 + }, + { + "epoch": 0.9369646386997584, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8436331748962402, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6883974075317383, + "num_tokens": 220849305.0, + "step": 8532 + }, + { + "epoch": 0.9370744564023721, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.546381950378418, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6862431764602661, + "num_tokens": 220875563.0, + "step": 8533 + }, + { + "epoch": 0.9371842741049857, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6138265132904053, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7030495405197144, + "num_tokens": 220898919.0, + "step": 8534 + }, + { + "epoch": 0.9372940918075994, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4879705905914307, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.6894878149032593, + "num_tokens": 220925388.0, + "step": 8535 + }, + { + "epoch": 0.937403909510213, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8351335525512695, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7006985545158386, + "num_tokens": 220946587.0, + "step": 8536 + }, + { + "epoch": 0.9375137272128267, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.673456907272339, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6940286755561829, + "num_tokens": 220970097.0, + "step": 8537 + }, + { + "epoch": 0.9376235449154404, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4650356769561768, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6890205144882202, + "num_tokens": 220999343.0, + "step": 8538 + }, + { + "epoch": 0.9377333626180541, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.568605661392212, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7006266117095947, + "num_tokens": 221027252.0, + "step": 8539 + }, + { + "epoch": 0.9378431803206677, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7811527252197266, + "learning_rate": 1e-06, + "loss": 1.1153, + "mean_token_accuracy": 0.6962467432022095, + "num_tokens": 221052625.0, + "step": 8540 + }, + { + "epoch": 0.9379529980232814, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.695356845855713, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7010100483894348, + "num_tokens": 221076412.0, + "step": 8541 + }, + { + "epoch": 0.938062815725895, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.508274555206299, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7167019844055176, + "num_tokens": 221104149.0, + "step": 8542 + }, + { + "epoch": 0.9381726334285087, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4051616191864014, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6958448886871338, + "num_tokens": 221132434.0, + "step": 8543 + }, + { + "epoch": 0.9382824511311223, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.484072208404541, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7009897828102112, + "num_tokens": 221157135.0, + "step": 8544 + }, + { + "epoch": 0.9383922688337359, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7283458709716797, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7166303396224976, + "num_tokens": 221177932.0, + "step": 8545 + }, + { + "epoch": 0.9385020865363497, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6929984092712402, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7046339511871338, + "num_tokens": 221200677.0, + "step": 8546 + }, + { + "epoch": 0.9386119042389633, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4552812576293945, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7075628042221069, + "num_tokens": 221226510.0, + "step": 8547 + }, + { + "epoch": 0.938721721941577, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.585012674331665, + "learning_rate": 1e-06, + "loss": 1.1061, + "mean_token_accuracy": 0.6783428192138672, + "num_tokens": 221254321.0, + "step": 8548 + }, + { + "epoch": 0.9388315396441906, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.628007173538208, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.6998674869537354, + "num_tokens": 221277301.0, + "step": 8549 + }, + { + "epoch": 0.9389413573468043, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.769784927368164, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7216232419013977, + "num_tokens": 221298457.0, + "step": 8550 + }, + { + "epoch": 0.9390511750494179, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6533801555633545, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7062337398529053, + "num_tokens": 221323999.0, + "step": 8551 + }, + { + "epoch": 0.9391609927520316, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.525080680847168, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7139028906822205, + "num_tokens": 221350861.0, + "step": 8552 + }, + { + "epoch": 0.9392708104546453, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.571040630340576, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6927294135093689, + "num_tokens": 221379066.0, + "step": 8553 + }, + { + "epoch": 0.939380628157259, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 7.91457462310791, + "learning_rate": 1e-06, + "loss": 1.1573, + "mean_token_accuracy": 0.6692905426025391, + "num_tokens": 221405645.0, + "step": 8554 + }, + { + "epoch": 0.9394904458598726, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.355180501937866, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7009621262550354, + "num_tokens": 221434456.0, + "step": 8555 + }, + { + "epoch": 0.9396002635624863, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.684934616088867, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7194797992706299, + "num_tokens": 221455517.0, + "step": 8556 + }, + { + "epoch": 0.9397100812650999, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.817884683609009, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7045588493347168, + "num_tokens": 221475373.0, + "step": 8557 + }, + { + "epoch": 0.9398198989677136, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.532754898071289, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.7050241231918335, + "num_tokens": 221503028.0, + "step": 8558 + }, + { + "epoch": 0.9399297166703272, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.578155755996704, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.709377646446228, + "num_tokens": 221526144.0, + "step": 8559 + }, + { + "epoch": 0.940039534372941, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.761476755142212, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6909946203231812, + "num_tokens": 221547183.0, + "step": 8560 + }, + { + "epoch": 0.9401493520755546, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7175841331481934, + "learning_rate": 1e-06, + "loss": 0.8887, + "mean_token_accuracy": 0.7431519031524658, + "num_tokens": 221569164.0, + "step": 8561 + }, + { + "epoch": 0.9402591697781683, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5247373580932617, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7006372213363647, + "num_tokens": 221597868.0, + "step": 8562 + }, + { + "epoch": 0.9403689874807819, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5491018295288086, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6976353526115417, + "num_tokens": 221624517.0, + "step": 8563 + }, + { + "epoch": 0.9404788051833955, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4921326637268066, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6963664293289185, + "num_tokens": 221654198.0, + "step": 8564 + }, + { + "epoch": 0.9405886228860092, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7976319789886475, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7080259323120117, + "num_tokens": 221676354.0, + "step": 8565 + }, + { + "epoch": 0.9406984405886228, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2352442741394043, + "learning_rate": 1e-06, + "loss": 1.1216, + "mean_token_accuracy": 0.6751723289489746, + "num_tokens": 221711475.0, + "step": 8566 + }, + { + "epoch": 0.9408082582912366, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.505302906036377, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7141306400299072, + "num_tokens": 221738072.0, + "step": 8567 + }, + { + "epoch": 0.9409180759938502, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.435865640640259, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6875312328338623, + "num_tokens": 221765201.0, + "step": 8568 + }, + { + "epoch": 0.9410278936964639, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6274940967559814, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.700392484664917, + "num_tokens": 221789983.0, + "step": 8569 + }, + { + "epoch": 0.9411377113990775, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.3983325958251953, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7060083150863647, + "num_tokens": 221818675.0, + "step": 8570 + }, + { + "epoch": 0.9412475291016912, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4054136276245117, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.690497636795044, + "num_tokens": 221849966.0, + "step": 8571 + }, + { + "epoch": 0.9413573468043048, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.959815740585327, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7009116411209106, + "num_tokens": 221877967.0, + "step": 8572 + }, + { + "epoch": 0.9414671645069185, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.7772269248962402, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7253816723823547, + "num_tokens": 221901832.0, + "step": 8573 + }, + { + "epoch": 0.9415769822095322, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6699111461639404, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7000971436500549, + "num_tokens": 221926364.0, + "step": 8574 + }, + { + "epoch": 0.9416867999121459, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.704787492752075, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7168675065040588, + "num_tokens": 221949997.0, + "step": 8575 + }, + { + "epoch": 0.9417966176147595, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9926040172576904, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7106443047523499, + "num_tokens": 221972166.0, + "step": 8576 + }, + { + "epoch": 0.9419064353173732, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4498376846313477, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6976726055145264, + "num_tokens": 222000577.0, + "step": 8577 + }, + { + "epoch": 0.9420162530199868, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.361283779144287, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7098358273506165, + "num_tokens": 222028566.0, + "step": 8578 + }, + { + "epoch": 0.9421260707226005, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.558593273162842, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7214313745498657, + "num_tokens": 222052475.0, + "step": 8579 + }, + { + "epoch": 0.9422358884252141, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.2985892295837402, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6925725936889648, + "num_tokens": 222080988.0, + "step": 8580 + }, + { + "epoch": 0.9423457061278278, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.4923884868621826, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6988062858581543, + "num_tokens": 222108697.0, + "step": 8581 + }, + { + "epoch": 0.9424555238304415, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9522275924682617, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7165114283561707, + "num_tokens": 222126981.0, + "step": 8582 + }, + { + "epoch": 0.9425653415330552, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.412604331970215, + "learning_rate": 1e-06, + "loss": 1.1797, + "mean_token_accuracy": 0.6657226085662842, + "num_tokens": 222155767.0, + "step": 8583 + }, + { + "epoch": 0.9426751592356688, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.0291478633880615, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6941584944725037, + "num_tokens": 222177533.0, + "step": 8584 + }, + { + "epoch": 0.9427849769382824, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.5036935806274414, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7101442217826843, + "num_tokens": 222201844.0, + "step": 8585 + }, + { + "epoch": 0.9428947946408961, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.6992027759552, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7095096707344055, + "num_tokens": 222224495.0, + "step": 8586 + }, + { + "epoch": 0.9430046123435097, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 3.4608423709869385, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.6967509984970093, + "num_tokens": 222243560.0, + "step": 8587 + }, + { + "epoch": 0.9431144300461234, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.262453079223633, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7158569097518921, + "num_tokens": 222274129.0, + "step": 8588 + }, + { + "epoch": 0.9432242477487371, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.335665464401245, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7165317535400391, + "num_tokens": 222305237.0, + "step": 8589 + }, + { + "epoch": 0.9433340654513508, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.0287020206451416, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.6968749761581421, + "num_tokens": 222328774.0, + "step": 8590 + }, + { + "epoch": 0.9434438831539644, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.468893527984619, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7192296981811523, + "num_tokens": 222352974.0, + "step": 8591 + }, + { + "epoch": 0.9435537008565781, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5774335861206055, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6982649564743042, + "num_tokens": 222375481.0, + "step": 8592 + }, + { + "epoch": 0.9436635185591917, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6327621936798096, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6993627548217773, + "num_tokens": 222399146.0, + "step": 8593 + }, + { + "epoch": 0.9437733362618054, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.220104455947876, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6881221532821655, + "num_tokens": 222432584.0, + "step": 8594 + }, + { + "epoch": 0.943883153964419, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.8433899879455566, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6984133124351501, + "num_tokens": 222454366.0, + "step": 8595 + }, + { + "epoch": 0.9439929716670328, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.620640754699707, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7020420432090759, + "num_tokens": 222479174.0, + "step": 8596 + }, + { + "epoch": 0.9441027893696464, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.603825330734253, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7185218334197998, + "num_tokens": 222503908.0, + "step": 8597 + }, + { + "epoch": 0.9442126070722601, + "ewc_loss": 8.225440979003906e-06, + "grad_norm": 2.9253885746002197, + "learning_rate": 1e-06, + "loss": 0.887, + "mean_token_accuracy": 0.7364517450332642, + "num_tokens": 222520848.0, + "step": 8598 + }, + { + "epoch": 0.9443224247748737, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.848268985748291, + "learning_rate": 1e-06, + "loss": 0.9303, + "mean_token_accuracy": 0.7235378623008728, + "num_tokens": 222538364.0, + "step": 8599 + }, + { + "epoch": 0.9444322424774874, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.836885690689087, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6890139579772949, + "num_tokens": 222565875.0, + "step": 8600 + }, + { + "epoch": 0.944542060180101, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.8553359508514404, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6763778328895569, + "num_tokens": 222591050.0, + "step": 8601 + }, + { + "epoch": 0.9446518778827147, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.527407169342041, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7104004621505737, + "num_tokens": 222616289.0, + "step": 8602 + }, + { + "epoch": 0.9447616955853284, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3705217838287354, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6994603872299194, + "num_tokens": 222642934.0, + "step": 8603 + }, + { + "epoch": 0.944871513287942, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4905660152435303, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7269488573074341, + "num_tokens": 222669971.0, + "step": 8604 + }, + { + "epoch": 0.9449813309905557, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.744128465652466, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7106310129165649, + "num_tokens": 222693886.0, + "step": 8605 + }, + { + "epoch": 0.9450911486931693, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5634896755218506, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6968536376953125, + "num_tokens": 222718751.0, + "step": 8606 + }, + { + "epoch": 0.945200966395783, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.82950758934021, + "learning_rate": 1e-06, + "loss": 0.9356, + "mean_token_accuracy": 0.7214202284812927, + "num_tokens": 222739723.0, + "step": 8607 + }, + { + "epoch": 0.9453107840983966, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6766164302825928, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7097057104110718, + "num_tokens": 222764431.0, + "step": 8608 + }, + { + "epoch": 0.9454206018010103, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.77689790725708, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.713904857635498, + "num_tokens": 222786067.0, + "step": 8609 + }, + { + "epoch": 0.9455304195036239, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5470175743103027, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6828408241271973, + "num_tokens": 222811849.0, + "step": 8610 + }, + { + "epoch": 0.9456402372062377, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.641897201538086, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.684073805809021, + "num_tokens": 222837098.0, + "step": 8611 + }, + { + "epoch": 0.9457500549088513, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4998161792755127, + "learning_rate": 1e-06, + "loss": 1.1299, + "mean_token_accuracy": 0.6855617165565491, + "num_tokens": 222867209.0, + "step": 8612 + }, + { + "epoch": 0.945859872611465, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.587398052215576, + "learning_rate": 1e-06, + "loss": 1.0876, + "mean_token_accuracy": 0.6773913502693176, + "num_tokens": 222891024.0, + "step": 8613 + }, + { + "epoch": 0.9459696903140786, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.601421594619751, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7109078764915466, + "num_tokens": 222914495.0, + "step": 8614 + }, + { + "epoch": 0.9460795080166923, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.860955238342285, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7156057357788086, + "num_tokens": 222933415.0, + "step": 8615 + }, + { + "epoch": 0.9461893257193059, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6278045177459717, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.7022711038589478, + "num_tokens": 222956258.0, + "step": 8616 + }, + { + "epoch": 0.9462991434219196, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.781348466873169, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.713579535484314, + "num_tokens": 222976109.0, + "step": 8617 + }, + { + "epoch": 0.9464089611245333, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5989811420440674, + "learning_rate": 1e-06, + "loss": 1.0679, + "mean_token_accuracy": 0.6949638724327087, + "num_tokens": 223002419.0, + "step": 8618 + }, + { + "epoch": 0.946518778827147, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7959439754486084, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6969509124755859, + "num_tokens": 223029426.0, + "step": 8619 + }, + { + "epoch": 0.9466285965297606, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.476365327835083, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7195806503295898, + "num_tokens": 223055601.0, + "step": 8620 + }, + { + "epoch": 0.9467384142323743, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4557912349700928, + "learning_rate": 1e-06, + "loss": 1.108, + "mean_token_accuracy": 0.6829757690429688, + "num_tokens": 223086401.0, + "step": 8621 + }, + { + "epoch": 0.9468482319349879, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4767677783966064, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.6995205879211426, + "num_tokens": 223112709.0, + "step": 8622 + }, + { + "epoch": 0.9469580496376016, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7849762439727783, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7083526253700256, + "num_tokens": 223132726.0, + "step": 8623 + }, + { + "epoch": 0.9470678673402152, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.005666494369507, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7194019556045532, + "num_tokens": 223151442.0, + "step": 8624 + }, + { + "epoch": 0.947177685042829, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6556966304779053, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7015752792358398, + "num_tokens": 223174241.0, + "step": 8625 + }, + { + "epoch": 0.9472875027454426, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6283891201019287, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7159057855606079, + "num_tokens": 223196993.0, + "step": 8626 + }, + { + "epoch": 0.9473973204480562, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3287720680236816, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6955957412719727, + "num_tokens": 223226771.0, + "step": 8627 + }, + { + "epoch": 0.9475071381506699, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6836116313934326, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7082940936088562, + "num_tokens": 223249487.0, + "step": 8628 + }, + { + "epoch": 0.9476169558532835, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.523176431655884, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6921851634979248, + "num_tokens": 223275629.0, + "step": 8629 + }, + { + "epoch": 0.9477267735558972, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.524557590484619, + "learning_rate": 1e-06, + "loss": 1.1104, + "mean_token_accuracy": 0.6849324703216553, + "num_tokens": 223304025.0, + "step": 8630 + }, + { + "epoch": 0.9478365912585108, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.585080862045288, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.6992274522781372, + "num_tokens": 223333077.0, + "step": 8631 + }, + { + "epoch": 0.9479464089611246, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6439175605773926, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.694638729095459, + "num_tokens": 223356647.0, + "step": 8632 + }, + { + "epoch": 0.9480562266637382, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.0372209548950195, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.7217180132865906, + "num_tokens": 223374548.0, + "step": 8633 + }, + { + "epoch": 0.9481660443663519, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.900230884552002, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7143322229385376, + "num_tokens": 223400800.0, + "step": 8634 + }, + { + "epoch": 0.9482758620689655, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.998628854751587, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7216147780418396, + "num_tokens": 223419768.0, + "step": 8635 + }, + { + "epoch": 0.9483856797715792, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.543840169906616, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7118396759033203, + "num_tokens": 223446443.0, + "step": 8636 + }, + { + "epoch": 0.9484954974741928, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4606430530548096, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7118133902549744, + "num_tokens": 223473510.0, + "step": 8637 + }, + { + "epoch": 0.9486053151768065, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.513230562210083, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7311810255050659, + "num_tokens": 223503095.0, + "step": 8638 + }, + { + "epoch": 0.9487151328794201, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6604995727539062, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.691039502620697, + "num_tokens": 223532652.0, + "step": 8639 + }, + { + "epoch": 0.9488249505820339, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5645406246185303, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7102969288825989, + "num_tokens": 223560980.0, + "step": 8640 + }, + { + "epoch": 0.9489347682846475, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.3093013763427734, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7215380072593689, + "num_tokens": 223586348.0, + "step": 8641 + }, + { + "epoch": 0.9490445859872612, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.897291660308838, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7046570181846619, + "num_tokens": 223607708.0, + "step": 8642 + }, + { + "epoch": 0.9491544036898748, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8162782192230225, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7249400019645691, + "num_tokens": 223628312.0, + "step": 8643 + }, + { + "epoch": 0.9492642213924884, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.469996213912964, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7026103734970093, + "num_tokens": 223656073.0, + "step": 8644 + }, + { + "epoch": 0.9493740390951021, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.777247190475464, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7051098942756653, + "num_tokens": 223681909.0, + "step": 8645 + }, + { + "epoch": 0.9494838567977157, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.626021146774292, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6880723834037781, + "num_tokens": 223706150.0, + "step": 8646 + }, + { + "epoch": 0.9495936745003295, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.589556932449341, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7001066207885742, + "num_tokens": 223730524.0, + "step": 8647 + }, + { + "epoch": 0.9497034922029431, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.085463285446167, + "learning_rate": 1e-06, + "loss": 0.8894, + "mean_token_accuracy": 0.7304571866989136, + "num_tokens": 223748705.0, + "step": 8648 + }, + { + "epoch": 0.9498133099055568, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5219788551330566, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.6948104500770569, + "num_tokens": 223784072.0, + "step": 8649 + }, + { + "epoch": 0.9499231276081704, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.562948703765869, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7322891354560852, + "num_tokens": 223811086.0, + "step": 8650 + }, + { + "epoch": 0.9500329453107841, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7078592777252197, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.7098860740661621, + "num_tokens": 223835135.0, + "step": 8651 + }, + { + "epoch": 0.9501427630133977, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5286707878112793, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6847962737083435, + "num_tokens": 223861787.0, + "step": 8652 + }, + { + "epoch": 0.9502525807160114, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.375255823135376, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.6815968751907349, + "num_tokens": 223891162.0, + "step": 8653 + }, + { + "epoch": 0.9503623984186251, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6221797466278076, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.713138222694397, + "num_tokens": 223912433.0, + "step": 8654 + }, + { + "epoch": 0.9504722161212388, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.0883944034576416, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.681236743927002, + "num_tokens": 223938108.0, + "step": 8655 + }, + { + "epoch": 0.9505820338238524, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.549137830734253, + "learning_rate": 1e-06, + "loss": 1.1178, + "mean_token_accuracy": 0.6812366247177124, + "num_tokens": 223965637.0, + "step": 8656 + }, + { + "epoch": 0.9506918515264661, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.9955620765686035, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.707616925239563, + "num_tokens": 223985296.0, + "step": 8657 + }, + { + "epoch": 0.9508016692290797, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4477500915527344, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6930060386657715, + "num_tokens": 224013684.0, + "step": 8658 + }, + { + "epoch": 0.9509114869316934, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 7.869555473327637, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.6980381608009338, + "num_tokens": 224041925.0, + "step": 8659 + }, + { + "epoch": 0.951021304634307, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.926959276199341, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.6979184150695801, + "num_tokens": 224062807.0, + "step": 8660 + }, + { + "epoch": 0.9511311223369208, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.076063394546509, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6984933614730835, + "num_tokens": 224090444.0, + "step": 8661 + }, + { + "epoch": 0.9512409400395344, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.9657232761383057, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6890063285827637, + "num_tokens": 224115116.0, + "step": 8662 + }, + { + "epoch": 0.9513507577421481, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.448458671569824, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6852498650550842, + "num_tokens": 224144631.0, + "step": 8663 + }, + { + "epoch": 0.9514605754447617, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.463940143585205, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.7023431658744812, + "num_tokens": 224173130.0, + "step": 8664 + }, + { + "epoch": 0.9515703931473753, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7144265174865723, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7238479852676392, + "num_tokens": 224196040.0, + "step": 8665 + }, + { + "epoch": 0.951680210849989, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.484811305999756, + "learning_rate": 1e-06, + "loss": 0.9026, + "mean_token_accuracy": 0.7313041687011719, + "num_tokens": 224221596.0, + "step": 8666 + }, + { + "epoch": 0.9517900285526026, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.49444317817688, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7094485759735107, + "num_tokens": 224247333.0, + "step": 8667 + }, + { + "epoch": 0.9518998462552163, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5097997188568115, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7115166187286377, + "num_tokens": 224274178.0, + "step": 8668 + }, + { + "epoch": 0.95200966395783, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5578081607818604, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6923520565032959, + "num_tokens": 224304474.0, + "step": 8669 + }, + { + "epoch": 0.9521194816604437, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 4.6673583984375, + "learning_rate": 1e-06, + "loss": 1.1197, + "mean_token_accuracy": 0.6810792088508606, + "num_tokens": 224329214.0, + "step": 8670 + }, + { + "epoch": 0.9522292993630573, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6516849994659424, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.7010548710823059, + "num_tokens": 224353752.0, + "step": 8671 + }, + { + "epoch": 0.952339117065671, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.0340311527252197, + "learning_rate": 1e-06, + "loss": 0.9234, + "mean_token_accuracy": 0.7294212579727173, + "num_tokens": 224369261.0, + "step": 8672 + }, + { + "epoch": 0.9524489347682846, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5048413276672363, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7009958028793335, + "num_tokens": 224397273.0, + "step": 8673 + }, + { + "epoch": 0.9525587524708983, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.9257960319519043, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7051483392715454, + "num_tokens": 224419111.0, + "step": 8674 + }, + { + "epoch": 0.9526685701735119, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.506220817565918, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7079797983169556, + "num_tokens": 224445100.0, + "step": 8675 + }, + { + "epoch": 0.9527783878761257, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5518441200256348, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.7210856676101685, + "num_tokens": 224469047.0, + "step": 8676 + }, + { + "epoch": 0.9528882055787393, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7343554496765137, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7124914526939392, + "num_tokens": 224491089.0, + "step": 8677 + }, + { + "epoch": 0.952998023281353, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.519080877304077, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.6972004175186157, + "num_tokens": 224515134.0, + "step": 8678 + }, + { + "epoch": 0.9531078409839666, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3367624282836914, + "learning_rate": 1e-06, + "loss": 1.1381, + "mean_token_accuracy": 0.6720729470252991, + "num_tokens": 224548089.0, + "step": 8679 + }, + { + "epoch": 0.9532176586865803, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.504779577255249, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7040026187896729, + "num_tokens": 224571705.0, + "step": 8680 + }, + { + "epoch": 0.9533274763891939, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6401658058166504, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6836968660354614, + "num_tokens": 224596889.0, + "step": 8681 + }, + { + "epoch": 0.9534372940918076, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5823581218719482, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7097961902618408, + "num_tokens": 224632269.0, + "step": 8682 + }, + { + "epoch": 0.9535471117944213, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4232804775238037, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6940053105354309, + "num_tokens": 224659968.0, + "step": 8683 + }, + { + "epoch": 0.953656929497035, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.291726589202881, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.709818959236145, + "num_tokens": 224689653.0, + "step": 8684 + }, + { + "epoch": 0.9537667471996486, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.398808002471924, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6886624097824097, + "num_tokens": 224717652.0, + "step": 8685 + }, + { + "epoch": 0.9538765649022622, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.2298154830932617, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6862919330596924, + "num_tokens": 224751602.0, + "step": 8686 + }, + { + "epoch": 0.9539863826048759, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6221892833709717, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6906239986419678, + "num_tokens": 224775812.0, + "step": 8687 + }, + { + "epoch": 0.9540962003074895, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.570946455001831, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7178903222084045, + "num_tokens": 224798433.0, + "step": 8688 + }, + { + "epoch": 0.9542060180101032, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.526275157928467, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6908937096595764, + "num_tokens": 224826605.0, + "step": 8689 + }, + { + "epoch": 0.9543158357127169, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.545469045639038, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.719433605670929, + "num_tokens": 224850608.0, + "step": 8690 + }, + { + "epoch": 0.9544256534153306, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7589848041534424, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7019228935241699, + "num_tokens": 224872229.0, + "step": 8691 + }, + { + "epoch": 0.9545354711179442, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3767406940460205, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7166472673416138, + "num_tokens": 224899062.0, + "step": 8692 + }, + { + "epoch": 0.9546452888205579, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.8962271213531494, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7079847455024719, + "num_tokens": 224920953.0, + "step": 8693 + }, + { + "epoch": 0.9547551065231715, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.750413656234741, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6920148134231567, + "num_tokens": 224945827.0, + "step": 8694 + }, + { + "epoch": 0.9548649242257852, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.8057336807250977, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6977283358573914, + "num_tokens": 224975422.0, + "step": 8695 + }, + { + "epoch": 0.9549747419283988, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6370201110839844, + "learning_rate": 1e-06, + "loss": 1.0623, + "mean_token_accuracy": 0.6921578645706177, + "num_tokens": 225001592.0, + "step": 8696 + }, + { + "epoch": 0.9550845596310125, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.9519598484039307, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7154641151428223, + "num_tokens": 225022802.0, + "step": 8697 + }, + { + "epoch": 0.9551943773336262, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7353148460388184, + "learning_rate": 1e-06, + "loss": 1.1105, + "mean_token_accuracy": 0.6819480657577515, + "num_tokens": 225048842.0, + "step": 8698 + }, + { + "epoch": 0.9553041950362399, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3658695220947266, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.7019551992416382, + "num_tokens": 225080619.0, + "step": 8699 + }, + { + "epoch": 0.9554140127388535, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3628287315368652, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7055262327194214, + "num_tokens": 225110529.0, + "step": 8700 + }, + { + "epoch": 0.9555238304414672, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.45601224899292, + "learning_rate": 1e-06, + "loss": 0.899, + "mean_token_accuracy": 0.7316268086433411, + "num_tokens": 225136463.0, + "step": 8701 + }, + { + "epoch": 0.9556336481440808, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5824739933013916, + "learning_rate": 1e-06, + "loss": 1.0932, + "mean_token_accuracy": 0.6804680228233337, + "num_tokens": 225161023.0, + "step": 8702 + }, + { + "epoch": 0.9557434658466945, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.51599383354187, + "learning_rate": 1e-06, + "loss": 1.1711, + "mean_token_accuracy": 0.6661312580108643, + "num_tokens": 225188822.0, + "step": 8703 + }, + { + "epoch": 0.9558532835493081, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.405536651611328, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7002238035202026, + "num_tokens": 225219426.0, + "step": 8704 + }, + { + "epoch": 0.9559631012519219, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.835503101348877, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.704683244228363, + "num_tokens": 225241096.0, + "step": 8705 + }, + { + "epoch": 0.9560729189545355, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.634228229522705, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.6991748809814453, + "num_tokens": 225264747.0, + "step": 8706 + }, + { + "epoch": 0.9561827366571491, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.50221586227417, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7021227478981018, + "num_tokens": 225287683.0, + "step": 8707 + }, + { + "epoch": 0.9562925543597628, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.480727195739746, + "learning_rate": 1e-06, + "loss": 1.0953, + "mean_token_accuracy": 0.6852555871009827, + "num_tokens": 225316574.0, + "step": 8708 + }, + { + "epoch": 0.9564023720623764, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6596930027008057, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7042934894561768, + "num_tokens": 225343257.0, + "step": 8709 + }, + { + "epoch": 0.9565121897649901, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7410175800323486, + "learning_rate": 1e-06, + "loss": 0.8945, + "mean_token_accuracy": 0.7335928678512573, + "num_tokens": 225366372.0, + "step": 8710 + }, + { + "epoch": 0.9566220074676037, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3820583820343018, + "learning_rate": 1e-06, + "loss": 1.1268, + "mean_token_accuracy": 0.6843320727348328, + "num_tokens": 225397796.0, + "step": 8711 + }, + { + "epoch": 0.9567318251702175, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.348322629928589, + "learning_rate": 1e-06, + "loss": 1.1841, + "mean_token_accuracy": 0.6718393564224243, + "num_tokens": 225431624.0, + "step": 8712 + }, + { + "epoch": 0.9568416428728311, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.482670307159424, + "learning_rate": 1e-06, + "loss": 1.1481, + "mean_token_accuracy": 0.6744829416275024, + "num_tokens": 225459819.0, + "step": 8713 + }, + { + "epoch": 0.9569514605754448, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7904021739959717, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6997129917144775, + "num_tokens": 225487993.0, + "step": 8714 + }, + { + "epoch": 0.9570612782780584, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.541010856628418, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7119004130363464, + "num_tokens": 225513334.0, + "step": 8715 + }, + { + "epoch": 0.9571710959806721, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.695277690887451, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7057390213012695, + "num_tokens": 225537623.0, + "step": 8716 + }, + { + "epoch": 0.9572809136832857, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.814427375793457, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7020100355148315, + "num_tokens": 225556885.0, + "step": 8717 + }, + { + "epoch": 0.9573907313858994, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6192312240600586, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6966180801391602, + "num_tokens": 225582847.0, + "step": 8718 + }, + { + "epoch": 0.9575005490885131, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.2281036376953125, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7288236618041992, + "num_tokens": 225598872.0, + "step": 8719 + }, + { + "epoch": 0.9576103667911268, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.696237802505493, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6917893290519714, + "num_tokens": 225624823.0, + "step": 8720 + }, + { + "epoch": 0.9577201844937404, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6420907974243164, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6938742995262146, + "num_tokens": 225649755.0, + "step": 8721 + }, + { + "epoch": 0.9578300021963541, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.229787826538086, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6833682060241699, + "num_tokens": 225683143.0, + "step": 8722 + }, + { + "epoch": 0.9579398198989677, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3021817207336426, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7169503569602966, + "num_tokens": 225714652.0, + "step": 8723 + }, + { + "epoch": 0.9580496376015813, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6461682319641113, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7143757343292236, + "num_tokens": 225737649.0, + "step": 8724 + }, + { + "epoch": 0.958159455304195, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.506101608276367, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7117405533790588, + "num_tokens": 225762807.0, + "step": 8725 + }, + { + "epoch": 0.9582692730068088, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5112061500549316, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6797070503234863, + "num_tokens": 225790806.0, + "step": 8726 + }, + { + "epoch": 0.9583790907094224, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4011521339416504, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7062191963195801, + "num_tokens": 225818475.0, + "step": 8727 + }, + { + "epoch": 0.958488908412036, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 8.120174407958984, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7053879499435425, + "num_tokens": 225838758.0, + "step": 8728 + }, + { + "epoch": 0.9585987261146497, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.644796371459961, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7025044560432434, + "num_tokens": 225869813.0, + "step": 8729 + }, + { + "epoch": 0.9587085438172633, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6036441326141357, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7122675776481628, + "num_tokens": 225893794.0, + "step": 8730 + }, + { + "epoch": 0.958818361519877, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5025148391723633, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6916093230247498, + "num_tokens": 225920531.0, + "step": 8731 + }, + { + "epoch": 0.9589281792224906, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.318425178527832, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.70537930727005, + "num_tokens": 225951936.0, + "step": 8732 + }, + { + "epoch": 0.9590379969251043, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6653518676757812, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.709946870803833, + "num_tokens": 225974222.0, + "step": 8733 + }, + { + "epoch": 0.959147814627718, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6031477451324463, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7191587686538696, + "num_tokens": 225997213.0, + "step": 8734 + }, + { + "epoch": 0.9592576323303317, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.765772581100464, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.6972691416740417, + "num_tokens": 226017402.0, + "step": 8735 + }, + { + "epoch": 0.9593674500329453, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.540571689605713, + "learning_rate": 1e-06, + "loss": 1.1091, + "mean_token_accuracy": 0.6802592873573303, + "num_tokens": 226046124.0, + "step": 8736 + }, + { + "epoch": 0.959477267735559, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.458735466003418, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6954430937767029, + "num_tokens": 226072832.0, + "step": 8737 + }, + { + "epoch": 0.9595870854381726, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3238604068756104, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7072537541389465, + "num_tokens": 226099350.0, + "step": 8738 + }, + { + "epoch": 0.9596969031407863, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7435858249664307, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7044522166252136, + "num_tokens": 226123899.0, + "step": 8739 + }, + { + "epoch": 0.9598067208433999, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.411480665206909, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7061426639556885, + "num_tokens": 226151450.0, + "step": 8740 + }, + { + "epoch": 0.9599165385460137, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.38684344291687, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7125817537307739, + "num_tokens": 226180002.0, + "step": 8741 + }, + { + "epoch": 0.9600263562486273, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3162455558776855, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6918573379516602, + "num_tokens": 226207642.0, + "step": 8742 + }, + { + "epoch": 0.960136173951241, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7149195671081543, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.6995638608932495, + "num_tokens": 226233048.0, + "step": 8743 + }, + { + "epoch": 0.9602459916538546, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7261760234832764, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7202662825584412, + "num_tokens": 226254190.0, + "step": 8744 + }, + { + "epoch": 0.9603558093564682, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5355775356292725, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.6983537673950195, + "num_tokens": 226280390.0, + "step": 8745 + }, + { + "epoch": 0.9604656270590819, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3996052742004395, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6989673376083374, + "num_tokens": 226308489.0, + "step": 8746 + }, + { + "epoch": 0.9605754447616955, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4613895416259766, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7102797031402588, + "num_tokens": 226334100.0, + "step": 8747 + }, + { + "epoch": 0.9606852624643093, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6695055961608887, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7151148319244385, + "num_tokens": 226355150.0, + "step": 8748 + }, + { + "epoch": 0.9607950801669229, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.458331346511841, + "learning_rate": 1e-06, + "loss": 1.1293, + "mean_token_accuracy": 0.6779841184616089, + "num_tokens": 226385231.0, + "step": 8749 + }, + { + "epoch": 0.9609048978695366, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.2900230884552, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.706209659576416, + "num_tokens": 226416209.0, + "step": 8750 + }, + { + "epoch": 0.9610147155721502, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.845848798751831, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7072265148162842, + "num_tokens": 226435635.0, + "step": 8751 + }, + { + "epoch": 0.9611245332747639, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.2953426837921143, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6965044140815735, + "num_tokens": 226468132.0, + "step": 8752 + }, + { + "epoch": 0.9612343509773775, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5834968090057373, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7187145948410034, + "num_tokens": 226492763.0, + "step": 8753 + }, + { + "epoch": 0.9613441686799912, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5780375003814697, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7032531499862671, + "num_tokens": 226518240.0, + "step": 8754 + }, + { + "epoch": 0.9614539863826049, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.893629312515259, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7075881361961365, + "num_tokens": 226538071.0, + "step": 8755 + }, + { + "epoch": 0.9615638040852186, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.208726406097412, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7006076574325562, + "num_tokens": 226567951.0, + "step": 8756 + }, + { + "epoch": 0.9616736217878322, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.008945941925049, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7040683627128601, + "num_tokens": 226585283.0, + "step": 8757 + }, + { + "epoch": 0.9617834394904459, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.439549446105957, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.698688268661499, + "num_tokens": 226614720.0, + "step": 8758 + }, + { + "epoch": 0.9618932571930595, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.8308322429656982, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.7021824717521667, + "num_tokens": 226636638.0, + "step": 8759 + }, + { + "epoch": 0.9620030748956732, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.436084270477295, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6898807287216187, + "num_tokens": 226664693.0, + "step": 8760 + }, + { + "epoch": 0.9621128925982868, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5285680294036865, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.6718817353248596, + "num_tokens": 226691569.0, + "step": 8761 + }, + { + "epoch": 0.9622227103009005, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.143609046936035, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7202451825141907, + "num_tokens": 226708968.0, + "step": 8762 + }, + { + "epoch": 0.9623325280035142, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3545033931732178, + "learning_rate": 1e-06, + "loss": 1.0954, + "mean_token_accuracy": 0.6868436336517334, + "num_tokens": 226739371.0, + "step": 8763 + }, + { + "epoch": 0.9624423457061279, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6566202640533447, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6910269260406494, + "num_tokens": 226765112.0, + "step": 8764 + }, + { + "epoch": 0.9625521634087415, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4763646125793457, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6996359825134277, + "num_tokens": 226790665.0, + "step": 8765 + }, + { + "epoch": 0.9626619811113551, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3148274421691895, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.701961874961853, + "num_tokens": 226820430.0, + "step": 8766 + }, + { + "epoch": 0.9627717988139688, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.365849494934082, + "learning_rate": 1e-06, + "loss": 1.1163, + "mean_token_accuracy": 0.6877801418304443, + "num_tokens": 226852716.0, + "step": 8767 + }, + { + "epoch": 0.9628816165165824, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6371142864227295, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7092595100402832, + "num_tokens": 226880918.0, + "step": 8768 + }, + { + "epoch": 0.9629914342191961, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6233296394348145, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.6957032084465027, + "num_tokens": 226904758.0, + "step": 8769 + }, + { + "epoch": 0.9631012519218098, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4543261528015137, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6872096061706543, + "num_tokens": 226933464.0, + "step": 8770 + }, + { + "epoch": 0.9632110696244235, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.40079927444458, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.709869384765625, + "num_tokens": 226961442.0, + "step": 8771 + }, + { + "epoch": 0.9633208873270371, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6785078048706055, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6943870782852173, + "num_tokens": 226987718.0, + "step": 8772 + }, + { + "epoch": 0.9634307050296508, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.499321460723877, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6944442391395569, + "num_tokens": 227014707.0, + "step": 8773 + }, + { + "epoch": 0.9635405227322644, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.42642879486084, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.690449595451355, + "num_tokens": 227042658.0, + "step": 8774 + }, + { + "epoch": 0.9636503404348781, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.631911516189575, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7121114134788513, + "num_tokens": 227065980.0, + "step": 8775 + }, + { + "epoch": 0.9637601581374917, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4914743900299072, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.6812546253204346, + "num_tokens": 227092798.0, + "step": 8776 + }, + { + "epoch": 0.9638699758401055, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4618148803710938, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6929632425308228, + "num_tokens": 227119037.0, + "step": 8777 + }, + { + "epoch": 0.9639797935427191, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3628835678100586, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7046972513198853, + "num_tokens": 227147385.0, + "step": 8778 + }, + { + "epoch": 0.9640896112453328, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4768004417419434, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6947486400604248, + "num_tokens": 227175846.0, + "step": 8779 + }, + { + "epoch": 0.9641994289479464, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4163663387298584, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.688893735408783, + "num_tokens": 227203360.0, + "step": 8780 + }, + { + "epoch": 0.9643092466505601, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.557661771774292, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6953098177909851, + "num_tokens": 227228358.0, + "step": 8781 + }, + { + "epoch": 0.9644190643531737, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.426762580871582, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7190738320350647, + "num_tokens": 227255492.0, + "step": 8782 + }, + { + "epoch": 0.9645288820557874, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.729076623916626, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7082446217536926, + "num_tokens": 227276939.0, + "step": 8783 + }, + { + "epoch": 0.9646386997584011, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3656582832336426, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6925393342971802, + "num_tokens": 227305598.0, + "step": 8784 + }, + { + "epoch": 0.9647485174610148, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6518027782440186, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6967241764068604, + "num_tokens": 227329581.0, + "step": 8785 + }, + { + "epoch": 0.9648583351636284, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4690141677856445, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7033169269561768, + "num_tokens": 227354844.0, + "step": 8786 + }, + { + "epoch": 0.964968152866242, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.37679123878479, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.6961278319358826, + "num_tokens": 227383559.0, + "step": 8787 + }, + { + "epoch": 0.9650779705688557, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6423704624176025, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7111377120018005, + "num_tokens": 227407240.0, + "step": 8788 + }, + { + "epoch": 0.9651877882714693, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.301814556121826, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.7015628218650818, + "num_tokens": 227438972.0, + "step": 8789 + }, + { + "epoch": 0.965297605974083, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.524482011795044, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6985365748405457, + "num_tokens": 227466755.0, + "step": 8790 + }, + { + "epoch": 0.9654074236766966, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7808456420898438, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7288968563079834, + "num_tokens": 227486042.0, + "step": 8791 + }, + { + "epoch": 0.9655172413793104, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3971288204193115, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7257046103477478, + "num_tokens": 227513569.0, + "step": 8792 + }, + { + "epoch": 0.965627059081924, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.769181251525879, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6957038640975952, + "num_tokens": 227536632.0, + "step": 8793 + }, + { + "epoch": 0.9657368767845377, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5255305767059326, + "learning_rate": 1e-06, + "loss": 1.1032, + "mean_token_accuracy": 0.684505820274353, + "num_tokens": 227565842.0, + "step": 8794 + }, + { + "epoch": 0.9658466944871513, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3287301063537598, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6980822086334229, + "num_tokens": 227595085.0, + "step": 8795 + }, + { + "epoch": 0.965956512189765, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.599576711654663, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7072997093200684, + "num_tokens": 227619531.0, + "step": 8796 + }, + { + "epoch": 0.9660663298923786, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.766106128692627, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6902014017105103, + "num_tokens": 227643118.0, + "step": 8797 + }, + { + "epoch": 0.9661761475949923, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4715442657470703, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7081579566001892, + "num_tokens": 227668000.0, + "step": 8798 + }, + { + "epoch": 0.966285965297606, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5203633308410645, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7221738696098328, + "num_tokens": 227690524.0, + "step": 8799 + }, + { + "epoch": 0.9663957830002197, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5951991081237793, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7084947824478149, + "num_tokens": 227714620.0, + "step": 8800 + }, + { + "epoch": 0.9665056007028333, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8342721462249756, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7182711362838745, + "num_tokens": 227739464.0, + "step": 8801 + }, + { + "epoch": 0.966615418405447, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.552070140838623, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6940017342567444, + "num_tokens": 227764732.0, + "step": 8802 + }, + { + "epoch": 0.9667252361080606, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 5.300182342529297, + "learning_rate": 1e-06, + "loss": 1.0996, + "mean_token_accuracy": 0.677997350692749, + "num_tokens": 227795769.0, + "step": 8803 + }, + { + "epoch": 0.9668350538106742, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6372804641723633, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7094940543174744, + "num_tokens": 227820704.0, + "step": 8804 + }, + { + "epoch": 0.9669448715132879, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.281966209411621, + "learning_rate": 1e-06, + "loss": 1.1288, + "mean_token_accuracy": 0.6849744915962219, + "num_tokens": 227857518.0, + "step": 8805 + }, + { + "epoch": 0.9670546892159017, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.490079402923584, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.7145568132400513, + "num_tokens": 227881946.0, + "step": 8806 + }, + { + "epoch": 0.9671645069185153, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4178357124328613, + "learning_rate": 1e-06, + "loss": 1.11, + "mean_token_accuracy": 0.6825299263000488, + "num_tokens": 227912468.0, + "step": 8807 + }, + { + "epoch": 0.967274324621129, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5856881141662598, + "learning_rate": 1e-06, + "loss": 1.1328, + "mean_token_accuracy": 0.6744268536567688, + "num_tokens": 227938208.0, + "step": 8808 + }, + { + "epoch": 0.9673841423237426, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.540360927581787, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6884445548057556, + "num_tokens": 227964581.0, + "step": 8809 + }, + { + "epoch": 0.9674939600263562, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.426089286804199, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.7005220651626587, + "num_tokens": 227995051.0, + "step": 8810 + }, + { + "epoch": 0.9676037777289699, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.1921050548553467, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.6968210935592651, + "num_tokens": 228011822.0, + "step": 8811 + }, + { + "epoch": 0.9677135954315835, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.522660970687866, + "learning_rate": 1e-06, + "loss": 1.1459, + "mean_token_accuracy": 0.6655667424201965, + "num_tokens": 228041097.0, + "step": 8812 + }, + { + "epoch": 0.9678234131341973, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3270325660705566, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.6999762058258057, + "num_tokens": 228073745.0, + "step": 8813 + }, + { + "epoch": 0.9679332308368109, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.633211851119995, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.708055853843689, + "num_tokens": 228096488.0, + "step": 8814 + }, + { + "epoch": 0.9680430485394246, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.3514037132263184, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6892228126525879, + "num_tokens": 228126843.0, + "step": 8815 + }, + { + "epoch": 0.9681528662420382, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7394418716430664, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7208729982376099, + "num_tokens": 228148818.0, + "step": 8816 + }, + { + "epoch": 0.9682626839446519, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.685668468475342, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.7255151271820068, + "num_tokens": 228171522.0, + "step": 8817 + }, + { + "epoch": 0.9683725016472655, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.75545072555542, + "learning_rate": 1e-06, + "loss": 1.1042, + "mean_token_accuracy": 0.6840498447418213, + "num_tokens": 228197254.0, + "step": 8818 + }, + { + "epoch": 0.9684823193498792, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6515979766845703, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.6983136534690857, + "num_tokens": 228219818.0, + "step": 8819 + }, + { + "epoch": 0.9685921370524928, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.806302309036255, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7123472690582275, + "num_tokens": 228239719.0, + "step": 8820 + }, + { + "epoch": 0.9687019547551066, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.710476875305176, + "learning_rate": 1e-06, + "loss": 0.9478, + "mean_token_accuracy": 0.7177621126174927, + "num_tokens": 228262713.0, + "step": 8821 + }, + { + "epoch": 0.9688117724577202, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.423358201980591, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.7030892372131348, + "num_tokens": 228289415.0, + "step": 8822 + }, + { + "epoch": 0.9689215901603339, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4305002689361572, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7018699049949646, + "num_tokens": 228315429.0, + "step": 8823 + }, + { + "epoch": 0.9690314078629475, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.501267671585083, + "learning_rate": 1e-06, + "loss": 0.9095, + "mean_token_accuracy": 0.7345731258392334, + "num_tokens": 228339247.0, + "step": 8824 + }, + { + "epoch": 0.9691412255655611, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5960347652435303, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7178961634635925, + "num_tokens": 228364715.0, + "step": 8825 + }, + { + "epoch": 0.9692510432681748, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.8401424884796143, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.7207632064819336, + "num_tokens": 228386350.0, + "step": 8826 + }, + { + "epoch": 0.9693608609707884, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3291947841644287, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7028363943099976, + "num_tokens": 228418504.0, + "step": 8827 + }, + { + "epoch": 0.9694706786734022, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6361639499664307, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7108268737792969, + "num_tokens": 228443214.0, + "step": 8828 + }, + { + "epoch": 0.9695804963760158, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7813220024108887, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7100462317466736, + "num_tokens": 228465758.0, + "step": 8829 + }, + { + "epoch": 0.9696903140786295, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.093465566635132, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.709781289100647, + "num_tokens": 228492918.0, + "step": 8830 + }, + { + "epoch": 0.9698001317812431, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7673590183258057, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6872652173042297, + "num_tokens": 228517881.0, + "step": 8831 + }, + { + "epoch": 0.9699099494838568, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4666664600372314, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7078382968902588, + "num_tokens": 228548073.0, + "step": 8832 + }, + { + "epoch": 0.9700197671864704, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8678202629089355, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7115490436553955, + "num_tokens": 228568379.0, + "step": 8833 + }, + { + "epoch": 0.9701295848890841, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5516350269317627, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.7013221383094788, + "num_tokens": 228595878.0, + "step": 8834 + }, + { + "epoch": 0.9702394025916978, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.317293405532837, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6929587721824646, + "num_tokens": 228626734.0, + "step": 8835 + }, + { + "epoch": 0.9703492202943115, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4232308864593506, + "learning_rate": 1e-06, + "loss": 1.1025, + "mean_token_accuracy": 0.6848540902137756, + "num_tokens": 228657659.0, + "step": 8836 + }, + { + "epoch": 0.9704590379969251, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.695239305496216, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7049099802970886, + "num_tokens": 228679738.0, + "step": 8837 + }, + { + "epoch": 0.9705688556995388, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.540074348449707, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7041821479797363, + "num_tokens": 228706159.0, + "step": 8838 + }, + { + "epoch": 0.9706786734021524, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.9948859214782715, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.7039539813995361, + "num_tokens": 228727542.0, + "step": 8839 + }, + { + "epoch": 0.9707884911047661, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.616732597351074, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6879985928535461, + "num_tokens": 228751295.0, + "step": 8840 + }, + { + "epoch": 0.9708983088073797, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4611940383911133, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6948728561401367, + "num_tokens": 228778417.0, + "step": 8841 + }, + { + "epoch": 0.9710081265099935, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.2387681007385254, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6929071545600891, + "num_tokens": 228812038.0, + "step": 8842 + }, + { + "epoch": 0.9711179442126071, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.503553867340088, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.7004886865615845, + "num_tokens": 228839443.0, + "step": 8843 + }, + { + "epoch": 0.9712277619152208, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6026861667633057, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.714195966720581, + "num_tokens": 228863670.0, + "step": 8844 + }, + { + "epoch": 0.9713375796178344, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.2487270832061768, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6996411085128784, + "num_tokens": 228899048.0, + "step": 8845 + }, + { + "epoch": 0.971447397320448, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.63339900970459, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6783619523048401, + "num_tokens": 228925711.0, + "step": 8846 + }, + { + "epoch": 0.9715572150230617, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.643953323364258, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7090674042701721, + "num_tokens": 228948934.0, + "step": 8847 + }, + { + "epoch": 0.9716670327256753, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.331554412841797, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7068580389022827, + "num_tokens": 228975185.0, + "step": 8848 + }, + { + "epoch": 0.971776850428289, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.495335578918457, + "learning_rate": 1e-06, + "loss": 0.9278, + "mean_token_accuracy": 0.7217223048210144, + "num_tokens": 229001724.0, + "step": 8849 + }, + { + "epoch": 0.9718866681309027, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.055750608444214, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7203937768936157, + "num_tokens": 229023802.0, + "step": 8850 + }, + { + "epoch": 0.9719964858335164, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.471803903579712, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7166684865951538, + "num_tokens": 229053214.0, + "step": 8851 + }, + { + "epoch": 0.97210630353613, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5213160514831543, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7059683799743652, + "num_tokens": 229082556.0, + "step": 8852 + }, + { + "epoch": 0.9722161212387437, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 16.749778747558594, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.694473385810852, + "num_tokens": 229102678.0, + "step": 8853 + }, + { + "epoch": 0.9723259389413573, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.814429998397827, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7012195587158203, + "num_tokens": 229127059.0, + "step": 8854 + }, + { + "epoch": 0.972435756643971, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6720876693725586, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6957079172134399, + "num_tokens": 229151902.0, + "step": 8855 + }, + { + "epoch": 0.9725455743465846, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.724045515060425, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7149244546890259, + "num_tokens": 229174950.0, + "step": 8856 + }, + { + "epoch": 0.9726553920491984, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.800102710723877, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7103835344314575, + "num_tokens": 229196120.0, + "step": 8857 + }, + { + "epoch": 0.972765209751812, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.5248234272003174, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6943323016166687, + "num_tokens": 229223840.0, + "step": 8858 + }, + { + "epoch": 0.9728750274544257, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.846933126449585, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7081654071807861, + "num_tokens": 229243153.0, + "step": 8859 + }, + { + "epoch": 0.9729848451570393, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.634554386138916, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6950623393058777, + "num_tokens": 229270777.0, + "step": 8860 + }, + { + "epoch": 0.973094662859653, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.2283825874328613, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.7056981921195984, + "num_tokens": 229292593.0, + "step": 8861 + }, + { + "epoch": 0.9732044805622666, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.6569769382476807, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7066882848739624, + "num_tokens": 229317275.0, + "step": 8862 + }, + { + "epoch": 0.9733142982648803, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6213057041168213, + "learning_rate": 1e-06, + "loss": 1.1043, + "mean_token_accuracy": 0.6879698634147644, + "num_tokens": 229343841.0, + "step": 8863 + }, + { + "epoch": 0.973424115967494, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.664680242538452, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6881070137023926, + "num_tokens": 229368220.0, + "step": 8864 + }, + { + "epoch": 0.9735339336701077, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.7501444816589355, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.6855257749557495, + "num_tokens": 229389388.0, + "step": 8865 + }, + { + "epoch": 0.9736437513727213, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.439074993133545, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6908317804336548, + "num_tokens": 229419230.0, + "step": 8866 + }, + { + "epoch": 0.973753569075335, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.4350178241729736, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.685014009475708, + "num_tokens": 229448049.0, + "step": 8867 + }, + { + "epoch": 0.9738633867779486, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 3.0119173526763916, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6904045939445496, + "num_tokens": 229472943.0, + "step": 8868 + }, + { + "epoch": 0.9739732044805622, + "ewc_loss": 8.285045623779297e-06, + "grad_norm": 2.9393270015716553, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7167271375656128, + "num_tokens": 229492958.0, + "step": 8869 + }, + { + "epoch": 0.9740830221831759, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 16.67194938659668, + "learning_rate": 1e-06, + "loss": 0.8547, + "mean_token_accuracy": 0.7473580241203308, + "num_tokens": 229518904.0, + "step": 8870 + }, + { + "epoch": 0.9741928398857896, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.62894868850708, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7012492418289185, + "num_tokens": 229546570.0, + "step": 8871 + }, + { + "epoch": 0.9743026575884033, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.805638313293457, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7106595039367676, + "num_tokens": 229567965.0, + "step": 8872 + }, + { + "epoch": 0.9744124752910169, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.456691026687622, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.7025352716445923, + "num_tokens": 229596198.0, + "step": 8873 + }, + { + "epoch": 0.9745222929936306, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3605751991271973, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6991276144981384, + "num_tokens": 229626224.0, + "step": 8874 + }, + { + "epoch": 0.9746321106962442, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.488288402557373, + "learning_rate": 1e-06, + "loss": 1.0754, + "mean_token_accuracy": 0.6864051818847656, + "num_tokens": 229652706.0, + "step": 8875 + }, + { + "epoch": 0.9747419283988579, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4172840118408203, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6994426250457764, + "num_tokens": 229679203.0, + "step": 8876 + }, + { + "epoch": 0.9748517461014715, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5887598991394043, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6956045627593994, + "num_tokens": 229702096.0, + "step": 8877 + }, + { + "epoch": 0.9749615638040853, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.954697847366333, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6922401189804077, + "num_tokens": 229721871.0, + "step": 8878 + }, + { + "epoch": 0.9750713815066989, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7053685188293457, + "learning_rate": 1e-06, + "loss": 1.1072, + "mean_token_accuracy": 0.6809751987457275, + "num_tokens": 229745519.0, + "step": 8879 + }, + { + "epoch": 0.9751811992093126, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5576329231262207, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7120277881622314, + "num_tokens": 229771220.0, + "step": 8880 + }, + { + "epoch": 0.9752910169119262, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6284964084625244, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.6863491535186768, + "num_tokens": 229795634.0, + "step": 8881 + }, + { + "epoch": 0.9754008346145399, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.830425977706909, + "learning_rate": 1e-06, + "loss": 0.9225, + "mean_token_accuracy": 0.7297581434249878, + "num_tokens": 229815447.0, + "step": 8882 + }, + { + "epoch": 0.9755106523171535, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.342412233352661, + "learning_rate": 1e-06, + "loss": 1.1882, + "mean_token_accuracy": 0.6580880880355835, + "num_tokens": 229846624.0, + "step": 8883 + }, + { + "epoch": 0.9756204700197671, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.594245433807373, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.702965497970581, + "num_tokens": 229870063.0, + "step": 8884 + }, + { + "epoch": 0.9757302877223808, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.404893159866333, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7016671895980835, + "num_tokens": 229898350.0, + "step": 8885 + }, + { + "epoch": 0.9758401054249946, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.769272565841675, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6870819926261902, + "num_tokens": 229921538.0, + "step": 8886 + }, + { + "epoch": 0.9759499231276082, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.322800397872925, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6913422346115112, + "num_tokens": 229945119.0, + "step": 8887 + }, + { + "epoch": 0.9760597408302218, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7980356216430664, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6842542290687561, + "num_tokens": 229968920.0, + "step": 8888 + }, + { + "epoch": 0.9761695585328355, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3336119651794434, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7047079801559448, + "num_tokens": 230003802.0, + "step": 8889 + }, + { + "epoch": 0.9762793762354491, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.510756254196167, + "learning_rate": 1e-06, + "loss": 1.1152, + "mean_token_accuracy": 0.6927099227905273, + "num_tokens": 230030227.0, + "step": 8890 + }, + { + "epoch": 0.9763891939380628, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7098217010498047, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7081661820411682, + "num_tokens": 230053413.0, + "step": 8891 + }, + { + "epoch": 0.9764990116406764, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7666420936584473, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6977826356887817, + "num_tokens": 230076540.0, + "step": 8892 + }, + { + "epoch": 0.9766088293432902, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5760700702667236, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7049664855003357, + "num_tokens": 230102079.0, + "step": 8893 + }, + { + "epoch": 0.9767186470459038, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5430445671081543, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.6893956065177917, + "num_tokens": 230129033.0, + "step": 8894 + }, + { + "epoch": 0.9768284647485175, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.332289934158325, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6851065158843994, + "num_tokens": 230158669.0, + "step": 8895 + }, + { + "epoch": 0.9769382824511311, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7141549587249756, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7165150046348572, + "num_tokens": 230178979.0, + "step": 8896 + }, + { + "epoch": 0.9770481001537448, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5951056480407715, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.7073739171028137, + "num_tokens": 230203677.0, + "step": 8897 + }, + { + "epoch": 0.9771579178563584, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.861067533493042, + "learning_rate": 1e-06, + "loss": 0.8813, + "mean_token_accuracy": 0.7325892448425293, + "num_tokens": 230223685.0, + "step": 8898 + }, + { + "epoch": 0.9772677355589721, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.745565891265869, + "learning_rate": 1e-06, + "loss": 0.9554, + "mean_token_accuracy": 0.7216936349868774, + "num_tokens": 230243836.0, + "step": 8899 + }, + { + "epoch": 0.9773775532615858, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6298437118530273, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6856921911239624, + "num_tokens": 230267068.0, + "step": 8900 + }, + { + "epoch": 0.9774873709641995, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.573439359664917, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6902306079864502, + "num_tokens": 230293163.0, + "step": 8901 + }, + { + "epoch": 0.9775971886668131, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.2945926189422607, + "learning_rate": 1e-06, + "loss": 1.1212, + "mean_token_accuracy": 0.6794573068618774, + "num_tokens": 230325348.0, + "step": 8902 + }, + { + "epoch": 0.9777070063694268, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.806075096130371, + "learning_rate": 1e-06, + "loss": 0.9253, + "mean_token_accuracy": 0.7281960248947144, + "num_tokens": 230345616.0, + "step": 8903 + }, + { + "epoch": 0.9778168240720404, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.0578017234802246, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7102646231651306, + "num_tokens": 230366751.0, + "step": 8904 + }, + { + "epoch": 0.977926641774654, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5803091526031494, + "learning_rate": 1e-06, + "loss": 1.1194, + "mean_token_accuracy": 0.6790699362754822, + "num_tokens": 230395863.0, + "step": 8905 + }, + { + "epoch": 0.9780364594772677, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.043383836746216, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7188801765441895, + "num_tokens": 230414203.0, + "step": 8906 + }, + { + "epoch": 0.9781462771798815, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6457536220550537, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7085963487625122, + "num_tokens": 230440005.0, + "step": 8907 + }, + { + "epoch": 0.9782560948824951, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.742194175720215, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6947975754737854, + "num_tokens": 230464148.0, + "step": 8908 + }, + { + "epoch": 0.9783659125851087, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7576215267181396, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7072869539260864, + "num_tokens": 230484594.0, + "step": 8909 + }, + { + "epoch": 0.9784757302877224, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 16.679393768310547, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7045673131942749, + "num_tokens": 230510260.0, + "step": 8910 + }, + { + "epoch": 0.978585547990336, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 3.2681028842926025, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6906312108039856, + "num_tokens": 230539461.0, + "step": 8911 + }, + { + "epoch": 0.9786953656929497, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6146116256713867, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7005442380905151, + "num_tokens": 230566495.0, + "step": 8912 + }, + { + "epoch": 0.9788051833955633, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.2737534046173096, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6879573464393616, + "num_tokens": 230601093.0, + "step": 8913 + }, + { + "epoch": 0.978915001098177, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.2990782260894775, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7021418213844299, + "num_tokens": 230635869.0, + "step": 8914 + }, + { + "epoch": 0.9790248188007907, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.294614553451538, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6869853734970093, + "num_tokens": 230663962.0, + "step": 8915 + }, + { + "epoch": 0.9791346365034044, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.2550418376922607, + "learning_rate": 1e-06, + "loss": 1.1046, + "mean_token_accuracy": 0.6837860345840454, + "num_tokens": 230696912.0, + "step": 8916 + }, + { + "epoch": 0.979244454206018, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.369983434677124, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.6965289115905762, + "num_tokens": 230725370.0, + "step": 8917 + }, + { + "epoch": 0.9793542719086317, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.624126672744751, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7016907930374146, + "num_tokens": 230749997.0, + "step": 8918 + }, + { + "epoch": 0.9794640896112453, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4920730590820312, + "learning_rate": 1e-06, + "loss": 0.9126, + "mean_token_accuracy": 0.7325913310050964, + "num_tokens": 230774837.0, + "step": 8919 + }, + { + "epoch": 0.979573907313859, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.726107120513916, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7160907983779907, + "num_tokens": 230799299.0, + "step": 8920 + }, + { + "epoch": 0.9796837250164726, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.780803918838501, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6957517862319946, + "num_tokens": 230822987.0, + "step": 8921 + }, + { + "epoch": 0.9797935427190864, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8969950675964355, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7006524801254272, + "num_tokens": 230843790.0, + "step": 8922 + }, + { + "epoch": 0.9799033604217, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.573946237564087, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7161933183670044, + "num_tokens": 230866674.0, + "step": 8923 + }, + { + "epoch": 0.9800131781243137, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.440772771835327, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6923547387123108, + "num_tokens": 230896171.0, + "step": 8924 + }, + { + "epoch": 0.9801229958269273, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5752577781677246, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.6859041452407837, + "num_tokens": 230927449.0, + "step": 8925 + }, + { + "epoch": 0.980232813529541, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6881675720214844, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7075056433677673, + "num_tokens": 230949584.0, + "step": 8926 + }, + { + "epoch": 0.9803426312321546, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7909107208251953, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7180101275444031, + "num_tokens": 230971160.0, + "step": 8927 + }, + { + "epoch": 0.9804524489347682, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.463916540145874, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7063555717468262, + "num_tokens": 230998463.0, + "step": 8928 + }, + { + "epoch": 0.980562266637382, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.423936605453491, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7242671847343445, + "num_tokens": 231027874.0, + "step": 8929 + }, + { + "epoch": 0.9806720843399956, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.521554708480835, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6970819234848022, + "num_tokens": 231054479.0, + "step": 8930 + }, + { + "epoch": 0.9807819020426093, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.40120792388916, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6969350576400757, + "num_tokens": 231084304.0, + "step": 8931 + }, + { + "epoch": 0.9808917197452229, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.501610279083252, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7048733234405518, + "num_tokens": 231109727.0, + "step": 8932 + }, + { + "epoch": 0.9810015374478366, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.553743600845337, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.691767692565918, + "num_tokens": 231138001.0, + "step": 8933 + }, + { + "epoch": 0.9811113551504502, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7174267768859863, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6991959810256958, + "num_tokens": 231161019.0, + "step": 8934 + }, + { + "epoch": 0.9812211728530639, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6628453731536865, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7120881080627441, + "num_tokens": 231182904.0, + "step": 8935 + }, + { + "epoch": 0.9813309905556776, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7181639671325684, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7074073553085327, + "num_tokens": 231205834.0, + "step": 8936 + }, + { + "epoch": 0.9814408082582913, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.8826000690460205, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7103851437568665, + "num_tokens": 231229297.0, + "step": 8937 + }, + { + "epoch": 0.9815506259609049, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5061535835266113, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7094093561172485, + "num_tokens": 231253732.0, + "step": 8938 + }, + { + "epoch": 0.9816604436635186, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6464881896972656, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6937357783317566, + "num_tokens": 231281803.0, + "step": 8939 + }, + { + "epoch": 0.9817702613661322, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.359029769897461, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.6830611228942871, + "num_tokens": 231312159.0, + "step": 8940 + }, + { + "epoch": 0.9818800790687459, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5488169193267822, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.712042510509491, + "num_tokens": 231339780.0, + "step": 8941 + }, + { + "epoch": 0.9819898967713595, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4625675678253174, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.721699595451355, + "num_tokens": 231365522.0, + "step": 8942 + }, + { + "epoch": 0.9820997144739732, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5032451152801514, + "learning_rate": 1e-06, + "loss": 1.0753, + "mean_token_accuracy": 0.6885654330253601, + "num_tokens": 231392658.0, + "step": 8943 + }, + { + "epoch": 0.9822095321765869, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.303877115249634, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6929559111595154, + "num_tokens": 231421138.0, + "step": 8944 + }, + { + "epoch": 0.9823193498792006, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.807389259338379, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6946768164634705, + "num_tokens": 231449858.0, + "step": 8945 + }, + { + "epoch": 0.9824291675818142, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6738061904907227, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6922959089279175, + "num_tokens": 231471215.0, + "step": 8946 + }, + { + "epoch": 0.9825389852844278, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.687164545059204, + "learning_rate": 1e-06, + "loss": 0.8691, + "mean_token_accuracy": 0.7449284791946411, + "num_tokens": 231493172.0, + "step": 8947 + }, + { + "epoch": 0.9826488029870415, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.664498805999756, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.7062278985977173, + "num_tokens": 231517220.0, + "step": 8948 + }, + { + "epoch": 0.9827586206896551, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.469315767288208, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.6875973343849182, + "num_tokens": 231546201.0, + "step": 8949 + }, + { + "epoch": 0.9828684383922688, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6499640941619873, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7094579935073853, + "num_tokens": 231569750.0, + "step": 8950 + }, + { + "epoch": 0.9829782560948825, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.7844388484954834, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6981245279312134, + "num_tokens": 231590161.0, + "step": 8951 + }, + { + "epoch": 0.9830880737974962, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3909411430358887, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6977942585945129, + "num_tokens": 231621120.0, + "step": 8952 + }, + { + "epoch": 0.9831978915001098, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4770314693450928, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7002012133598328, + "num_tokens": 231648905.0, + "step": 8953 + }, + { + "epoch": 0.9833077092027235, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6224074363708496, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6830674409866333, + "num_tokens": 231676121.0, + "step": 8954 + }, + { + "epoch": 0.9834175269053371, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8114821910858154, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7062573432922363, + "num_tokens": 231703666.0, + "step": 8955 + }, + { + "epoch": 0.9835273446079508, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.671532154083252, + "learning_rate": 1e-06, + "loss": 1.1159, + "mean_token_accuracy": 0.6728217601776123, + "num_tokens": 231729719.0, + "step": 8956 + }, + { + "epoch": 0.9836371623105644, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5704617500305176, + "learning_rate": 1e-06, + "loss": 1.0999, + "mean_token_accuracy": 0.6904095411300659, + "num_tokens": 231756062.0, + "step": 8957 + }, + { + "epoch": 0.9837469800131782, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.465440034866333, + "learning_rate": 1e-06, + "loss": 1.1637, + "mean_token_accuracy": 0.6621737480163574, + "num_tokens": 231787721.0, + "step": 8958 + }, + { + "epoch": 0.9838567977157918, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.480393171310425, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6906365156173706, + "num_tokens": 231813870.0, + "step": 8959 + }, + { + "epoch": 0.9839666154184055, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4654223918914795, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7024780511856079, + "num_tokens": 231840917.0, + "step": 8960 + }, + { + "epoch": 0.9840764331210191, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6906652450561523, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7118617296218872, + "num_tokens": 231862843.0, + "step": 8961 + }, + { + "epoch": 0.9841862508236328, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5418875217437744, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.708077609539032, + "num_tokens": 231889043.0, + "step": 8962 + }, + { + "epoch": 0.9842960685262464, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4513251781463623, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.6993509531021118, + "num_tokens": 231916725.0, + "step": 8963 + }, + { + "epoch": 0.98440588622886, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.382805347442627, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7179663181304932, + "num_tokens": 231946104.0, + "step": 8964 + }, + { + "epoch": 0.9845157039314738, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.773228645324707, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.6989294290542603, + "num_tokens": 231967785.0, + "step": 8965 + }, + { + "epoch": 0.9846255216340875, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4581332206726074, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.696682333946228, + "num_tokens": 231994741.0, + "step": 8966 + }, + { + "epoch": 0.9847353393367011, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.819030284881592, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6882352828979492, + "num_tokens": 232016422.0, + "step": 8967 + }, + { + "epoch": 0.9848451570393147, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.577414035797119, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6973241567611694, + "num_tokens": 232040771.0, + "step": 8968 + }, + { + "epoch": 0.9849549747419284, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4503867626190186, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.713580846786499, + "num_tokens": 232066750.0, + "step": 8969 + }, + { + "epoch": 0.985064792444542, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5672364234924316, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7029334306716919, + "num_tokens": 232090449.0, + "step": 8970 + }, + { + "epoch": 0.9851746101471557, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.266160011291504, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.6843005418777466, + "num_tokens": 232122063.0, + "step": 8971 + }, + { + "epoch": 0.9852844278497693, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4074623584747314, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7106167674064636, + "num_tokens": 232149959.0, + "step": 8972 + }, + { + "epoch": 0.9853942455523831, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 3.00302791595459, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.721467912197113, + "num_tokens": 232169056.0, + "step": 8973 + }, + { + "epoch": 0.9855040632549967, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.593634605407715, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7034155130386353, + "num_tokens": 232194796.0, + "step": 8974 + }, + { + "epoch": 0.9856138809576104, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.390843391418457, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7041194438934326, + "num_tokens": 232224152.0, + "step": 8975 + }, + { + "epoch": 0.985723698660224, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.59537935256958, + "learning_rate": 1e-06, + "loss": 1.1117, + "mean_token_accuracy": 0.6817777752876282, + "num_tokens": 232248823.0, + "step": 8976 + }, + { + "epoch": 0.9858335163628377, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.308422088623047, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6909939050674438, + "num_tokens": 232278960.0, + "step": 8977 + }, + { + "epoch": 0.9859433340654513, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7945632934570312, + "learning_rate": 1e-06, + "loss": 0.9399, + "mean_token_accuracy": 0.7231791019439697, + "num_tokens": 232299002.0, + "step": 8978 + }, + { + "epoch": 0.986053151768065, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.9447925090789795, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7254580855369568, + "num_tokens": 232323783.0, + "step": 8979 + }, + { + "epoch": 0.9861629694706787, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6374568939208984, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.723077654838562, + "num_tokens": 232345943.0, + "step": 8980 + }, + { + "epoch": 0.9862727871732924, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6568691730499268, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6961148381233215, + "num_tokens": 232370383.0, + "step": 8981 + }, + { + "epoch": 0.986382604875906, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.466545581817627, + "learning_rate": 1e-06, + "loss": 1.0969, + "mean_token_accuracy": 0.6842951774597168, + "num_tokens": 232399776.0, + "step": 8982 + }, + { + "epoch": 0.9864924225785197, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.412532091140747, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.698223888874054, + "num_tokens": 232429891.0, + "step": 8983 + }, + { + "epoch": 0.9866022402811333, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6289432048797607, + "learning_rate": 1e-06, + "loss": 1.1433, + "mean_token_accuracy": 0.6717208623886108, + "num_tokens": 232456867.0, + "step": 8984 + }, + { + "epoch": 0.986712057983747, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5453155040740967, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7252258062362671, + "num_tokens": 232481328.0, + "step": 8985 + }, + { + "epoch": 0.9868218756863606, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5079452991485596, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7114995121955872, + "num_tokens": 232504646.0, + "step": 8986 + }, + { + "epoch": 0.9869316933889744, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.349630117416382, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6925856471061707, + "num_tokens": 232535282.0, + "step": 8987 + }, + { + "epoch": 0.987041511091588, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.514681577682495, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7057375311851501, + "num_tokens": 232561180.0, + "step": 8988 + }, + { + "epoch": 0.9871513287942016, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.390669584274292, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.7020339369773865, + "num_tokens": 232588832.0, + "step": 8989 + }, + { + "epoch": 0.9872611464968153, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3746554851531982, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.6911739110946655, + "num_tokens": 232616528.0, + "step": 8990 + }, + { + "epoch": 0.9873709641994289, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5728583335876465, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6961543560028076, + "num_tokens": 232643620.0, + "step": 8991 + }, + { + "epoch": 0.9874807819020426, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.3018431663513184, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.7021905183792114, + "num_tokens": 232672758.0, + "step": 8992 + }, + { + "epoch": 0.9875905996046562, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 4.332113265991211, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6961331367492676, + "num_tokens": 232698818.0, + "step": 8993 + }, + { + "epoch": 0.98770041730727, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5155508518218994, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.695263147354126, + "num_tokens": 232723684.0, + "step": 8994 + }, + { + "epoch": 0.9878102350098836, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.607243061065674, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7141016721725464, + "num_tokens": 232748578.0, + "step": 8995 + }, + { + "epoch": 0.9879200527124973, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.405121088027954, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7048490047454834, + "num_tokens": 232777194.0, + "step": 8996 + }, + { + "epoch": 0.9880298704151109, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.351101875305176, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.691210150718689, + "num_tokens": 232806509.0, + "step": 8997 + }, + { + "epoch": 0.9881396881177246, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.750908374786377, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6902700662612915, + "num_tokens": 232829602.0, + "step": 8998 + }, + { + "epoch": 0.9882495058203382, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6527438163757324, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7023506164550781, + "num_tokens": 232853101.0, + "step": 8999 + }, + { + "epoch": 0.9883593235229519, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.615279197692871, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7100528478622437, + "num_tokens": 232876847.0, + "step": 9000 + }, + { + "epoch": 0.9884691412255655, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.673403024673462, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7238279581069946, + "num_tokens": 232897886.0, + "step": 9001 + }, + { + "epoch": 0.9885789589281793, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5427441596984863, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.689775824546814, + "num_tokens": 232923874.0, + "step": 9002 + }, + { + "epoch": 0.9886887766307929, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.544684886932373, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7039198875427246, + "num_tokens": 232948175.0, + "step": 9003 + }, + { + "epoch": 0.9887985943334066, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.68418025970459, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.6913637518882751, + "num_tokens": 232970566.0, + "step": 9004 + }, + { + "epoch": 0.9889084120360202, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6802995204925537, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.6870517730712891, + "num_tokens": 232993237.0, + "step": 9005 + }, + { + "epoch": 0.9890182297386338, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.250614643096924, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6874802112579346, + "num_tokens": 233027525.0, + "step": 9006 + }, + { + "epoch": 0.9891280474412475, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4368197917938232, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6976174116134644, + "num_tokens": 233061174.0, + "step": 9007 + }, + { + "epoch": 0.9892378651438611, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.806053876876831, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6872508525848389, + "num_tokens": 233087169.0, + "step": 9008 + }, + { + "epoch": 0.9893476828464749, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.488892078399658, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.693500816822052, + "num_tokens": 233113169.0, + "step": 9009 + }, + { + "epoch": 0.9894575005490885, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3306949138641357, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6995317339897156, + "num_tokens": 233145830.0, + "step": 9010 + }, + { + "epoch": 0.9895673182517022, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.64847731590271, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6891316771507263, + "num_tokens": 233172642.0, + "step": 9011 + }, + { + "epoch": 0.9896771359543158, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6367757320404053, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7193282246589661, + "num_tokens": 233194751.0, + "step": 9012 + }, + { + "epoch": 0.9897869536569295, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.313142776489258, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6996941566467285, + "num_tokens": 233225997.0, + "step": 9013 + }, + { + "epoch": 0.9898967713595431, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4511709213256836, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6994511485099792, + "num_tokens": 233252009.0, + "step": 9014 + }, + { + "epoch": 0.9900065890621568, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4094698429107666, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7156925797462463, + "num_tokens": 233278741.0, + "step": 9015 + }, + { + "epoch": 0.9901164067647705, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7790558338165283, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7104246020317078, + "num_tokens": 233298654.0, + "step": 9016 + }, + { + "epoch": 0.9902262244673842, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.378061056137085, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7042632102966309, + "num_tokens": 233327394.0, + "step": 9017 + }, + { + "epoch": 0.9903360421699978, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5166876316070557, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7033188343048096, + "num_tokens": 233352462.0, + "step": 9018 + }, + { + "epoch": 0.9904458598726115, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6901919841766357, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7105200886726379, + "num_tokens": 233374329.0, + "step": 9019 + }, + { + "epoch": 0.9905556775752251, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4502737522125244, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7113538384437561, + "num_tokens": 233399704.0, + "step": 9020 + }, + { + "epoch": 0.9906654952778388, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.769827127456665, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.69749516248703, + "num_tokens": 233422103.0, + "step": 9021 + }, + { + "epoch": 0.9907753129804524, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.844501495361328, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7005001306533813, + "num_tokens": 233444041.0, + "step": 9022 + }, + { + "epoch": 0.9908851306830662, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.402045488357544, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7116712927818298, + "num_tokens": 233472497.0, + "step": 9023 + }, + { + "epoch": 0.9909949483856798, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7984981536865234, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.7000823020935059, + "num_tokens": 233495761.0, + "step": 9024 + }, + { + "epoch": 0.9911047660882935, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.2888100147247314, + "learning_rate": 1e-06, + "loss": 1.1273, + "mean_token_accuracy": 0.675392746925354, + "num_tokens": 233526528.0, + "step": 9025 + }, + { + "epoch": 0.9912145837909071, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.662351369857788, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7051457166671753, + "num_tokens": 233549552.0, + "step": 9026 + }, + { + "epoch": 0.9913244014935207, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.598522663116455, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7022555470466614, + "num_tokens": 233575314.0, + "step": 9027 + }, + { + "epoch": 0.9914342191961344, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6347570419311523, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6962919235229492, + "num_tokens": 233600910.0, + "step": 9028 + }, + { + "epoch": 0.991544036898748, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6343352794647217, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.6948952674865723, + "num_tokens": 233624970.0, + "step": 9029 + }, + { + "epoch": 0.9916538546013618, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4347972869873047, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.694396436214447, + "num_tokens": 233656091.0, + "step": 9030 + }, + { + "epoch": 0.9917636723039754, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.666213035583496, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7066087126731873, + "num_tokens": 233678512.0, + "step": 9031 + }, + { + "epoch": 0.9918734900065891, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4626338481903076, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7044302225112915, + "num_tokens": 233704187.0, + "step": 9032 + }, + { + "epoch": 0.9919833077092027, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5375540256500244, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6981123089790344, + "num_tokens": 233731448.0, + "step": 9033 + }, + { + "epoch": 0.9920931254118164, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5715322494506836, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7051079273223877, + "num_tokens": 233755151.0, + "step": 9034 + }, + { + "epoch": 0.99220294311443, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.25368595123291, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7074010968208313, + "num_tokens": 233789917.0, + "step": 9035 + }, + { + "epoch": 0.9923127608170437, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.7381911277770996, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7083677053451538, + "num_tokens": 233812965.0, + "step": 9036 + }, + { + "epoch": 0.9924225785196573, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5005922317504883, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.7022910118103027, + "num_tokens": 233842025.0, + "step": 9037 + }, + { + "epoch": 0.9925323962222711, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.329099416732788, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7114503979682922, + "num_tokens": 233870113.0, + "step": 9038 + }, + { + "epoch": 0.9926422139248847, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4295294284820557, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7033213973045349, + "num_tokens": 233897948.0, + "step": 9039 + }, + { + "epoch": 0.9927520316274984, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5996744632720947, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6867139339447021, + "num_tokens": 233922903.0, + "step": 9040 + }, + { + "epoch": 0.992861849330112, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4105870723724365, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7274429202079773, + "num_tokens": 233949903.0, + "step": 9041 + }, + { + "epoch": 0.9929716670327257, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.487905502319336, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6956619620323181, + "num_tokens": 233980714.0, + "step": 9042 + }, + { + "epoch": 0.9930814847353393, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4952707290649414, + "learning_rate": 1e-06, + "loss": 1.1285, + "mean_token_accuracy": 0.6798232793807983, + "num_tokens": 234011634.0, + "step": 9043 + }, + { + "epoch": 0.993191302437953, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8335421085357666, + "learning_rate": 1e-06, + "loss": 1.1095, + "mean_token_accuracy": 0.6804021596908569, + "num_tokens": 234035125.0, + "step": 9044 + }, + { + "epoch": 0.9933011201405667, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.7484331130981445, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6924439668655396, + "num_tokens": 234057417.0, + "step": 9045 + }, + { + "epoch": 0.9934109378431804, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.8972370624542236, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6883835792541504, + "num_tokens": 234078743.0, + "step": 9046 + }, + { + "epoch": 0.993520755545794, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.47983717918396, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7239792346954346, + "num_tokens": 234105817.0, + "step": 9047 + }, + { + "epoch": 0.9936305732484076, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 3.426170587539673, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6878138780593872, + "num_tokens": 234132511.0, + "step": 9048 + }, + { + "epoch": 0.9937403909510213, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6045851707458496, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7150084972381592, + "num_tokens": 234155713.0, + "step": 9049 + }, + { + "epoch": 0.9938502086536349, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4118714332580566, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7072734832763672, + "num_tokens": 234184946.0, + "step": 9050 + }, + { + "epoch": 0.9939600263562486, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6670162677764893, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7076234817504883, + "num_tokens": 234206926.0, + "step": 9051 + }, + { + "epoch": 0.9940698440588623, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.496696949005127, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7086935043334961, + "num_tokens": 234231917.0, + "step": 9052 + }, + { + "epoch": 0.994179661761476, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.59708309173584, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6911649703979492, + "num_tokens": 234257913.0, + "step": 9053 + }, + { + "epoch": 0.9942894794640896, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.61295747756958, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7170289754867554, + "num_tokens": 234281030.0, + "step": 9054 + }, + { + "epoch": 0.9943992971667033, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.683936834335327, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7079269886016846, + "num_tokens": 234304409.0, + "step": 9055 + }, + { + "epoch": 0.9945091148693169, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.426457643508911, + "learning_rate": 1e-06, + "loss": 1.143, + "mean_token_accuracy": 0.6669899821281433, + "num_tokens": 234336413.0, + "step": 9056 + }, + { + "epoch": 0.9946189325719306, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5410754680633545, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7102052569389343, + "num_tokens": 234360815.0, + "step": 9057 + }, + { + "epoch": 0.9947287502745442, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.4974215030670166, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7110408544540405, + "num_tokens": 234389038.0, + "step": 9058 + }, + { + "epoch": 0.994838567977158, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6923489570617676, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6925739645957947, + "num_tokens": 234410875.0, + "step": 9059 + }, + { + "epoch": 0.9949483856797716, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6347358226776123, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7158383131027222, + "num_tokens": 234436468.0, + "step": 9060 + }, + { + "epoch": 0.9950582033823853, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.692215919494629, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7037110328674316, + "num_tokens": 234462935.0, + "step": 9061 + }, + { + "epoch": 0.9951680210849989, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5317797660827637, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.7038524746894836, + "num_tokens": 234496078.0, + "step": 9062 + }, + { + "epoch": 0.9952778387876126, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4504764080047607, + "learning_rate": 1e-06, + "loss": 1.1204, + "mean_token_accuracy": 0.682834804058075, + "num_tokens": 234524273.0, + "step": 9063 + }, + { + "epoch": 0.9953876564902262, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6367969512939453, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7112154960632324, + "num_tokens": 234549620.0, + "step": 9064 + }, + { + "epoch": 0.9954974741928398, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4848527908325195, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7119823694229126, + "num_tokens": 234576957.0, + "step": 9065 + }, + { + "epoch": 0.9956072918954535, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.678899049758911, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6938538551330566, + "num_tokens": 234601466.0, + "step": 9066 + }, + { + "epoch": 0.9957171095980673, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.3381032943725586, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.72412109375, + "num_tokens": 234632198.0, + "step": 9067 + }, + { + "epoch": 0.9958269273006809, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.296142101287842, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7069320678710938, + "num_tokens": 234662027.0, + "step": 9068 + }, + { + "epoch": 0.9959367450032945, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.372476100921631, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7056038975715637, + "num_tokens": 234688572.0, + "step": 9069 + }, + { + "epoch": 0.9960465627059082, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5584003925323486, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6827483177185059, + "num_tokens": 234716877.0, + "step": 9070 + }, + { + "epoch": 0.9961563804085218, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.293670654296875, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6954936981201172, + "num_tokens": 234746332.0, + "step": 9071 + }, + { + "epoch": 0.9962661981111355, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.661332845687866, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7080330848693848, + "num_tokens": 234769472.0, + "step": 9072 + }, + { + "epoch": 0.9963760158137491, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5736870765686035, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6961509585380554, + "num_tokens": 234794215.0, + "step": 9073 + }, + { + "epoch": 0.9964858335163629, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.367698907852173, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7084023356437683, + "num_tokens": 234821761.0, + "step": 9074 + }, + { + "epoch": 0.9965956512189765, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.5232553482055664, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7095603346824646, + "num_tokens": 234848284.0, + "step": 9075 + }, + { + "epoch": 0.9967054689215902, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.8676161766052246, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7089582085609436, + "num_tokens": 234866900.0, + "step": 9076 + }, + { + "epoch": 0.9968152866242038, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.658673048019409, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7133668065071106, + "num_tokens": 234890696.0, + "step": 9077 + }, + { + "epoch": 0.9969251043268175, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 3.103557825088501, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7102358937263489, + "num_tokens": 234909194.0, + "step": 9078 + }, + { + "epoch": 0.9970349220294311, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3218834400177, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6933138370513916, + "num_tokens": 234938499.0, + "step": 9079 + }, + { + "epoch": 0.9971447397320448, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.982844114303589, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6967285871505737, + "num_tokens": 234956871.0, + "step": 9080 + }, + { + "epoch": 0.9972545574346585, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6247425079345703, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.718307614326477, + "num_tokens": 234983142.0, + "step": 9081 + }, + { + "epoch": 0.9973643751372722, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.3460726737976074, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.6965059638023376, + "num_tokens": 235015256.0, + "step": 9082 + }, + { + "epoch": 0.9974741928398858, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.410440444946289, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7198591232299805, + "num_tokens": 235039088.0, + "step": 9083 + }, + { + "epoch": 0.9975840105424995, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.767058849334717, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6963745355606079, + "num_tokens": 235059458.0, + "step": 9084 + }, + { + "epoch": 0.9976938282451131, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.6228041648864746, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7148083448410034, + "num_tokens": 235080777.0, + "step": 9085 + }, + { + "epoch": 0.9978036459477267, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.3629207611083984, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7241326570510864, + "num_tokens": 235106971.0, + "step": 9086 + }, + { + "epoch": 0.9979134636503404, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.879371166229248, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7103170156478882, + "num_tokens": 235126459.0, + "step": 9087 + }, + { + "epoch": 0.9980232813529542, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5532352924346924, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.6998815536499023, + "num_tokens": 235154795.0, + "step": 9088 + }, + { + "epoch": 0.9981330990555678, + "ewc_loss": 8.344650268554688e-06, + "grad_norm": 2.851618528366089, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6953777074813843, + "num_tokens": 235177941.0, + "step": 9089 + }, + { + "epoch": 0.9982429167581814, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4431238174438477, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6950851082801819, + "num_tokens": 235205466.0, + "step": 9090 + }, + { + "epoch": 0.9983527344607951, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 3.1366498470306396, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6954653263092041, + "num_tokens": 235228890.0, + "step": 9091 + }, + { + "epoch": 0.9984625521634087, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6761786937713623, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.71773761510849, + "num_tokens": 235250592.0, + "step": 9092 + }, + { + "epoch": 0.9985723698660224, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.3855693340301514, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7115536332130432, + "num_tokens": 235278580.0, + "step": 9093 + }, + { + "epoch": 0.998682187568636, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5964515209198, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7173973321914673, + "num_tokens": 235301987.0, + "step": 9094 + }, + { + "epoch": 0.9987920052712497, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.7477385997772217, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.691556453704834, + "num_tokens": 235328062.0, + "step": 9095 + }, + { + "epoch": 0.9989018229738634, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.755077362060547, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7213848829269409, + "num_tokens": 235350562.0, + "step": 9096 + }, + { + "epoch": 0.9990116406764771, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.2902066707611084, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7017645835876465, + "num_tokens": 235379626.0, + "step": 9097 + }, + { + "epoch": 0.9991214583790907, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6507668495178223, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.7037311792373657, + "num_tokens": 235405941.0, + "step": 9098 + }, + { + "epoch": 0.9992312760817044, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4439074993133545, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.702721118927002, + "num_tokens": 235432729.0, + "step": 9099 + }, + { + "epoch": 0.999341093784318, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.372424840927124, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6956814527511597, + "num_tokens": 235463599.0, + "step": 9100 + }, + { + "epoch": 0.9994509114869317, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5948195457458496, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7013354301452637, + "num_tokens": 235486921.0, + "step": 9101 + }, + { + "epoch": 0.9995607291895453, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.2253501415252686, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6883425712585449, + "num_tokens": 235519812.0, + "step": 9102 + }, + { + "epoch": 0.9996705468921591, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.591062545776367, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7208840847015381, + "num_tokens": 235543456.0, + "step": 9103 + }, + { + "epoch": 0.9997803645947727, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4080872535705566, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.688496470451355, + "num_tokens": 235572016.0, + "step": 9104 + }, + { + "epoch": 0.9998901822973864, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.7002522945404053, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7135059833526611, + "num_tokens": 235594912.0, + "step": 9105 + }, + { + "epoch": 1.0, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.61596417427063, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7024730443954468, + "num_tokens": 235617245.0, + "step": 9106 + }, + { + "epoch": 1.0001098177026138, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.3313992023468018, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6932922601699829, + "num_tokens": 235647042.0, + "step": 9107 + }, + { + "epoch": 1.0002196354052273, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.3075850009918213, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7070137858390808, + "num_tokens": 235675620.0, + "step": 9108 + }, + { + "epoch": 1.000329453107841, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.709362030029297, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7022684812545776, + "num_tokens": 235699190.0, + "step": 9109 + }, + { + "epoch": 1.0004392708104546, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.351767063140869, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7061500549316406, + "num_tokens": 235729568.0, + "step": 9110 + }, + { + "epoch": 1.0005490885130683, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5553996562957764, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7154310941696167, + "num_tokens": 235753975.0, + "step": 9111 + }, + { + "epoch": 1.0006589062156819, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.684183359146118, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7176209688186646, + "num_tokens": 235774114.0, + "step": 9112 + }, + { + "epoch": 1.0007687239182956, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6147780418395996, + "learning_rate": 1e-06, + "loss": 1.1047, + "mean_token_accuracy": 0.6839469075202942, + "num_tokens": 235799505.0, + "step": 9113 + }, + { + "epoch": 1.0008785416209094, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.4285850524902344, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7070563435554504, + "num_tokens": 235828000.0, + "step": 9114 + }, + { + "epoch": 1.000988359323523, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.682978630065918, + "learning_rate": 1e-06, + "loss": 0.9091, + "mean_token_accuracy": 0.7356003522872925, + "num_tokens": 235851026.0, + "step": 9115 + }, + { + "epoch": 1.0010981770261367, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5617034435272217, + "learning_rate": 1e-06, + "loss": 1.1306, + "mean_token_accuracy": 0.6754534244537354, + "num_tokens": 235880806.0, + "step": 9116 + }, + { + "epoch": 1.0012079947287502, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.507479667663574, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6927350759506226, + "num_tokens": 235913442.0, + "step": 9117 + }, + { + "epoch": 1.001317812431364, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.654540777206421, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7192772030830383, + "num_tokens": 235936145.0, + "step": 9118 + }, + { + "epoch": 1.0014276301339775, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.858234405517578, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7131801843643188, + "num_tokens": 235958222.0, + "step": 9119 + }, + { + "epoch": 1.0015374478365913, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6733758449554443, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7148212790489197, + "num_tokens": 235980934.0, + "step": 9120 + }, + { + "epoch": 1.001647265539205, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.6761090755462646, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6961804628372192, + "num_tokens": 236005418.0, + "step": 9121 + }, + { + "epoch": 1.0017570832418186, + "ewc_loss": 8.404254913330078e-06, + "grad_norm": 2.5804452896118164, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.715036153793335, + "num_tokens": 236031486.0, + "step": 9122 + }, + { + "epoch": 1.0018669009444323, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 3.1302008628845215, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6858034133911133, + "num_tokens": 236058222.0, + "step": 9123 + }, + { + "epoch": 1.0019767186470458, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.732313871383667, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7227035164833069, + "num_tokens": 236082865.0, + "step": 9124 + }, + { + "epoch": 1.0020865363496596, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.277679681777954, + "learning_rate": 1e-06, + "loss": 1.1157, + "mean_token_accuracy": 0.6856452226638794, + "num_tokens": 236118019.0, + "step": 9125 + }, + { + "epoch": 1.0021963540522731, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.8698501586914062, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7078222036361694, + "num_tokens": 236140259.0, + "step": 9126 + }, + { + "epoch": 1.002306171754887, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.4760403633117676, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6960604190826416, + "num_tokens": 236171193.0, + "step": 9127 + }, + { + "epoch": 1.0024159894575007, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.589277982711792, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.703703761100769, + "num_tokens": 236198553.0, + "step": 9128 + }, + { + "epoch": 1.0025258071601142, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.8482093811035156, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6876196265220642, + "num_tokens": 236225340.0, + "step": 9129 + }, + { + "epoch": 1.002635624862728, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.727003812789917, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7109514474868774, + "num_tokens": 236249433.0, + "step": 9130 + }, + { + "epoch": 1.0027454425653415, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.432248592376709, + "learning_rate": 1e-06, + "loss": 1.1225, + "mean_token_accuracy": 0.6765111684799194, + "num_tokens": 236278592.0, + "step": 9131 + }, + { + "epoch": 1.0028552602679552, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 3.1608219146728516, + "learning_rate": 1e-06, + "loss": 0.9504, + "mean_token_accuracy": 0.7254428863525391, + "num_tokens": 236296550.0, + "step": 9132 + }, + { + "epoch": 1.0029650779705688, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.5340216159820557, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7037818431854248, + "num_tokens": 236325169.0, + "step": 9133 + }, + { + "epoch": 1.0030748956731825, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.3939192295074463, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6847714185714722, + "num_tokens": 236355195.0, + "step": 9134 + }, + { + "epoch": 1.0031847133757963, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.3377606868743896, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.6853891611099243, + "num_tokens": 236392651.0, + "step": 9135 + }, + { + "epoch": 1.0032945310784098, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.298266887664795, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7008485794067383, + "num_tokens": 236429123.0, + "step": 9136 + }, + { + "epoch": 1.0034043487810236, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.853123903274536, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7114049792289734, + "num_tokens": 236458505.0, + "step": 9137 + }, + { + "epoch": 1.0035141664836371, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.8130011558532715, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7092854976654053, + "num_tokens": 236483365.0, + "step": 9138 + }, + { + "epoch": 1.0036239841862509, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.6294479370117188, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7097375392913818, + "num_tokens": 236507918.0, + "step": 9139 + }, + { + "epoch": 1.0037338018888644, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.324506998062134, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6940125226974487, + "num_tokens": 236542471.0, + "step": 9140 + }, + { + "epoch": 1.0038436195914782, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.754270315170288, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7141003608703613, + "num_tokens": 236566275.0, + "step": 9141 + }, + { + "epoch": 1.0039534372940917, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 3.1074836254119873, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.715189516544342, + "num_tokens": 236588724.0, + "step": 9142 + }, + { + "epoch": 1.0040632549967055, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.337584972381592, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6842679977416992, + "num_tokens": 236619468.0, + "step": 9143 + }, + { + "epoch": 1.0041730726993192, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.9544918537139893, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7023231983184814, + "num_tokens": 236641105.0, + "step": 9144 + }, + { + "epoch": 1.0042828904019327, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.7418291568756104, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7118765711784363, + "num_tokens": 236664417.0, + "step": 9145 + }, + { + "epoch": 1.0043927081045465, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.8083460330963135, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7126177549362183, + "num_tokens": 236685394.0, + "step": 9146 + }, + { + "epoch": 1.00450252580716, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.7039475440979004, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7246948480606079, + "num_tokens": 236709099.0, + "step": 9147 + }, + { + "epoch": 1.0046123435097738, + "ewc_loss": 8.463859558105469e-06, + "grad_norm": 2.5200765132904053, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.6925252676010132, + "num_tokens": 236736328.0, + "step": 9148 + }, + { + "epoch": 1.0047221612123873, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 3.0634829998016357, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.711860716342926, + "num_tokens": 236762359.0, + "step": 9149 + }, + { + "epoch": 1.004831978915001, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.468510627746582, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6908750534057617, + "num_tokens": 236788664.0, + "step": 9150 + }, + { + "epoch": 1.0049417966176148, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.5459818840026855, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7163349390029907, + "num_tokens": 236811440.0, + "step": 9151 + }, + { + "epoch": 1.0050516143202284, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.3161556720733643, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6884485483169556, + "num_tokens": 236844805.0, + "step": 9152 + }, + { + "epoch": 1.0051614320228421, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.5052855014801025, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7183597683906555, + "num_tokens": 236869318.0, + "step": 9153 + }, + { + "epoch": 1.0052712497254557, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.1901400089263916, + "learning_rate": 1e-06, + "loss": 0.9202, + "mean_token_accuracy": 0.7294556498527527, + "num_tokens": 236901039.0, + "step": 9154 + }, + { + "epoch": 1.0053810674280694, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.4688427448272705, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6898595094680786, + "num_tokens": 236928639.0, + "step": 9155 + }, + { + "epoch": 1.005490885130683, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.6239826679229736, + "learning_rate": 1e-06, + "loss": 1.0909, + "mean_token_accuracy": 0.6838253736495972, + "num_tokens": 236955654.0, + "step": 9156 + }, + { + "epoch": 1.0056007028332967, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.5187110900878906, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7198320627212524, + "num_tokens": 236980378.0, + "step": 9157 + }, + { + "epoch": 1.0057105205359105, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.5205652713775635, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6923582553863525, + "num_tokens": 237008385.0, + "step": 9158 + }, + { + "epoch": 1.005820338238524, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.540040969848633, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6993685960769653, + "num_tokens": 237034555.0, + "step": 9159 + }, + { + "epoch": 1.0059301559411378, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.4501185417175293, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6945441961288452, + "num_tokens": 237066324.0, + "step": 9160 + }, + { + "epoch": 1.0060399736437513, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.446596145629883, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.696304440498352, + "num_tokens": 237094989.0, + "step": 9161 + }, + { + "epoch": 1.006149791346365, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.8658933639526367, + "learning_rate": 1e-06, + "loss": 1.0766, + "mean_token_accuracy": 0.6835328340530396, + "num_tokens": 237120765.0, + "step": 9162 + }, + { + "epoch": 1.0062596090489786, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.4349772930145264, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.6985567212104797, + "num_tokens": 237151511.0, + "step": 9163 + }, + { + "epoch": 1.0063694267515924, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.63270902633667, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7035248875617981, + "num_tokens": 237176336.0, + "step": 9164 + }, + { + "epoch": 1.0064792444542061, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.4238698482513428, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6817352175712585, + "num_tokens": 237207576.0, + "step": 9165 + }, + { + "epoch": 1.0065890621568196, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.5845844745635986, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7014086842536926, + "num_tokens": 237235137.0, + "step": 9166 + }, + { + "epoch": 1.0066988798594334, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.7169485092163086, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7159761190414429, + "num_tokens": 237255124.0, + "step": 9167 + }, + { + "epoch": 1.006808697562047, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.9126975536346436, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.685187578201294, + "num_tokens": 237278364.0, + "step": 9168 + }, + { + "epoch": 1.0069185152646607, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.543928384780884, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7207942008972168, + "num_tokens": 237302500.0, + "step": 9169 + }, + { + "epoch": 1.0070283329672742, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.5304551124572754, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7199811339378357, + "num_tokens": 237327908.0, + "step": 9170 + }, + { + "epoch": 1.007138150669888, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.62841796875, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6942415237426758, + "num_tokens": 237355523.0, + "step": 9171 + }, + { + "epoch": 1.0072479683725017, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.520019769668579, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6999298334121704, + "num_tokens": 237381925.0, + "step": 9172 + }, + { + "epoch": 1.0073577860751153, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.7093019485473633, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7037447690963745, + "num_tokens": 237408436.0, + "step": 9173 + }, + { + "epoch": 1.007467603777729, + "ewc_loss": 8.52346420288086e-06, + "grad_norm": 2.514390468597412, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7102633714675903, + "num_tokens": 237434340.0, + "step": 9174 + }, + { + "epoch": 1.0075774214803426, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.45241379737854, + "learning_rate": 1e-06, + "loss": 0.9155, + "mean_token_accuracy": 0.7301563024520874, + "num_tokens": 237459680.0, + "step": 9175 + }, + { + "epoch": 1.0076872391829563, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.397993803024292, + "learning_rate": 1e-06, + "loss": 1.0984, + "mean_token_accuracy": 0.6817789673805237, + "num_tokens": 237492224.0, + "step": 9176 + }, + { + "epoch": 1.0077970568855699, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.4804415702819824, + "learning_rate": 1e-06, + "loss": 0.8819, + "mean_token_accuracy": 0.7342002391815186, + "num_tokens": 237516302.0, + "step": 9177 + }, + { + "epoch": 1.0079068745881836, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.5352272987365723, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7030494809150696, + "num_tokens": 237540954.0, + "step": 9178 + }, + { + "epoch": 1.0080166922907974, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.591918468475342, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7024491429328918, + "num_tokens": 237565679.0, + "step": 9179 + }, + { + "epoch": 1.008126509993411, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.481797933578491, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6946821212768555, + "num_tokens": 237592971.0, + "step": 9180 + }, + { + "epoch": 1.0082363276960247, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.5539073944091797, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.710906445980072, + "num_tokens": 237618973.0, + "step": 9181 + }, + { + "epoch": 1.0083461453986382, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.4986236095428467, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.6973118782043457, + "num_tokens": 237649304.0, + "step": 9182 + }, + { + "epoch": 1.008455963101252, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.79512619972229, + "learning_rate": 1e-06, + "loss": 1.1214, + "mean_token_accuracy": 0.6703318357467651, + "num_tokens": 237676775.0, + "step": 9183 + }, + { + "epoch": 1.0085657808038655, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.547825813293457, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.6960501670837402, + "num_tokens": 237707669.0, + "step": 9184 + }, + { + "epoch": 1.0086755985064793, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.731173515319824, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7206987142562866, + "num_tokens": 237732335.0, + "step": 9185 + }, + { + "epoch": 1.008785416209093, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.6955201625823975, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7048386335372925, + "num_tokens": 237755742.0, + "step": 9186 + }, + { + "epoch": 1.0088952339117065, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.610396146774292, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7075382471084595, + "num_tokens": 237779404.0, + "step": 9187 + }, + { + "epoch": 1.0090050516143203, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.851796865463257, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7166867852210999, + "num_tokens": 237801424.0, + "step": 9188 + }, + { + "epoch": 1.0091148693169338, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.8364458084106445, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.704699695110321, + "num_tokens": 237824471.0, + "step": 9189 + }, + { + "epoch": 1.0092246870195476, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.3459384441375732, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6820058822631836, + "num_tokens": 237856973.0, + "step": 9190 + }, + { + "epoch": 1.0093345047221611, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.647700309753418, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7009218335151672, + "num_tokens": 237886112.0, + "step": 9191 + }, + { + "epoch": 1.0094443224247749, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.5911543369293213, + "learning_rate": 1e-06, + "loss": 1.1486, + "mean_token_accuracy": 0.6749281287193298, + "num_tokens": 237912563.0, + "step": 9192 + }, + { + "epoch": 1.0095541401273886, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.7922348976135254, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7176303863525391, + "num_tokens": 237931329.0, + "step": 9193 + }, + { + "epoch": 1.0096639578300022, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 3.0854368209838867, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7145422697067261, + "num_tokens": 237949637.0, + "step": 9194 + }, + { + "epoch": 1.009773775532616, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.927173137664795, + "learning_rate": 1e-06, + "loss": 0.9127, + "mean_token_accuracy": 0.7288757562637329, + "num_tokens": 237968165.0, + "step": 9195 + }, + { + "epoch": 1.0098835932352295, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.6307640075683594, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6970092058181763, + "num_tokens": 237993937.0, + "step": 9196 + }, + { + "epoch": 1.0099934109378432, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.7922165393829346, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7106245160102844, + "num_tokens": 238017976.0, + "step": 9197 + }, + { + "epoch": 1.0101032286404568, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 4.212005138397217, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6852738857269287, + "num_tokens": 238049863.0, + "step": 9198 + }, + { + "epoch": 1.0102130463430705, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.505969285964966, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7179452776908875, + "num_tokens": 238079151.0, + "step": 9199 + }, + { + "epoch": 1.010322864045684, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.580036163330078, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7156861424446106, + "num_tokens": 238103439.0, + "step": 9200 + }, + { + "epoch": 1.0104326817482978, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.453970432281494, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6943501234054565, + "num_tokens": 238132132.0, + "step": 9201 + }, + { + "epoch": 1.0105424994509116, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.8371245861053467, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.722649097442627, + "num_tokens": 238152749.0, + "step": 9202 + }, + { + "epoch": 1.010652317153525, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.442343235015869, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7188365459442139, + "num_tokens": 238180616.0, + "step": 9203 + }, + { + "epoch": 1.0107621348561389, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.812176465988159, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7027196884155273, + "num_tokens": 238205218.0, + "step": 9204 + }, + { + "epoch": 1.0108719525587524, + "ewc_loss": 8.58306884765625e-06, + "grad_norm": 2.640758991241455, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7041135430335999, + "num_tokens": 238228805.0, + "step": 9205 + }, + { + "epoch": 1.0109817702613662, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.583390951156616, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6922154426574707, + "num_tokens": 238258713.0, + "step": 9206 + }, + { + "epoch": 1.0110915879639797, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.83848237991333, + "learning_rate": 1e-06, + "loss": 0.8786, + "mean_token_accuracy": 0.7447546720504761, + "num_tokens": 238277651.0, + "step": 9207 + }, + { + "epoch": 1.0112014056665934, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.430418014526367, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.687736988067627, + "num_tokens": 238306098.0, + "step": 9208 + }, + { + "epoch": 1.0113112233692072, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.761279582977295, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7218281030654907, + "num_tokens": 238327071.0, + "step": 9209 + }, + { + "epoch": 1.0114210410718207, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5001492500305176, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6937524676322937, + "num_tokens": 238353322.0, + "step": 9210 + }, + { + "epoch": 1.0115308587744345, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.444317102432251, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7125688195228577, + "num_tokens": 238382018.0, + "step": 9211 + }, + { + "epoch": 1.011640676477048, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.799339532852173, + "learning_rate": 1e-06, + "loss": 0.9278, + "mean_token_accuracy": 0.7217398881912231, + "num_tokens": 238401017.0, + "step": 9212 + }, + { + "epoch": 1.0117504941796618, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5296969413757324, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7049847841262817, + "num_tokens": 238428954.0, + "step": 9213 + }, + { + "epoch": 1.0118603118822753, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5354459285736084, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7269191741943359, + "num_tokens": 238455499.0, + "step": 9214 + }, + { + "epoch": 1.011970129584889, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4780895709991455, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7048821449279785, + "num_tokens": 238481219.0, + "step": 9215 + }, + { + "epoch": 1.0120799472875028, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5304481983184814, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.726413905620575, + "num_tokens": 238506894.0, + "step": 9216 + }, + { + "epoch": 1.0121897649901164, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5736398696899414, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7117689847946167, + "num_tokens": 238534255.0, + "step": 9217 + }, + { + "epoch": 1.0122995826927301, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.795023202896118, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7132528424263, + "num_tokens": 238555269.0, + "step": 9218 + }, + { + "epoch": 1.0124094003953437, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.219905376434326, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6952834129333496, + "num_tokens": 238588256.0, + "step": 9219 + }, + { + "epoch": 1.0125192180979574, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.243232250213623, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6934914588928223, + "num_tokens": 238622800.0, + "step": 9220 + }, + { + "epoch": 1.012629035800571, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.742992639541626, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7151431441307068, + "num_tokens": 238645351.0, + "step": 9221 + }, + { + "epoch": 1.0127388535031847, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6167378425598145, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6942615509033203, + "num_tokens": 238668786.0, + "step": 9222 + }, + { + "epoch": 1.0128486712057985, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4705896377563477, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7180283665657043, + "num_tokens": 238694597.0, + "step": 9223 + }, + { + "epoch": 1.012958488908412, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.719722270965576, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7176133990287781, + "num_tokens": 238719112.0, + "step": 9224 + }, + { + "epoch": 1.0130683066110258, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.1798813343048096, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7103108763694763, + "num_tokens": 238751808.0, + "step": 9225 + }, + { + "epoch": 1.0131781243136393, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 3.316101551055908, + "learning_rate": 1e-06, + "loss": 0.9555, + "mean_token_accuracy": 0.7133285999298096, + "num_tokens": 238776031.0, + "step": 9226 + }, + { + "epoch": 1.013287942016253, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8351802825927734, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7013604640960693, + "num_tokens": 238796867.0, + "step": 9227 + }, + { + "epoch": 1.0133977597188666, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.9874000549316406, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.6991028785705566, + "num_tokens": 238822315.0, + "step": 9228 + }, + { + "epoch": 1.0135075774214803, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.7454333305358887, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.700084924697876, + "num_tokens": 238847142.0, + "step": 9229 + }, + { + "epoch": 1.013617395124094, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.627149820327759, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6990561485290527, + "num_tokens": 238877845.0, + "step": 9230 + }, + { + "epoch": 1.0137272128267076, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4282093048095703, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6880751848220825, + "num_tokens": 238904482.0, + "step": 9231 + }, + { + "epoch": 1.0138370305293214, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.585160255432129, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6945425868034363, + "num_tokens": 238928874.0, + "step": 9232 + }, + { + "epoch": 1.013946848231935, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.7250094413757324, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6966300010681152, + "num_tokens": 238955467.0, + "step": 9233 + }, + { + "epoch": 1.0140566659345487, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6557254791259766, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7265546321868896, + "num_tokens": 238977613.0, + "step": 9234 + }, + { + "epoch": 1.0141664836371622, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.637439250946045, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6955525875091553, + "num_tokens": 239002498.0, + "step": 9235 + }, + { + "epoch": 1.014276301339776, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.248800277709961, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.6920284628868103, + "num_tokens": 239038191.0, + "step": 9236 + }, + { + "epoch": 1.0143861190423897, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.55615234375, + "learning_rate": 1e-06, + "loss": 1.1131, + "mean_token_accuracy": 0.6785634756088257, + "num_tokens": 239067476.0, + "step": 9237 + }, + { + "epoch": 1.0144959367450033, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.618636131286621, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7113521695137024, + "num_tokens": 239094249.0, + "step": 9238 + }, + { + "epoch": 1.014605754447617, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.57869029045105, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7040724754333496, + "num_tokens": 239119765.0, + "step": 9239 + }, + { + "epoch": 1.0147155721502306, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.679612636566162, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.703458845615387, + "num_tokens": 239147445.0, + "step": 9240 + }, + { + "epoch": 1.0148253898528443, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8744194507598877, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7183418273925781, + "num_tokens": 239169710.0, + "step": 9241 + }, + { + "epoch": 1.0149352075554579, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.746922492980957, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7205579280853271, + "num_tokens": 239191689.0, + "step": 9242 + }, + { + "epoch": 1.0150450252580716, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4112157821655273, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6888788938522339, + "num_tokens": 239220203.0, + "step": 9243 + }, + { + "epoch": 1.0151548429606854, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5379815101623535, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.728290855884552, + "num_tokens": 239247003.0, + "step": 9244 + }, + { + "epoch": 1.015264660663299, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.7506816387176514, + "learning_rate": 1e-06, + "loss": 0.9299, + "mean_token_accuracy": 0.7229015827178955, + "num_tokens": 239266673.0, + "step": 9245 + }, + { + "epoch": 1.0153744783659127, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.479522228240967, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6977411508560181, + "num_tokens": 239291722.0, + "step": 9246 + }, + { + "epoch": 1.0154842960685262, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.701082944869995, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7113086581230164, + "num_tokens": 239315226.0, + "step": 9247 + }, + { + "epoch": 1.01559411377114, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6626930236816406, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7156071662902832, + "num_tokens": 239341129.0, + "step": 9248 + }, + { + "epoch": 1.0157039314737535, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5602760314941406, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7069435119628906, + "num_tokens": 239366717.0, + "step": 9249 + }, + { + "epoch": 1.0158137491763672, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.496872663497925, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7030002474784851, + "num_tokens": 239392989.0, + "step": 9250 + }, + { + "epoch": 1.015923566878981, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 3.2656524181365967, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6944388747215271, + "num_tokens": 239424703.0, + "step": 9251 + }, + { + "epoch": 1.0160333845815945, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.615335464477539, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6808422207832336, + "num_tokens": 239454364.0, + "step": 9252 + }, + { + "epoch": 1.0161432022842083, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5261449813842773, + "learning_rate": 1e-06, + "loss": 1.0903, + "mean_token_accuracy": 0.6833829879760742, + "num_tokens": 239482975.0, + "step": 9253 + }, + { + "epoch": 1.0162530199868218, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.42914080619812, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.6847073435783386, + "num_tokens": 239515619.0, + "step": 9254 + }, + { + "epoch": 1.0163628376894356, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.702660322189331, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6959878206253052, + "num_tokens": 239537698.0, + "step": 9255 + }, + { + "epoch": 1.0164726553920491, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6375269889831543, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7117549180984497, + "num_tokens": 239562114.0, + "step": 9256 + }, + { + "epoch": 1.0165824730946629, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.430455446243286, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6870823502540588, + "num_tokens": 239590906.0, + "step": 9257 + }, + { + "epoch": 1.0166922907972764, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5637011528015137, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.724394679069519, + "num_tokens": 239615166.0, + "step": 9258 + }, + { + "epoch": 1.0168021084998902, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.424384117126465, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7011590003967285, + "num_tokens": 239641273.0, + "step": 9259 + }, + { + "epoch": 1.016911926202504, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.506880760192871, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.6988356113433838, + "num_tokens": 239669932.0, + "step": 9260 + }, + { + "epoch": 1.0170217439051175, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8974006175994873, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7123868465423584, + "num_tokens": 239691431.0, + "step": 9261 + }, + { + "epoch": 1.0171315616077312, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 3.002899169921875, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7005685567855835, + "num_tokens": 239714438.0, + "step": 9262 + }, + { + "epoch": 1.0172413793103448, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.387660026550293, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7061599493026733, + "num_tokens": 239741908.0, + "step": 9263 + }, + { + "epoch": 1.0173511970129585, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4516561031341553, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.7011966705322266, + "num_tokens": 239772040.0, + "step": 9264 + }, + { + "epoch": 1.017461014715572, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5051939487457275, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.723910927772522, + "num_tokens": 239798973.0, + "step": 9265 + }, + { + "epoch": 1.0175708324181858, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5613889694213867, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7160570025444031, + "num_tokens": 239826141.0, + "step": 9266 + }, + { + "epoch": 1.0176806501207996, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6440980434417725, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6914621591567993, + "num_tokens": 239852736.0, + "step": 9267 + }, + { + "epoch": 1.017790467823413, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8537681102752686, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.725949764251709, + "num_tokens": 239872701.0, + "step": 9268 + }, + { + "epoch": 1.0179002855260268, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4103806018829346, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6932566165924072, + "num_tokens": 239904145.0, + "step": 9269 + }, + { + "epoch": 1.0180101032286404, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5991358757019043, + "learning_rate": 1e-06, + "loss": 0.8714, + "mean_token_accuracy": 0.7363591194152832, + "num_tokens": 239928765.0, + "step": 9270 + }, + { + "epoch": 1.0181199209312541, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.674899101257324, + "learning_rate": 1e-06, + "loss": 1.0989, + "mean_token_accuracy": 0.6851342916488647, + "num_tokens": 239953182.0, + "step": 9271 + }, + { + "epoch": 1.0182297386338677, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.815295696258545, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7057808041572571, + "num_tokens": 239975609.0, + "step": 9272 + }, + { + "epoch": 1.0183395563364814, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.328425168991089, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7133455276489258, + "num_tokens": 240006532.0, + "step": 9273 + }, + { + "epoch": 1.0184493740390952, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.31561541557312, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7183834314346313, + "num_tokens": 240037463.0, + "step": 9274 + }, + { + "epoch": 1.0185591917417087, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.56841778755188, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6873657703399658, + "num_tokens": 240068775.0, + "step": 9275 + }, + { + "epoch": 1.0186690094443225, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.321892499923706, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6949768662452698, + "num_tokens": 240098457.0, + "step": 9276 + }, + { + "epoch": 1.018778827146936, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.480137825012207, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6958465576171875, + "num_tokens": 240126114.0, + "step": 9277 + }, + { + "epoch": 1.0188886448495498, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 3.2326149940490723, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7162259221076965, + "num_tokens": 240149000.0, + "step": 9278 + }, + { + "epoch": 1.0189984625521633, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.336754560470581, + "learning_rate": 1e-06, + "loss": 1.0959, + "mean_token_accuracy": 0.6799222230911255, + "num_tokens": 240184417.0, + "step": 9279 + }, + { + "epoch": 1.019108280254777, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 80.5406265258789, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6987904906272888, + "num_tokens": 240213251.0, + "step": 9280 + }, + { + "epoch": 1.0192180979573908, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.802518606185913, + "learning_rate": 1e-06, + "loss": 1.1088, + "mean_token_accuracy": 0.6787576675415039, + "num_tokens": 240243902.0, + "step": 9281 + }, + { + "epoch": 1.0193279156600044, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.666311264038086, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7082971334457397, + "num_tokens": 240271686.0, + "step": 9282 + }, + { + "epoch": 1.0194377333626181, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.550830364227295, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6892967224121094, + "num_tokens": 240298105.0, + "step": 9283 + }, + { + "epoch": 1.0195475510652316, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.634335517883301, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7110274434089661, + "num_tokens": 240323329.0, + "step": 9284 + }, + { + "epoch": 1.0196573687678454, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5821168422698975, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7024513483047485, + "num_tokens": 240348312.0, + "step": 9285 + }, + { + "epoch": 1.019767186470459, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.501190662384033, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7025573253631592, + "num_tokens": 240373960.0, + "step": 9286 + }, + { + "epoch": 1.0198770041730727, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.7496743202209473, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.7123627662658691, + "num_tokens": 240395472.0, + "step": 9287 + }, + { + "epoch": 1.0199868218756865, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4230873584747314, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6898316144943237, + "num_tokens": 240427402.0, + "step": 9288 + }, + { + "epoch": 1.0200966395783, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.2161383628845215, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7151569724082947, + "num_tokens": 240461485.0, + "step": 9289 + }, + { + "epoch": 1.0202064572809137, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6387131214141846, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7093642950057983, + "num_tokens": 240488154.0, + "step": 9290 + }, + { + "epoch": 1.0203162749835273, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5492451190948486, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7086516618728638, + "num_tokens": 240515041.0, + "step": 9291 + }, + { + "epoch": 1.020426092686141, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5101265907287598, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7111479043960571, + "num_tokens": 240543212.0, + "step": 9292 + }, + { + "epoch": 1.0205359103887546, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4637088775634766, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7155736088752747, + "num_tokens": 240577429.0, + "step": 9293 + }, + { + "epoch": 1.0206457280913683, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.564614772796631, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.6996300220489502, + "num_tokens": 240606022.0, + "step": 9294 + }, + { + "epoch": 1.020755545793982, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.7178115844726562, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.6978853344917297, + "num_tokens": 240627360.0, + "step": 9295 + }, + { + "epoch": 1.0208653634965956, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4830353260040283, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6951159238815308, + "num_tokens": 240655953.0, + "step": 9296 + }, + { + "epoch": 1.0209751811992094, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5297350883483887, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.70756995677948, + "num_tokens": 240681901.0, + "step": 9297 + }, + { + "epoch": 1.021084998901823, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.534512996673584, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7000285983085632, + "num_tokens": 240708896.0, + "step": 9298 + }, + { + "epoch": 1.0211948166044367, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.570974826812744, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7201510071754456, + "num_tokens": 240734080.0, + "step": 9299 + }, + { + "epoch": 1.0213046343070502, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.772608995437622, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.6968571543693542, + "num_tokens": 240757776.0, + "step": 9300 + }, + { + "epoch": 1.021414452009664, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.9053711891174316, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7083410024642944, + "num_tokens": 240779967.0, + "step": 9301 + }, + { + "epoch": 1.0215242697122777, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.575593948364258, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.697573184967041, + "num_tokens": 240811222.0, + "step": 9302 + }, + { + "epoch": 1.0216340874148913, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5847771167755127, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7027367353439331, + "num_tokens": 240839514.0, + "step": 9303 + }, + { + "epoch": 1.021743905117505, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.3679683208465576, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6876683831214905, + "num_tokens": 240869492.0, + "step": 9304 + }, + { + "epoch": 1.0218537228201185, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8416037559509277, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6862489581108093, + "num_tokens": 240903330.0, + "step": 9305 + }, + { + "epoch": 1.0219635405227323, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4625415802001953, + "learning_rate": 1e-06, + "loss": 0.8875, + "mean_token_accuracy": 0.7383625507354736, + "num_tokens": 240931670.0, + "step": 9306 + }, + { + "epoch": 1.0220733582253458, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5977487564086914, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7230097055435181, + "num_tokens": 240955765.0, + "step": 9307 + }, + { + "epoch": 1.0221831759279596, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4515933990478516, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7088181972503662, + "num_tokens": 240984698.0, + "step": 9308 + }, + { + "epoch": 1.0222929936305734, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6790246963500977, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7041392922401428, + "num_tokens": 241011486.0, + "step": 9309 + }, + { + "epoch": 1.022402811333187, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.63873553276062, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6953161954879761, + "num_tokens": 241038036.0, + "step": 9310 + }, + { + "epoch": 1.0225126290358006, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.616889238357544, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7105993628501892, + "num_tokens": 241062846.0, + "step": 9311 + }, + { + "epoch": 1.0226224467384142, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.6715779304504395, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7169821858406067, + "num_tokens": 241089714.0, + "step": 9312 + }, + { + "epoch": 1.022732264441028, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.328763008117676, + "learning_rate": 1e-06, + "loss": 1.0753, + "mean_token_accuracy": 0.689636766910553, + "num_tokens": 241122272.0, + "step": 9313 + }, + { + "epoch": 1.0228420821436415, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.779064893722534, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7128951549530029, + "num_tokens": 241143105.0, + "step": 9314 + }, + { + "epoch": 1.0229518998462552, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5682027339935303, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6938862204551697, + "num_tokens": 241170861.0, + "step": 9315 + }, + { + "epoch": 1.023061717548869, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8216726779937744, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7128028869628906, + "num_tokens": 241193185.0, + "step": 9316 + }, + { + "epoch": 1.0231715352514825, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.427197217941284, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7088777422904968, + "num_tokens": 241222262.0, + "step": 9317 + }, + { + "epoch": 1.0232813529540963, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.748990297317505, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.6984169483184814, + "num_tokens": 241245067.0, + "step": 9318 + }, + { + "epoch": 1.0233911706567098, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5245606899261475, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7049679756164551, + "num_tokens": 241272467.0, + "step": 9319 + }, + { + "epoch": 1.0235009883593236, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.5559747219085693, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7045019268989563, + "num_tokens": 241297534.0, + "step": 9320 + }, + { + "epoch": 1.023610806061937, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.4055569171905518, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.703552782535553, + "num_tokens": 241327611.0, + "step": 9321 + }, + { + "epoch": 1.0237206237645509, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.413984537124634, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.6996774673461914, + "num_tokens": 241356755.0, + "step": 9322 + }, + { + "epoch": 1.0238304414671644, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.715402126312256, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.709556519985199, + "num_tokens": 241378369.0, + "step": 9323 + }, + { + "epoch": 1.0239402591697782, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.8470234870910645, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7067469358444214, + "num_tokens": 241404176.0, + "step": 9324 + }, + { + "epoch": 1.024050076872392, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.653294324874878, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7124407291412354, + "num_tokens": 241429475.0, + "step": 9325 + }, + { + "epoch": 1.0241598945750054, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.853264570236206, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7030950784683228, + "num_tokens": 241450642.0, + "step": 9326 + }, + { + "epoch": 1.0242697122776192, + "ewc_loss": 8.64267349243164e-06, + "grad_norm": 2.384910821914673, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7240326404571533, + "num_tokens": 241480796.0, + "step": 9327 + }, + { + "epoch": 1.0243795299802327, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 3.1298656463623047, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.7196589708328247, + "num_tokens": 241498114.0, + "step": 9328 + }, + { + "epoch": 1.0244893476828465, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7300312519073486, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7216897010803223, + "num_tokens": 241519122.0, + "step": 9329 + }, + { + "epoch": 1.02459916538546, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.416530132293701, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6976289749145508, + "num_tokens": 241550097.0, + "step": 9330 + }, + { + "epoch": 1.0247089830880738, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5078518390655518, + "learning_rate": 1e-06, + "loss": 1.093, + "mean_token_accuracy": 0.6820309162139893, + "num_tokens": 241578237.0, + "step": 9331 + }, + { + "epoch": 1.0248188007906875, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7415053844451904, + "learning_rate": 1e-06, + "loss": 0.9069, + "mean_token_accuracy": 0.7278949022293091, + "num_tokens": 241599538.0, + "step": 9332 + }, + { + "epoch": 1.024928618493301, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7857866287231445, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7123048901557922, + "num_tokens": 241623433.0, + "step": 9333 + }, + { + "epoch": 1.0250384361959148, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6006312370300293, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7090921998023987, + "num_tokens": 241647005.0, + "step": 9334 + }, + { + "epoch": 1.0251482538985284, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.4907734394073486, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6944632530212402, + "num_tokens": 241676177.0, + "step": 9335 + }, + { + "epoch": 1.0252580716011421, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.3801982402801514, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7139304876327515, + "num_tokens": 241705183.0, + "step": 9336 + }, + { + "epoch": 1.0253678893037557, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.8288700580596924, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.6961815357208252, + "num_tokens": 241728831.0, + "step": 9337 + }, + { + "epoch": 1.0254777070063694, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.339590072631836, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7200304269790649, + "num_tokens": 241757350.0, + "step": 9338 + }, + { + "epoch": 1.0255875247089832, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 3.3812930583953857, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6978240013122559, + "num_tokens": 241792208.0, + "step": 9339 + }, + { + "epoch": 1.0256973424115967, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5432522296905518, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7132179737091064, + "num_tokens": 241820190.0, + "step": 9340 + }, + { + "epoch": 1.0258071601142105, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5107901096343994, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7090665698051453, + "num_tokens": 241844640.0, + "step": 9341 + }, + { + "epoch": 1.025916977816824, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.700287342071533, + "learning_rate": 1e-06, + "loss": 0.9353, + "mean_token_accuracy": 0.7225502133369446, + "num_tokens": 241866014.0, + "step": 9342 + }, + { + "epoch": 1.0260267955194378, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.3750483989715576, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6823297739028931, + "num_tokens": 241896370.0, + "step": 9343 + }, + { + "epoch": 1.0261366132220513, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.479012966156006, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6955105066299438, + "num_tokens": 241921445.0, + "step": 9344 + }, + { + "epoch": 1.026246430924665, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.841165542602539, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.715241551399231, + "num_tokens": 241942105.0, + "step": 9345 + }, + { + "epoch": 1.0263562486272788, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.412980318069458, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.689354419708252, + "num_tokens": 241974155.0, + "step": 9346 + }, + { + "epoch": 1.0264660663298923, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 3.4599525928497314, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7031710147857666, + "num_tokens": 242000788.0, + "step": 9347 + }, + { + "epoch": 1.026575884032506, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7331643104553223, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.713236927986145, + "num_tokens": 242028371.0, + "step": 9348 + }, + { + "epoch": 1.0266857017351196, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6637589931488037, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7154453992843628, + "num_tokens": 242056593.0, + "step": 9349 + }, + { + "epoch": 1.0267955194377334, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.535154104232788, + "learning_rate": 1e-06, + "loss": 0.9254, + "mean_token_accuracy": 0.7298099994659424, + "num_tokens": 242083641.0, + "step": 9350 + }, + { + "epoch": 1.026905337140347, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5211145877838135, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.720598578453064, + "num_tokens": 242109187.0, + "step": 9351 + }, + { + "epoch": 1.0270151548429607, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7523016929626465, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7062892913818359, + "num_tokens": 242131308.0, + "step": 9352 + }, + { + "epoch": 1.0271249725455744, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5801992416381836, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.7035238146781921, + "num_tokens": 242158139.0, + "step": 9353 + }, + { + "epoch": 1.027234790248188, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.524829149246216, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.6966732144355774, + "num_tokens": 242185517.0, + "step": 9354 + }, + { + "epoch": 1.0273446079508017, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.60131573677063, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7039816975593567, + "num_tokens": 242212944.0, + "step": 9355 + }, + { + "epoch": 1.0274544256534153, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.514340877532959, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7066928148269653, + "num_tokens": 242237921.0, + "step": 9356 + }, + { + "epoch": 1.027564243356029, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5764682292938232, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6951391100883484, + "num_tokens": 242264084.0, + "step": 9357 + }, + { + "epoch": 1.0276740610586426, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.682234764099121, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.6846003532409668, + "num_tokens": 242289515.0, + "step": 9358 + }, + { + "epoch": 1.0277838787612563, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.657193899154663, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7207481861114502, + "num_tokens": 242312676.0, + "step": 9359 + }, + { + "epoch": 1.02789369646387, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.516178607940674, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6955803632736206, + "num_tokens": 242341108.0, + "step": 9360 + }, + { + "epoch": 1.0280035141664836, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.720708131790161, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.704828143119812, + "num_tokens": 242364261.0, + "step": 9361 + }, + { + "epoch": 1.0281133318690974, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.73209547996521, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7112423777580261, + "num_tokens": 242385200.0, + "step": 9362 + }, + { + "epoch": 1.028223149571711, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.2698118686676025, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6916154026985168, + "num_tokens": 242421340.0, + "step": 9363 + }, + { + "epoch": 1.0283329672743247, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 20.531108856201172, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.693212628364563, + "num_tokens": 242449861.0, + "step": 9364 + }, + { + "epoch": 1.0284427849769382, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6397459506988525, + "learning_rate": 1e-06, + "loss": 0.9264, + "mean_token_accuracy": 0.72459876537323, + "num_tokens": 242476627.0, + "step": 9365 + }, + { + "epoch": 1.028552602679552, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.722705364227295, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.6979719400405884, + "num_tokens": 242502244.0, + "step": 9366 + }, + { + "epoch": 1.0286624203821657, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.8064401149749756, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7009512782096863, + "num_tokens": 242528568.0, + "step": 9367 + }, + { + "epoch": 1.0287722380847792, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.4820258617401123, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7228847742080688, + "num_tokens": 242554618.0, + "step": 9368 + }, + { + "epoch": 1.028882055787393, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6942546367645264, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7049319744110107, + "num_tokens": 242576525.0, + "step": 9369 + }, + { + "epoch": 1.0289918734900065, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.760288953781128, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6882368326187134, + "num_tokens": 242604868.0, + "step": 9370 + }, + { + "epoch": 1.0291016911926203, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.619985342025757, + "learning_rate": 1e-06, + "loss": 0.9267, + "mean_token_accuracy": 0.7141355276107788, + "num_tokens": 242627934.0, + "step": 9371 + }, + { + "epoch": 1.0292115088952338, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.814854145050049, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7163550853729248, + "num_tokens": 242648332.0, + "step": 9372 + }, + { + "epoch": 1.0293213265978476, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.513414144515991, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7131476402282715, + "num_tokens": 242675591.0, + "step": 9373 + }, + { + "epoch": 1.0294311443004613, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7416579723358154, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6987096667289734, + "num_tokens": 242699493.0, + "step": 9374 + }, + { + "epoch": 1.0295409620030749, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.494260311126709, + "learning_rate": 1e-06, + "loss": 1.1195, + "mean_token_accuracy": 0.6721985340118408, + "num_tokens": 242726766.0, + "step": 9375 + }, + { + "epoch": 1.0296507797056886, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.19758677482605, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.682970404624939, + "num_tokens": 242761048.0, + "step": 9376 + }, + { + "epoch": 1.0297605974083022, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.8616182804107666, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6882184743881226, + "num_tokens": 242785961.0, + "step": 9377 + }, + { + "epoch": 1.029870415110916, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5700738430023193, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.7004491090774536, + "num_tokens": 242813003.0, + "step": 9378 + }, + { + "epoch": 1.0299802328135295, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.3889520168304443, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.6868538856506348, + "num_tokens": 242844550.0, + "step": 9379 + }, + { + "epoch": 1.0300900505161432, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.7938966751098633, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6997530460357666, + "num_tokens": 242868020.0, + "step": 9380 + }, + { + "epoch": 1.030199868218757, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.8653066158294678, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6938989758491516, + "num_tokens": 242890243.0, + "step": 9381 + }, + { + "epoch": 1.0303096859213705, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.526287317276001, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7095459699630737, + "num_tokens": 242915586.0, + "step": 9382 + }, + { + "epoch": 1.0304195036239843, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.4826273918151855, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6975200176239014, + "num_tokens": 242943042.0, + "step": 9383 + }, + { + "epoch": 1.0305293213265978, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.535191059112549, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7097688913345337, + "num_tokens": 242968983.0, + "step": 9384 + }, + { + "epoch": 1.0306391390292116, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.451322555541992, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.709115207195282, + "num_tokens": 242999537.0, + "step": 9385 + }, + { + "epoch": 1.030748956731825, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.4816298484802246, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.69585120677948, + "num_tokens": 243029698.0, + "step": 9386 + }, + { + "epoch": 1.0308587744344389, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6675825119018555, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7115679979324341, + "num_tokens": 243057058.0, + "step": 9387 + }, + { + "epoch": 1.0309685921370524, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.8071281909942627, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7097693085670471, + "num_tokens": 243083376.0, + "step": 9388 + }, + { + "epoch": 1.0310784098396661, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.711005687713623, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7051522731781006, + "num_tokens": 243106844.0, + "step": 9389 + }, + { + "epoch": 1.03118822754228, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.3906099796295166, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.7010265588760376, + "num_tokens": 243139308.0, + "step": 9390 + }, + { + "epoch": 1.0312980452448934, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5270071029663086, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.6966975927352905, + "num_tokens": 243165509.0, + "step": 9391 + }, + { + "epoch": 1.0314078629475072, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.633017063140869, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7159867882728577, + "num_tokens": 243187638.0, + "step": 9392 + }, + { + "epoch": 1.0315176806501207, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.854278564453125, + "learning_rate": 1e-06, + "loss": 0.9454, + "mean_token_accuracy": 0.7176925539970398, + "num_tokens": 243207999.0, + "step": 9393 + }, + { + "epoch": 1.0316274983527345, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5487382411956787, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6960539221763611, + "num_tokens": 243236555.0, + "step": 9394 + }, + { + "epoch": 1.031737316055348, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.663104295730591, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7244709134101868, + "num_tokens": 243262821.0, + "step": 9395 + }, + { + "epoch": 1.0318471337579618, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.563922882080078, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6979860067367554, + "num_tokens": 243292156.0, + "step": 9396 + }, + { + "epoch": 1.0319569514605755, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6123669147491455, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.700953483581543, + "num_tokens": 243317821.0, + "step": 9397 + }, + { + "epoch": 1.032066769163189, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.587334156036377, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6945843696594238, + "num_tokens": 243345859.0, + "step": 9398 + }, + { + "epoch": 1.0321765868658028, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.649355411529541, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.694961428642273, + "num_tokens": 243372271.0, + "step": 9399 + }, + { + "epoch": 1.0322864045684164, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.4478774070739746, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6827927827835083, + "num_tokens": 243399628.0, + "step": 9400 + }, + { + "epoch": 1.0323962222710301, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 3.074826955795288, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.7254928350448608, + "num_tokens": 243419002.0, + "step": 9401 + }, + { + "epoch": 1.0325060399736437, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6082186698913574, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6986963748931885, + "num_tokens": 243449824.0, + "step": 9402 + }, + { + "epoch": 1.0326158576762574, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.6911208629608154, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7022521495819092, + "num_tokens": 243474777.0, + "step": 9403 + }, + { + "epoch": 1.0327256753788712, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.2897732257843018, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6978201866149902, + "num_tokens": 243508762.0, + "step": 9404 + }, + { + "epoch": 1.0328354930814847, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.8497509956359863, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7234846353530884, + "num_tokens": 243528461.0, + "step": 9405 + }, + { + "epoch": 1.0329453107840985, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.5900070667266846, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.7021089196205139, + "num_tokens": 243556317.0, + "step": 9406 + }, + { + "epoch": 1.033055128486712, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.466275930404663, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7077407836914062, + "num_tokens": 243585250.0, + "step": 9407 + }, + { + "epoch": 1.0331649461893258, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.9352376461029053, + "learning_rate": 1e-06, + "loss": 0.9416, + "mean_token_accuracy": 0.7252576351165771, + "num_tokens": 243606557.0, + "step": 9408 + }, + { + "epoch": 1.0332747638919393, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.745496988296509, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.6989490985870361, + "num_tokens": 243632472.0, + "step": 9409 + }, + { + "epoch": 1.033384581594553, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.4259955883026123, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7152543067932129, + "num_tokens": 243660567.0, + "step": 9410 + }, + { + "epoch": 1.0334943992971668, + "ewc_loss": 8.702278137207031e-06, + "grad_norm": 2.9052629470825195, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7093380689620972, + "num_tokens": 243681101.0, + "step": 9411 + }, + { + "epoch": 1.0336042169997803, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.424833059310913, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6982727646827698, + "num_tokens": 243710120.0, + "step": 9412 + }, + { + "epoch": 1.033714034702394, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.863358736038208, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7234725952148438, + "num_tokens": 243731564.0, + "step": 9413 + }, + { + "epoch": 1.0338238524050076, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.8361611366271973, + "learning_rate": 1e-06, + "loss": 0.8917, + "mean_token_accuracy": 0.7367129325866699, + "num_tokens": 243752083.0, + "step": 9414 + }, + { + "epoch": 1.0339336701076214, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.57919979095459, + "learning_rate": 1e-06, + "loss": 0.8838, + "mean_token_accuracy": 0.7405416369438171, + "num_tokens": 243776098.0, + "step": 9415 + }, + { + "epoch": 1.034043487810235, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7599942684173584, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7240888476371765, + "num_tokens": 243798831.0, + "step": 9416 + }, + { + "epoch": 1.0341533055128487, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.456751585006714, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.6991718411445618, + "num_tokens": 243826736.0, + "step": 9417 + }, + { + "epoch": 1.0342631232154624, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 3.118307113647461, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7101550102233887, + "num_tokens": 243850406.0, + "step": 9418 + }, + { + "epoch": 1.034372940918076, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 3.117790699005127, + "learning_rate": 1e-06, + "loss": 0.8914, + "mean_token_accuracy": 0.7366858124732971, + "num_tokens": 243866665.0, + "step": 9419 + }, + { + "epoch": 1.0344827586206897, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.6979572772979736, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7109395265579224, + "num_tokens": 243891260.0, + "step": 9420 + }, + { + "epoch": 1.0345925763233033, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.63718843460083, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7131961584091187, + "num_tokens": 243912863.0, + "step": 9421 + }, + { + "epoch": 1.034702394025917, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.3874101638793945, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7110669016838074, + "num_tokens": 243941500.0, + "step": 9422 + }, + { + "epoch": 1.0348122117285306, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.637725591659546, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6914756298065186, + "num_tokens": 243965604.0, + "step": 9423 + }, + { + "epoch": 1.0349220294311443, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.6493449211120605, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7145180106163025, + "num_tokens": 243989569.0, + "step": 9424 + }, + { + "epoch": 1.035031847133758, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.605595588684082, + "learning_rate": 1e-06, + "loss": 0.9475, + "mean_token_accuracy": 0.7186610102653503, + "num_tokens": 244018389.0, + "step": 9425 + }, + { + "epoch": 1.0351416648363716, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7202823162078857, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7078142166137695, + "num_tokens": 244042053.0, + "step": 9426 + }, + { + "epoch": 1.0352514825389854, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4296212196350098, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7075719237327576, + "num_tokens": 244072302.0, + "step": 9427 + }, + { + "epoch": 1.035361300241599, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.8059229850769043, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6964691877365112, + "num_tokens": 244095758.0, + "step": 9428 + }, + { + "epoch": 1.0354711179442126, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5328850746154785, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7096336483955383, + "num_tokens": 244124923.0, + "step": 9429 + }, + { + "epoch": 1.0355809356468262, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5553762912750244, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7027485370635986, + "num_tokens": 244148357.0, + "step": 9430 + }, + { + "epoch": 1.03569075334944, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5023703575134277, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.708471417427063, + "num_tokens": 244175301.0, + "step": 9431 + }, + { + "epoch": 1.0358005710520537, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.572511911392212, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7099527716636658, + "num_tokens": 244201386.0, + "step": 9432 + }, + { + "epoch": 1.0359103887546672, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4923105239868164, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6980889439582825, + "num_tokens": 244228633.0, + "step": 9433 + }, + { + "epoch": 1.036020206457281, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5670840740203857, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.7031396627426147, + "num_tokens": 244256930.0, + "step": 9434 + }, + { + "epoch": 1.0361300241598945, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5809972286224365, + "learning_rate": 1e-06, + "loss": 0.881, + "mean_token_accuracy": 0.7333120107650757, + "num_tokens": 244278992.0, + "step": 9435 + }, + { + "epoch": 1.0362398418625083, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.705909490585327, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7051385045051575, + "num_tokens": 244307073.0, + "step": 9436 + }, + { + "epoch": 1.0363496595651218, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5752689838409424, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7026277780532837, + "num_tokens": 244333029.0, + "step": 9437 + }, + { + "epoch": 1.0364594772677356, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.741851568222046, + "learning_rate": 1e-06, + "loss": 0.9196, + "mean_token_accuracy": 0.72881019115448, + "num_tokens": 244354581.0, + "step": 9438 + }, + { + "epoch": 1.036569294970349, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.9639217853546143, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7040904760360718, + "num_tokens": 244376640.0, + "step": 9439 + }, + { + "epoch": 1.0366791126729629, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.53560471534729, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6856223940849304, + "num_tokens": 244403385.0, + "step": 9440 + }, + { + "epoch": 1.0367889303755766, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.568523645401001, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6952446699142456, + "num_tokens": 244431391.0, + "step": 9441 + }, + { + "epoch": 1.0368987480781902, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.771094799041748, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6988496780395508, + "num_tokens": 244458610.0, + "step": 9442 + }, + { + "epoch": 1.037008565780804, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.6453685760498047, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6951740980148315, + "num_tokens": 244490252.0, + "step": 9443 + }, + { + "epoch": 1.0371183834834174, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5231075286865234, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7176920175552368, + "num_tokens": 244516220.0, + "step": 9444 + }, + { + "epoch": 1.0372282011860312, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5530436038970947, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7073109149932861, + "num_tokens": 244544253.0, + "step": 9445 + }, + { + "epoch": 1.037338018888645, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7220919132232666, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.706502377986908, + "num_tokens": 244567126.0, + "step": 9446 + }, + { + "epoch": 1.0374478365912585, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.401811122894287, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7079118490219116, + "num_tokens": 244595282.0, + "step": 9447 + }, + { + "epoch": 1.0375576542938723, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.852734327316284, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.6899052858352661, + "num_tokens": 244617865.0, + "step": 9448 + }, + { + "epoch": 1.0376674719964858, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4190421104431152, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7040862441062927, + "num_tokens": 244645296.0, + "step": 9449 + }, + { + "epoch": 1.0377772896990995, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 3.034667730331421, + "learning_rate": 1e-06, + "loss": 0.9151, + "mean_token_accuracy": 0.7265640497207642, + "num_tokens": 244663788.0, + "step": 9450 + }, + { + "epoch": 1.037887107401713, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7142410278320312, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7056829333305359, + "num_tokens": 244685999.0, + "step": 9451 + }, + { + "epoch": 1.0379969251043268, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.64095139503479, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.6943832039833069, + "num_tokens": 244713331.0, + "step": 9452 + }, + { + "epoch": 1.0381067428069404, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.405704975128174, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7179228067398071, + "num_tokens": 244740153.0, + "step": 9453 + }, + { + "epoch": 1.0382165605095541, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7947373390197754, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7112246751785278, + "num_tokens": 244763602.0, + "step": 9454 + }, + { + "epoch": 1.0383263782121679, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5063939094543457, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7146180868148804, + "num_tokens": 244793724.0, + "step": 9455 + }, + { + "epoch": 1.0384361959147814, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5385775566101074, + "learning_rate": 1e-06, + "loss": 0.9404, + "mean_token_accuracy": 0.7290315628051758, + "num_tokens": 244818177.0, + "step": 9456 + }, + { + "epoch": 1.0385460136173952, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.3142387866973877, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7186224460601807, + "num_tokens": 244848557.0, + "step": 9457 + }, + { + "epoch": 1.0386558313200087, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5711629390716553, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7165291905403137, + "num_tokens": 244874269.0, + "step": 9458 + }, + { + "epoch": 1.0387656490226225, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.77724027633667, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7145557403564453, + "num_tokens": 244894552.0, + "step": 9459 + }, + { + "epoch": 1.038875466725236, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.715029239654541, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7220966219902039, + "num_tokens": 244919365.0, + "step": 9460 + }, + { + "epoch": 1.0389852844278498, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.589574098587036, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7012566924095154, + "num_tokens": 244943450.0, + "step": 9461 + }, + { + "epoch": 1.0390951021304635, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7630062103271484, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7054919004440308, + "num_tokens": 244966856.0, + "step": 9462 + }, + { + "epoch": 1.039204919833077, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.335249900817871, + "learning_rate": 1e-06, + "loss": 1.1166, + "mean_token_accuracy": 0.6843622922897339, + "num_tokens": 244998059.0, + "step": 9463 + }, + { + "epoch": 1.0393147375356908, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5817489624023438, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6983237266540527, + "num_tokens": 245022171.0, + "step": 9464 + }, + { + "epoch": 1.0394245552383043, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.6772677898406982, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7182583808898926, + "num_tokens": 245049263.0, + "step": 9465 + }, + { + "epoch": 1.039534372940918, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.635617733001709, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7289080023765564, + "num_tokens": 245073793.0, + "step": 9466 + }, + { + "epoch": 1.0396441906435316, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.3490617275238037, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7081241607666016, + "num_tokens": 245103029.0, + "step": 9467 + }, + { + "epoch": 1.0397540083461454, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.683133363723755, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7034175395965576, + "num_tokens": 245128495.0, + "step": 9468 + }, + { + "epoch": 1.0398638260487592, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5136518478393555, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7138255834579468, + "num_tokens": 245156033.0, + "step": 9469 + }, + { + "epoch": 1.0399736437513727, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.9801185131073, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7264401912689209, + "num_tokens": 245174744.0, + "step": 9470 + }, + { + "epoch": 1.0400834614539864, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.6296730041503906, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7014397382736206, + "num_tokens": 245200011.0, + "step": 9471 + }, + { + "epoch": 1.0401932791566, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.4205901622772217, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7182316184043884, + "num_tokens": 245227165.0, + "step": 9472 + }, + { + "epoch": 1.0403030968592137, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4748952388763428, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6992548704147339, + "num_tokens": 245257985.0, + "step": 9473 + }, + { + "epoch": 1.0404129145618273, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7250027656555176, + "learning_rate": 1e-06, + "loss": 0.8805, + "mean_token_accuracy": 0.7399650812149048, + "num_tokens": 245279876.0, + "step": 9474 + }, + { + "epoch": 1.040522732264441, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.7161545753479004, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6930235624313354, + "num_tokens": 245305272.0, + "step": 9475 + }, + { + "epoch": 1.0406325499670548, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.614527463912964, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.704373836517334, + "num_tokens": 245328051.0, + "step": 9476 + }, + { + "epoch": 1.0407423676696683, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.404632091522217, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.7032246589660645, + "num_tokens": 245358157.0, + "step": 9477 + }, + { + "epoch": 1.040852185372282, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4807522296905518, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6960514783859253, + "num_tokens": 245384741.0, + "step": 9478 + }, + { + "epoch": 1.0409620030748956, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.623032331466675, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6965134143829346, + "num_tokens": 245409463.0, + "step": 9479 + }, + { + "epoch": 1.0410718207775094, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.5677947998046875, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7233759760856628, + "num_tokens": 245438723.0, + "step": 9480 + }, + { + "epoch": 1.041181638480123, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.88137149810791, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7209092378616333, + "num_tokens": 245458902.0, + "step": 9481 + }, + { + "epoch": 1.0412914561827367, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4621052742004395, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6939899325370789, + "num_tokens": 245485163.0, + "step": 9482 + }, + { + "epoch": 1.0414012738853504, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.4654204845428467, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7195490598678589, + "num_tokens": 245511526.0, + "step": 9483 + }, + { + "epoch": 1.041511091587964, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.794851064682007, + "learning_rate": 1e-06, + "loss": 0.95, + "mean_token_accuracy": 0.7137377262115479, + "num_tokens": 245533940.0, + "step": 9484 + }, + { + "epoch": 1.0416209092905777, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.860424757003784, + "learning_rate": 1e-06, + "loss": 0.8642, + "mean_token_accuracy": 0.7376213073730469, + "num_tokens": 245553366.0, + "step": 9485 + }, + { + "epoch": 1.0417307269931912, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.860466480255127, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7220311164855957, + "num_tokens": 245573331.0, + "step": 9486 + }, + { + "epoch": 1.041840544695805, + "ewc_loss": 8.761882781982422e-06, + "grad_norm": 2.987109899520874, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7289833426475525, + "num_tokens": 245595441.0, + "step": 9487 + }, + { + "epoch": 1.0419503623984185, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.81503963470459, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7162501811981201, + "num_tokens": 245616215.0, + "step": 9488 + }, + { + "epoch": 1.0420601801010323, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.7458441257476807, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7007151246070862, + "num_tokens": 245641474.0, + "step": 9489 + }, + { + "epoch": 1.042169997803646, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.690011739730835, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7011226415634155, + "num_tokens": 245666626.0, + "step": 9490 + }, + { + "epoch": 1.0422798155062596, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.585209369659424, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7089830636978149, + "num_tokens": 245693609.0, + "step": 9491 + }, + { + "epoch": 1.0423896332088733, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.848222255706787, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7064985036849976, + "num_tokens": 245715909.0, + "step": 9492 + }, + { + "epoch": 1.0424994509114869, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.673348903656006, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7137041687965393, + "num_tokens": 245743609.0, + "step": 9493 + }, + { + "epoch": 1.0426092686141006, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.833275556564331, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.7074673175811768, + "num_tokens": 245765807.0, + "step": 9494 + }, + { + "epoch": 1.0427190863167142, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3878278732299805, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6883199214935303, + "num_tokens": 245798404.0, + "step": 9495 + }, + { + "epoch": 1.042828904019328, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.8279178142547607, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.7009402513504028, + "num_tokens": 245820725.0, + "step": 9496 + }, + { + "epoch": 1.0429387217219417, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.644052267074585, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6988949775695801, + "num_tokens": 245844551.0, + "step": 9497 + }, + { + "epoch": 1.0430485394245552, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.525352954864502, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7238794565200806, + "num_tokens": 245870890.0, + "step": 9498 + }, + { + "epoch": 1.043158357127169, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.681370258331299, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.6984779834747314, + "num_tokens": 245896044.0, + "step": 9499 + }, + { + "epoch": 1.0432681748297825, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5117712020874023, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6875686049461365, + "num_tokens": 245923223.0, + "step": 9500 + }, + { + "epoch": 1.0433779925323963, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5237972736358643, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6972631216049194, + "num_tokens": 245950175.0, + "step": 9501 + }, + { + "epoch": 1.0434878102350098, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6068882942199707, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7145912051200867, + "num_tokens": 245974500.0, + "step": 9502 + }, + { + "epoch": 1.0435976279376236, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.395009994506836, + "learning_rate": 1e-06, + "loss": 1.0841, + "mean_token_accuracy": 0.6907899379730225, + "num_tokens": 246005891.0, + "step": 9503 + }, + { + "epoch": 1.043707445640237, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.7051732540130615, + "learning_rate": 1e-06, + "loss": 1.1058, + "mean_token_accuracy": 0.678139865398407, + "num_tokens": 246030961.0, + "step": 9504 + }, + { + "epoch": 1.0438172633428509, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6884191036224365, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6999038457870483, + "num_tokens": 246054915.0, + "step": 9505 + }, + { + "epoch": 1.0439270810454646, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.524657964706421, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7159501314163208, + "num_tokens": 246081890.0, + "step": 9506 + }, + { + "epoch": 1.0440368987480781, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.4757754802703857, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.708400547504425, + "num_tokens": 246110456.0, + "step": 9507 + }, + { + "epoch": 1.044146716450692, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3547186851501465, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6879585385322571, + "num_tokens": 246139574.0, + "step": 9508 + }, + { + "epoch": 1.0442565341533054, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 16.69608497619629, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6935572624206543, + "num_tokens": 246165445.0, + "step": 9509 + }, + { + "epoch": 1.0443663518559192, + "ewc_loss": 8.940696716308594e-06, + "grad_norm": 2.9134151935577393, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7182474136352539, + "num_tokens": 246185870.0, + "step": 9510 + }, + { + "epoch": 1.0444761695585327, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3224167823791504, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6881712675094604, + "num_tokens": 246216822.0, + "step": 9511 + }, + { + "epoch": 1.0445859872611465, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7355568408966064, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7183545827865601, + "num_tokens": 246239351.0, + "step": 9512 + }, + { + "epoch": 1.0446958049637602, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.7207329273223877, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7220906615257263, + "num_tokens": 246262271.0, + "step": 9513 + }, + { + "epoch": 1.0448056226663738, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.7467243671417236, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7068777680397034, + "num_tokens": 246286562.0, + "step": 9514 + }, + { + "epoch": 1.0449154403689875, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.6434736251831055, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.7011839747428894, + "num_tokens": 246310265.0, + "step": 9515 + }, + { + "epoch": 1.045025258071601, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.774764060974121, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6913347244262695, + "num_tokens": 246335195.0, + "step": 9516 + }, + { + "epoch": 1.0451350757742148, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.6981661319732666, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7149778604507446, + "num_tokens": 246360099.0, + "step": 9517 + }, + { + "epoch": 1.0452448934768284, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.6015241146087646, + "learning_rate": 1e-06, + "loss": 1.1198, + "mean_token_accuracy": 0.6764643788337708, + "num_tokens": 246385857.0, + "step": 9518 + }, + { + "epoch": 1.0453547111794421, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.4456663131713867, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7272678017616272, + "num_tokens": 246411918.0, + "step": 9519 + }, + { + "epoch": 1.0454645288820559, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.7245073318481445, + "learning_rate": 1e-06, + "loss": 0.9107, + "mean_token_accuracy": 0.7296228408813477, + "num_tokens": 246433859.0, + "step": 9520 + }, + { + "epoch": 1.0455743465846694, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.825122594833374, + "learning_rate": 1e-06, + "loss": 0.9431, + "mean_token_accuracy": 0.7238461971282959, + "num_tokens": 246455155.0, + "step": 9521 + }, + { + "epoch": 1.0456841642872832, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3721799850463867, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.703603982925415, + "num_tokens": 246482223.0, + "step": 9522 + }, + { + "epoch": 1.0457939819898967, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5555851459503174, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.6957113742828369, + "num_tokens": 246507535.0, + "step": 9523 + }, + { + "epoch": 1.0459037996925105, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.8103668689727783, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7073748111724854, + "num_tokens": 246530389.0, + "step": 9524 + }, + { + "epoch": 1.046013617395124, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.4440994262695312, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6993718147277832, + "num_tokens": 246558410.0, + "step": 9525 + }, + { + "epoch": 1.0461234350977378, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5329749584198, + "learning_rate": 1e-06, + "loss": 0.9373, + "mean_token_accuracy": 0.7199673652648926, + "num_tokens": 246586971.0, + "step": 9526 + }, + { + "epoch": 1.0462332528003515, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.549590826034546, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7106056213378906, + "num_tokens": 246611921.0, + "step": 9527 + }, + { + "epoch": 1.046343070502965, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.644412040710449, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7091602683067322, + "num_tokens": 246637183.0, + "step": 9528 + }, + { + "epoch": 1.0464528882055788, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.5266664028167725, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.6982626914978027, + "num_tokens": 246661123.0, + "step": 9529 + }, + { + "epoch": 1.0465627059081923, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.6118898391723633, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7178503274917603, + "num_tokens": 246684376.0, + "step": 9530 + }, + { + "epoch": 1.046672523610806, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.8902440071105957, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7126332521438599, + "num_tokens": 246703163.0, + "step": 9531 + }, + { + "epoch": 1.0467823413134196, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.659010410308838, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7192197442054749, + "num_tokens": 246727436.0, + "step": 9532 + }, + { + "epoch": 1.0468921590160334, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.6120200157165527, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7110628485679626, + "num_tokens": 246754963.0, + "step": 9533 + }, + { + "epoch": 1.0470019767186471, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.8347020149230957, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7198954820632935, + "num_tokens": 246777203.0, + "step": 9534 + }, + { + "epoch": 1.0471117944212607, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.787753105163574, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.734501838684082, + "num_tokens": 246800178.0, + "step": 9535 + }, + { + "epoch": 1.0472216121238744, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 3.9103684425354004, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.6957424879074097, + "num_tokens": 246827037.0, + "step": 9536 + }, + { + "epoch": 1.047331429826488, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.862882375717163, + "learning_rate": 1e-06, + "loss": 0.8861, + "mean_token_accuracy": 0.7373325228691101, + "num_tokens": 246848866.0, + "step": 9537 + }, + { + "epoch": 1.0474412475291017, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7150967121124268, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7133533954620361, + "num_tokens": 246869487.0, + "step": 9538 + }, + { + "epoch": 1.0475510652317153, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.672891616821289, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6862456798553467, + "num_tokens": 246897217.0, + "step": 9539 + }, + { + "epoch": 1.047660882934329, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5388166904449463, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.693896472454071, + "num_tokens": 246927339.0, + "step": 9540 + }, + { + "epoch": 1.0477707006369428, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3741376399993896, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7106078863143921, + "num_tokens": 246957855.0, + "step": 9541 + }, + { + "epoch": 1.0478805183395563, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.637526512145996, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.721144437789917, + "num_tokens": 246979698.0, + "step": 9542 + }, + { + "epoch": 1.04799033604217, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4334845542907715, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.7309222221374512, + "num_tokens": 247007272.0, + "step": 9543 + }, + { + "epoch": 1.0481001537447836, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.7722833156585693, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.724611222743988, + "num_tokens": 247030032.0, + "step": 9544 + }, + { + "epoch": 1.0482099714473974, + "ewc_loss": 8.821487426757812e-06, + "grad_norm": 2.620866537094116, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7170925736427307, + "num_tokens": 247053901.0, + "step": 9545 + }, + { + "epoch": 1.048319789150011, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6936190128326416, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7019287347793579, + "num_tokens": 247076976.0, + "step": 9546 + }, + { + "epoch": 1.0484296068526247, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.8870067596435547, + "learning_rate": 1e-06, + "loss": 0.9204, + "mean_token_accuracy": 0.7256742119789124, + "num_tokens": 247097690.0, + "step": 9547 + }, + { + "epoch": 1.0485394245552384, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6896941661834717, + "learning_rate": 1e-06, + "loss": 0.9064, + "mean_token_accuracy": 0.7265797853469849, + "num_tokens": 247121076.0, + "step": 9548 + }, + { + "epoch": 1.048649242257852, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.675849437713623, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7194845676422119, + "num_tokens": 247145963.0, + "step": 9549 + }, + { + "epoch": 1.0487590599604657, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.686141014099121, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.6997427940368652, + "num_tokens": 247172368.0, + "step": 9550 + }, + { + "epoch": 1.0488688776630792, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.441066265106201, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6882126331329346, + "num_tokens": 247201180.0, + "step": 9551 + }, + { + "epoch": 1.048978695365693, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3182601928710938, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6951431632041931, + "num_tokens": 247232563.0, + "step": 9552 + }, + { + "epoch": 1.0490885130683065, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6144697666168213, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6857938766479492, + "num_tokens": 247257790.0, + "step": 9553 + }, + { + "epoch": 1.0491983307709203, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.700516700744629, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7079583406448364, + "num_tokens": 247279700.0, + "step": 9554 + }, + { + "epoch": 1.0493081484735338, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7274298667907715, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7262809872627258, + "num_tokens": 247303222.0, + "step": 9555 + }, + { + "epoch": 1.0494179661761476, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6686673164367676, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7241264581680298, + "num_tokens": 247325782.0, + "step": 9556 + }, + { + "epoch": 1.0495277838787613, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.493795394897461, + "learning_rate": 1e-06, + "loss": 1.0937, + "mean_token_accuracy": 0.6848869323730469, + "num_tokens": 247356404.0, + "step": 9557 + }, + { + "epoch": 1.0496376015813749, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3270955085754395, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6904979944229126, + "num_tokens": 247389537.0, + "step": 9558 + }, + { + "epoch": 1.0497474192839886, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.8444080352783203, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7019830942153931, + "num_tokens": 247412214.0, + "step": 9559 + }, + { + "epoch": 1.0498572369866022, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7611987590789795, + "learning_rate": 1e-06, + "loss": 0.8805, + "mean_token_accuracy": 0.7381659746170044, + "num_tokens": 247432745.0, + "step": 9560 + }, + { + "epoch": 1.049967054689216, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5292913913726807, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7098056077957153, + "num_tokens": 247457197.0, + "step": 9561 + }, + { + "epoch": 1.0500768723918297, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.9149258136749268, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6874686479568481, + "num_tokens": 247479184.0, + "step": 9562 + }, + { + "epoch": 1.0501866900944432, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.811126470565796, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7271600961685181, + "num_tokens": 247502682.0, + "step": 9563 + }, + { + "epoch": 1.050296507797057, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.511068820953369, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6951290369033813, + "num_tokens": 247531199.0, + "step": 9564 + }, + { + "epoch": 1.0504063254996705, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6613285541534424, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6874762177467346, + "num_tokens": 247560239.0, + "step": 9565 + }, + { + "epoch": 1.0505161432022843, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.62432599067688, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6925148367881775, + "num_tokens": 247584598.0, + "step": 9566 + }, + { + "epoch": 1.0506259609048978, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.587311267852783, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7142826318740845, + "num_tokens": 247610185.0, + "step": 9567 + }, + { + "epoch": 1.0507357786075116, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.749553918838501, + "learning_rate": 1e-06, + "loss": 0.8655, + "mean_token_accuracy": 0.7422047853469849, + "num_tokens": 247632298.0, + "step": 9568 + }, + { + "epoch": 1.050845596310125, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4532179832458496, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.7208881974220276, + "num_tokens": 247659916.0, + "step": 9569 + }, + { + "epoch": 1.0509554140127388, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5967555046081543, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7150911092758179, + "num_tokens": 247688480.0, + "step": 9570 + }, + { + "epoch": 1.0510652317153526, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5700674057006836, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7043671011924744, + "num_tokens": 247716386.0, + "step": 9571 + }, + { + "epoch": 1.0511750494179661, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6267635822296143, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7023890018463135, + "num_tokens": 247741248.0, + "step": 9572 + }, + { + "epoch": 1.05128486712058, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.577092170715332, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7090961337089539, + "num_tokens": 247767135.0, + "step": 9573 + }, + { + "epoch": 1.0513946848231934, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4690351486206055, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7062066197395325, + "num_tokens": 247793866.0, + "step": 9574 + }, + { + "epoch": 1.0515045025258072, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4161629676818848, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6995054483413696, + "num_tokens": 247826488.0, + "step": 9575 + }, + { + "epoch": 1.0516143202284207, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.899871349334717, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6927025318145752, + "num_tokens": 247848265.0, + "step": 9576 + }, + { + "epoch": 1.0517241379310345, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4206459522247314, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.7031325101852417, + "num_tokens": 247878027.0, + "step": 9577 + }, + { + "epoch": 1.0518339556336482, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 3.0410385131835938, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7190773487091064, + "num_tokens": 247899160.0, + "step": 9578 + }, + { + "epoch": 1.0519437733362618, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.860487222671509, + "learning_rate": 1e-06, + "loss": 1.1354, + "mean_token_accuracy": 0.6719493865966797, + "num_tokens": 247924094.0, + "step": 9579 + }, + { + "epoch": 1.0520535910388755, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.2725605964660645, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6918469071388245, + "num_tokens": 247958200.0, + "step": 9580 + }, + { + "epoch": 1.052163408741489, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.700650453567505, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.7042083144187927, + "num_tokens": 247982276.0, + "step": 9581 + }, + { + "epoch": 1.0522732264441028, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.514256238937378, + "learning_rate": 1e-06, + "loss": 0.979, + "mean_token_accuracy": 0.7119971513748169, + "num_tokens": 248010082.0, + "step": 9582 + }, + { + "epoch": 1.0523830441467164, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.718230962753296, + "learning_rate": 1e-06, + "loss": 0.9475, + "mean_token_accuracy": 0.7222599983215332, + "num_tokens": 248035005.0, + "step": 9583 + }, + { + "epoch": 1.05249286184933, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.509427785873413, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6991221904754639, + "num_tokens": 248061281.0, + "step": 9584 + }, + { + "epoch": 1.0526026795519439, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6089470386505127, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.6965288519859314, + "num_tokens": 248089291.0, + "step": 9585 + }, + { + "epoch": 1.0527124972545574, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3906502723693848, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.709385871887207, + "num_tokens": 248118085.0, + "step": 9586 + }, + { + "epoch": 1.0528223149571712, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7785472869873047, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7039098739624023, + "num_tokens": 248141100.0, + "step": 9587 + }, + { + "epoch": 1.0529321326597847, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6050894260406494, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7107218503952026, + "num_tokens": 248165329.0, + "step": 9588 + }, + { + "epoch": 1.0530419503623984, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.391321897506714, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6921300888061523, + "num_tokens": 248196463.0, + "step": 9589 + }, + { + "epoch": 1.053151768065012, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.727189779281616, + "learning_rate": 1e-06, + "loss": 0.9252, + "mean_token_accuracy": 0.7280768156051636, + "num_tokens": 248219419.0, + "step": 9590 + }, + { + "epoch": 1.0532615857676257, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.97330379486084, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6881081461906433, + "num_tokens": 248242837.0, + "step": 9591 + }, + { + "epoch": 1.0533714034702395, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.694525957107544, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7037703394889832, + "num_tokens": 248267252.0, + "step": 9592 + }, + { + "epoch": 1.053481221172853, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.3512847423553467, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6860800981521606, + "num_tokens": 248303356.0, + "step": 9593 + }, + { + "epoch": 1.0535910388754668, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7137186527252197, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6958000063896179, + "num_tokens": 248331023.0, + "step": 9594 + }, + { + "epoch": 1.0537008565780803, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.327012538909912, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.71086585521698, + "num_tokens": 248360940.0, + "step": 9595 + }, + { + "epoch": 1.053810674280694, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.341779947280884, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7070047855377197, + "num_tokens": 248390650.0, + "step": 9596 + }, + { + "epoch": 1.0539204919833076, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5029940605163574, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7066822648048401, + "num_tokens": 248417494.0, + "step": 9597 + }, + { + "epoch": 1.0540303096859214, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.318617105484009, + "learning_rate": 1e-06, + "loss": 0.9302, + "mean_token_accuracy": 0.7261205911636353, + "num_tokens": 248444723.0, + "step": 9598 + }, + { + "epoch": 1.0541401273885351, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.605431079864502, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7003599405288696, + "num_tokens": 248466685.0, + "step": 9599 + }, + { + "epoch": 1.0542499450911487, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.962186574935913, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7167654633522034, + "num_tokens": 248488630.0, + "step": 9600 + }, + { + "epoch": 1.0543597627937624, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.412281036376953, + "learning_rate": 1e-06, + "loss": 1.124, + "mean_token_accuracy": 0.6734404563903809, + "num_tokens": 248518374.0, + "step": 9601 + }, + { + "epoch": 1.054469580496376, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6214168071746826, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.703514039516449, + "num_tokens": 248542665.0, + "step": 9602 + }, + { + "epoch": 1.0545793981989897, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.2830400466918945, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7172276973724365, + "num_tokens": 248576368.0, + "step": 9603 + }, + { + "epoch": 1.0546892159016032, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4617931842803955, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6849172711372375, + "num_tokens": 248606520.0, + "step": 9604 + }, + { + "epoch": 1.054799033604217, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.391077756881714, + "learning_rate": 1e-06, + "loss": 1.1062, + "mean_token_accuracy": 0.6783623695373535, + "num_tokens": 248634950.0, + "step": 9605 + }, + { + "epoch": 1.0549088513068308, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5422375202178955, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7078447341918945, + "num_tokens": 248663549.0, + "step": 9606 + }, + { + "epoch": 1.0550186690094443, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 3.0008504390716553, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.6992594599723816, + "num_tokens": 248688326.0, + "step": 9607 + }, + { + "epoch": 1.055128486712058, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6332011222839355, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7006871700286865, + "num_tokens": 248714511.0, + "step": 9608 + }, + { + "epoch": 1.0552383044146716, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4452223777770996, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7089811563491821, + "num_tokens": 248742304.0, + "step": 9609 + }, + { + "epoch": 1.0553481221172853, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.471078634262085, + "learning_rate": 1e-06, + "loss": 0.9167, + "mean_token_accuracy": 0.7272127270698547, + "num_tokens": 248768175.0, + "step": 9610 + }, + { + "epoch": 1.0554579398198989, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4795095920562744, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7008333206176758, + "num_tokens": 248798423.0, + "step": 9611 + }, + { + "epoch": 1.0555677575225126, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.5210342407226562, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7010289430618286, + "num_tokens": 248823545.0, + "step": 9612 + }, + { + "epoch": 1.0556775752251264, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.9199860095977783, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.7259178161621094, + "num_tokens": 248842430.0, + "step": 9613 + }, + { + "epoch": 1.05578739292774, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.48172664642334, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7007764577865601, + "num_tokens": 248870096.0, + "step": 9614 + }, + { + "epoch": 1.0558972106303537, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.820143938064575, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7004141807556152, + "num_tokens": 248893581.0, + "step": 9615 + }, + { + "epoch": 1.0560070283329672, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.580603837966919, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.6961522698402405, + "num_tokens": 248919001.0, + "step": 9616 + }, + { + "epoch": 1.056116846035581, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.507002115249634, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6978554725646973, + "num_tokens": 248946510.0, + "step": 9617 + }, + { + "epoch": 1.0562266637381945, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.720639705657959, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.698974072933197, + "num_tokens": 248978612.0, + "step": 9618 + }, + { + "epoch": 1.0563364814408083, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.4924306869506836, + "learning_rate": 1e-06, + "loss": 1.1039, + "mean_token_accuracy": 0.6782655715942383, + "num_tokens": 249008011.0, + "step": 9619 + }, + { + "epoch": 1.0564462991434218, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6662724018096924, + "learning_rate": 1e-06, + "loss": 0.9526, + "mean_token_accuracy": 0.7181878089904785, + "num_tokens": 249031806.0, + "step": 9620 + }, + { + "epoch": 1.0565561168460356, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.8978872299194336, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7063616514205933, + "num_tokens": 249052702.0, + "step": 9621 + }, + { + "epoch": 1.0566659345486493, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.196593999862671, + "learning_rate": 1e-06, + "loss": 1.0999, + "mean_token_accuracy": 0.6769787073135376, + "num_tokens": 249087582.0, + "step": 9622 + }, + { + "epoch": 1.0567757522512629, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7667760848999023, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6985613703727722, + "num_tokens": 249111026.0, + "step": 9623 + }, + { + "epoch": 1.0568855699538766, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.7773585319519043, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.7034980058670044, + "num_tokens": 249133904.0, + "step": 9624 + }, + { + "epoch": 1.0569953876564901, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 3.007694959640503, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7062259912490845, + "num_tokens": 249153840.0, + "step": 9625 + }, + { + "epoch": 1.057105205359104, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.8012402057647705, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.6975566148757935, + "num_tokens": 249177898.0, + "step": 9626 + }, + { + "epoch": 1.0572150230617177, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.8243319988250732, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7114441990852356, + "num_tokens": 249199392.0, + "step": 9627 + }, + { + "epoch": 1.0573248407643312, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.6224288940429688, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7008349895477295, + "num_tokens": 249227251.0, + "step": 9628 + }, + { + "epoch": 1.057434658466945, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.8577868938446045, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7282429933547974, + "num_tokens": 249248156.0, + "step": 9629 + }, + { + "epoch": 1.0575444761695585, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.854522943496704, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6861687302589417, + "num_tokens": 249271245.0, + "step": 9630 + }, + { + "epoch": 1.0576542938721722, + "ewc_loss": 8.881092071533203e-06, + "grad_norm": 2.665635108947754, + "learning_rate": 1e-06, + "loss": 1.0975, + "mean_token_accuracy": 0.6845496892929077, + "num_tokens": 249297210.0, + "step": 9631 + }, + { + "epoch": 1.0577641115747858, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.904539108276367, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7195894718170166, + "num_tokens": 249317807.0, + "step": 9632 + }, + { + "epoch": 1.0578739292773995, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4558417797088623, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.706700325012207, + "num_tokens": 249347204.0, + "step": 9633 + }, + { + "epoch": 1.057983746980013, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4486918449401855, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7031112909317017, + "num_tokens": 249377018.0, + "step": 9634 + }, + { + "epoch": 1.0580935646826268, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.433476686477661, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7136794328689575, + "num_tokens": 249404706.0, + "step": 9635 + }, + { + "epoch": 1.0582033823852406, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6167755126953125, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7133466005325317, + "num_tokens": 249428576.0, + "step": 9636 + }, + { + "epoch": 1.0583132000878541, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5449604988098145, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7238044738769531, + "num_tokens": 249455495.0, + "step": 9637 + }, + { + "epoch": 1.0584230177904679, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.922428607940674, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7204651832580566, + "num_tokens": 249474722.0, + "step": 9638 + }, + { + "epoch": 1.0585328354930814, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6006083488464355, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7082412242889404, + "num_tokens": 249501296.0, + "step": 9639 + }, + { + "epoch": 1.0586426531956952, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.949235439300537, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7234389781951904, + "num_tokens": 249521551.0, + "step": 9640 + }, + { + "epoch": 1.0587524708983087, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6023976802825928, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7151514887809753, + "num_tokens": 249546353.0, + "step": 9641 + }, + { + "epoch": 1.0588622886009225, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4965314865112305, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6976557970046997, + "num_tokens": 249572895.0, + "step": 9642 + }, + { + "epoch": 1.0589721063035362, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.673112630844116, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.7283962965011597, + "num_tokens": 249596783.0, + "step": 9643 + }, + { + "epoch": 1.0590819240061498, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6908411979675293, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7045985460281372, + "num_tokens": 249623279.0, + "step": 9644 + }, + { + "epoch": 1.0591917417087635, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6904594898223877, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7033339738845825, + "num_tokens": 249651574.0, + "step": 9645 + }, + { + "epoch": 1.059301559411377, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5505714416503906, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.697359561920166, + "num_tokens": 249677631.0, + "step": 9646 + }, + { + "epoch": 1.0594113771139908, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.908704996109009, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6967423558235168, + "num_tokens": 249703995.0, + "step": 9647 + }, + { + "epoch": 1.0595211948166043, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6487505435943604, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.710239589214325, + "num_tokens": 249728629.0, + "step": 9648 + }, + { + "epoch": 1.059631012519218, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4092836380004883, + "learning_rate": 1e-06, + "loss": 1.1105, + "mean_token_accuracy": 0.6910574436187744, + "num_tokens": 249757537.0, + "step": 9649 + }, + { + "epoch": 1.0597408302218319, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.701148509979248, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.707126259803772, + "num_tokens": 249779905.0, + "step": 9650 + }, + { + "epoch": 1.0598506479244454, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.506336212158203, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7149844765663147, + "num_tokens": 249806155.0, + "step": 9651 + }, + { + "epoch": 1.0599604656270591, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3735814094543457, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.6957947015762329, + "num_tokens": 249836082.0, + "step": 9652 + }, + { + "epoch": 1.0600702833296727, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6406800746917725, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7067839503288269, + "num_tokens": 249860640.0, + "step": 9653 + }, + { + "epoch": 1.0601801010322864, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.729374885559082, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6993762254714966, + "num_tokens": 249884934.0, + "step": 9654 + }, + { + "epoch": 1.0602899187349, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7186033725738525, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.702828049659729, + "num_tokens": 249910064.0, + "step": 9655 + }, + { + "epoch": 1.0603997364375137, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.534200429916382, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6900765299797058, + "num_tokens": 249937239.0, + "step": 9656 + }, + { + "epoch": 1.0605095541401275, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.0949079990386963, + "learning_rate": 1e-06, + "loss": 0.9129, + "mean_token_accuracy": 0.730945348739624, + "num_tokens": 249955880.0, + "step": 9657 + }, + { + "epoch": 1.060619371842741, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3044519424438477, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6889353394508362, + "num_tokens": 249988295.0, + "step": 9658 + }, + { + "epoch": 1.0607291895453548, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.708653688430786, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7120864391326904, + "num_tokens": 250014371.0, + "step": 9659 + }, + { + "epoch": 1.0608390072479683, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7154459953308105, + "learning_rate": 1e-06, + "loss": 0.9195, + "mean_token_accuracy": 0.7209802865982056, + "num_tokens": 250035108.0, + "step": 9660 + }, + { + "epoch": 1.060948824950582, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7947065830230713, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7114943265914917, + "num_tokens": 250057924.0, + "step": 9661 + }, + { + "epoch": 1.0610586426531956, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4459354877471924, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7014898061752319, + "num_tokens": 250088872.0, + "step": 9662 + }, + { + "epoch": 1.0611684603558094, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.0197629928588867, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.6838839054107666, + "num_tokens": 250110129.0, + "step": 9663 + }, + { + "epoch": 1.0612782780584231, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.713299512863159, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7032997608184814, + "num_tokens": 250132915.0, + "step": 9664 + }, + { + "epoch": 1.0613880957610367, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.609727621078491, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.701197624206543, + "num_tokens": 250158548.0, + "step": 9665 + }, + { + "epoch": 1.0614979134636504, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6040279865264893, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6999702453613281, + "num_tokens": 250183701.0, + "step": 9666 + }, + { + "epoch": 1.061607731166264, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3837368488311768, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7034392952919006, + "num_tokens": 250212131.0, + "step": 9667 + }, + { + "epoch": 1.0617175488688777, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.643016815185547, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7238219380378723, + "num_tokens": 250235131.0, + "step": 9668 + }, + { + "epoch": 1.0618273665714912, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7048916816711426, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7076252698898315, + "num_tokens": 250258718.0, + "step": 9669 + }, + { + "epoch": 1.061937184274105, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.491995334625244, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7321715950965881, + "num_tokens": 250283341.0, + "step": 9670 + }, + { + "epoch": 1.0620470019767188, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6385326385498047, + "learning_rate": 1e-06, + "loss": 0.9185, + "mean_token_accuracy": 0.7250074148178101, + "num_tokens": 250304962.0, + "step": 9671 + }, + { + "epoch": 1.0621568196793323, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5274417400360107, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6882108449935913, + "num_tokens": 250332787.0, + "step": 9672 + }, + { + "epoch": 1.062266637381946, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3358302116394043, + "learning_rate": 1e-06, + "loss": 1.1005, + "mean_token_accuracy": 0.6810191869735718, + "num_tokens": 250368623.0, + "step": 9673 + }, + { + "epoch": 1.0623764550845596, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.727036952972412, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7090696096420288, + "num_tokens": 250392124.0, + "step": 9674 + }, + { + "epoch": 1.0624862727871733, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6005005836486816, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7082246541976929, + "num_tokens": 250417204.0, + "step": 9675 + }, + { + "epoch": 1.0625960904897869, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6950833797454834, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6933737993240356, + "num_tokens": 250444684.0, + "step": 9676 + }, + { + "epoch": 1.0627059081924006, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6689341068267822, + "learning_rate": 1e-06, + "loss": 1.099, + "mean_token_accuracy": 0.6937993764877319, + "num_tokens": 250469887.0, + "step": 9677 + }, + { + "epoch": 1.0628157258950144, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4238836765289307, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7126317620277405, + "num_tokens": 250498534.0, + "step": 9678 + }, + { + "epoch": 1.062925543597628, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4688072204589844, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.7294974327087402, + "num_tokens": 250525802.0, + "step": 9679 + }, + { + "epoch": 1.0630353613002417, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.606959819793701, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7145810723304749, + "num_tokens": 250553585.0, + "step": 9680 + }, + { + "epoch": 1.0631451790028552, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.686319351196289, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7215830087661743, + "num_tokens": 250576137.0, + "step": 9681 + }, + { + "epoch": 1.063254996705469, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4748618602752686, + "learning_rate": 1e-06, + "loss": 1.099, + "mean_token_accuracy": 0.6835082173347473, + "num_tokens": 250606474.0, + "step": 9682 + }, + { + "epoch": 1.0633648144080825, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.567140817642212, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7088268995285034, + "num_tokens": 250636724.0, + "step": 9683 + }, + { + "epoch": 1.0634746321106963, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3981809616088867, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7001552581787109, + "num_tokens": 250665159.0, + "step": 9684 + }, + { + "epoch": 1.0635844498133098, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6986420154571533, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7239498496055603, + "num_tokens": 250687852.0, + "step": 9685 + }, + { + "epoch": 1.0636942675159236, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4184224605560303, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7212074995040894, + "num_tokens": 250717529.0, + "step": 9686 + }, + { + "epoch": 1.0638040852185373, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8440792560577393, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.6993954181671143, + "num_tokens": 250738915.0, + "step": 9687 + }, + { + "epoch": 1.0639139029211508, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6976993083953857, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7182983160018921, + "num_tokens": 250761103.0, + "step": 9688 + }, + { + "epoch": 1.0640237206237646, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.722092628479004, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7202246785163879, + "num_tokens": 250784246.0, + "step": 9689 + }, + { + "epoch": 1.0641335383263781, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6476612091064453, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7124125361442566, + "num_tokens": 250808922.0, + "step": 9690 + }, + { + "epoch": 1.064243356028992, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4807674884796143, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6897176504135132, + "num_tokens": 250837921.0, + "step": 9691 + }, + { + "epoch": 1.0643531737316057, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.664232015609741, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7096472382545471, + "num_tokens": 250864218.0, + "step": 9692 + }, + { + "epoch": 1.0644629914342192, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.559457302093506, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7211718559265137, + "num_tokens": 250889667.0, + "step": 9693 + }, + { + "epoch": 1.064572809136833, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.9813051223754883, + "learning_rate": 1e-06, + "loss": 0.868, + "mean_token_accuracy": 0.7371492981910706, + "num_tokens": 250906367.0, + "step": 9694 + }, + { + "epoch": 1.0646826268394465, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4886057376861572, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7093566060066223, + "num_tokens": 250935459.0, + "step": 9695 + }, + { + "epoch": 1.0647924445420602, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6569764614105225, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7141780853271484, + "num_tokens": 250960205.0, + "step": 9696 + }, + { + "epoch": 1.0649022622446738, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.511584997177124, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7060986757278442, + "num_tokens": 250986395.0, + "step": 9697 + }, + { + "epoch": 1.0650120799472875, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.773845672607422, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.698116660118103, + "num_tokens": 251010206.0, + "step": 9698 + }, + { + "epoch": 1.065121897649901, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.489323377609253, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7002452611923218, + "num_tokens": 251037764.0, + "step": 9699 + }, + { + "epoch": 1.0652317153525148, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.539754867553711, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7105574607849121, + "num_tokens": 251065632.0, + "step": 9700 + }, + { + "epoch": 1.0653415330551286, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8153185844421387, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.6968303918838501, + "num_tokens": 251087567.0, + "step": 9701 + }, + { + "epoch": 1.065451350757742, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.9968764781951904, + "learning_rate": 1e-06, + "loss": 0.9082, + "mean_token_accuracy": 0.7287547588348389, + "num_tokens": 251105169.0, + "step": 9702 + }, + { + "epoch": 1.0655611684603559, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5156173706054688, + "learning_rate": 1e-06, + "loss": 0.8389, + "mean_token_accuracy": 0.7482607960700989, + "num_tokens": 251129897.0, + "step": 9703 + }, + { + "epoch": 1.0656709861629694, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8909614086151123, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7174862623214722, + "num_tokens": 251152836.0, + "step": 9704 + }, + { + "epoch": 1.0657808038655832, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5997908115386963, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.707795262336731, + "num_tokens": 251178610.0, + "step": 9705 + }, + { + "epoch": 1.0658906215681967, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.601092576980591, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7030458450317383, + "num_tokens": 251203642.0, + "step": 9706 + }, + { + "epoch": 1.0660004392708105, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5707194805145264, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.70556640625, + "num_tokens": 251228970.0, + "step": 9707 + }, + { + "epoch": 1.0661102569734242, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4796805381774902, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7103519439697266, + "num_tokens": 251256481.0, + "step": 9708 + }, + { + "epoch": 1.0662200746760377, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.503258466720581, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6870185136795044, + "num_tokens": 251284419.0, + "step": 9709 + }, + { + "epoch": 1.0663298923786515, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.540382146835327, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7067421674728394, + "num_tokens": 251308932.0, + "step": 9710 + }, + { + "epoch": 1.066439710081265, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.414851427078247, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6934007406234741, + "num_tokens": 251339253.0, + "step": 9711 + }, + { + "epoch": 1.0665495277838788, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.892656087875366, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7161473035812378, + "num_tokens": 251360860.0, + "step": 9712 + }, + { + "epoch": 1.0666593454864923, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5816292762756348, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7153167724609375, + "num_tokens": 251385661.0, + "step": 9713 + }, + { + "epoch": 1.066769163189106, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.728609085083008, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7051849365234375, + "num_tokens": 251411174.0, + "step": 9714 + }, + { + "epoch": 1.0668789808917198, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8916049003601074, + "learning_rate": 1e-06, + "loss": 0.9726, + "mean_token_accuracy": 0.7117257118225098, + "num_tokens": 251433780.0, + "step": 9715 + }, + { + "epoch": 1.0669887985943334, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.80558443069458, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7009752988815308, + "num_tokens": 251457958.0, + "step": 9716 + }, + { + "epoch": 1.0670986162969471, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5716116428375244, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7084741592407227, + "num_tokens": 251483937.0, + "step": 9717 + }, + { + "epoch": 1.0672084339995607, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6900928020477295, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7112595438957214, + "num_tokens": 251506303.0, + "step": 9718 + }, + { + "epoch": 1.0673182517021744, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6916186809539795, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.6982800960540771, + "num_tokens": 251531605.0, + "step": 9719 + }, + { + "epoch": 1.067428069404788, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4398884773254395, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6803958415985107, + "num_tokens": 251559470.0, + "step": 9720 + }, + { + "epoch": 1.0675378871074017, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7885172367095947, + "learning_rate": 1e-06, + "loss": 0.9164, + "mean_token_accuracy": 0.7343608140945435, + "num_tokens": 251580254.0, + "step": 9721 + }, + { + "epoch": 1.0676477048100155, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5568339824676514, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7035797238349915, + "num_tokens": 251606712.0, + "step": 9722 + }, + { + "epoch": 1.067757522512629, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.768277406692505, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6927007436752319, + "num_tokens": 251632873.0, + "step": 9723 + }, + { + "epoch": 1.0678673402152428, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4632742404937744, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6928722262382507, + "num_tokens": 251662345.0, + "step": 9724 + }, + { + "epoch": 1.0679771579178563, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.223949670791626, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7099207043647766, + "num_tokens": 251694926.0, + "step": 9725 + }, + { + "epoch": 1.06808697562047, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7070746421813965, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.708143413066864, + "num_tokens": 251717795.0, + "step": 9726 + }, + { + "epoch": 1.0681967933230836, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.548969268798828, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7125266790390015, + "num_tokens": 251744985.0, + "step": 9727 + }, + { + "epoch": 1.0683066110256974, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 5.14599084854126, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.6993403434753418, + "num_tokens": 251769990.0, + "step": 9728 + }, + { + "epoch": 1.068416428728311, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.959456205368042, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6914573311805725, + "num_tokens": 251793256.0, + "step": 9729 + }, + { + "epoch": 1.0685262464309246, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4898183345794678, + "learning_rate": 1e-06, + "loss": 1.1074, + "mean_token_accuracy": 0.680086612701416, + "num_tokens": 251824616.0, + "step": 9730 + }, + { + "epoch": 1.0686360641335384, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7595479488372803, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6901623010635376, + "num_tokens": 251849164.0, + "step": 9731 + }, + { + "epoch": 1.068745881836152, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4093220233917236, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7213815450668335, + "num_tokens": 251879989.0, + "step": 9732 + }, + { + "epoch": 1.0688556995387657, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.576659679412842, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.7054483890533447, + "num_tokens": 251906896.0, + "step": 9733 + }, + { + "epoch": 1.0689655172413792, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4735593795776367, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6971257925033569, + "num_tokens": 251935270.0, + "step": 9734 + }, + { + "epoch": 1.069075334943993, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.688992500305176, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7158694267272949, + "num_tokens": 251958585.0, + "step": 9735 + }, + { + "epoch": 1.0691851526466065, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4734699726104736, + "learning_rate": 1e-06, + "loss": 1.1238, + "mean_token_accuracy": 0.6737487316131592, + "num_tokens": 251989265.0, + "step": 9736 + }, + { + "epoch": 1.0692949703492203, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5221996307373047, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.7273446917533875, + "num_tokens": 252015010.0, + "step": 9737 + }, + { + "epoch": 1.069404788051834, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.567962646484375, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7157619595527649, + "num_tokens": 252038183.0, + "step": 9738 + }, + { + "epoch": 1.0695146057544476, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7820868492126465, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.6923305988311768, + "num_tokens": 252061193.0, + "step": 9739 + }, + { + "epoch": 1.0696244234570613, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3866922855377197, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7042658925056458, + "num_tokens": 252091708.0, + "step": 9740 + }, + { + "epoch": 1.0697342411596749, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8699839115142822, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7281833291053772, + "num_tokens": 252110662.0, + "step": 9741 + }, + { + "epoch": 1.0698440588622886, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4619879722595215, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7105433344841003, + "num_tokens": 252138537.0, + "step": 9742 + }, + { + "epoch": 1.0699538765649024, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6092071533203125, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7061197757720947, + "num_tokens": 252164974.0, + "step": 9743 + }, + { + "epoch": 1.070063694267516, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.570326089859009, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7104496955871582, + "num_tokens": 252189644.0, + "step": 9744 + }, + { + "epoch": 1.0701735119701297, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5319764614105225, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6862661838531494, + "num_tokens": 252215816.0, + "step": 9745 + }, + { + "epoch": 1.0702833296727432, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.445967674255371, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7074906826019287, + "num_tokens": 252242546.0, + "step": 9746 + }, + { + "epoch": 1.070393147375357, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.950789213180542, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7087770700454712, + "num_tokens": 252261287.0, + "step": 9747 + }, + { + "epoch": 1.0705029650779705, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.538536310195923, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7124679088592529, + "num_tokens": 252288336.0, + "step": 9748 + }, + { + "epoch": 1.0706127827805842, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.448829412460327, + "learning_rate": 1e-06, + "loss": 0.9423, + "mean_token_accuracy": 0.7182761430740356, + "num_tokens": 252315179.0, + "step": 9749 + }, + { + "epoch": 1.0707226004831978, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.493532419204712, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6987082958221436, + "num_tokens": 252341554.0, + "step": 9750 + }, + { + "epoch": 1.0708324181858115, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.765465021133423, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6953551769256592, + "num_tokens": 252369884.0, + "step": 9751 + }, + { + "epoch": 1.0709422358884253, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.554436683654785, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7101317048072815, + "num_tokens": 252395945.0, + "step": 9752 + }, + { + "epoch": 1.0710520535910388, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5917742252349854, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7100000381469727, + "num_tokens": 252421213.0, + "step": 9753 + }, + { + "epoch": 1.0711618712936526, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5835494995117188, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7038211226463318, + "num_tokens": 252444980.0, + "step": 9754 + }, + { + "epoch": 1.0712716889962661, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.64579701423645, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7012415528297424, + "num_tokens": 252472709.0, + "step": 9755 + }, + { + "epoch": 1.0713815066988799, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4494946002960205, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7060905694961548, + "num_tokens": 252502923.0, + "step": 9756 + }, + { + "epoch": 1.0714913244014934, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6457760334014893, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6992812752723694, + "num_tokens": 252530288.0, + "step": 9757 + }, + { + "epoch": 1.0716011421041072, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.622058391571045, + "learning_rate": 1e-06, + "loss": 1.1193, + "mean_token_accuracy": 0.6849831938743591, + "num_tokens": 252561438.0, + "step": 9758 + }, + { + "epoch": 1.071710959806721, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4525489807128906, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6814888715744019, + "num_tokens": 252591279.0, + "step": 9759 + }, + { + "epoch": 1.0718207775093345, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6558783054351807, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7128086090087891, + "num_tokens": 252621008.0, + "step": 9760 + }, + { + "epoch": 1.0719305952119482, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6872828006744385, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7020397782325745, + "num_tokens": 252647364.0, + "step": 9761 + }, + { + "epoch": 1.0720404129145618, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.789135217666626, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.712383508682251, + "num_tokens": 252670834.0, + "step": 9762 + }, + { + "epoch": 1.0721502306171755, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.862436294555664, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6927831172943115, + "num_tokens": 252695647.0, + "step": 9763 + }, + { + "epoch": 1.072260048319789, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.485352039337158, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7082672715187073, + "num_tokens": 252725425.0, + "step": 9764 + }, + { + "epoch": 1.0723698660224028, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4397120475769043, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6883455514907837, + "num_tokens": 252755765.0, + "step": 9765 + }, + { + "epoch": 1.0724796837250166, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4227712154388428, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7104088068008423, + "num_tokens": 252784490.0, + "step": 9766 + }, + { + "epoch": 1.07258950142763, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.478609323501587, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7120554447174072, + "num_tokens": 252812843.0, + "step": 9767 + }, + { + "epoch": 1.0726993191302439, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.9255495071411133, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7095257043838501, + "num_tokens": 252834022.0, + "step": 9768 + }, + { + "epoch": 1.0728091368328574, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4410769939422607, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7094285488128662, + "num_tokens": 252860473.0, + "step": 9769 + }, + { + "epoch": 1.0729189545354711, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4632270336151123, + "learning_rate": 1e-06, + "loss": 1.1044, + "mean_token_accuracy": 0.6805774569511414, + "num_tokens": 252892540.0, + "step": 9770 + }, + { + "epoch": 1.0730287722380847, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.404114246368408, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7088723182678223, + "num_tokens": 252925170.0, + "step": 9771 + }, + { + "epoch": 1.0731385899406984, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.0083980560302734, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7116387486457825, + "num_tokens": 252946526.0, + "step": 9772 + }, + { + "epoch": 1.0732484076433122, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6795222759246826, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6880035996437073, + "num_tokens": 252974024.0, + "step": 9773 + }, + { + "epoch": 1.0733582253459257, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.746265411376953, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7100626230239868, + "num_tokens": 252997552.0, + "step": 9774 + }, + { + "epoch": 1.0734680430485395, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6091256141662598, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7060015797615051, + "num_tokens": 253024018.0, + "step": 9775 + }, + { + "epoch": 1.073577860751153, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.375284194946289, + "learning_rate": 1e-06, + "loss": 0.9194, + "mean_token_accuracy": 0.7338951826095581, + "num_tokens": 253059439.0, + "step": 9776 + }, + { + "epoch": 1.0736876784537668, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.364215612411499, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6962676644325256, + "num_tokens": 253092060.0, + "step": 9777 + }, + { + "epoch": 1.0737974961563803, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5989620685577393, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.709570050239563, + "num_tokens": 253116732.0, + "step": 9778 + }, + { + "epoch": 1.073907313858994, + "ewc_loss": 8.940696716308594e-06, + "grad_norm": 2.9128329753875732, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.6970022916793823, + "num_tokens": 253137444.0, + "step": 9779 + }, + { + "epoch": 1.0740171315616078, + "ewc_loss": 8.940696716308594e-06, + "grad_norm": 2.623878002166748, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7016616463661194, + "num_tokens": 253163728.0, + "step": 9780 + }, + { + "epoch": 1.0741269492642214, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.9125664234161377, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.70574951171875, + "num_tokens": 253185007.0, + "step": 9781 + }, + { + "epoch": 1.0742367669668351, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7661454677581787, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7061786651611328, + "num_tokens": 253210039.0, + "step": 9782 + }, + { + "epoch": 1.0743465846694487, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7194440364837646, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7015847563743591, + "num_tokens": 253233730.0, + "step": 9783 + }, + { + "epoch": 1.0744564023720624, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8817012310028076, + "learning_rate": 1e-06, + "loss": 0.9116, + "mean_token_accuracy": 0.7336907982826233, + "num_tokens": 253253245.0, + "step": 9784 + }, + { + "epoch": 1.074566220074676, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7419774532318115, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7291431427001953, + "num_tokens": 253276720.0, + "step": 9785 + }, + { + "epoch": 1.0746760377772897, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.49804425239563, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.7006176710128784, + "num_tokens": 253304939.0, + "step": 9786 + }, + { + "epoch": 1.0747858554799035, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.499232053756714, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6906828284263611, + "num_tokens": 253337224.0, + "step": 9787 + }, + { + "epoch": 1.074895673182517, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.776273012161255, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7072637677192688, + "num_tokens": 253361449.0, + "step": 9788 + }, + { + "epoch": 1.0750054908851308, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8532497882843018, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.720809280872345, + "num_tokens": 253382157.0, + "step": 9789 + }, + { + "epoch": 1.0751153085877443, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3295438289642334, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6979438066482544, + "num_tokens": 253416618.0, + "step": 9790 + }, + { + "epoch": 1.075225126290358, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8860416412353516, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7183535099029541, + "num_tokens": 253437656.0, + "step": 9791 + }, + { + "epoch": 1.0753349439929716, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5213942527770996, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6972111463546753, + "num_tokens": 253465676.0, + "step": 9792 + }, + { + "epoch": 1.0754447616955853, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.9913127422332764, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7059897780418396, + "num_tokens": 253488247.0, + "step": 9793 + }, + { + "epoch": 1.075554579398199, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.5174217224121094, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7024657726287842, + "num_tokens": 253509582.0, + "step": 9794 + }, + { + "epoch": 1.0756643971008126, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.757193088531494, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6974654197692871, + "num_tokens": 253533499.0, + "step": 9795 + }, + { + "epoch": 1.0757742148034264, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.590446710586548, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7095266580581665, + "num_tokens": 253559942.0, + "step": 9796 + }, + { + "epoch": 1.07588403250604, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.659188985824585, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7105149030685425, + "num_tokens": 253587048.0, + "step": 9797 + }, + { + "epoch": 1.0759938502086537, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7842841148376465, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.7000950574874878, + "num_tokens": 253614210.0, + "step": 9798 + }, + { + "epoch": 1.0761036679112672, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4868712425231934, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7061887979507446, + "num_tokens": 253645388.0, + "step": 9799 + }, + { + "epoch": 1.076213485613881, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.913595199584961, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7185366749763489, + "num_tokens": 253666976.0, + "step": 9800 + }, + { + "epoch": 1.0763233033164945, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.510338306427002, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6838419437408447, + "num_tokens": 253695884.0, + "step": 9801 + }, + { + "epoch": 1.0764331210191083, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 20.52887535095215, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6875320672988892, + "num_tokens": 253730963.0, + "step": 9802 + }, + { + "epoch": 1.076542938721722, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.190955400466919, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7115331888198853, + "num_tokens": 253754933.0, + "step": 9803 + }, + { + "epoch": 1.0766527564243356, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6709213256835938, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6946687698364258, + "num_tokens": 253782586.0, + "step": 9804 + }, + { + "epoch": 1.0767625741269493, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.268676519393921, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.714454174041748, + "num_tokens": 253816135.0, + "step": 9805 + }, + { + "epoch": 1.0768723918295628, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.546339988708496, + "learning_rate": 1e-06, + "loss": 0.9401, + "mean_token_accuracy": 0.7260897159576416, + "num_tokens": 253840466.0, + "step": 9806 + }, + { + "epoch": 1.0769822095321766, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5670485496520996, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7118515968322754, + "num_tokens": 253866512.0, + "step": 9807 + }, + { + "epoch": 1.0770920272347904, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.8367655277252197, + "learning_rate": 1e-06, + "loss": 0.9356, + "mean_token_accuracy": 0.7276763319969177, + "num_tokens": 253890883.0, + "step": 9808 + }, + { + "epoch": 1.077201844937404, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.788236618041992, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7143646478652954, + "num_tokens": 253913297.0, + "step": 9809 + }, + { + "epoch": 1.0773116626400177, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4377663135528564, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.7001408338546753, + "num_tokens": 253941360.0, + "step": 9810 + }, + { + "epoch": 1.0774214803426312, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3549489974975586, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6820766925811768, + "num_tokens": 253971636.0, + "step": 9811 + }, + { + "epoch": 1.077531298045245, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8334264755249023, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7076693177223206, + "num_tokens": 253996545.0, + "step": 9812 + }, + { + "epoch": 1.0776411157478585, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.267817735671997, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6891905665397644, + "num_tokens": 254030517.0, + "step": 9813 + }, + { + "epoch": 1.0777509334504722, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.75757098197937, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.6999765634536743, + "num_tokens": 254052332.0, + "step": 9814 + }, + { + "epoch": 1.0778607511530858, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.70735764503479, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7077283263206482, + "num_tokens": 254075027.0, + "step": 9815 + }, + { + "epoch": 1.0779705688556995, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5352256298065186, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6927664875984192, + "num_tokens": 254105596.0, + "step": 9816 + }, + { + "epoch": 1.0780803865583133, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6836533546447754, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7050189971923828, + "num_tokens": 254129912.0, + "step": 9817 + }, + { + "epoch": 1.0781902042609268, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.591395378112793, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7249528765678406, + "num_tokens": 254152853.0, + "step": 9818 + }, + { + "epoch": 1.0783000219635406, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.625781536102295, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7039914131164551, + "num_tokens": 254179367.0, + "step": 9819 + }, + { + "epoch": 1.0784098396661541, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6156535148620605, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6895205974578857, + "num_tokens": 254205615.0, + "step": 9820 + }, + { + "epoch": 1.0785196573687679, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6287190914154053, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7198729515075684, + "num_tokens": 254232185.0, + "step": 9821 + }, + { + "epoch": 1.0786294750713814, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6392762660980225, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6956347227096558, + "num_tokens": 254258918.0, + "step": 9822 + }, + { + "epoch": 1.0787392927739952, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8102331161499023, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.7214186787605286, + "num_tokens": 254281886.0, + "step": 9823 + }, + { + "epoch": 1.078849110476609, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6476681232452393, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6945513486862183, + "num_tokens": 254312053.0, + "step": 9824 + }, + { + "epoch": 1.0789589281792225, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6581027507781982, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7174637913703918, + "num_tokens": 254335024.0, + "step": 9825 + }, + { + "epoch": 1.0790687458818362, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5405232906341553, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7013373970985413, + "num_tokens": 254361962.0, + "step": 9826 + }, + { + "epoch": 1.0791785635844497, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.2764041423797607, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7000597715377808, + "num_tokens": 254396842.0, + "step": 9827 + }, + { + "epoch": 1.0792883812870635, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5419085025787354, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7021694183349609, + "num_tokens": 254422731.0, + "step": 9828 + }, + { + "epoch": 1.079398198989677, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5002405643463135, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7187914252281189, + "num_tokens": 254448617.0, + "step": 9829 + }, + { + "epoch": 1.0795080166922908, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8515334129333496, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6943100690841675, + "num_tokens": 254468811.0, + "step": 9830 + }, + { + "epoch": 1.0796178343949046, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.756535053253174, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7011247277259827, + "num_tokens": 254492675.0, + "step": 9831 + }, + { + "epoch": 1.079727652097518, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 6.870559215545654, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7125465273857117, + "num_tokens": 254512410.0, + "step": 9832 + }, + { + "epoch": 1.0798374698001318, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.871717691421509, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.7048990726470947, + "num_tokens": 254535716.0, + "step": 9833 + }, + { + "epoch": 1.0799472875027454, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8459014892578125, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7183420658111572, + "num_tokens": 254557037.0, + "step": 9834 + }, + { + "epoch": 1.0800571052053591, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7218143939971924, + "learning_rate": 1e-06, + "loss": 1.1032, + "mean_token_accuracy": 0.6796256899833679, + "num_tokens": 254584259.0, + "step": 9835 + }, + { + "epoch": 1.0801669229079727, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5935137271881104, + "learning_rate": 1e-06, + "loss": 1.1296, + "mean_token_accuracy": 0.6700707674026489, + "num_tokens": 254614845.0, + "step": 9836 + }, + { + "epoch": 1.0802767406105864, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5670182704925537, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7139739394187927, + "num_tokens": 254640490.0, + "step": 9837 + }, + { + "epoch": 1.0803865583132002, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.551907777786255, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7094403505325317, + "num_tokens": 254667390.0, + "step": 9838 + }, + { + "epoch": 1.0804963760158137, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.9469387531280518, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.6921496391296387, + "num_tokens": 254690296.0, + "step": 9839 + }, + { + "epoch": 1.0806061937184275, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7346625328063965, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6905145645141602, + "num_tokens": 254715809.0, + "step": 9840 + }, + { + "epoch": 1.080716011421041, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5192618370056152, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.7382827401161194, + "num_tokens": 254741448.0, + "step": 9841 + }, + { + "epoch": 1.0808258291236548, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5673041343688965, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6940117478370667, + "num_tokens": 254767857.0, + "step": 9842 + }, + { + "epoch": 1.0809356468262683, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.790924310684204, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.709288477897644, + "num_tokens": 254792749.0, + "step": 9843 + }, + { + "epoch": 1.081045464528882, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3973512649536133, + "learning_rate": 1e-06, + "loss": 1.1247, + "mean_token_accuracy": 0.6836121082305908, + "num_tokens": 254827509.0, + "step": 9844 + }, + { + "epoch": 1.0811552822314958, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.594691038131714, + "learning_rate": 1e-06, + "loss": 0.8941, + "mean_token_accuracy": 0.7329821586608887, + "num_tokens": 254849998.0, + "step": 9845 + }, + { + "epoch": 1.0812650999341094, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6450998783111572, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6953825950622559, + "num_tokens": 254873692.0, + "step": 9846 + }, + { + "epoch": 1.081374917636723, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4959867000579834, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.715613603591919, + "num_tokens": 254901458.0, + "step": 9847 + }, + { + "epoch": 1.0814847353393366, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.064063787460327, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.7064536213874817, + "num_tokens": 254922737.0, + "step": 9848 + }, + { + "epoch": 1.0815945530419504, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7600505352020264, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6948814392089844, + "num_tokens": 254951216.0, + "step": 9849 + }, + { + "epoch": 1.081704370744564, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7294254302978516, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7033710479736328, + "num_tokens": 254975725.0, + "step": 9850 + }, + { + "epoch": 1.0818141884471777, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3057823181152344, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6899305582046509, + "num_tokens": 255012158.0, + "step": 9851 + }, + { + "epoch": 1.0819240061497912, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.127002239227295, + "learning_rate": 1e-06, + "loss": 0.9037, + "mean_token_accuracy": 0.7247551679611206, + "num_tokens": 255028102.0, + "step": 9852 + }, + { + "epoch": 1.082033823852405, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.611255168914795, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7149789333343506, + "num_tokens": 255053098.0, + "step": 9853 + }, + { + "epoch": 1.0821436415550187, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.633294105529785, + "learning_rate": 1e-06, + "loss": 0.8914, + "mean_token_accuracy": 0.7380079030990601, + "num_tokens": 255079204.0, + "step": 9854 + }, + { + "epoch": 1.0822534592576323, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.490571975708008, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7124736309051514, + "num_tokens": 255109320.0, + "step": 9855 + }, + { + "epoch": 1.082363276960246, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3043551445007324, + "learning_rate": 1e-06, + "loss": 1.127, + "mean_token_accuracy": 0.6869714260101318, + "num_tokens": 255142634.0, + "step": 9856 + }, + { + "epoch": 1.0824730946628596, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7127861976623535, + "learning_rate": 1e-06, + "loss": 0.9057, + "mean_token_accuracy": 0.7313887476921082, + "num_tokens": 255166313.0, + "step": 9857 + }, + { + "epoch": 1.0825829123654733, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5926828384399414, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7040544152259827, + "num_tokens": 255194625.0, + "step": 9858 + }, + { + "epoch": 1.082692730068087, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5178489685058594, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7216498255729675, + "num_tokens": 255220850.0, + "step": 9859 + }, + { + "epoch": 1.0828025477707006, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9545631408691406, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7118825912475586, + "num_tokens": 255243028.0, + "step": 9860 + }, + { + "epoch": 1.0829123654733144, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.702529191970825, + "learning_rate": 1e-06, + "loss": 1.0974, + "mean_token_accuracy": 0.6830834150314331, + "num_tokens": 255272468.0, + "step": 9861 + }, + { + "epoch": 1.083022183175928, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7743399143218994, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7080299854278564, + "num_tokens": 255296927.0, + "step": 9862 + }, + { + "epoch": 1.0831320008785417, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.403273582458496, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7309861183166504, + "num_tokens": 255323354.0, + "step": 9863 + }, + { + "epoch": 1.0832418185811552, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4073007106781006, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.68647700548172, + "num_tokens": 255356309.0, + "step": 9864 + }, + { + "epoch": 1.083351636283769, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5471365451812744, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7043982744216919, + "num_tokens": 255383162.0, + "step": 9865 + }, + { + "epoch": 1.0834614539863825, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6489529609680176, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6920474767684937, + "num_tokens": 255408460.0, + "step": 9866 + }, + { + "epoch": 1.0835712716889963, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.112856864929199, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7064160108566284, + "num_tokens": 255434169.0, + "step": 9867 + }, + { + "epoch": 1.08368108939161, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.861521005630493, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6951581239700317, + "num_tokens": 255454970.0, + "step": 9868 + }, + { + "epoch": 1.0837909070942235, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8936655521392822, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7264754772186279, + "num_tokens": 255475839.0, + "step": 9869 + }, + { + "epoch": 1.0839007247968373, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.629112720489502, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6866151690483093, + "num_tokens": 255505850.0, + "step": 9870 + }, + { + "epoch": 1.0840105424994508, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4342453479766846, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.688994288444519, + "num_tokens": 255536149.0, + "step": 9871 + }, + { + "epoch": 1.0841203602020646, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6710519790649414, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.7115187644958496, + "num_tokens": 255561334.0, + "step": 9872 + }, + { + "epoch": 1.0842301779046783, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5671041011810303, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7004795074462891, + "num_tokens": 255589566.0, + "step": 9873 + }, + { + "epoch": 1.0843399956072919, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6067609786987305, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7273237109184265, + "num_tokens": 255613867.0, + "step": 9874 + }, + { + "epoch": 1.0844498133099056, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.805453300476074, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7133432030677795, + "num_tokens": 255634796.0, + "step": 9875 + }, + { + "epoch": 1.0845596310125192, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.3779361248016357, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6903843879699707, + "num_tokens": 255666912.0, + "step": 9876 + }, + { + "epoch": 1.084669448715133, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.475470781326294, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6992467641830444, + "num_tokens": 255693497.0, + "step": 9877 + }, + { + "epoch": 1.0847792664177465, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7362802028656006, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.6957345604896545, + "num_tokens": 255719485.0, + "step": 9878 + }, + { + "epoch": 1.0848890841203602, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.527109384536743, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7062912583351135, + "num_tokens": 255743656.0, + "step": 9879 + }, + { + "epoch": 1.0849989018229738, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.593871831893921, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.6976860165596008, + "num_tokens": 255774053.0, + "step": 9880 + }, + { + "epoch": 1.0851087195255875, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.118816614151001, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6930316686630249, + "num_tokens": 255794035.0, + "step": 9881 + }, + { + "epoch": 1.0852185372282013, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5266740322113037, + "learning_rate": 1e-06, + "loss": 0.8819, + "mean_token_accuracy": 0.7366557717323303, + "num_tokens": 255819485.0, + "step": 9882 + }, + { + "epoch": 1.0853283549308148, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6898627281188965, + "learning_rate": 1e-06, + "loss": 1.1191, + "mean_token_accuracy": 0.679430365562439, + "num_tokens": 255844570.0, + "step": 9883 + }, + { + "epoch": 1.0854381726334286, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5397493839263916, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.684199333190918, + "num_tokens": 255873762.0, + "step": 9884 + }, + { + "epoch": 1.085547990336042, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.824610471725464, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7077957391738892, + "num_tokens": 255896096.0, + "step": 9885 + }, + { + "epoch": 1.0856578080386559, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3449816703796387, + "learning_rate": 1e-06, + "loss": 1.1063, + "mean_token_accuracy": 0.6768643856048584, + "num_tokens": 255927815.0, + "step": 9886 + }, + { + "epoch": 1.0857676257412694, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6498937606811523, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.7191457748413086, + "num_tokens": 255949988.0, + "step": 9887 + }, + { + "epoch": 1.0858774434438832, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8853774070739746, + "learning_rate": 1e-06, + "loss": 0.8688, + "mean_token_accuracy": 0.7347022891044617, + "num_tokens": 255968301.0, + "step": 9888 + }, + { + "epoch": 1.085987261146497, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7756073474884033, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7086824178695679, + "num_tokens": 255991431.0, + "step": 9889 + }, + { + "epoch": 1.0860970788491104, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.702530860900879, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6844430565834045, + "num_tokens": 256023317.0, + "step": 9890 + }, + { + "epoch": 1.0862068965517242, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.1045806407928467, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7032251358032227, + "num_tokens": 256042239.0, + "step": 9891 + }, + { + "epoch": 1.0863167142543377, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.573812484741211, + "learning_rate": 1e-06, + "loss": 0.9237, + "mean_token_accuracy": 0.7292917966842651, + "num_tokens": 256069235.0, + "step": 9892 + }, + { + "epoch": 1.0864265319569515, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.532201051712036, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7112549543380737, + "num_tokens": 256094727.0, + "step": 9893 + }, + { + "epoch": 1.086536349659565, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6507151126861572, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7195253372192383, + "num_tokens": 256116410.0, + "step": 9894 + }, + { + "epoch": 1.0866461673621788, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.65315842628479, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7182355523109436, + "num_tokens": 256140523.0, + "step": 9895 + }, + { + "epoch": 1.0867559850647925, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3718509674072266, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7198803424835205, + "num_tokens": 256173084.0, + "step": 9896 + }, + { + "epoch": 1.086865802767406, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.116515636444092, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6961039304733276, + "num_tokens": 256197211.0, + "step": 9897 + }, + { + "epoch": 1.0869756204700198, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.677616596221924, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7145518064498901, + "num_tokens": 256219446.0, + "step": 9898 + }, + { + "epoch": 1.0870854381726334, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.5429935455322266, + "learning_rate": 1e-06, + "loss": 0.9069, + "mean_token_accuracy": 0.7289310693740845, + "num_tokens": 256240081.0, + "step": 9899 + }, + { + "epoch": 1.0871952558752471, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4713985919952393, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7174938917160034, + "num_tokens": 256268252.0, + "step": 9900 + }, + { + "epoch": 1.0873050735778607, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.798440933227539, + "learning_rate": 1e-06, + "loss": 0.8915, + "mean_token_accuracy": 0.735307514667511, + "num_tokens": 256288236.0, + "step": 9901 + }, + { + "epoch": 1.0874148912804744, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9368836879730225, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7204228639602661, + "num_tokens": 256307306.0, + "step": 9902 + }, + { + "epoch": 1.0875247089830882, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4904770851135254, + "learning_rate": 1e-06, + "loss": 1.0914, + "mean_token_accuracy": 0.6820788383483887, + "num_tokens": 256336305.0, + "step": 9903 + }, + { + "epoch": 1.0876345266857017, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.634387254714966, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.724793553352356, + "num_tokens": 256360289.0, + "step": 9904 + }, + { + "epoch": 1.0877443443883155, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.930957555770874, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7136549949645996, + "num_tokens": 256390078.0, + "step": 9905 + }, + { + "epoch": 1.087854162090929, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3643414974212646, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6987018585205078, + "num_tokens": 256420136.0, + "step": 9906 + }, + { + "epoch": 1.0879639797935428, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.480048179626465, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7096803188323975, + "num_tokens": 256448256.0, + "step": 9907 + }, + { + "epoch": 1.0880737974961563, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.40706467628479, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6956948041915894, + "num_tokens": 256477398.0, + "step": 9908 + }, + { + "epoch": 1.08818361519877, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9076225757598877, + "learning_rate": 1e-06, + "loss": 0.9353, + "mean_token_accuracy": 0.7240455150604248, + "num_tokens": 256499772.0, + "step": 9909 + }, + { + "epoch": 1.0882934329013838, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6332192420959473, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7036714553833008, + "num_tokens": 256528746.0, + "step": 9910 + }, + { + "epoch": 1.0884032506039973, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9144487380981445, + "learning_rate": 1e-06, + "loss": 1.1363, + "mean_token_accuracy": 0.6708941459655762, + "num_tokens": 256553700.0, + "step": 9911 + }, + { + "epoch": 1.088513068306611, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.479647159576416, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.712678074836731, + "num_tokens": 256581267.0, + "step": 9912 + }, + { + "epoch": 1.0886228860092246, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6052238941192627, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6898319721221924, + "num_tokens": 256606458.0, + "step": 9913 + }, + { + "epoch": 1.0887327037118384, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.6553266048431396, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7121931910514832, + "num_tokens": 256631026.0, + "step": 9914 + }, + { + "epoch": 1.088842521414452, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.4445700645446777, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6965848207473755, + "num_tokens": 256659438.0, + "step": 9915 + }, + { + "epoch": 1.0889523391170657, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.606731414794922, + "learning_rate": 1e-06, + "loss": 1.1125, + "mean_token_accuracy": 0.6756211519241333, + "num_tokens": 256690266.0, + "step": 9916 + }, + { + "epoch": 1.0890621568196792, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.7906486988067627, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7201292514801025, + "num_tokens": 256719995.0, + "step": 9917 + }, + { + "epoch": 1.089171974522293, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.4702823162078857, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7022688984870911, + "num_tokens": 256741027.0, + "step": 9918 + }, + { + "epoch": 1.0892817922249067, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.787482261657715, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6776108741760254, + "num_tokens": 256764856.0, + "step": 9919 + }, + { + "epoch": 1.0893916099275203, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5226247310638428, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6919838190078735, + "num_tokens": 256791611.0, + "step": 9920 + }, + { + "epoch": 1.089501427630134, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.5885367393493652, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7010668516159058, + "num_tokens": 256816799.0, + "step": 9921 + }, + { + "epoch": 1.0896112453327476, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4633052349090576, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6839486360549927, + "num_tokens": 256844587.0, + "step": 9922 + }, + { + "epoch": 1.0897210630353613, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4402341842651367, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7049399018287659, + "num_tokens": 256869769.0, + "step": 9923 + }, + { + "epoch": 1.089830880737975, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3682382106781006, + "learning_rate": 1e-06, + "loss": 1.1167, + "mean_token_accuracy": 0.6820991039276123, + "num_tokens": 256902427.0, + "step": 9924 + }, + { + "epoch": 1.0899406984405886, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6699063777923584, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7045471668243408, + "num_tokens": 256928302.0, + "step": 9925 + }, + { + "epoch": 1.0900505161432024, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9222569465637207, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6945717334747314, + "num_tokens": 256949129.0, + "step": 9926 + }, + { + "epoch": 1.090160333845816, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.490722894668579, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7039523124694824, + "num_tokens": 256979446.0, + "step": 9927 + }, + { + "epoch": 1.0902701515484297, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.620757579803467, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7069903612136841, + "num_tokens": 257006538.0, + "step": 9928 + }, + { + "epoch": 1.0903799692510432, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6266443729400635, + "learning_rate": 1e-06, + "loss": 0.9407, + "mean_token_accuracy": 0.722461462020874, + "num_tokens": 257030709.0, + "step": 9929 + }, + { + "epoch": 1.090489786953657, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6637139320373535, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7164580821990967, + "num_tokens": 257054215.0, + "step": 9930 + }, + { + "epoch": 1.0905996046562705, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.1793711185455322, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7091240882873535, + "num_tokens": 257080815.0, + "step": 9931 + }, + { + "epoch": 1.0907094223588842, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6092610359191895, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7156557440757751, + "num_tokens": 257105877.0, + "step": 9932 + }, + { + "epoch": 1.090819240061498, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.811145544052124, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7125505208969116, + "num_tokens": 257128601.0, + "step": 9933 + }, + { + "epoch": 1.0909290577641115, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8256311416625977, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7182943820953369, + "num_tokens": 257149597.0, + "step": 9934 + }, + { + "epoch": 1.0910388754667253, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9940853118896484, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6972455382347107, + "num_tokens": 257171312.0, + "step": 9935 + }, + { + "epoch": 1.0911486931693388, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4525740146636963, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.695974588394165, + "num_tokens": 257204631.0, + "step": 9936 + }, + { + "epoch": 1.0912585108719526, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7398793697357178, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6935054659843445, + "num_tokens": 257228960.0, + "step": 9937 + }, + { + "epoch": 1.0913683285745663, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.768500804901123, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7156279683113098, + "num_tokens": 257251455.0, + "step": 9938 + }, + { + "epoch": 1.0914781462771799, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6589431762695312, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7058858871459961, + "num_tokens": 257277132.0, + "step": 9939 + }, + { + "epoch": 1.0915879639797936, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.038905382156372, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.7015970945358276, + "num_tokens": 257301807.0, + "step": 9940 + }, + { + "epoch": 1.0916977816824072, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3510053157806396, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.700218677520752, + "num_tokens": 257333641.0, + "step": 9941 + }, + { + "epoch": 1.091807599385021, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.464862108230591, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7170202136039734, + "num_tokens": 257361424.0, + "step": 9942 + }, + { + "epoch": 1.0919174170876345, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8670756816864014, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7003673911094666, + "num_tokens": 257383212.0, + "step": 9943 + }, + { + "epoch": 1.0920272347902482, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.556473970413208, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7000927329063416, + "num_tokens": 257412798.0, + "step": 9944 + }, + { + "epoch": 1.0921370524928617, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.642570972442627, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6987794637680054, + "num_tokens": 257440412.0, + "step": 9945 + }, + { + "epoch": 1.0922468701954755, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.56897234916687, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7104103565216064, + "num_tokens": 257468699.0, + "step": 9946 + }, + { + "epoch": 1.0923566878980893, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.414520263671875, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6939702033996582, + "num_tokens": 257497606.0, + "step": 9947 + }, + { + "epoch": 1.0924665056007028, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8619964122772217, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.719773530960083, + "num_tokens": 257520184.0, + "step": 9948 + }, + { + "epoch": 1.0925763233033166, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.442589044570923, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6844114661216736, + "num_tokens": 257550058.0, + "step": 9949 + }, + { + "epoch": 1.09268614100593, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.516620635986328, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6977840065956116, + "num_tokens": 257577836.0, + "step": 9950 + }, + { + "epoch": 1.0927959587085438, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.846987009048462, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6943176984786987, + "num_tokens": 257614681.0, + "step": 9951 + }, + { + "epoch": 1.0929057764111574, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6766841411590576, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7057783007621765, + "num_tokens": 257638113.0, + "step": 9952 + }, + { + "epoch": 1.0930155941137711, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5938520431518555, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7185530066490173, + "num_tokens": 257661459.0, + "step": 9953 + }, + { + "epoch": 1.093125411816385, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3135547637939453, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6890577077865601, + "num_tokens": 257694517.0, + "step": 9954 + }, + { + "epoch": 1.0932352295189984, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.450802803039551, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7180510759353638, + "num_tokens": 257721198.0, + "step": 9955 + }, + { + "epoch": 1.0933450472216122, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7209970951080322, + "learning_rate": 1e-06, + "loss": 0.907, + "mean_token_accuracy": 0.7301859855651855, + "num_tokens": 257741074.0, + "step": 9956 + }, + { + "epoch": 1.0934548649242257, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.787144422531128, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6939966678619385, + "num_tokens": 257767076.0, + "step": 9957 + }, + { + "epoch": 1.0935646826268395, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4501771926879883, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6822197437286377, + "num_tokens": 257798867.0, + "step": 9958 + }, + { + "epoch": 1.093674500329453, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.868936777114868, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7185858488082886, + "num_tokens": 257821000.0, + "step": 9959 + }, + { + "epoch": 1.0937843180320668, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.794898509979248, + "learning_rate": 1e-06, + "loss": 0.9178, + "mean_token_accuracy": 0.7270879745483398, + "num_tokens": 257845751.0, + "step": 9960 + }, + { + "epoch": 1.0938941357346805, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5943892002105713, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7105897068977356, + "num_tokens": 257870734.0, + "step": 9961 + }, + { + "epoch": 1.094003953437294, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.82881236076355, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6992024183273315, + "num_tokens": 257892104.0, + "step": 9962 + }, + { + "epoch": 1.0941137711399078, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.674436092376709, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6927148103713989, + "num_tokens": 257918169.0, + "step": 9963 + }, + { + "epoch": 1.0942235888425214, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.902186155319214, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7125758528709412, + "num_tokens": 257939361.0, + "step": 9964 + }, + { + "epoch": 1.0943334065451351, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7700891494750977, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7081989645957947, + "num_tokens": 257962990.0, + "step": 9965 + }, + { + "epoch": 1.0944432242477486, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 16.802350997924805, + "learning_rate": 1e-06, + "loss": 0.8873, + "mean_token_accuracy": 0.7367945909500122, + "num_tokens": 257980096.0, + "step": 9966 + }, + { + "epoch": 1.0945530419503624, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6431500911712646, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7240512371063232, + "num_tokens": 258007207.0, + "step": 9967 + }, + { + "epoch": 1.0946628596529762, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3691458702087402, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.690566897392273, + "num_tokens": 258043075.0, + "step": 9968 + }, + { + "epoch": 1.0947726773555897, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4587652683258057, + "learning_rate": 1e-06, + "loss": 0.9373, + "mean_token_accuracy": 0.7271510362625122, + "num_tokens": 258073828.0, + "step": 9969 + }, + { + "epoch": 1.0948824950582035, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.713322401046753, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6910154819488525, + "num_tokens": 258100626.0, + "step": 9970 + }, + { + "epoch": 1.094992312760817, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6280720233917236, + "learning_rate": 1e-06, + "loss": 0.8844, + "mean_token_accuracy": 0.7414026260375977, + "num_tokens": 258125951.0, + "step": 9971 + }, + { + "epoch": 1.0951021304634307, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3680708408355713, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7249700427055359, + "num_tokens": 258151984.0, + "step": 9972 + }, + { + "epoch": 1.0952119481660443, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6368408203125, + "learning_rate": 1e-06, + "loss": 1.0765, + "mean_token_accuracy": 0.6867786645889282, + "num_tokens": 258177525.0, + "step": 9973 + }, + { + "epoch": 1.095321765868658, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.759701728820801, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7020421028137207, + "num_tokens": 258203254.0, + "step": 9974 + }, + { + "epoch": 1.0954315835712718, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.606902837753296, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6986035108566284, + "num_tokens": 258230192.0, + "step": 9975 + }, + { + "epoch": 1.0955414012738853, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8725459575653076, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6897522211074829, + "num_tokens": 258253607.0, + "step": 9976 + }, + { + "epoch": 1.095651218976499, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.787688732147217, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7075657844543457, + "num_tokens": 258276578.0, + "step": 9977 + }, + { + "epoch": 1.0957610366791126, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6940910816192627, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7035287618637085, + "num_tokens": 258302613.0, + "step": 9978 + }, + { + "epoch": 1.0958708543817264, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3886020183563232, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6942127346992493, + "num_tokens": 258337946.0, + "step": 9979 + }, + { + "epoch": 1.09598067208434, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4143502712249756, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6913654804229736, + "num_tokens": 258369426.0, + "step": 9980 + }, + { + "epoch": 1.0960904897869537, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8260388374328613, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7096465826034546, + "num_tokens": 258391493.0, + "step": 9981 + }, + { + "epoch": 1.0962003074895672, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4770939350128174, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7043672204017639, + "num_tokens": 258417199.0, + "step": 9982 + }, + { + "epoch": 1.096310125192181, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8533449172973633, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.6987323760986328, + "num_tokens": 258438100.0, + "step": 9983 + }, + { + "epoch": 1.0964199428947947, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5010035037994385, + "learning_rate": 1e-06, + "loss": 0.923, + "mean_token_accuracy": 0.7321527004241943, + "num_tokens": 258463474.0, + "step": 9984 + }, + { + "epoch": 1.0965297605974083, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.684422016143799, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.691794216632843, + "num_tokens": 258488155.0, + "step": 9985 + }, + { + "epoch": 1.096639578300022, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.633363962173462, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7115035057067871, + "num_tokens": 258511403.0, + "step": 9986 + }, + { + "epoch": 1.0967493960026355, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.802478075027466, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6917198300361633, + "num_tokens": 258533843.0, + "step": 9987 + }, + { + "epoch": 1.0968592137052493, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5936758518218994, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7095382213592529, + "num_tokens": 258559380.0, + "step": 9988 + }, + { + "epoch": 1.096969031407863, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5366950035095215, + "learning_rate": 1e-06, + "loss": 1.1563, + "mean_token_accuracy": 0.6665436625480652, + "num_tokens": 258589868.0, + "step": 9989 + }, + { + "epoch": 1.0970788491104766, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6082727909088135, + "learning_rate": 1e-06, + "loss": 0.9137, + "mean_token_accuracy": 0.7270431518554688, + "num_tokens": 258611281.0, + "step": 9990 + }, + { + "epoch": 1.0971886668130904, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.545213460922241, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6858200430870056, + "num_tokens": 258639167.0, + "step": 9991 + }, + { + "epoch": 1.0972984845157039, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5628931522369385, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7074991464614868, + "num_tokens": 258665202.0, + "step": 9992 + }, + { + "epoch": 1.0974083022183176, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.829010009765625, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7184916734695435, + "num_tokens": 258686204.0, + "step": 9993 + }, + { + "epoch": 1.0975181199209312, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.532639503479004, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7181117534637451, + "num_tokens": 258712260.0, + "step": 9994 + }, + { + "epoch": 1.097627937623545, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9541261196136475, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7121660709381104, + "num_tokens": 258732364.0, + "step": 9995 + }, + { + "epoch": 1.0977377553261585, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5829975605010986, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7354661226272583, + "num_tokens": 258758137.0, + "step": 9996 + }, + { + "epoch": 1.0978475730287722, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7044591903686523, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7057286500930786, + "num_tokens": 258781894.0, + "step": 9997 + }, + { + "epoch": 1.097957390731386, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.1858723163604736, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7165650725364685, + "num_tokens": 258816433.0, + "step": 9998 + }, + { + "epoch": 1.0980672084339995, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5975944995880127, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6903259754180908, + "num_tokens": 258842293.0, + "step": 9999 + }, + { + "epoch": 1.0981770261366133, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.644788980484009, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.6994036436080933, + "num_tokens": 258868498.0, + "step": 10000 + }, + { + "epoch": 1.0982868438392268, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6295783519744873, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7019633054733276, + "num_tokens": 258894056.0, + "step": 10001 + }, + { + "epoch": 1.0983966615418406, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9211177825927734, + "learning_rate": 1e-06, + "loss": 0.8919, + "mean_token_accuracy": 0.7331207394599915, + "num_tokens": 258913204.0, + "step": 10002 + }, + { + "epoch": 1.098506479244454, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 16.59241485595703, + "learning_rate": 1e-06, + "loss": 0.8767, + "mean_token_accuracy": 0.7379392385482788, + "num_tokens": 258934560.0, + "step": 10003 + }, + { + "epoch": 1.0986162969470679, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8864309787750244, + "learning_rate": 1e-06, + "loss": 0.9365, + "mean_token_accuracy": 0.7236426472663879, + "num_tokens": 258957202.0, + "step": 10004 + }, + { + "epoch": 1.0987261146496816, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.0431511402130127, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7141962051391602, + "num_tokens": 258978611.0, + "step": 10005 + }, + { + "epoch": 1.0988359323522952, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.661461591720581, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.705278754234314, + "num_tokens": 259004896.0, + "step": 10006 + }, + { + "epoch": 1.098945750054909, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.9223453998565674, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7011986970901489, + "num_tokens": 259029899.0, + "step": 10007 + }, + { + "epoch": 1.0990555677575224, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.450871706008911, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7162997126579285, + "num_tokens": 259058605.0, + "step": 10008 + }, + { + "epoch": 1.0991653854601362, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5363099575042725, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7055182456970215, + "num_tokens": 259082524.0, + "step": 10009 + }, + { + "epoch": 1.0992752031627497, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6067333221435547, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7123545408248901, + "num_tokens": 259107190.0, + "step": 10010 + }, + { + "epoch": 1.0993850208653635, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3321359157562256, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7078428268432617, + "num_tokens": 259140725.0, + "step": 10011 + }, + { + "epoch": 1.0994948385679773, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.621131658554077, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6869875192642212, + "num_tokens": 259166114.0, + "step": 10012 + }, + { + "epoch": 1.0996046562705908, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.702155113220215, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.708291232585907, + "num_tokens": 259188512.0, + "step": 10013 + }, + { + "epoch": 1.0997144739732045, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6278793811798096, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6970021724700928, + "num_tokens": 259213322.0, + "step": 10014 + }, + { + "epoch": 1.099824291675818, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 3.0101401805877686, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7010802030563354, + "num_tokens": 259244247.0, + "step": 10015 + }, + { + "epoch": 1.0999341093784318, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.523473024368286, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7159441113471985, + "num_tokens": 259272335.0, + "step": 10016 + }, + { + "epoch": 1.1000439270810454, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5184433460235596, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.713194727897644, + "num_tokens": 259299824.0, + "step": 10017 + }, + { + "epoch": 1.1001537447836591, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.547950029373169, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7163786888122559, + "num_tokens": 259322970.0, + "step": 10018 + }, + { + "epoch": 1.1002635624862729, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.445345163345337, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7030236721038818, + "num_tokens": 259352116.0, + "step": 10019 + }, + { + "epoch": 1.1003733801888864, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.543551206588745, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6872169375419617, + "num_tokens": 259383382.0, + "step": 10020 + }, + { + "epoch": 1.1004831978915002, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.3194808959960938, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7077302932739258, + "num_tokens": 259412079.0, + "step": 10021 + }, + { + "epoch": 1.1005930155941137, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.35420823097229, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6994372606277466, + "num_tokens": 259445070.0, + "step": 10022 + }, + { + "epoch": 1.1007028332967275, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5871903896331787, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7071545124053955, + "num_tokens": 259473069.0, + "step": 10023 + }, + { + "epoch": 1.100812650999341, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5985379219055176, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7034918069839478, + "num_tokens": 259498640.0, + "step": 10024 + }, + { + "epoch": 1.1009224687019548, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6362736225128174, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.683610737323761, + "num_tokens": 259525249.0, + "step": 10025 + }, + { + "epoch": 1.1010322864045685, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4213016033172607, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6828645467758179, + "num_tokens": 259558002.0, + "step": 10026 + }, + { + "epoch": 1.101142104107182, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.331660509109497, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7140526175498962, + "num_tokens": 259587005.0, + "step": 10027 + }, + { + "epoch": 1.1012519218097958, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 3.3050968647003174, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6907711625099182, + "num_tokens": 259608555.0, + "step": 10028 + }, + { + "epoch": 1.1013617395124093, + "ewc_loss": 9.000301361083984e-06, + "grad_norm": 2.56785249710083, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7037840485572815, + "num_tokens": 259632527.0, + "step": 10029 + }, + { + "epoch": 1.101471557215023, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.560781717300415, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7267924547195435, + "num_tokens": 259657114.0, + "step": 10030 + }, + { + "epoch": 1.1015813749176366, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.54081654548645, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7075552940368652, + "num_tokens": 259687056.0, + "step": 10031 + }, + { + "epoch": 1.1016911926202504, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6724376678466797, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7199200391769409, + "num_tokens": 259711041.0, + "step": 10032 + }, + { + "epoch": 1.1018010103228641, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5905816555023193, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6946724653244019, + "num_tokens": 259735799.0, + "step": 10033 + }, + { + "epoch": 1.1019108280254777, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7006022930145264, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7084981799125671, + "num_tokens": 259759245.0, + "step": 10034 + }, + { + "epoch": 1.1020206457280914, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5209031105041504, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7000694870948792, + "num_tokens": 259786430.0, + "step": 10035 + }, + { + "epoch": 1.102130463430705, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5424599647521973, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6920029520988464, + "num_tokens": 259812205.0, + "step": 10036 + }, + { + "epoch": 1.1022402811333187, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.738919496536255, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7098519802093506, + "num_tokens": 259834090.0, + "step": 10037 + }, + { + "epoch": 1.1023500988359323, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.655233860015869, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.7059714794158936, + "num_tokens": 259859684.0, + "step": 10038 + }, + { + "epoch": 1.102459916538546, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.4224464893341064, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7174294590950012, + "num_tokens": 259887694.0, + "step": 10039 + }, + { + "epoch": 1.1025697342411598, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.520378589630127, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6968443393707275, + "num_tokens": 259915361.0, + "step": 10040 + }, + { + "epoch": 1.1026795519437733, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.8085696697235107, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.697749137878418, + "num_tokens": 259939604.0, + "step": 10041 + }, + { + "epoch": 1.102789369646387, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.60386323928833, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6919102668762207, + "num_tokens": 259964570.0, + "step": 10042 + }, + { + "epoch": 1.1028991873490006, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.579296827316284, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7186926603317261, + "num_tokens": 259988233.0, + "step": 10043 + }, + { + "epoch": 1.1030090050516144, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.509004592895508, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.6928226351737976, + "num_tokens": 260016806.0, + "step": 10044 + }, + { + "epoch": 1.103118822754228, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.552464246749878, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.6971828937530518, + "num_tokens": 260044011.0, + "step": 10045 + }, + { + "epoch": 1.1032286404568417, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.7850730419158936, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7101424932479858, + "num_tokens": 260068734.0, + "step": 10046 + }, + { + "epoch": 1.1033384581594552, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.542039155960083, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7139307856559753, + "num_tokens": 260093825.0, + "step": 10047 + }, + { + "epoch": 1.103448275862069, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5203909873962402, + "learning_rate": 1e-06, + "loss": 1.1043, + "mean_token_accuracy": 0.6801591515541077, + "num_tokens": 260122935.0, + "step": 10048 + }, + { + "epoch": 1.1035580935646827, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.5834403038024902, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7178674936294556, + "num_tokens": 260148211.0, + "step": 10049 + }, + { + "epoch": 1.1036679112672962, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.753289222717285, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7146753072738647, + "num_tokens": 260171463.0, + "step": 10050 + }, + { + "epoch": 1.10377772896991, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6525166034698486, + "learning_rate": 1e-06, + "loss": 0.9462, + "mean_token_accuracy": 0.7211393117904663, + "num_tokens": 260194896.0, + "step": 10051 + }, + { + "epoch": 1.1038875466725235, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 16.73272705078125, + "learning_rate": 1e-06, + "loss": 1.0816, + "mean_token_accuracy": 0.6878657937049866, + "num_tokens": 260218383.0, + "step": 10052 + }, + { + "epoch": 1.1039973643751373, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6495907306671143, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.696992814540863, + "num_tokens": 260244991.0, + "step": 10053 + }, + { + "epoch": 1.104107182077751, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.613663911819458, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7102330327033997, + "num_tokens": 260271284.0, + "step": 10054 + }, + { + "epoch": 1.1042169997803646, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8113017082214355, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6913134455680847, + "num_tokens": 260294880.0, + "step": 10055 + }, + { + "epoch": 1.1043268174829783, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8498222827911377, + "learning_rate": 1e-06, + "loss": 1.1102, + "mean_token_accuracy": 0.6757810115814209, + "num_tokens": 260322375.0, + "step": 10056 + }, + { + "epoch": 1.1044366351855919, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4751055240631104, + "learning_rate": 1e-06, + "loss": 1.0956, + "mean_token_accuracy": 0.6842539310455322, + "num_tokens": 260352339.0, + "step": 10057 + }, + { + "epoch": 1.1045464528882056, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.755932569503784, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.7019757032394409, + "num_tokens": 260376081.0, + "step": 10058 + }, + { + "epoch": 1.1046562705908192, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.666459321975708, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6978892683982849, + "num_tokens": 260400360.0, + "step": 10059 + }, + { + "epoch": 1.104766088293433, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.544990062713623, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7097005844116211, + "num_tokens": 260427736.0, + "step": 10060 + }, + { + "epoch": 1.1048759059960465, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.628627300262451, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.7002317309379578, + "num_tokens": 260451174.0, + "step": 10061 + }, + { + "epoch": 1.1049857236986602, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 3.2507011890411377, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7019776701927185, + "num_tokens": 260474631.0, + "step": 10062 + }, + { + "epoch": 1.105095541401274, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.669743299484253, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7197920680046082, + "num_tokens": 260499111.0, + "step": 10063 + }, + { + "epoch": 1.1052053591038875, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.859241008758545, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7147156000137329, + "num_tokens": 260520362.0, + "step": 10064 + }, + { + "epoch": 1.1053151768065013, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.6387574672698975, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7181749939918518, + "num_tokens": 260547511.0, + "step": 10065 + }, + { + "epoch": 1.1054249945091148, + "ewc_loss": 9.059906005859375e-06, + "grad_norm": 2.686002016067505, + "learning_rate": 1e-06, + "loss": 0.8979, + "mean_token_accuracy": 0.7322515249252319, + "num_tokens": 260569000.0, + "step": 10066 + }, + { + "epoch": 1.1055348122117286, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.2340645790100098, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7127599716186523, + "num_tokens": 260602601.0, + "step": 10067 + }, + { + "epoch": 1.105644629914342, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.520986557006836, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6923457384109497, + "num_tokens": 260631014.0, + "step": 10068 + }, + { + "epoch": 1.1057544476169558, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.613035202026367, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7132546901702881, + "num_tokens": 260656564.0, + "step": 10069 + }, + { + "epoch": 1.1058642653195696, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.761706590652466, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7147183418273926, + "num_tokens": 260678529.0, + "step": 10070 + }, + { + "epoch": 1.1059740830221831, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.075089931488037, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7182748317718506, + "num_tokens": 260703650.0, + "step": 10071 + }, + { + "epoch": 1.106083900724797, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7628254890441895, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7061066627502441, + "num_tokens": 260725587.0, + "step": 10072 + }, + { + "epoch": 1.1061937184274104, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5260915756225586, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.725081741809845, + "num_tokens": 260749706.0, + "step": 10073 + }, + { + "epoch": 1.1063035361300242, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.479315757751465, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.726057767868042, + "num_tokens": 260775255.0, + "step": 10074 + }, + { + "epoch": 1.1064133538326377, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 4.3528242111206055, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7066600322723389, + "num_tokens": 260803395.0, + "step": 10075 + }, + { + "epoch": 1.1065231715352515, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8145763874053955, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7060739994049072, + "num_tokens": 260827511.0, + "step": 10076 + }, + { + "epoch": 1.1066329892378652, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.46930193901062, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7077381014823914, + "num_tokens": 260857517.0, + "step": 10077 + }, + { + "epoch": 1.1067428069404788, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.812246799468994, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7135358452796936, + "num_tokens": 260881444.0, + "step": 10078 + }, + { + "epoch": 1.1068526246430925, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4955687522888184, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7128664255142212, + "num_tokens": 260907718.0, + "step": 10079 + }, + { + "epoch": 1.106962442345706, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5002341270446777, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7025388479232788, + "num_tokens": 260934543.0, + "step": 10080 + }, + { + "epoch": 1.1070722600483198, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6887857913970947, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6952501535415649, + "num_tokens": 260960660.0, + "step": 10081 + }, + { + "epoch": 1.1071820777509334, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.549041509628296, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7095363736152649, + "num_tokens": 260988653.0, + "step": 10082 + }, + { + "epoch": 1.1072918954535471, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5292301177978516, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.69754558801651, + "num_tokens": 261017363.0, + "step": 10083 + }, + { + "epoch": 1.1074017131561609, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4584391117095947, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7008624076843262, + "num_tokens": 261044320.0, + "step": 10084 + }, + { + "epoch": 1.1075115308587744, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6316564083099365, + "learning_rate": 1e-06, + "loss": 0.9536, + "mean_token_accuracy": 0.7293497920036316, + "num_tokens": 261069403.0, + "step": 10085 + }, + { + "epoch": 1.1076213485613882, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.02612566947937, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7116760015487671, + "num_tokens": 261088395.0, + "step": 10086 + }, + { + "epoch": 1.1077311662640017, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.713038206100464, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7223938703536987, + "num_tokens": 261113142.0, + "step": 10087 + }, + { + "epoch": 1.1078409839666155, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7922754287719727, + "learning_rate": 1e-06, + "loss": 1.1075, + "mean_token_accuracy": 0.6810241937637329, + "num_tokens": 261135076.0, + "step": 10088 + }, + { + "epoch": 1.107950801669229, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.643862724304199, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.6986724138259888, + "num_tokens": 261159556.0, + "step": 10089 + }, + { + "epoch": 1.1080606193718427, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.741135835647583, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7130901217460632, + "num_tokens": 261185178.0, + "step": 10090 + }, + { + "epoch": 1.1081704370744565, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5427005290985107, + "learning_rate": 1e-06, + "loss": 0.9326, + "mean_token_accuracy": 0.7329370975494385, + "num_tokens": 261210986.0, + "step": 10091 + }, + { + "epoch": 1.10828025477707, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5091373920440674, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.692078709602356, + "num_tokens": 261242510.0, + "step": 10092 + }, + { + "epoch": 1.1083900724796838, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6353025436401367, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7054199576377869, + "num_tokens": 261267792.0, + "step": 10093 + }, + { + "epoch": 1.1084998901822973, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.795644998550415, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7162981033325195, + "num_tokens": 261293488.0, + "step": 10094 + }, + { + "epoch": 1.108609707884911, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.85207462310791, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7113569974899292, + "num_tokens": 261313720.0, + "step": 10095 + }, + { + "epoch": 1.1087195255875246, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 80.56085968017578, + "learning_rate": 1e-06, + "loss": 0.921, + "mean_token_accuracy": 0.7271932363510132, + "num_tokens": 261331486.0, + "step": 10096 + }, + { + "epoch": 1.1088293432901384, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.99526309967041, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7081326246261597, + "num_tokens": 261353896.0, + "step": 10097 + }, + { + "epoch": 1.108939160992752, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.9981930255889893, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.700637936592102, + "num_tokens": 261376889.0, + "step": 10098 + }, + { + "epoch": 1.1090489786953657, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.0510926246643066, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7092570066452026, + "num_tokens": 261395160.0, + "step": 10099 + }, + { + "epoch": 1.1091587963979794, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6778407096862793, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7115901708602905, + "num_tokens": 261421634.0, + "step": 10100 + }, + { + "epoch": 1.109268614100593, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5584142208099365, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7063525915145874, + "num_tokens": 261450939.0, + "step": 10101 + }, + { + "epoch": 1.1093784318032067, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.800107955932617, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7110263109207153, + "num_tokens": 261474246.0, + "step": 10102 + }, + { + "epoch": 1.1094882495058203, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5081918239593506, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7264265418052673, + "num_tokens": 261499381.0, + "step": 10103 + }, + { + "epoch": 1.109598067208434, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3608124256134033, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7149228453636169, + "num_tokens": 261529173.0, + "step": 10104 + }, + { + "epoch": 1.1097078849110478, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3368825912475586, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7023009061813354, + "num_tokens": 261562491.0, + "step": 10105 + }, + { + "epoch": 1.1098177026136613, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.645660400390625, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7101207971572876, + "num_tokens": 261587022.0, + "step": 10106 + }, + { + "epoch": 1.109927520316275, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.623414993286133, + "learning_rate": 1e-06, + "loss": 1.0937, + "mean_token_accuracy": 0.6973042488098145, + "num_tokens": 261613628.0, + "step": 10107 + }, + { + "epoch": 1.1100373380188886, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.702791690826416, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7138781547546387, + "num_tokens": 261634952.0, + "step": 10108 + }, + { + "epoch": 1.1101471557215024, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7388486862182617, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7176175117492676, + "num_tokens": 261657788.0, + "step": 10109 + }, + { + "epoch": 1.110256973424116, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4196557998657227, + "learning_rate": 1e-06, + "loss": 1.1706, + "mean_token_accuracy": 0.6628755927085876, + "num_tokens": 261691464.0, + "step": 10110 + }, + { + "epoch": 1.1103667911267296, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6621148586273193, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7219957113265991, + "num_tokens": 261713940.0, + "step": 10111 + }, + { + "epoch": 1.1104766088293432, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.458756446838379, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.7197401523590088, + "num_tokens": 261739358.0, + "step": 10112 + }, + { + "epoch": 1.110586426531957, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.662191390991211, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7206160426139832, + "num_tokens": 261765350.0, + "step": 10113 + }, + { + "epoch": 1.1106962442345707, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7537002563476562, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7058526277542114, + "num_tokens": 261790069.0, + "step": 10114 + }, + { + "epoch": 1.1108060619371842, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4840662479400635, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7097647786140442, + "num_tokens": 261815604.0, + "step": 10115 + }, + { + "epoch": 1.110915879639798, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5567891597747803, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6919640898704529, + "num_tokens": 261841623.0, + "step": 10116 + }, + { + "epoch": 1.1110256973424115, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.884789228439331, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6860815286636353, + "num_tokens": 261862501.0, + "step": 10117 + }, + { + "epoch": 1.1111355150450253, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6426119804382324, + "learning_rate": 1e-06, + "loss": 1.0857, + "mean_token_accuracy": 0.6903830766677856, + "num_tokens": 261890576.0, + "step": 10118 + }, + { + "epoch": 1.111245332747639, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.63472318649292, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.722524881362915, + "num_tokens": 261915171.0, + "step": 10119 + }, + { + "epoch": 1.1113551504502526, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4971792697906494, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7218896150588989, + "num_tokens": 261942007.0, + "step": 10120 + }, + { + "epoch": 1.1114649681528663, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.779933214187622, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6945335268974304, + "num_tokens": 261966752.0, + "step": 10121 + }, + { + "epoch": 1.1115747858554799, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.791141986846924, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.7017461657524109, + "num_tokens": 261990830.0, + "step": 10122 + }, + { + "epoch": 1.1116846035580936, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7132067680358887, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6948338747024536, + "num_tokens": 262015820.0, + "step": 10123 + }, + { + "epoch": 1.1117944212607072, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.958233594894409, + "learning_rate": 1e-06, + "loss": 0.9364, + "mean_token_accuracy": 0.7260481119155884, + "num_tokens": 262034653.0, + "step": 10124 + }, + { + "epoch": 1.111904238963321, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.507916212081909, + "learning_rate": 1e-06, + "loss": 1.1378, + "mean_token_accuracy": 0.6734597086906433, + "num_tokens": 262065968.0, + "step": 10125 + }, + { + "epoch": 1.1120140566659344, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3260838985443115, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6923213005065918, + "num_tokens": 262102326.0, + "step": 10126 + }, + { + "epoch": 1.1121238743685482, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4499995708465576, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6929518580436707, + "num_tokens": 262129269.0, + "step": 10127 + }, + { + "epoch": 1.112233692071162, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.393846273422241, + "learning_rate": 1e-06, + "loss": 1.1457, + "mean_token_accuracy": 0.6705413460731506, + "num_tokens": 262159214.0, + "step": 10128 + }, + { + "epoch": 1.1123435097737755, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4970593452453613, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6897668838500977, + "num_tokens": 262187605.0, + "step": 10129 + }, + { + "epoch": 1.1124533274763893, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.629671573638916, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6892380714416504, + "num_tokens": 262215300.0, + "step": 10130 + }, + { + "epoch": 1.1125631451790028, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6295065879821777, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.710914134979248, + "num_tokens": 262237667.0, + "step": 10131 + }, + { + "epoch": 1.1126729628816165, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.749783515930176, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7078157663345337, + "num_tokens": 262260847.0, + "step": 10132 + }, + { + "epoch": 1.11278278058423, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4236223697662354, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7247762084007263, + "num_tokens": 262288164.0, + "step": 10133 + }, + { + "epoch": 1.1128925982868438, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4947879314422607, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7051742076873779, + "num_tokens": 262315357.0, + "step": 10134 + }, + { + "epoch": 1.1130024159894576, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5655689239501953, + "learning_rate": 1e-06, + "loss": 0.8705, + "mean_token_accuracy": 0.7379988431930542, + "num_tokens": 262340817.0, + "step": 10135 + }, + { + "epoch": 1.1131122336920711, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.534754753112793, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7111661434173584, + "num_tokens": 262369774.0, + "step": 10136 + }, + { + "epoch": 1.1132220513946849, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6534125804901123, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6934347152709961, + "num_tokens": 262395648.0, + "step": 10137 + }, + { + "epoch": 1.1133318690972984, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4579968452453613, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7075774669647217, + "num_tokens": 262421124.0, + "step": 10138 + }, + { + "epoch": 1.1134416867999122, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4240379333496094, + "learning_rate": 1e-06, + "loss": 0.921, + "mean_token_accuracy": 0.726233720779419, + "num_tokens": 262446550.0, + "step": 10139 + }, + { + "epoch": 1.1135515045025257, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5191597938537598, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7224173545837402, + "num_tokens": 262471616.0, + "step": 10140 + }, + { + "epoch": 1.1136613222051395, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.1272313594818115, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.708747923374176, + "num_tokens": 262489414.0, + "step": 10141 + }, + { + "epoch": 1.1137711399077532, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3626725673675537, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6932946443557739, + "num_tokens": 262526085.0, + "step": 10142 + }, + { + "epoch": 1.1138809576103668, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6550633907318115, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.703597903251648, + "num_tokens": 262550686.0, + "step": 10143 + }, + { + "epoch": 1.1139907753129805, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7194623947143555, + "learning_rate": 1e-06, + "loss": 1.0832, + "mean_token_accuracy": 0.6906154155731201, + "num_tokens": 262575657.0, + "step": 10144 + }, + { + "epoch": 1.114100593015594, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.58443284034729, + "learning_rate": 1e-06, + "loss": 1.1134, + "mean_token_accuracy": 0.6790956258773804, + "num_tokens": 262603127.0, + "step": 10145 + }, + { + "epoch": 1.1142104107182078, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.760843515396118, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6852431893348694, + "num_tokens": 262628529.0, + "step": 10146 + }, + { + "epoch": 1.1143202284208213, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6567540168762207, + "learning_rate": 1e-06, + "loss": 1.1409, + "mean_token_accuracy": 0.6734139919281006, + "num_tokens": 262656454.0, + "step": 10147 + }, + { + "epoch": 1.114430046123435, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3844454288482666, + "learning_rate": 1e-06, + "loss": 0.9417, + "mean_token_accuracy": 0.7290560007095337, + "num_tokens": 262685153.0, + "step": 10148 + }, + { + "epoch": 1.1145398638260489, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.432319164276123, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.6907467842102051, + "num_tokens": 262713813.0, + "step": 10149 + }, + { + "epoch": 1.1146496815286624, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.426011800765991, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7122401595115662, + "num_tokens": 262743376.0, + "step": 10150 + }, + { + "epoch": 1.1147594992312762, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6624274253845215, + "learning_rate": 1e-06, + "loss": 0.9288, + "mean_token_accuracy": 0.7293631434440613, + "num_tokens": 262770450.0, + "step": 10151 + }, + { + "epoch": 1.1148693169338897, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.708101749420166, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6798940896987915, + "num_tokens": 262795402.0, + "step": 10152 + }, + { + "epoch": 1.1149791346365034, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.0525362491607666, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6891171932220459, + "num_tokens": 262818722.0, + "step": 10153 + }, + { + "epoch": 1.115088952339117, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.624652147293091, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7288146018981934, + "num_tokens": 262841738.0, + "step": 10154 + }, + { + "epoch": 1.1151987700417307, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5371391773223877, + "learning_rate": 1e-06, + "loss": 0.9188, + "mean_token_accuracy": 0.7314615249633789, + "num_tokens": 262865964.0, + "step": 10155 + }, + { + "epoch": 1.1153085877443445, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.0634844303131104, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6918042898178101, + "num_tokens": 262885839.0, + "step": 10156 + }, + { + "epoch": 1.115418405446958, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6960084438323975, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7116997838020325, + "num_tokens": 262908507.0, + "step": 10157 + }, + { + "epoch": 1.1155282231495718, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.819849729537964, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7098972797393799, + "num_tokens": 262931525.0, + "step": 10158 + }, + { + "epoch": 1.1156380408521853, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.941171884536743, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.6937327980995178, + "num_tokens": 262954414.0, + "step": 10159 + }, + { + "epoch": 1.115747858554799, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.750676155090332, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.6927155256271362, + "num_tokens": 262976802.0, + "step": 10160 + }, + { + "epoch": 1.1158576762574126, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4818499088287354, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7082792520523071, + "num_tokens": 263006436.0, + "step": 10161 + }, + { + "epoch": 1.1159674939600264, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.745121479034424, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6907604336738586, + "num_tokens": 263033693.0, + "step": 10162 + }, + { + "epoch": 1.11607731166264, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8742194175720215, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7072548866271973, + "num_tokens": 263055731.0, + "step": 10163 + }, + { + "epoch": 1.1161871293652537, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.7287817001342773, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.6973758935928345, + "num_tokens": 263081564.0, + "step": 10164 + }, + { + "epoch": 1.1162969470678674, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.598327398300171, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.711033821105957, + "num_tokens": 263110244.0, + "step": 10165 + }, + { + "epoch": 1.116406764770481, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7325921058654785, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6933588981628418, + "num_tokens": 263135014.0, + "step": 10166 + }, + { + "epoch": 1.1165165824730947, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.761558771133423, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6861637830734253, + "num_tokens": 263156727.0, + "step": 10167 + }, + { + "epoch": 1.1166264001757082, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6996920108795166, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.7106939554214478, + "num_tokens": 263183522.0, + "step": 10168 + }, + { + "epoch": 1.116736217878322, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.693943738937378, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7221246957778931, + "num_tokens": 263208367.0, + "step": 10169 + }, + { + "epoch": 1.1168460355809358, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.9081954956054688, + "learning_rate": 1e-06, + "loss": 0.8813, + "mean_token_accuracy": 0.7387689352035522, + "num_tokens": 263227569.0, + "step": 10170 + }, + { + "epoch": 1.1169558532835493, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5032896995544434, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7109955549240112, + "num_tokens": 263255747.0, + "step": 10171 + }, + { + "epoch": 1.117065670986163, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8114302158355713, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7086092829704285, + "num_tokens": 263277067.0, + "step": 10172 + }, + { + "epoch": 1.1171754886887766, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.82582688331604, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7130378484725952, + "num_tokens": 263300135.0, + "step": 10173 + }, + { + "epoch": 1.1172853063913903, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.0342600345611572, + "learning_rate": 1e-06, + "loss": 0.9235, + "mean_token_accuracy": 0.7223986387252808, + "num_tokens": 263318790.0, + "step": 10174 + }, + { + "epoch": 1.1173951240940039, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.4391603469848633, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7035185098648071, + "num_tokens": 263348357.0, + "step": 10175 + }, + { + "epoch": 1.1175049417966176, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.660362958908081, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7259613275527954, + "num_tokens": 263372010.0, + "step": 10176 + }, + { + "epoch": 1.1176147594992312, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.752699851989746, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7018643617630005, + "num_tokens": 263397378.0, + "step": 10177 + }, + { + "epoch": 1.117724577201845, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4765398502349854, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7089353203773499, + "num_tokens": 263422840.0, + "step": 10178 + }, + { + "epoch": 1.1178343949044587, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.603233575820923, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7110307216644287, + "num_tokens": 263447140.0, + "step": 10179 + }, + { + "epoch": 1.1179442126070722, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3934884071350098, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6921727657318115, + "num_tokens": 263477950.0, + "step": 10180 + }, + { + "epoch": 1.118054030309686, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.2429587841033936, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6881407499313354, + "num_tokens": 263511333.0, + "step": 10181 + }, + { + "epoch": 1.1181638480122995, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.482888698577881, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7024447917938232, + "num_tokens": 263539008.0, + "step": 10182 + }, + { + "epoch": 1.1182736657149133, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.9104230403900146, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.717115581035614, + "num_tokens": 263559854.0, + "step": 10183 + }, + { + "epoch": 1.1183834834175268, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.4266443252563477, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7069206833839417, + "num_tokens": 263583069.0, + "step": 10184 + }, + { + "epoch": 1.1184933011201406, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8565480709075928, + "learning_rate": 1e-06, + "loss": 0.8844, + "mean_token_accuracy": 0.7428411245346069, + "num_tokens": 263603230.0, + "step": 10185 + }, + { + "epoch": 1.1186031188227543, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4678258895874023, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6981212496757507, + "num_tokens": 263632745.0, + "step": 10186 + }, + { + "epoch": 1.1187129365253679, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6130330562591553, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.697472095489502, + "num_tokens": 263656588.0, + "step": 10187 + }, + { + "epoch": 1.1188227542279816, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.847306489944458, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7184872627258301, + "num_tokens": 263677098.0, + "step": 10188 + }, + { + "epoch": 1.1189325719305951, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7360048294067383, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.691511332988739, + "num_tokens": 263700576.0, + "step": 10189 + }, + { + "epoch": 1.119042389633209, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5450503826141357, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7025048732757568, + "num_tokens": 263728753.0, + "step": 10190 + }, + { + "epoch": 1.1191522073358224, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5280587673187256, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.7019339799880981, + "num_tokens": 263754501.0, + "step": 10191 + }, + { + "epoch": 1.1192620250384362, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.878948211669922, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6942234039306641, + "num_tokens": 263781638.0, + "step": 10192 + }, + { + "epoch": 1.11937184274105, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5766496658325195, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.689898669719696, + "num_tokens": 263806214.0, + "step": 10193 + }, + { + "epoch": 1.1194816604436635, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4528939723968506, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7000874280929565, + "num_tokens": 263834287.0, + "step": 10194 + }, + { + "epoch": 1.1195914781462772, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5812923908233643, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.702842116355896, + "num_tokens": 263861180.0, + "step": 10195 + }, + { + "epoch": 1.1197012958488908, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7435507774353027, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7144253849983215, + "num_tokens": 263883046.0, + "step": 10196 + }, + { + "epoch": 1.1198111135515045, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4783689975738525, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7130801677703857, + "num_tokens": 263910091.0, + "step": 10197 + }, + { + "epoch": 1.119920931254118, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3799548149108887, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.716065526008606, + "num_tokens": 263937454.0, + "step": 10198 + }, + { + "epoch": 1.1200307489567318, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5716798305511475, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7091549634933472, + "num_tokens": 263963185.0, + "step": 10199 + }, + { + "epoch": 1.1201405666593456, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7999656200408936, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7150975465774536, + "num_tokens": 263985334.0, + "step": 10200 + }, + { + "epoch": 1.1202503843619591, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5093135833740234, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6860382556915283, + "num_tokens": 264013050.0, + "step": 10201 + }, + { + "epoch": 1.1203602020645729, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.690641403198242, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.6941035985946655, + "num_tokens": 264040065.0, + "step": 10202 + }, + { + "epoch": 1.1204700197671864, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4530012607574463, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6964719295501709, + "num_tokens": 264067824.0, + "step": 10203 + }, + { + "epoch": 1.1205798374698002, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5122876167297363, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.6914100050926208, + "num_tokens": 264094078.0, + "step": 10204 + }, + { + "epoch": 1.1206896551724137, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.553107738494873, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6969702243804932, + "num_tokens": 264122206.0, + "step": 10205 + }, + { + "epoch": 1.1207994728750275, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7887585163116455, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6901852488517761, + "num_tokens": 264146490.0, + "step": 10206 + }, + { + "epoch": 1.1209092905776412, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 4.113560676574707, + "learning_rate": 1e-06, + "loss": 0.9189, + "mean_token_accuracy": 0.7288574576377869, + "num_tokens": 264162769.0, + "step": 10207 + }, + { + "epoch": 1.1210191082802548, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6819968223571777, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7145870923995972, + "num_tokens": 264185118.0, + "step": 10208 + }, + { + "epoch": 1.1211289259828685, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.335973024368286, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7030303478240967, + "num_tokens": 264217125.0, + "step": 10209 + }, + { + "epoch": 1.121238743685482, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5090415477752686, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7188242077827454, + "num_tokens": 264243982.0, + "step": 10210 + }, + { + "epoch": 1.1213485613880958, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6967668533325195, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6925169825553894, + "num_tokens": 264268132.0, + "step": 10211 + }, + { + "epoch": 1.1214583790907093, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.825845718383789, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7088987827301025, + "num_tokens": 264290913.0, + "step": 10212 + }, + { + "epoch": 1.121568196793323, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.450075387954712, + "learning_rate": 1e-06, + "loss": 0.9532, + "mean_token_accuracy": 0.7201351523399353, + "num_tokens": 264317797.0, + "step": 10213 + }, + { + "epoch": 1.1216780144959368, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.0632171630859375, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7037425637245178, + "num_tokens": 264343062.0, + "step": 10214 + }, + { + "epoch": 1.1217878321985504, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8276185989379883, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7017669677734375, + "num_tokens": 264365747.0, + "step": 10215 + }, + { + "epoch": 1.1218976499011641, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.750917434692383, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.695590078830719, + "num_tokens": 264389155.0, + "step": 10216 + }, + { + "epoch": 1.1220074676037777, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6225924491882324, + "learning_rate": 1e-06, + "loss": 0.8594, + "mean_token_accuracy": 0.7513223886489868, + "num_tokens": 264412724.0, + "step": 10217 + }, + { + "epoch": 1.1221172853063914, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8786978721618652, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6922628879547119, + "num_tokens": 264437737.0, + "step": 10218 + }, + { + "epoch": 1.122227103009005, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.619220733642578, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6879182457923889, + "num_tokens": 264464141.0, + "step": 10219 + }, + { + "epoch": 1.1223369207116187, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.587606191635132, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7001456618309021, + "num_tokens": 264489951.0, + "step": 10220 + }, + { + "epoch": 1.1224467384142325, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.335181713104248, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6912247538566589, + "num_tokens": 264522420.0, + "step": 10221 + }, + { + "epoch": 1.122556556116846, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.525944471359253, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6993871331214905, + "num_tokens": 264550110.0, + "step": 10222 + }, + { + "epoch": 1.1226663738194598, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.556140661239624, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6902092695236206, + "num_tokens": 264577054.0, + "step": 10223 + }, + { + "epoch": 1.1227761915220733, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7509605884552, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7028380632400513, + "num_tokens": 264602938.0, + "step": 10224 + }, + { + "epoch": 1.122886009224687, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.561155080795288, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.7064788341522217, + "num_tokens": 264629059.0, + "step": 10225 + }, + { + "epoch": 1.1229958269273006, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.2769906520843506, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7177606225013733, + "num_tokens": 264655928.0, + "step": 10226 + }, + { + "epoch": 1.1231056446299144, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6301634311676025, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7171765565872192, + "num_tokens": 264681093.0, + "step": 10227 + }, + { + "epoch": 1.123215462332528, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.697211503982544, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7198376059532166, + "num_tokens": 264705936.0, + "step": 10228 + }, + { + "epoch": 1.1233252800351416, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.604313373565674, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.6993869543075562, + "num_tokens": 264731489.0, + "step": 10229 + }, + { + "epoch": 1.1234350977377554, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.729322671890259, + "learning_rate": 1e-06, + "loss": 0.9327, + "mean_token_accuracy": 0.7245201468467712, + "num_tokens": 264753490.0, + "step": 10230 + }, + { + "epoch": 1.123544915440369, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.609299898147583, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.7012256383895874, + "num_tokens": 264779633.0, + "step": 10231 + }, + { + "epoch": 1.1236547331429827, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.62070631980896, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6962434649467468, + "num_tokens": 264804496.0, + "step": 10232 + }, + { + "epoch": 1.1237645508455962, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.74545955657959, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6814857721328735, + "num_tokens": 264832883.0, + "step": 10233 + }, + { + "epoch": 1.12387436854821, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.559878349304199, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7207428812980652, + "num_tokens": 264858624.0, + "step": 10234 + }, + { + "epoch": 1.1239841862508237, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.9217000007629395, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7001996636390686, + "num_tokens": 264880895.0, + "step": 10235 + }, + { + "epoch": 1.1240940039534373, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.542457342147827, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.687525749206543, + "num_tokens": 264907135.0, + "step": 10236 + }, + { + "epoch": 1.124203821656051, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.571134567260742, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.7008742690086365, + "num_tokens": 264934646.0, + "step": 10237 + }, + { + "epoch": 1.1243136393586646, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7406206130981445, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7093302607536316, + "num_tokens": 264963994.0, + "step": 10238 + }, + { + "epoch": 1.1244234570612783, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.183465003967285, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7114253640174866, + "num_tokens": 264988809.0, + "step": 10239 + }, + { + "epoch": 1.1245332747638919, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.843726634979248, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7152801752090454, + "num_tokens": 265011918.0, + "step": 10240 + }, + { + "epoch": 1.1246430924665056, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.653273820877075, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7032074332237244, + "num_tokens": 265040220.0, + "step": 10241 + }, + { + "epoch": 1.1247529101691192, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.634606122970581, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7208045721054077, + "num_tokens": 265066840.0, + "step": 10242 + }, + { + "epoch": 1.124862727871733, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.8642971515655518, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7160043716430664, + "num_tokens": 265090591.0, + "step": 10243 + }, + { + "epoch": 1.1249725455743467, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6062164306640625, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.6974483728408813, + "num_tokens": 265117912.0, + "step": 10244 + }, + { + "epoch": 1.1250823632769602, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6130473613739014, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7019702196121216, + "num_tokens": 265143335.0, + "step": 10245 + }, + { + "epoch": 1.125192180979574, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.658684730529785, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6822192668914795, + "num_tokens": 265168875.0, + "step": 10246 + }, + { + "epoch": 1.1253019986821875, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5874056816101074, + "learning_rate": 1e-06, + "loss": 1.1151, + "mean_token_accuracy": 0.68611079454422, + "num_tokens": 265195136.0, + "step": 10247 + }, + { + "epoch": 1.1254118163848013, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4292800426483154, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7010164260864258, + "num_tokens": 265222875.0, + "step": 10248 + }, + { + "epoch": 1.125521634087415, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.0049941539764404, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.711336612701416, + "num_tokens": 265242173.0, + "step": 10249 + }, + { + "epoch": 1.1256314517900285, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7663230895996094, + "learning_rate": 1e-06, + "loss": 0.9189, + "mean_token_accuracy": 0.7258427143096924, + "num_tokens": 265265667.0, + "step": 10250 + }, + { + "epoch": 1.1257412694926423, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.57672119140625, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6853410005569458, + "num_tokens": 265295514.0, + "step": 10251 + }, + { + "epoch": 1.1258510871952558, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.731520891189575, + "learning_rate": 1e-06, + "loss": 0.9251, + "mean_token_accuracy": 0.7246304750442505, + "num_tokens": 265317261.0, + "step": 10252 + }, + { + "epoch": 1.1259609048978696, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.710753917694092, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6912029981613159, + "num_tokens": 265342770.0, + "step": 10253 + }, + { + "epoch": 1.1260707226004831, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.98089599609375, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7062269449234009, + "num_tokens": 265363346.0, + "step": 10254 + }, + { + "epoch": 1.1261805403030969, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.527369737625122, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.7024067640304565, + "num_tokens": 265391417.0, + "step": 10255 + }, + { + "epoch": 1.1262903580057104, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.371622085571289, + "learning_rate": 1e-06, + "loss": 0.8982, + "mean_token_accuracy": 0.7310612201690674, + "num_tokens": 265412658.0, + "step": 10256 + }, + { + "epoch": 1.1264001757083242, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.688154458999634, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7131551504135132, + "num_tokens": 265439582.0, + "step": 10257 + }, + { + "epoch": 1.126509993410938, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.5290720462799072, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7039358615875244, + "num_tokens": 265466946.0, + "step": 10258 + }, + { + "epoch": 1.1266198111135515, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.292539596557617, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7006104588508606, + "num_tokens": 265499854.0, + "step": 10259 + }, + { + "epoch": 1.1267296288161652, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.374347686767578, + "learning_rate": 1e-06, + "loss": 0.9416, + "mean_token_accuracy": 0.7179521322250366, + "num_tokens": 265523938.0, + "step": 10260 + }, + { + "epoch": 1.1268394465187788, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3364546298980713, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.706820011138916, + "num_tokens": 265560468.0, + "step": 10261 + }, + { + "epoch": 1.1269492642213925, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.765911340713501, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.6956290602684021, + "num_tokens": 265583667.0, + "step": 10262 + }, + { + "epoch": 1.127059081924006, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.571176767349243, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.7255560159683228, + "num_tokens": 265606595.0, + "step": 10263 + }, + { + "epoch": 1.1271688996266198, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4690706729888916, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6963794827461243, + "num_tokens": 265633660.0, + "step": 10264 + }, + { + "epoch": 1.1272787173292333, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.465843439102173, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6999642848968506, + "num_tokens": 265659487.0, + "step": 10265 + }, + { + "epoch": 1.127388535031847, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.460739850997925, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6961293816566467, + "num_tokens": 265688273.0, + "step": 10266 + }, + { + "epoch": 1.1274983527344609, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.673133134841919, + "learning_rate": 1e-06, + "loss": 1.0971, + "mean_token_accuracy": 0.6850900053977966, + "num_tokens": 265713972.0, + "step": 10267 + }, + { + "epoch": 1.1276081704370744, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.644583225250244, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6832401156425476, + "num_tokens": 265739755.0, + "step": 10268 + }, + { + "epoch": 1.1277179881396882, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7951905727386475, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6852478384971619, + "num_tokens": 265764051.0, + "step": 10269 + }, + { + "epoch": 1.1278278058423017, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.531865119934082, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7023073434829712, + "num_tokens": 265789653.0, + "step": 10270 + }, + { + "epoch": 1.1279376235449154, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4245851039886475, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7172056436538696, + "num_tokens": 265816462.0, + "step": 10271 + }, + { + "epoch": 1.1280474412475292, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.734952449798584, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7218456268310547, + "num_tokens": 265840306.0, + "step": 10272 + }, + { + "epoch": 1.1281572589501427, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.368647575378418, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7174017429351807, + "num_tokens": 265870347.0, + "step": 10273 + }, + { + "epoch": 1.1282670766527565, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7860755920410156, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6941179633140564, + "num_tokens": 265894363.0, + "step": 10274 + }, + { + "epoch": 1.12837689435537, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.490036964416504, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7029751539230347, + "num_tokens": 265921549.0, + "step": 10275 + }, + { + "epoch": 1.1284867120579838, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.572141408920288, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7109938859939575, + "num_tokens": 265948626.0, + "step": 10276 + }, + { + "epoch": 1.1285965297605973, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4842095375061035, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6950218677520752, + "num_tokens": 265983453.0, + "step": 10277 + }, + { + "epoch": 1.128706347463211, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.825507879257202, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7196462154388428, + "num_tokens": 266005444.0, + "step": 10278 + }, + { + "epoch": 1.1288161651658246, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.464860677719116, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7249915599822998, + "num_tokens": 266031241.0, + "step": 10279 + }, + { + "epoch": 1.1289259828684384, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7763173580169678, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7145522832870483, + "num_tokens": 266053070.0, + "step": 10280 + }, + { + "epoch": 1.1290358005710521, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.644591808319092, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6887275576591492, + "num_tokens": 266076940.0, + "step": 10281 + }, + { + "epoch": 1.1291456182736657, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.3018367290496826, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6933969259262085, + "num_tokens": 266111542.0, + "step": 10282 + }, + { + "epoch": 1.1292554359762794, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.394421100616455, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.7358098030090332, + "num_tokens": 266137534.0, + "step": 10283 + }, + { + "epoch": 1.129365253678893, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6706666946411133, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7204695343971252, + "num_tokens": 266161658.0, + "step": 10284 + }, + { + "epoch": 1.1294750713815067, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7258455753326416, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7049884796142578, + "num_tokens": 266184514.0, + "step": 10285 + }, + { + "epoch": 1.1295848890841205, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.588367462158203, + "learning_rate": 1e-06, + "loss": 1.0905, + "mean_token_accuracy": 0.6803560853004456, + "num_tokens": 266214113.0, + "step": 10286 + }, + { + "epoch": 1.129694706786734, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6531269550323486, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7169109582901001, + "num_tokens": 266234590.0, + "step": 10287 + }, + { + "epoch": 1.1298045244893478, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.349112033843994, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7278271913528442, + "num_tokens": 266262165.0, + "step": 10288 + }, + { + "epoch": 1.1299143421919613, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.544373035430908, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.717992901802063, + "num_tokens": 266287184.0, + "step": 10289 + }, + { + "epoch": 1.130024159894575, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.9996769428253174, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7165781259536743, + "num_tokens": 266305758.0, + "step": 10290 + }, + { + "epoch": 1.1301339775971886, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.387573480606079, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.710303544998169, + "num_tokens": 266335294.0, + "step": 10291 + }, + { + "epoch": 1.1302437952998023, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4742612838745117, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.7025274038314819, + "num_tokens": 266363544.0, + "step": 10292 + }, + { + "epoch": 1.1303536130024159, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.717525005340576, + "learning_rate": 1e-06, + "loss": 1.1082, + "mean_token_accuracy": 0.6764928102493286, + "num_tokens": 266389150.0, + "step": 10293 + }, + { + "epoch": 1.1304634307050296, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.469217538833618, + "learning_rate": 1e-06, + "loss": 1.1099, + "mean_token_accuracy": 0.685776948928833, + "num_tokens": 266418930.0, + "step": 10294 + }, + { + "epoch": 1.1305732484076434, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.554226875305176, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6846388578414917, + "num_tokens": 266445519.0, + "step": 10295 + }, + { + "epoch": 1.130683066110257, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0018258094787598, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7074488997459412, + "num_tokens": 266465326.0, + "step": 10296 + }, + { + "epoch": 1.1307928838128707, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4393534660339355, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.6986587047576904, + "num_tokens": 266498034.0, + "step": 10297 + }, + { + "epoch": 1.1309027015154842, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.750966787338257, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6981719732284546, + "num_tokens": 266523244.0, + "step": 10298 + }, + { + "epoch": 1.131012519218098, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3247838020324707, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.6877577304840088, + "num_tokens": 266556348.0, + "step": 10299 + }, + { + "epoch": 1.1311223369207117, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4391703605651855, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.698658287525177, + "num_tokens": 266585163.0, + "step": 10300 + }, + { + "epoch": 1.1312321546233253, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5880978107452393, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7246780395507812, + "num_tokens": 266608597.0, + "step": 10301 + }, + { + "epoch": 1.131341972325939, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7439961433410645, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7131113409996033, + "num_tokens": 266631533.0, + "step": 10302 + }, + { + "epoch": 1.1314517900285526, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4982457160949707, + "learning_rate": 1e-06, + "loss": 1.1344, + "mean_token_accuracy": 0.6759428977966309, + "num_tokens": 266662321.0, + "step": 10303 + }, + { + "epoch": 1.1315616077311663, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9994874000549316, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7092651724815369, + "num_tokens": 266681531.0, + "step": 10304 + }, + { + "epoch": 1.1316714254337799, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.687809944152832, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7180761098861694, + "num_tokens": 266709453.0, + "step": 10305 + }, + { + "epoch": 1.1317812431363936, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5338895320892334, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.7183347344398499, + "num_tokens": 266735508.0, + "step": 10306 + }, + { + "epoch": 1.1318910608390071, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8368239402770996, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6902129054069519, + "num_tokens": 266758362.0, + "step": 10307 + }, + { + "epoch": 1.132000878541621, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.477694511413574, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6956707835197449, + "num_tokens": 266784633.0, + "step": 10308 + }, + { + "epoch": 1.1321106962442347, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5850472450256348, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.7025900483131409, + "num_tokens": 266811149.0, + "step": 10309 + }, + { + "epoch": 1.1322205139468482, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0813066959381104, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7220358848571777, + "num_tokens": 266829030.0, + "step": 10310 + }, + { + "epoch": 1.132330331649462, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5344882011413574, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7109605073928833, + "num_tokens": 266859845.0, + "step": 10311 + }, + { + "epoch": 1.1324401493520755, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.60709547996521, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7045658826828003, + "num_tokens": 266884529.0, + "step": 10312 + }, + { + "epoch": 1.1325499670546892, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.532776117324829, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.694025993347168, + "num_tokens": 266910291.0, + "step": 10313 + }, + { + "epoch": 1.132659784757303, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4194419384002686, + "learning_rate": 1e-06, + "loss": 0.9135, + "mean_token_accuracy": 0.7251721620559692, + "num_tokens": 266937059.0, + "step": 10314 + }, + { + "epoch": 1.1327696024599165, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5807571411132812, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6881141066551208, + "num_tokens": 266964277.0, + "step": 10315 + }, + { + "epoch": 1.1328794201625303, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.371279716491699, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6968271136283875, + "num_tokens": 266996463.0, + "step": 10316 + }, + { + "epoch": 1.1329892378651438, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7181549072265625, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7026564478874207, + "num_tokens": 267019702.0, + "step": 10317 + }, + { + "epoch": 1.1330990555677576, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.776172399520874, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.721447765827179, + "num_tokens": 267040614.0, + "step": 10318 + }, + { + "epoch": 1.1332088732703711, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7162702083587646, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7105785608291626, + "num_tokens": 267067908.0, + "step": 10319 + }, + { + "epoch": 1.1333186909729849, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5323736667633057, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7020114660263062, + "num_tokens": 267093389.0, + "step": 10320 + }, + { + "epoch": 1.1334285086755984, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3905646800994873, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.6994792819023132, + "num_tokens": 267123727.0, + "step": 10321 + }, + { + "epoch": 1.1335383263782122, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.4539027214050293, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6941584348678589, + "num_tokens": 267153359.0, + "step": 10322 + }, + { + "epoch": 1.133648144080826, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 3.267051935195923, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7071913480758667, + "num_tokens": 267180260.0, + "step": 10323 + }, + { + "epoch": 1.1337579617834395, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.459975242614746, + "learning_rate": 1e-06, + "loss": 1.1288, + "mean_token_accuracy": 0.6775594353675842, + "num_tokens": 267214943.0, + "step": 10324 + }, + { + "epoch": 1.1338677794860532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6987009048461914, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.703782856464386, + "num_tokens": 267238721.0, + "step": 10325 + }, + { + "epoch": 1.1339775971886668, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.452216863632202, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7100750207901001, + "num_tokens": 267269105.0, + "step": 10326 + }, + { + "epoch": 1.1340874148912805, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5560779571533203, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6999883651733398, + "num_tokens": 267295469.0, + "step": 10327 + }, + { + "epoch": 1.134197232593894, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.4850704669952393, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.6920183897018433, + "num_tokens": 267323179.0, + "step": 10328 + }, + { + "epoch": 1.1343070502965078, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.750608205795288, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7097489833831787, + "num_tokens": 267347769.0, + "step": 10329 + }, + { + "epoch": 1.1344168679991213, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5610556602478027, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7043333053588867, + "num_tokens": 267373815.0, + "step": 10330 + }, + { + "epoch": 1.134526685701735, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5914077758789062, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7061486840248108, + "num_tokens": 267396441.0, + "step": 10331 + }, + { + "epoch": 1.1346365034043489, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.6448490619659424, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7088348865509033, + "num_tokens": 267422653.0, + "step": 10332 + }, + { + "epoch": 1.1347463211069624, + "ewc_loss": 9.119510650634766e-06, + "grad_norm": 2.7237753868103027, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.713712215423584, + "num_tokens": 267445598.0, + "step": 10333 + }, + { + "epoch": 1.1348561388095761, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.655876398086548, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7189583778381348, + "num_tokens": 267471406.0, + "step": 10334 + }, + { + "epoch": 1.1349659565121897, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 3.4095072746276855, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7026065587997437, + "num_tokens": 267494400.0, + "step": 10335 + }, + { + "epoch": 1.1350757742148034, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4193084239959717, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7191699147224426, + "num_tokens": 267524579.0, + "step": 10336 + }, + { + "epoch": 1.1351855919174172, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.4232373237609863, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7242056131362915, + "num_tokens": 267555767.0, + "step": 10337 + }, + { + "epoch": 1.1352954096200307, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.204131841659546, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7030652165412903, + "num_tokens": 267589006.0, + "step": 10338 + }, + { + "epoch": 1.1354052273226445, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.7146594524383545, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7097307443618774, + "num_tokens": 267611517.0, + "step": 10339 + }, + { + "epoch": 1.135515045025258, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.2990541458129883, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6875326633453369, + "num_tokens": 267644383.0, + "step": 10340 + }, + { + "epoch": 1.1356248627278718, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.2173304557800293, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6908519864082336, + "num_tokens": 267667731.0, + "step": 10341 + }, + { + "epoch": 1.1357346804304853, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 3.3313817977905273, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6990833282470703, + "num_tokens": 267692049.0, + "step": 10342 + }, + { + "epoch": 1.135844498133099, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.445495128631592, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7139885425567627, + "num_tokens": 267719834.0, + "step": 10343 + }, + { + "epoch": 1.1359543158357126, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.9528353214263916, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.722631573677063, + "num_tokens": 267741202.0, + "step": 10344 + }, + { + "epoch": 1.1360641335383264, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.4244964122772217, + "learning_rate": 1e-06, + "loss": 1.118, + "mean_token_accuracy": 0.6853956580162048, + "num_tokens": 267771467.0, + "step": 10345 + }, + { + "epoch": 1.1361739512409401, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.796826124191284, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7182095646858215, + "num_tokens": 267792675.0, + "step": 10346 + }, + { + "epoch": 1.1362837689435537, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.584502935409546, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6935307383537292, + "num_tokens": 267821926.0, + "step": 10347 + }, + { + "epoch": 1.1363935866461674, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.394920825958252, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6994131803512573, + "num_tokens": 267852223.0, + "step": 10348 + }, + { + "epoch": 1.136503404348781, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6726341247558594, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.699006199836731, + "num_tokens": 267876621.0, + "step": 10349 + }, + { + "epoch": 1.1366132220513947, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.354945182800293, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7011346817016602, + "num_tokens": 267908585.0, + "step": 10350 + }, + { + "epoch": 1.1367230397540085, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.447124719619751, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.7146564722061157, + "num_tokens": 267940588.0, + "step": 10351 + }, + { + "epoch": 1.136832857456622, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6499698162078857, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7082210183143616, + "num_tokens": 267965779.0, + "step": 10352 + }, + { + "epoch": 1.1369426751592357, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.567091226577759, + "learning_rate": 1e-06, + "loss": 1.0982, + "mean_token_accuracy": 0.6837549805641174, + "num_tokens": 267992571.0, + "step": 10353 + }, + { + "epoch": 1.1370524928618493, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.4571714401245117, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.7011829614639282, + "num_tokens": 268019872.0, + "step": 10354 + }, + { + "epoch": 1.137162310564463, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6807456016540527, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7128667831420898, + "num_tokens": 268043544.0, + "step": 10355 + }, + { + "epoch": 1.1372721282670766, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.542534589767456, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.7020647525787354, + "num_tokens": 268069948.0, + "step": 10356 + }, + { + "epoch": 1.1373819459696903, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.9647703170776367, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7009829878807068, + "num_tokens": 268089960.0, + "step": 10357 + }, + { + "epoch": 1.1374917636723039, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.889298439025879, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7180220484733582, + "num_tokens": 268111988.0, + "step": 10358 + }, + { + "epoch": 1.1376015813749176, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.564612627029419, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7171427011489868, + "num_tokens": 268137247.0, + "step": 10359 + }, + { + "epoch": 1.1377113990775314, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5707881450653076, + "learning_rate": 1e-06, + "loss": 0.8854, + "mean_token_accuracy": 0.7333654165267944, + "num_tokens": 268161393.0, + "step": 10360 + }, + { + "epoch": 1.137821216780145, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.619900703430176, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6965507864952087, + "num_tokens": 268190923.0, + "step": 10361 + }, + { + "epoch": 1.1379310344827587, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.907597541809082, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6902881264686584, + "num_tokens": 268214444.0, + "step": 10362 + }, + { + "epoch": 1.1380408521853722, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.9331698417663574, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7063401341438293, + "num_tokens": 268236428.0, + "step": 10363 + }, + { + "epoch": 1.138150669887986, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9103901386260986, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7320184707641602, + "num_tokens": 268258604.0, + "step": 10364 + }, + { + "epoch": 1.1382604875905997, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5795302391052246, + "learning_rate": 1e-06, + "loss": 0.8911, + "mean_token_accuracy": 0.7402318716049194, + "num_tokens": 268285344.0, + "step": 10365 + }, + { + "epoch": 1.1383703052932133, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5144710540771484, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6945761442184448, + "num_tokens": 268320968.0, + "step": 10366 + }, + { + "epoch": 1.138480122995827, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5732064247131348, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.717888593673706, + "num_tokens": 268347096.0, + "step": 10367 + }, + { + "epoch": 1.1385899406984406, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.696486234664917, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7135440707206726, + "num_tokens": 268373691.0, + "step": 10368 + }, + { + "epoch": 1.1386997584010543, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6779699325561523, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6942175626754761, + "num_tokens": 268398350.0, + "step": 10369 + }, + { + "epoch": 1.1388095761036678, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6008334159851074, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7137483954429626, + "num_tokens": 268422393.0, + "step": 10370 + }, + { + "epoch": 1.1389193938062816, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.601306676864624, + "learning_rate": 1e-06, + "loss": 0.9001, + "mean_token_accuracy": 0.7320100665092468, + "num_tokens": 268447765.0, + "step": 10371 + }, + { + "epoch": 1.1390292115088951, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.550086498260498, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7134066820144653, + "num_tokens": 268476834.0, + "step": 10372 + }, + { + "epoch": 1.139139029211509, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.706630229949951, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7163823843002319, + "num_tokens": 268501276.0, + "step": 10373 + }, + { + "epoch": 1.1392488469141226, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.530583620071411, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7114052772521973, + "num_tokens": 268533575.0, + "step": 10374 + }, + { + "epoch": 1.1393586646167362, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.745044469833374, + "learning_rate": 1e-06, + "loss": 0.9416, + "mean_token_accuracy": 0.7258802056312561, + "num_tokens": 268554506.0, + "step": 10375 + }, + { + "epoch": 1.13946848231935, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.615525007247925, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7077229022979736, + "num_tokens": 268579127.0, + "step": 10376 + }, + { + "epoch": 1.1395783000219635, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.619004964828491, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6973422169685364, + "num_tokens": 268602082.0, + "step": 10377 + }, + { + "epoch": 1.1396881177245772, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.703148126602173, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6895382404327393, + "num_tokens": 268629322.0, + "step": 10378 + }, + { + "epoch": 1.1397979354271908, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6352381706237793, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7259308695793152, + "num_tokens": 268654858.0, + "step": 10379 + }, + { + "epoch": 1.1399077531298045, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.899818181991577, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7108772993087769, + "num_tokens": 268674996.0, + "step": 10380 + }, + { + "epoch": 1.1400175708324183, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.674753427505493, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.695928692817688, + "num_tokens": 268701396.0, + "step": 10381 + }, + { + "epoch": 1.1401273885350318, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.894268035888672, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.7058062553405762, + "num_tokens": 268725730.0, + "step": 10382 + }, + { + "epoch": 1.1402372062376456, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.438180685043335, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7241292595863342, + "num_tokens": 268756266.0, + "step": 10383 + }, + { + "epoch": 1.140347023940259, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.643285036087036, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.7228521108627319, + "num_tokens": 268779905.0, + "step": 10384 + }, + { + "epoch": 1.1404568416428729, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5133414268493652, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7163711190223694, + "num_tokens": 268805544.0, + "step": 10385 + }, + { + "epoch": 1.1405666593454864, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.636965751647949, + "learning_rate": 1e-06, + "loss": 0.9193, + "mean_token_accuracy": 0.729207456111908, + "num_tokens": 268828418.0, + "step": 10386 + }, + { + "epoch": 1.1406764770481002, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1694295406341553, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7041831016540527, + "num_tokens": 268846625.0, + "step": 10387 + }, + { + "epoch": 1.140786294750714, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.513786792755127, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7029281258583069, + "num_tokens": 268871966.0, + "step": 10388 + }, + { + "epoch": 1.1408961124533274, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4265658855438232, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7309284210205078, + "num_tokens": 268897336.0, + "step": 10389 + }, + { + "epoch": 1.1410059301559412, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.696883201599121, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6947947144508362, + "num_tokens": 268924918.0, + "step": 10390 + }, + { + "epoch": 1.1411157478585547, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.716583013534546, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7070351839065552, + "num_tokens": 268950848.0, + "step": 10391 + }, + { + "epoch": 1.1412255655611685, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6057286262512207, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6987286806106567, + "num_tokens": 268976844.0, + "step": 10392 + }, + { + "epoch": 1.141335383263782, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.558187484741211, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7124563455581665, + "num_tokens": 269004861.0, + "step": 10393 + }, + { + "epoch": 1.1414452009663958, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.814236879348755, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7133871912956238, + "num_tokens": 269025965.0, + "step": 10394 + }, + { + "epoch": 1.1415550186690093, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6532182693481445, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.718651533126831, + "num_tokens": 269048403.0, + "step": 10395 + }, + { + "epoch": 1.141664836371623, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.531402111053467, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.6818650960922241, + "num_tokens": 269077585.0, + "step": 10396 + }, + { + "epoch": 1.1417746540742368, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5392255783081055, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.7313381433486938, + "num_tokens": 269108294.0, + "step": 10397 + }, + { + "epoch": 1.1418844717768504, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5351648330688477, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6996714472770691, + "num_tokens": 269135641.0, + "step": 10398 + }, + { + "epoch": 1.1419942894794641, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.563721179962158, + "learning_rate": 1e-06, + "loss": 1.0833, + "mean_token_accuracy": 0.6853992938995361, + "num_tokens": 269165374.0, + "step": 10399 + }, + { + "epoch": 1.1421041071820777, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.788421869277954, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.704033613204956, + "num_tokens": 269187117.0, + "step": 10400 + }, + { + "epoch": 1.1422139248846914, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5900540351867676, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7168141007423401, + "num_tokens": 269213528.0, + "step": 10401 + }, + { + "epoch": 1.1423237425873052, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.596893787384033, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7108283042907715, + "num_tokens": 269237503.0, + "step": 10402 + }, + { + "epoch": 1.1424335602899187, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.162416696548462, + "learning_rate": 1e-06, + "loss": 0.9417, + "mean_token_accuracy": 0.7259162664413452, + "num_tokens": 269263747.0, + "step": 10403 + }, + { + "epoch": 1.1425433779925325, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7319846153259277, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7054230570793152, + "num_tokens": 269286290.0, + "step": 10404 + }, + { + "epoch": 1.142653195695146, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.655521869659424, + "learning_rate": 1e-06, + "loss": 0.9169, + "mean_token_accuracy": 0.7283275127410889, + "num_tokens": 269306380.0, + "step": 10405 + }, + { + "epoch": 1.1427630133977598, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.661259651184082, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7068991661071777, + "num_tokens": 269330612.0, + "step": 10406 + }, + { + "epoch": 1.1428728311003733, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6659913063049316, + "learning_rate": 1e-06, + "loss": 0.9135, + "mean_token_accuracy": 0.7293840646743774, + "num_tokens": 269352725.0, + "step": 10407 + }, + { + "epoch": 1.142982648802987, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.232927083969116, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.7018283605575562, + "num_tokens": 269385926.0, + "step": 10408 + }, + { + "epoch": 1.1430924665056006, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4869320392608643, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6889600157737732, + "num_tokens": 269411889.0, + "step": 10409 + }, + { + "epoch": 1.1432022842082143, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7166712284088135, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6918675303459167, + "num_tokens": 269436910.0, + "step": 10410 + }, + { + "epoch": 1.143312101910828, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.513322591781616, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.6999027132987976, + "num_tokens": 269464438.0, + "step": 10411 + }, + { + "epoch": 1.1434219196134416, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4213156700134277, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6989235877990723, + "num_tokens": 269491963.0, + "step": 10412 + }, + { + "epoch": 1.1435317373160554, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.342249631881714, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.70075523853302, + "num_tokens": 269521966.0, + "step": 10413 + }, + { + "epoch": 1.143641555018669, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2749176025390625, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7008539438247681, + "num_tokens": 269554796.0, + "step": 10414 + }, + { + "epoch": 1.1437513727212827, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.58185076713562, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.692611575126648, + "num_tokens": 269580105.0, + "step": 10415 + }, + { + "epoch": 1.1438611904238964, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6045422554016113, + "learning_rate": 1e-06, + "loss": 1.1173, + "mean_token_accuracy": 0.6764437556266785, + "num_tokens": 269608531.0, + "step": 10416 + }, + { + "epoch": 1.14397100812651, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.463923215866089, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.693783700466156, + "num_tokens": 269638760.0, + "step": 10417 + }, + { + "epoch": 1.1440808258291237, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4796698093414307, + "learning_rate": 1e-06, + "loss": 1.0839, + "mean_token_accuracy": 0.6845040321350098, + "num_tokens": 269666361.0, + "step": 10418 + }, + { + "epoch": 1.1441906435317373, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.504239320755005, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.702252984046936, + "num_tokens": 269692763.0, + "step": 10419 + }, + { + "epoch": 1.144300461234351, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3176448345184326, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.688668429851532, + "num_tokens": 269727061.0, + "step": 10420 + }, + { + "epoch": 1.1444102789369646, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.520301103591919, + "learning_rate": 1e-06, + "loss": 0.9003, + "mean_token_accuracy": 0.7319802045822144, + "num_tokens": 269752838.0, + "step": 10421 + }, + { + "epoch": 1.1445200966395783, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.428898811340332, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.6942896842956543, + "num_tokens": 269783618.0, + "step": 10422 + }, + { + "epoch": 1.1446299143421919, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.440298080444336, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.703474760055542, + "num_tokens": 269813083.0, + "step": 10423 + }, + { + "epoch": 1.1447397320448056, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5413763523101807, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7091308832168579, + "num_tokens": 269841125.0, + "step": 10424 + }, + { + "epoch": 1.1448495497474194, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5000014305114746, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7289918661117554, + "num_tokens": 269865835.0, + "step": 10425 + }, + { + "epoch": 1.144959367450033, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9570131301879883, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.701276957988739, + "num_tokens": 269888757.0, + "step": 10426 + }, + { + "epoch": 1.1450691851526467, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.810392141342163, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.6968530416488647, + "num_tokens": 269912248.0, + "step": 10427 + }, + { + "epoch": 1.1451790028552602, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.736894369125366, + "learning_rate": 1e-06, + "loss": 0.9335, + "mean_token_accuracy": 0.7238225936889648, + "num_tokens": 269936794.0, + "step": 10428 + }, + { + "epoch": 1.145288820557874, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5491292476654053, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7076377868652344, + "num_tokens": 269962722.0, + "step": 10429 + }, + { + "epoch": 1.1453986382604877, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.74499249458313, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7093144655227661, + "num_tokens": 269986040.0, + "step": 10430 + }, + { + "epoch": 1.1455084559631012, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.343946933746338, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7037849426269531, + "num_tokens": 270018188.0, + "step": 10431 + }, + { + "epoch": 1.145618273665715, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.513028621673584, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7193707823753357, + "num_tokens": 270045447.0, + "step": 10432 + }, + { + "epoch": 1.1457280913683285, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.764918565750122, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6936033964157104, + "num_tokens": 270073881.0, + "step": 10433 + }, + { + "epoch": 1.1458379090709423, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8307950496673584, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7129954099655151, + "num_tokens": 270097232.0, + "step": 10434 + }, + { + "epoch": 1.1459477267735558, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7310197353363037, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6983054876327515, + "num_tokens": 270122091.0, + "step": 10435 + }, + { + "epoch": 1.1460575444761696, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6441779136657715, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6881012916564941, + "num_tokens": 270147807.0, + "step": 10436 + }, + { + "epoch": 1.1461673621787831, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8241918087005615, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7135250568389893, + "num_tokens": 270169830.0, + "step": 10437 + }, + { + "epoch": 1.1462771798813969, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.716923475265503, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7148181200027466, + "num_tokens": 270194328.0, + "step": 10438 + }, + { + "epoch": 1.1463869975840106, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5709762573242188, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7058535814285278, + "num_tokens": 270219357.0, + "step": 10439 + }, + { + "epoch": 1.1464968152866242, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8419063091278076, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7008614540100098, + "num_tokens": 270241779.0, + "step": 10440 + }, + { + "epoch": 1.146606632989238, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8953680992126465, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.716209352016449, + "num_tokens": 270261733.0, + "step": 10441 + }, + { + "epoch": 1.1467164506918515, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7473230361938477, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7079815864562988, + "num_tokens": 270283333.0, + "step": 10442 + }, + { + "epoch": 1.1468262683944652, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3882791996002197, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7013065218925476, + "num_tokens": 270315335.0, + "step": 10443 + }, + { + "epoch": 1.1469360860970788, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4526901245117188, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7205955982208252, + "num_tokens": 270342964.0, + "step": 10444 + }, + { + "epoch": 1.1470459037996925, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.899146556854248, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.687057375907898, + "num_tokens": 270368653.0, + "step": 10445 + }, + { + "epoch": 1.147155721502306, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1777713298797607, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7000501155853271, + "num_tokens": 270393362.0, + "step": 10446 + }, + { + "epoch": 1.1472655392049198, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.669691324234009, + "learning_rate": 1e-06, + "loss": 0.9285, + "mean_token_accuracy": 0.7207767367362976, + "num_tokens": 270416873.0, + "step": 10447 + }, + { + "epoch": 1.1473753569075336, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6533761024475098, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7117941975593567, + "num_tokens": 270440103.0, + "step": 10448 + }, + { + "epoch": 1.147485174610147, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.741661548614502, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.711426317691803, + "num_tokens": 270462076.0, + "step": 10449 + }, + { + "epoch": 1.1475949923127609, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6314826011657715, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.6904641389846802, + "num_tokens": 270488307.0, + "step": 10450 + }, + { + "epoch": 1.1477048100153744, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.817251443862915, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7117327451705933, + "num_tokens": 270517987.0, + "step": 10451 + }, + { + "epoch": 1.1478146277179881, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5820839405059814, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.6944355368614197, + "num_tokens": 270543828.0, + "step": 10452 + }, + { + "epoch": 1.147924445420602, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6397767066955566, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6943211555480957, + "num_tokens": 270570706.0, + "step": 10453 + }, + { + "epoch": 1.1480342631232154, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5192348957061768, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7247968912124634, + "num_tokens": 270596739.0, + "step": 10454 + }, + { + "epoch": 1.1481440808258292, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3882699012756348, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.725269079208374, + "num_tokens": 270623788.0, + "step": 10455 + }, + { + "epoch": 1.1482538985284427, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6307992935180664, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7207039594650269, + "num_tokens": 270649744.0, + "step": 10456 + }, + { + "epoch": 1.1483637162310565, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.743863821029663, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7190742492675781, + "num_tokens": 270671889.0, + "step": 10457 + }, + { + "epoch": 1.14847353393367, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8001434803009033, + "learning_rate": 1e-06, + "loss": 0.8457, + "mean_token_accuracy": 0.7442384958267212, + "num_tokens": 270692169.0, + "step": 10458 + }, + { + "epoch": 1.1485833516362838, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.500278949737549, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7156662940979004, + "num_tokens": 270716963.0, + "step": 10459 + }, + { + "epoch": 1.1486931693388973, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7471837997436523, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.70623779296875, + "num_tokens": 270747232.0, + "step": 10460 + }, + { + "epoch": 1.148802987041511, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3821356296539307, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.695410966873169, + "num_tokens": 270777490.0, + "step": 10461 + }, + { + "epoch": 1.1489128047441248, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7145256996154785, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7139243483543396, + "num_tokens": 270799558.0, + "step": 10462 + }, + { + "epoch": 1.1490226224467384, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5789456367492676, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6956031322479248, + "num_tokens": 270824945.0, + "step": 10463 + }, + { + "epoch": 1.1491324401493521, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2672102451324463, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6954221725463867, + "num_tokens": 270859452.0, + "step": 10464 + }, + { + "epoch": 1.1492422578519657, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3773934841156006, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.709860622882843, + "num_tokens": 270889401.0, + "step": 10465 + }, + { + "epoch": 1.1493520755545794, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9795563220977783, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7156891226768494, + "num_tokens": 270911780.0, + "step": 10466 + }, + { + "epoch": 1.1494618932571932, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4549336433410645, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7032537460327148, + "num_tokens": 270939541.0, + "step": 10467 + }, + { + "epoch": 1.1495717109598067, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7957162857055664, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7047924399375916, + "num_tokens": 270961764.0, + "step": 10468 + }, + { + "epoch": 1.1496815286624205, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9470789432525635, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7158000469207764, + "num_tokens": 270982189.0, + "step": 10469 + }, + { + "epoch": 1.149791346365034, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.176785945892334, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7093142867088318, + "num_tokens": 271000362.0, + "step": 10470 + }, + { + "epoch": 1.1499011640676478, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.761253833770752, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7170428037643433, + "num_tokens": 271023383.0, + "step": 10471 + }, + { + "epoch": 1.1500109817702613, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.426537275314331, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.7208107709884644, + "num_tokens": 271052828.0, + "step": 10472 + }, + { + "epoch": 1.150120799472875, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5268235206604004, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.6911188364028931, + "num_tokens": 271081617.0, + "step": 10473 + }, + { + "epoch": 1.1502306171754886, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6076114177703857, + "learning_rate": 1e-06, + "loss": 1.1401, + "mean_token_accuracy": 0.6704352498054504, + "num_tokens": 271110567.0, + "step": 10474 + }, + { + "epoch": 1.1503404348781023, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6338260173797607, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6968234181404114, + "num_tokens": 271134094.0, + "step": 10475 + }, + { + "epoch": 1.150450252580716, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.816800355911255, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7291507720947266, + "num_tokens": 271159825.0, + "step": 10476 + }, + { + "epoch": 1.1505600702833296, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8263916969299316, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7207242250442505, + "num_tokens": 271179892.0, + "step": 10477 + }, + { + "epoch": 1.1506698879859434, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.516552448272705, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.7023702263832092, + "num_tokens": 271206416.0, + "step": 10478 + }, + { + "epoch": 1.150779705688557, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.458327293395996, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.681816577911377, + "num_tokens": 271235255.0, + "step": 10479 + }, + { + "epoch": 1.1508895233911707, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.723611831665039, + "learning_rate": 1e-06, + "loss": 1.1497, + "mean_token_accuracy": 0.6693849563598633, + "num_tokens": 271273431.0, + "step": 10480 + }, + { + "epoch": 1.1509993410937844, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6714394092559814, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6927691102027893, + "num_tokens": 271296536.0, + "step": 10481 + }, + { + "epoch": 1.151109158796398, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5418803691864014, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6821600198745728, + "num_tokens": 271331104.0, + "step": 10482 + }, + { + "epoch": 1.1512189764990117, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.526559829711914, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7060423493385315, + "num_tokens": 271358864.0, + "step": 10483 + }, + { + "epoch": 1.1513287942016253, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4111745357513428, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6958681344985962, + "num_tokens": 271385636.0, + "step": 10484 + }, + { + "epoch": 1.151438611904239, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.766510009765625, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7145841717720032, + "num_tokens": 271406751.0, + "step": 10485 + }, + { + "epoch": 1.1515484296068526, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.669729709625244, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6925089359283447, + "num_tokens": 271433759.0, + "step": 10486 + }, + { + "epoch": 1.1516582473094663, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8801286220550537, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7234289646148682, + "num_tokens": 271453057.0, + "step": 10487 + }, + { + "epoch": 1.1517680650120798, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6089274883270264, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7104921936988831, + "num_tokens": 271478297.0, + "step": 10488 + }, + { + "epoch": 1.1518778827146936, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7471745014190674, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7129393815994263, + "num_tokens": 271500563.0, + "step": 10489 + }, + { + "epoch": 1.1519877004173074, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.406017780303955, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6913805603981018, + "num_tokens": 271533148.0, + "step": 10490 + }, + { + "epoch": 1.152097518119921, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8270719051361084, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.7012882828712463, + "num_tokens": 271561841.0, + "step": 10491 + }, + { + "epoch": 1.1522073358225347, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8991334438323975, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7219411134719849, + "num_tokens": 271583226.0, + "step": 10492 + }, + { + "epoch": 1.1523171535251482, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7201192378997803, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7196890115737915, + "num_tokens": 271605953.0, + "step": 10493 + }, + { + "epoch": 1.152426971227762, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5645158290863037, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7081912159919739, + "num_tokens": 271632154.0, + "step": 10494 + }, + { + "epoch": 1.1525367889303757, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.917616844177246, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7112340927124023, + "num_tokens": 271653841.0, + "step": 10495 + }, + { + "epoch": 1.1526466066329892, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5646581649780273, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7097549438476562, + "num_tokens": 271677977.0, + "step": 10496 + }, + { + "epoch": 1.152756424335603, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7259373664855957, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7063846588134766, + "num_tokens": 271701221.0, + "step": 10497 + }, + { + "epoch": 1.1528662420382165, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3828179836273193, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6853842735290527, + "num_tokens": 271732023.0, + "step": 10498 + }, + { + "epoch": 1.1529760597408303, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.659592866897583, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7131522297859192, + "num_tokens": 271757810.0, + "step": 10499 + }, + { + "epoch": 1.1530858774434438, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.139589309692383, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7143515348434448, + "num_tokens": 271775593.0, + "step": 10500 + }, + { + "epoch": 1.1531956951460576, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5574328899383545, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7165743112564087, + "num_tokens": 271802380.0, + "step": 10501 + }, + { + "epoch": 1.153305512848671, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.487436532974243, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.7016593217849731, + "num_tokens": 271829159.0, + "step": 10502 + }, + { + "epoch": 1.1534153305512849, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.92253041267395, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.7000631093978882, + "num_tokens": 271853961.0, + "step": 10503 + }, + { + "epoch": 1.1535251482538986, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.638514280319214, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.720082700252533, + "num_tokens": 271880431.0, + "step": 10504 + }, + { + "epoch": 1.1536349659565122, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4539945125579834, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.709669828414917, + "num_tokens": 271908730.0, + "step": 10505 + }, + { + "epoch": 1.153744783659126, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6939938068389893, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.724124550819397, + "num_tokens": 271931465.0, + "step": 10506 + }, + { + "epoch": 1.1538546013617395, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7976865768432617, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7108021378517151, + "num_tokens": 271957167.0, + "step": 10507 + }, + { + "epoch": 1.1539644190643532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.467556953430176, + "learning_rate": 1e-06, + "loss": 0.9217, + "mean_token_accuracy": 0.7224500775337219, + "num_tokens": 271982253.0, + "step": 10508 + }, + { + "epoch": 1.1540742367669667, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4293274879455566, + "learning_rate": 1e-06, + "loss": 1.156, + "mean_token_accuracy": 0.6708328723907471, + "num_tokens": 272014376.0, + "step": 10509 + }, + { + "epoch": 1.1541840544695805, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4856693744659424, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7090401649475098, + "num_tokens": 272043291.0, + "step": 10510 + }, + { + "epoch": 1.154293872172194, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.458824634552002, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6885778307914734, + "num_tokens": 272071838.0, + "step": 10511 + }, + { + "epoch": 1.1544036898748078, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4419023990631104, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7010394334793091, + "num_tokens": 272099074.0, + "step": 10512 + }, + { + "epoch": 1.1545135075774215, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9959397315979004, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7166883945465088, + "num_tokens": 272121508.0, + "step": 10513 + }, + { + "epoch": 1.154623325280035, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8465611934661865, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7227258682250977, + "num_tokens": 272145371.0, + "step": 10514 + }, + { + "epoch": 1.1547331429826488, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.793736219406128, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6938151717185974, + "num_tokens": 272170042.0, + "step": 10515 + }, + { + "epoch": 1.1548429606852624, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4811952114105225, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7232504487037659, + "num_tokens": 272197860.0, + "step": 10516 + }, + { + "epoch": 1.1549527783878761, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.99600887298584, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7158366441726685, + "num_tokens": 272219986.0, + "step": 10517 + }, + { + "epoch": 1.15506259609049, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6517364978790283, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7013760805130005, + "num_tokens": 272247393.0, + "step": 10518 + }, + { + "epoch": 1.1551724137931034, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.836303234100342, + "learning_rate": 1e-06, + "loss": 0.9265, + "mean_token_accuracy": 0.7280051708221436, + "num_tokens": 272266856.0, + "step": 10519 + }, + { + "epoch": 1.1552822314957172, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5268497467041016, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6930916905403137, + "num_tokens": 272296842.0, + "step": 10520 + }, + { + "epoch": 1.1553920491983307, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.607105016708374, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6812984943389893, + "num_tokens": 272324349.0, + "step": 10521 + }, + { + "epoch": 1.1555018669009445, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4507362842559814, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7224512696266174, + "num_tokens": 272351169.0, + "step": 10522 + }, + { + "epoch": 1.155611684603558, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.691549777984619, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.6956608295440674, + "num_tokens": 272375508.0, + "step": 10523 + }, + { + "epoch": 1.1557215023061718, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2496626377105713, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7070462703704834, + "num_tokens": 272405700.0, + "step": 10524 + }, + { + "epoch": 1.1558313200087853, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.507993698120117, + "learning_rate": 1e-06, + "loss": 1.0989, + "mean_token_accuracy": 0.6789424419403076, + "num_tokens": 272435400.0, + "step": 10525 + }, + { + "epoch": 1.155941137711399, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6652328968048096, + "learning_rate": 1e-06, + "loss": 1.1847, + "mean_token_accuracy": 0.6676697731018066, + "num_tokens": 272462424.0, + "step": 10526 + }, + { + "epoch": 1.1560509554140128, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.51259708404541, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6823776364326477, + "num_tokens": 272492794.0, + "step": 10527 + }, + { + "epoch": 1.1561607731166264, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0016140937805176, + "learning_rate": 1e-06, + "loss": 0.8303, + "mean_token_accuracy": 0.754952609539032, + "num_tokens": 272515654.0, + "step": 10528 + }, + { + "epoch": 1.15627059081924, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.674644947052002, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7077012062072754, + "num_tokens": 272540293.0, + "step": 10529 + }, + { + "epoch": 1.1563804085218536, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.549499273300171, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7117658853530884, + "num_tokens": 272570385.0, + "step": 10530 + }, + { + "epoch": 1.1564902262244674, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7560770511627197, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.70199054479599, + "num_tokens": 272591512.0, + "step": 10531 + }, + { + "epoch": 1.1566000439270812, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8156402111053467, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7080048322677612, + "num_tokens": 272614670.0, + "step": 10532 + }, + { + "epoch": 1.1567098616296947, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5190787315368652, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7044953107833862, + "num_tokens": 272640736.0, + "step": 10533 + }, + { + "epoch": 1.1568196793323084, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5560121536254883, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7058630585670471, + "num_tokens": 272665179.0, + "step": 10534 + }, + { + "epoch": 1.156929497034922, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.388188123703003, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7180229425430298, + "num_tokens": 272693745.0, + "step": 10535 + }, + { + "epoch": 1.1570393147375357, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.827463388442993, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7163171768188477, + "num_tokens": 272714068.0, + "step": 10536 + }, + { + "epoch": 1.1571491324401493, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6581242084503174, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6932098269462585, + "num_tokens": 272737718.0, + "step": 10537 + }, + { + "epoch": 1.157258950142763, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.713550090789795, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7135511636734009, + "num_tokens": 272760243.0, + "step": 10538 + }, + { + "epoch": 1.1573687678453766, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4366793632507324, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.697550892829895, + "num_tokens": 272788577.0, + "step": 10539 + }, + { + "epoch": 1.1574785855479903, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.728586435317993, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.6809226274490356, + "num_tokens": 272816076.0, + "step": 10540 + }, + { + "epoch": 1.157588403250604, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8473000526428223, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7084749937057495, + "num_tokens": 272838322.0, + "step": 10541 + }, + { + "epoch": 1.1576982209532176, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.652672290802002, + "learning_rate": 1e-06, + "loss": 0.9132, + "mean_token_accuracy": 0.7274296283721924, + "num_tokens": 272864169.0, + "step": 10542 + }, + { + "epoch": 1.1578080386558314, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.553245782852173, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7141213417053223, + "num_tokens": 272888136.0, + "step": 10543 + }, + { + "epoch": 1.157917856358445, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7108635902404785, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6960341930389404, + "num_tokens": 272912632.0, + "step": 10544 + }, + { + "epoch": 1.1580276740610587, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6496942043304443, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7004276514053345, + "num_tokens": 272939621.0, + "step": 10545 + }, + { + "epoch": 1.1581374917636724, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7497708797454834, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6968551874160767, + "num_tokens": 272963177.0, + "step": 10546 + }, + { + "epoch": 1.158247309466286, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6577584743499756, + "learning_rate": 1e-06, + "loss": 0.9263, + "mean_token_accuracy": 0.722606897354126, + "num_tokens": 272987430.0, + "step": 10547 + }, + { + "epoch": 1.1583571271688997, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7679247856140137, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.718953013420105, + "num_tokens": 273008216.0, + "step": 10548 + }, + { + "epoch": 1.1584669448715132, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0853540897369385, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6982354521751404, + "num_tokens": 273031222.0, + "step": 10549 + }, + { + "epoch": 1.158576762574127, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.770190954208374, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7179939150810242, + "num_tokens": 273053142.0, + "step": 10550 + }, + { + "epoch": 1.1586865802767405, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5939273834228516, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.692717969417572, + "num_tokens": 273080271.0, + "step": 10551 + }, + { + "epoch": 1.1587963979793543, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.631060838699341, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7118979692459106, + "num_tokens": 273103267.0, + "step": 10552 + }, + { + "epoch": 1.1589062156819678, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.469771146774292, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.7022780776023865, + "num_tokens": 273129129.0, + "step": 10553 + }, + { + "epoch": 1.1590160333845816, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.264890670776367, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7190983295440674, + "num_tokens": 273159544.0, + "step": 10554 + }, + { + "epoch": 1.1591258510871953, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.447911024093628, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.7087422609329224, + "num_tokens": 273186077.0, + "step": 10555 + }, + { + "epoch": 1.1592356687898089, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5848734378814697, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7069390416145325, + "num_tokens": 273213808.0, + "step": 10556 + }, + { + "epoch": 1.1593454864924226, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.660491943359375, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7155864238739014, + "num_tokens": 273242014.0, + "step": 10557 + }, + { + "epoch": 1.1594553041950362, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4458882808685303, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.6986992359161377, + "num_tokens": 273272011.0, + "step": 10558 + }, + { + "epoch": 1.15956512189765, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.709777593612671, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6838314533233643, + "num_tokens": 273296193.0, + "step": 10559 + }, + { + "epoch": 1.1596749396002635, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4652304649353027, + "learning_rate": 1e-06, + "loss": 0.9995, + "mean_token_accuracy": 0.7072622179985046, + "num_tokens": 273324154.0, + "step": 10560 + }, + { + "epoch": 1.1597847573028772, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.365347146987915, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6904928684234619, + "num_tokens": 273356852.0, + "step": 10561 + }, + { + "epoch": 1.159894575005491, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4993550777435303, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6898784041404724, + "num_tokens": 273386090.0, + "step": 10562 + }, + { + "epoch": 1.1600043927081045, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6532437801361084, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7041749954223633, + "num_tokens": 273416319.0, + "step": 10563 + }, + { + "epoch": 1.1601142104107183, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6882855892181396, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.6966021656990051, + "num_tokens": 273439730.0, + "step": 10564 + }, + { + "epoch": 1.1602240281133318, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.664069890975952, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6888983845710754, + "num_tokens": 273465595.0, + "step": 10565 + }, + { + "epoch": 1.1603338458159456, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3488783836364746, + "learning_rate": 1e-06, + "loss": 1.1293, + "mean_token_accuracy": 0.674056351184845, + "num_tokens": 273496619.0, + "step": 10566 + }, + { + "epoch": 1.160443663518559, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.935584306716919, + "learning_rate": 1e-06, + "loss": 0.9576, + "mean_token_accuracy": 0.7207428216934204, + "num_tokens": 273517232.0, + "step": 10567 + }, + { + "epoch": 1.1605534812211729, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8189616203308105, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.699373722076416, + "num_tokens": 273543123.0, + "step": 10568 + }, + { + "epoch": 1.1606632989237866, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5593743324279785, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.704692006111145, + "num_tokens": 273573824.0, + "step": 10569 + }, + { + "epoch": 1.1607731166264001, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7671029567718506, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7209560871124268, + "num_tokens": 273596774.0, + "step": 10570 + }, + { + "epoch": 1.160882934329014, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2605478763580322, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.707262396812439, + "num_tokens": 273625480.0, + "step": 10571 + }, + { + "epoch": 1.1609927520316274, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4844372272491455, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7079814076423645, + "num_tokens": 273650325.0, + "step": 10572 + }, + { + "epoch": 1.1611025697342412, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.753683567047119, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6845179796218872, + "num_tokens": 273673384.0, + "step": 10573 + }, + { + "epoch": 1.1612123874368547, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7965822219848633, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.7075995206832886, + "num_tokens": 273696548.0, + "step": 10574 + }, + { + "epoch": 1.1613222051394685, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5388975143432617, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.7010158896446228, + "num_tokens": 273725958.0, + "step": 10575 + }, + { + "epoch": 1.161432022842082, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.386411428451538, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6943574547767639, + "num_tokens": 273758568.0, + "step": 10576 + }, + { + "epoch": 1.1615418405446958, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.440157890319824, + "learning_rate": 1e-06, + "loss": 0.8931, + "mean_token_accuracy": 0.7396960258483887, + "num_tokens": 273786322.0, + "step": 10577 + }, + { + "epoch": 1.1616516582473095, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.492615222930908, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7066953778266907, + "num_tokens": 273812749.0, + "step": 10578 + }, + { + "epoch": 1.161761475949923, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.763957977294922, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7162158489227295, + "num_tokens": 273836812.0, + "step": 10579 + }, + { + "epoch": 1.1618712936525368, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.585563898086548, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.6782712340354919, + "num_tokens": 273861639.0, + "step": 10580 + }, + { + "epoch": 1.1619811113551504, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5527384281158447, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7056962251663208, + "num_tokens": 273890785.0, + "step": 10581 + }, + { + "epoch": 1.1620909290577641, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4374449253082275, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6987323760986328, + "num_tokens": 273920468.0, + "step": 10582 + }, + { + "epoch": 1.1622007467603779, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.816462516784668, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7303560972213745, + "num_tokens": 273946197.0, + "step": 10583 + }, + { + "epoch": 1.1623105644629914, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.41145396232605, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7009169459342957, + "num_tokens": 273974399.0, + "step": 10584 + }, + { + "epoch": 1.1624203821656052, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5614664554595947, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7126421928405762, + "num_tokens": 274001424.0, + "step": 10585 + }, + { + "epoch": 1.1625301998682187, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7161285877227783, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6926448941230774, + "num_tokens": 274024276.0, + "step": 10586 + }, + { + "epoch": 1.1626400175708325, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.600616455078125, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6953359842300415, + "num_tokens": 274048851.0, + "step": 10587 + }, + { + "epoch": 1.162749835273446, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7628138065338135, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.70304274559021, + "num_tokens": 274071445.0, + "step": 10588 + }, + { + "epoch": 1.1628596529760598, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5315492153167725, + "learning_rate": 1e-06, + "loss": 0.8383, + "mean_token_accuracy": 0.7482730150222778, + "num_tokens": 274098455.0, + "step": 10589 + }, + { + "epoch": 1.1629694706786733, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5733745098114014, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7301666736602783, + "num_tokens": 274124178.0, + "step": 10590 + }, + { + "epoch": 1.163079288381287, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6938834190368652, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7073116302490234, + "num_tokens": 274151069.0, + "step": 10591 + }, + { + "epoch": 1.1631891060839008, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.733417510986328, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6981633305549622, + "num_tokens": 274173561.0, + "step": 10592 + }, + { + "epoch": 1.1632989237865143, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.583625316619873, + "learning_rate": 1e-06, + "loss": 1.1249, + "mean_token_accuracy": 0.6728301048278809, + "num_tokens": 274205144.0, + "step": 10593 + }, + { + "epoch": 1.163408741489128, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.6182830333709717, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6855261325836182, + "num_tokens": 274231929.0, + "step": 10594 + }, + { + "epoch": 1.1635185591917416, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.624345541000366, + "learning_rate": 1e-06, + "loss": 0.9192, + "mean_token_accuracy": 0.7252732515335083, + "num_tokens": 274256158.0, + "step": 10595 + }, + { + "epoch": 1.1636283768943554, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6124229431152344, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.6876250505447388, + "num_tokens": 274283026.0, + "step": 10596 + }, + { + "epoch": 1.1637381945969691, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7997145652770996, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6889005899429321, + "num_tokens": 274309775.0, + "step": 10597 + }, + { + "epoch": 1.1638480122995827, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6707403659820557, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7067905068397522, + "num_tokens": 274334540.0, + "step": 10598 + }, + { + "epoch": 1.1639578300021964, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.780437707901001, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.704041063785553, + "num_tokens": 274357535.0, + "step": 10599 + }, + { + "epoch": 1.16406764770481, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.53132700920105, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7088254690170288, + "num_tokens": 274382477.0, + "step": 10600 + }, + { + "epoch": 1.1641774654074237, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5496866703033447, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7164963483810425, + "num_tokens": 274408585.0, + "step": 10601 + }, + { + "epoch": 1.1642872831100373, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4194719791412354, + "learning_rate": 1e-06, + "loss": 0.8752, + "mean_token_accuracy": 0.7377384305000305, + "num_tokens": 274437220.0, + "step": 10602 + }, + { + "epoch": 1.164397100812651, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4796717166900635, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7053639888763428, + "num_tokens": 274469802.0, + "step": 10603 + }, + { + "epoch": 1.1645069185152646, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3679943084716797, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7021927833557129, + "num_tokens": 274499495.0, + "step": 10604 + }, + { + "epoch": 1.1646167362178783, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5930705070495605, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7179669737815857, + "num_tokens": 274522048.0, + "step": 10605 + }, + { + "epoch": 1.164726553920492, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.517491102218628, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.684837281703949, + "num_tokens": 274550397.0, + "step": 10606 + }, + { + "epoch": 1.1648363716231056, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5913352966308594, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7133195996284485, + "num_tokens": 274572289.0, + "step": 10607 + }, + { + "epoch": 1.1649461893257194, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5977768898010254, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6957567930221558, + "num_tokens": 274598271.0, + "step": 10608 + }, + { + "epoch": 1.165056007028333, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4134275913238525, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7066037058830261, + "num_tokens": 274626039.0, + "step": 10609 + }, + { + "epoch": 1.1651658247309467, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5402657985687256, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.69306880235672, + "num_tokens": 274657055.0, + "step": 10610 + }, + { + "epoch": 1.1652756424335604, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4625742435455322, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7114717960357666, + "num_tokens": 274685455.0, + "step": 10611 + }, + { + "epoch": 1.165385460136174, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6486098766326904, + "learning_rate": 1e-06, + "loss": 1.133, + "mean_token_accuracy": 0.6833140254020691, + "num_tokens": 274709761.0, + "step": 10612 + }, + { + "epoch": 1.1654952778387877, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.71856951713562, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7116836905479431, + "num_tokens": 274734396.0, + "step": 10613 + }, + { + "epoch": 1.1656050955414012, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3767380714416504, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7010552883148193, + "num_tokens": 274768069.0, + "step": 10614 + }, + { + "epoch": 1.165714913244015, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.669238567352295, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.711761474609375, + "num_tokens": 274800029.0, + "step": 10615 + }, + { + "epoch": 1.1658247309466285, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0535836219787598, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7162281274795532, + "num_tokens": 274819886.0, + "step": 10616 + }, + { + "epoch": 1.1659345486492423, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.589709997177124, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6975980997085571, + "num_tokens": 274848395.0, + "step": 10617 + }, + { + "epoch": 1.1660443663518558, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.536546468734741, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7053995132446289, + "num_tokens": 274872749.0, + "step": 10618 + }, + { + "epoch": 1.1661541840544696, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.520632266998291, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7191479206085205, + "num_tokens": 274897661.0, + "step": 10619 + }, + { + "epoch": 1.1662640017570833, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6118218898773193, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.693758487701416, + "num_tokens": 274922900.0, + "step": 10620 + }, + { + "epoch": 1.1663738194596969, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3486649990081787, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7064515948295593, + "num_tokens": 274953199.0, + "step": 10621 + }, + { + "epoch": 1.1664836371623106, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7501957416534424, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7151541709899902, + "num_tokens": 274975312.0, + "step": 10622 + }, + { + "epoch": 1.1665934548649242, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4728150367736816, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6856210827827454, + "num_tokens": 275005756.0, + "step": 10623 + }, + { + "epoch": 1.166703272567538, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.853145122528076, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.71638023853302, + "num_tokens": 275029331.0, + "step": 10624 + }, + { + "epoch": 1.1668130902701515, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.99882173538208, + "learning_rate": 1e-06, + "loss": 1.1693, + "mean_token_accuracy": 0.6669309139251709, + "num_tokens": 275054582.0, + "step": 10625 + }, + { + "epoch": 1.1669229079727652, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6371941566467285, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7284400463104248, + "num_tokens": 275079544.0, + "step": 10626 + }, + { + "epoch": 1.167032725675379, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.528440475463867, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6962310075759888, + "num_tokens": 275109495.0, + "step": 10627 + }, + { + "epoch": 1.1671425433779925, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.620075225830078, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7030530571937561, + "num_tokens": 275132812.0, + "step": 10628 + }, + { + "epoch": 1.1672523610806063, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4056780338287354, + "learning_rate": 1e-06, + "loss": 1.1091, + "mean_token_accuracy": 0.6828510165214539, + "num_tokens": 275161256.0, + "step": 10629 + }, + { + "epoch": 1.1673621787832198, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4508488178253174, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7179523706436157, + "num_tokens": 275186318.0, + "step": 10630 + }, + { + "epoch": 1.1674719964858336, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0955941677093506, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6855013370513916, + "num_tokens": 275207258.0, + "step": 10631 + }, + { + "epoch": 1.167581814188447, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.625683069229126, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7049537301063538, + "num_tokens": 275233345.0, + "step": 10632 + }, + { + "epoch": 1.1676916318910608, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.674030065536499, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7193776369094849, + "num_tokens": 275258457.0, + "step": 10633 + }, + { + "epoch": 1.1678014495936746, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.915179967880249, + "learning_rate": 1e-06, + "loss": 0.9376, + "mean_token_accuracy": 0.7203332781791687, + "num_tokens": 275278010.0, + "step": 10634 + }, + { + "epoch": 1.1679112672962881, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.62355899810791, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7113813161849976, + "num_tokens": 275304279.0, + "step": 10635 + }, + { + "epoch": 1.168021084998902, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.732043504714966, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7039518356323242, + "num_tokens": 275326567.0, + "step": 10636 + }, + { + "epoch": 1.1681309027015154, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2352139949798584, + "learning_rate": 1e-06, + "loss": 1.1035, + "mean_token_accuracy": 0.6789398193359375, + "num_tokens": 275363466.0, + "step": 10637 + }, + { + "epoch": 1.1682407204041292, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.837693929672241, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7222229242324829, + "num_tokens": 275383405.0, + "step": 10638 + }, + { + "epoch": 1.1683505381067427, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5196237564086914, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.714873194694519, + "num_tokens": 275410995.0, + "step": 10639 + }, + { + "epoch": 1.1684603558093565, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7883458137512207, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6833043098449707, + "num_tokens": 275442348.0, + "step": 10640 + }, + { + "epoch": 1.16857017351197, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4701054096221924, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6945623159408569, + "num_tokens": 275473597.0, + "step": 10641 + }, + { + "epoch": 1.1686799912145838, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5554895401000977, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6884869337081909, + "num_tokens": 275505412.0, + "step": 10642 + }, + { + "epoch": 1.1687898089171975, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7791969776153564, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7018694877624512, + "num_tokens": 275528121.0, + "step": 10643 + }, + { + "epoch": 1.168899626619811, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6100637912750244, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7254292368888855, + "num_tokens": 275549898.0, + "step": 10644 + }, + { + "epoch": 1.1690094443224248, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3775131702423096, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.681243896484375, + "num_tokens": 275579992.0, + "step": 10645 + }, + { + "epoch": 1.1691192620250384, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.798678159713745, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7001558542251587, + "num_tokens": 275602864.0, + "step": 10646 + }, + { + "epoch": 1.169229079727652, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.327763557434082, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.718547523021698, + "num_tokens": 275631977.0, + "step": 10647 + }, + { + "epoch": 1.1693388974302659, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6113386154174805, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7110723257064819, + "num_tokens": 275661055.0, + "step": 10648 + }, + { + "epoch": 1.1694487151328794, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5204362869262695, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7216355800628662, + "num_tokens": 275686776.0, + "step": 10649 + }, + { + "epoch": 1.1695585328354932, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.2005300521850586, + "learning_rate": 1e-06, + "loss": 0.9049, + "mean_token_accuracy": 0.7314760684967041, + "num_tokens": 275715084.0, + "step": 10650 + }, + { + "epoch": 1.1696683505381067, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.402770519256592, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.6843522787094116, + "num_tokens": 275748446.0, + "step": 10651 + }, + { + "epoch": 1.1697781682407205, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.905689001083374, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6928072571754456, + "num_tokens": 275770040.0, + "step": 10652 + }, + { + "epoch": 1.169887985943334, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8487701416015625, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7085068225860596, + "num_tokens": 275793328.0, + "step": 10653 + }, + { + "epoch": 1.1699978036459477, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3992183208465576, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6973897218704224, + "num_tokens": 275823134.0, + "step": 10654 + }, + { + "epoch": 1.1701076213485613, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8876006603240967, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.692943811416626, + "num_tokens": 275851797.0, + "step": 10655 + }, + { + "epoch": 1.170217439051175, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7211711406707764, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7047375440597534, + "num_tokens": 275874996.0, + "step": 10656 + }, + { + "epoch": 1.1703272567537888, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.535325050354004, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.6989555358886719, + "num_tokens": 275901105.0, + "step": 10657 + }, + { + "epoch": 1.1704370744564023, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5476622581481934, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7029959559440613, + "num_tokens": 275928391.0, + "step": 10658 + }, + { + "epoch": 1.170546892159016, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5870513916015625, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7093295454978943, + "num_tokens": 275955151.0, + "step": 10659 + }, + { + "epoch": 1.1706567098616296, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.825922727584839, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7036899328231812, + "num_tokens": 275977107.0, + "step": 10660 + }, + { + "epoch": 1.1707665275642434, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6890056133270264, + "learning_rate": 1e-06, + "loss": 1.1235, + "mean_token_accuracy": 0.674236536026001, + "num_tokens": 276006547.0, + "step": 10661 + }, + { + "epoch": 1.1708763452668571, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.456007719039917, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7052599787712097, + "num_tokens": 276034604.0, + "step": 10662 + }, + { + "epoch": 1.1709861629694707, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.642864227294922, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6903446912765503, + "num_tokens": 276062945.0, + "step": 10663 + }, + { + "epoch": 1.1710959806720844, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9744138717651367, + "learning_rate": 1e-06, + "loss": 0.9303, + "mean_token_accuracy": 0.731511652469635, + "num_tokens": 276081219.0, + "step": 10664 + }, + { + "epoch": 1.171205798374698, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.673748254776001, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6916077136993408, + "num_tokens": 276105780.0, + "step": 10665 + }, + { + "epoch": 1.1713156160773117, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.620654344558716, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7222328782081604, + "num_tokens": 276127673.0, + "step": 10666 + }, + { + "epoch": 1.1714254337799253, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.505462408065796, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7056787014007568, + "num_tokens": 276155202.0, + "step": 10667 + }, + { + "epoch": 1.171535251482539, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.551327705383301, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.6985954642295837, + "num_tokens": 276182164.0, + "step": 10668 + }, + { + "epoch": 1.1716450691851525, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5985734462738037, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7212268114089966, + "num_tokens": 276210888.0, + "step": 10669 + }, + { + "epoch": 1.1717548868877663, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.372814893722534, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7151215076446533, + "num_tokens": 276238745.0, + "step": 10670 + }, + { + "epoch": 1.17186470459038, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8837225437164307, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7137791514396667, + "num_tokens": 276262017.0, + "step": 10671 + }, + { + "epoch": 1.1719745222929936, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4945948123931885, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.6996532678604126, + "num_tokens": 276291389.0, + "step": 10672 + }, + { + "epoch": 1.1720843399956073, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7666289806365967, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7113474011421204, + "num_tokens": 276314507.0, + "step": 10673 + }, + { + "epoch": 1.1721941576982209, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.493084669113159, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7044259309768677, + "num_tokens": 276341334.0, + "step": 10674 + }, + { + "epoch": 1.1723039754008346, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.469101667404175, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7009899616241455, + "num_tokens": 276366493.0, + "step": 10675 + }, + { + "epoch": 1.1724137931034484, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8299756050109863, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7091358304023743, + "num_tokens": 276388674.0, + "step": 10676 + }, + { + "epoch": 1.172523610806062, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6488494873046875, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6962127685546875, + "num_tokens": 276412824.0, + "step": 10677 + }, + { + "epoch": 1.1726334285086757, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5991082191467285, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7107871770858765, + "num_tokens": 276440248.0, + "step": 10678 + }, + { + "epoch": 1.1727432462112892, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.442171573638916, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7124074697494507, + "num_tokens": 276466961.0, + "step": 10679 + }, + { + "epoch": 1.172853063913903, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6217525005340576, + "learning_rate": 1e-06, + "loss": 0.8786, + "mean_token_accuracy": 0.734515368938446, + "num_tokens": 276488568.0, + "step": 10680 + }, + { + "epoch": 1.1729628816165165, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.544404983520508, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7099525332450867, + "num_tokens": 276514628.0, + "step": 10681 + }, + { + "epoch": 1.1730726993191303, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.37245774269104, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.7028656601905823, + "num_tokens": 276546001.0, + "step": 10682 + }, + { + "epoch": 1.1731825170217438, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.674494504928589, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7131167650222778, + "num_tokens": 276567144.0, + "step": 10683 + }, + { + "epoch": 1.1732923347243576, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.269998788833618, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6925387382507324, + "num_tokens": 276600189.0, + "step": 10684 + }, + { + "epoch": 1.1734021524269713, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.552258014678955, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.709419846534729, + "num_tokens": 276626814.0, + "step": 10685 + }, + { + "epoch": 1.1735119701295849, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.440463066101074, + "learning_rate": 1e-06, + "loss": 1.1454, + "mean_token_accuracy": 0.6713917851448059, + "num_tokens": 276660157.0, + "step": 10686 + }, + { + "epoch": 1.1736217878321986, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.498758316040039, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6986716985702515, + "num_tokens": 276687209.0, + "step": 10687 + }, + { + "epoch": 1.1737316055348122, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.242495059967041, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7116020917892456, + "num_tokens": 276719689.0, + "step": 10688 + }, + { + "epoch": 1.173841423237426, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.455037832260132, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7112314701080322, + "num_tokens": 276750718.0, + "step": 10689 + }, + { + "epoch": 1.1739512409400394, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1543121337890625, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7028055191040039, + "num_tokens": 276775344.0, + "step": 10690 + }, + { + "epoch": 1.1740610586426532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6994330883026123, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6821528673171997, + "num_tokens": 276802480.0, + "step": 10691 + }, + { + "epoch": 1.1741708763452667, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6199610233306885, + "learning_rate": 1e-06, + "loss": 0.8923, + "mean_token_accuracy": 0.7326846122741699, + "num_tokens": 276825093.0, + "step": 10692 + }, + { + "epoch": 1.1742806940478805, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5984277725219727, + "learning_rate": 1e-06, + "loss": 1.1158, + "mean_token_accuracy": 0.6785027980804443, + "num_tokens": 276859244.0, + "step": 10693 + }, + { + "epoch": 1.1743905117504942, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.80403733253479, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7078158855438232, + "num_tokens": 276883396.0, + "step": 10694 + }, + { + "epoch": 1.1745003294531078, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.631171464920044, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7200202345848083, + "num_tokens": 276910340.0, + "step": 10695 + }, + { + "epoch": 1.1746101471557215, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8564274311065674, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7142738103866577, + "num_tokens": 276935607.0, + "step": 10696 + }, + { + "epoch": 1.174719964858335, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6937825679779053, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7026742100715637, + "num_tokens": 276958407.0, + "step": 10697 + }, + { + "epoch": 1.1748297825609488, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5419981479644775, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.70009446144104, + "num_tokens": 276984274.0, + "step": 10698 + }, + { + "epoch": 1.1749396002635626, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 8.888590812683105, + "learning_rate": 1e-06, + "loss": 0.9099, + "mean_token_accuracy": 0.7284575700759888, + "num_tokens": 277003924.0, + "step": 10699 + }, + { + "epoch": 1.1750494179661761, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.800511121749878, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.6968799829483032, + "num_tokens": 277029326.0, + "step": 10700 + }, + { + "epoch": 1.1751592356687899, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6653497219085693, + "learning_rate": 1e-06, + "loss": 0.9099, + "mean_token_accuracy": 0.73656165599823, + "num_tokens": 277052982.0, + "step": 10701 + }, + { + "epoch": 1.1752690533714034, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4025869369506836, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7117506861686707, + "num_tokens": 277082856.0, + "step": 10702 + }, + { + "epoch": 1.1753788710740172, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.518097400665283, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6839749813079834, + "num_tokens": 277110580.0, + "step": 10703 + }, + { + "epoch": 1.1754886887766307, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7778689861297607, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7088330984115601, + "num_tokens": 277133087.0, + "step": 10704 + }, + { + "epoch": 1.1755985064792445, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.691798686981201, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7055967450141907, + "num_tokens": 277157403.0, + "step": 10705 + }, + { + "epoch": 1.175708324181858, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.525172710418701, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.7000522613525391, + "num_tokens": 277183739.0, + "step": 10706 + }, + { + "epoch": 1.1758181418844718, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 5.358803749084473, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6959091424942017, + "num_tokens": 277210821.0, + "step": 10707 + }, + { + "epoch": 1.1759279595870855, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4739298820495605, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7173020839691162, + "num_tokens": 277236725.0, + "step": 10708 + }, + { + "epoch": 1.176037777289699, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5594677925109863, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.6987677216529846, + "num_tokens": 277268483.0, + "step": 10709 + }, + { + "epoch": 1.1761475949923128, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.385798692703247, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7146068215370178, + "num_tokens": 277296117.0, + "step": 10710 + }, + { + "epoch": 1.1762574126949263, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.672731637954712, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7117035388946533, + "num_tokens": 277318543.0, + "step": 10711 + }, + { + "epoch": 1.17636723039754, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.481356143951416, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.6988157033920288, + "num_tokens": 277347866.0, + "step": 10712 + }, + { + "epoch": 1.1764770481001539, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4180774688720703, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6834016442298889, + "num_tokens": 277376341.0, + "step": 10713 + }, + { + "epoch": 1.1765868658027674, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3565242290496826, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7199920415878296, + "num_tokens": 277403427.0, + "step": 10714 + }, + { + "epoch": 1.1766966835053811, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.357030153274536, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7131534814834595, + "num_tokens": 277430711.0, + "step": 10715 + }, + { + "epoch": 1.1768065012079947, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6396825313568115, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6845632791519165, + "num_tokens": 277460092.0, + "step": 10716 + }, + { + "epoch": 1.1769163189106084, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.4150285720825195, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6880320906639099, + "num_tokens": 277486876.0, + "step": 10717 + }, + { + "epoch": 1.177026136613222, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.58823299407959, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7000343799591064, + "num_tokens": 277510917.0, + "step": 10718 + }, + { + "epoch": 1.1771359543158357, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.5337719917297363, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6946895122528076, + "num_tokens": 277537077.0, + "step": 10719 + }, + { + "epoch": 1.1772457720184493, + "ewc_loss": 9.179115295410156e-06, + "grad_norm": 2.7521512508392334, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7066874504089355, + "num_tokens": 277559338.0, + "step": 10720 + }, + { + "epoch": 1.177355589721063, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9028449058532715, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7228555679321289, + "num_tokens": 277580893.0, + "step": 10721 + }, + { + "epoch": 1.1774654074236768, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8455681800842285, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6958094835281372, + "num_tokens": 277602487.0, + "step": 10722 + }, + { + "epoch": 1.1775752251262903, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5204317569732666, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6993178725242615, + "num_tokens": 277633982.0, + "step": 10723 + }, + { + "epoch": 1.177685042828904, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8488285541534424, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7143616080284119, + "num_tokens": 277656418.0, + "step": 10724 + }, + { + "epoch": 1.1777948605315176, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6751739978790283, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6905050873756409, + "num_tokens": 277681706.0, + "step": 10725 + }, + { + "epoch": 1.1779046782341314, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.697575092315674, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7178064584732056, + "num_tokens": 277704022.0, + "step": 10726 + }, + { + "epoch": 1.1780144959367451, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.257641315460205, + "learning_rate": 1e-06, + "loss": 0.9071, + "mean_token_accuracy": 0.7350003123283386, + "num_tokens": 277722483.0, + "step": 10727 + }, + { + "epoch": 1.1781243136393587, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9893970489501953, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7111895084381104, + "num_tokens": 277744425.0, + "step": 10728 + }, + { + "epoch": 1.1782341313419724, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.4925482273101807, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7153352499008179, + "num_tokens": 277762468.0, + "step": 10729 + }, + { + "epoch": 1.178343949044586, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5874218940734863, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6943441033363342, + "num_tokens": 277789603.0, + "step": 10730 + }, + { + "epoch": 1.1784537667471997, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5020875930786133, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6860015392303467, + "num_tokens": 277816454.0, + "step": 10731 + }, + { + "epoch": 1.1785635844498132, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4254300594329834, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7221850156784058, + "num_tokens": 277844616.0, + "step": 10732 + }, + { + "epoch": 1.178673402152427, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5917975902557373, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.6978499889373779, + "num_tokens": 277870836.0, + "step": 10733 + }, + { + "epoch": 1.1787832198550405, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.64294695854187, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7130830883979797, + "num_tokens": 277894579.0, + "step": 10734 + }, + { + "epoch": 1.1788930375576543, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.753787040710449, + "learning_rate": 1e-06, + "loss": 1.1311, + "mean_token_accuracy": 0.6758246421813965, + "num_tokens": 277924162.0, + "step": 10735 + }, + { + "epoch": 1.179002855260268, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3139235973358154, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.706607460975647, + "num_tokens": 277952941.0, + "step": 10736 + }, + { + "epoch": 1.1791126729628816, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.749168872833252, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6925254464149475, + "num_tokens": 277977943.0, + "step": 10737 + }, + { + "epoch": 1.1792224906654953, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6299808025360107, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.6959086656570435, + "num_tokens": 278000702.0, + "step": 10738 + }, + { + "epoch": 1.1793323083681089, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7973835468292236, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7095142006874084, + "num_tokens": 278023195.0, + "step": 10739 + }, + { + "epoch": 1.1794421260707226, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.700871229171753, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7137160897254944, + "num_tokens": 278051203.0, + "step": 10740 + }, + { + "epoch": 1.1795519437733364, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.556321144104004, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7109702825546265, + "num_tokens": 278076421.0, + "step": 10741 + }, + { + "epoch": 1.17966176147595, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6422293186187744, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.699699878692627, + "num_tokens": 278101100.0, + "step": 10742 + }, + { + "epoch": 1.1797715791785637, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6876754760742188, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6902581453323364, + "num_tokens": 278126360.0, + "step": 10743 + }, + { + "epoch": 1.1798813968811772, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4232735633850098, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7087142467498779, + "num_tokens": 278152646.0, + "step": 10744 + }, + { + "epoch": 1.179991214583791, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.799403429031372, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6961421370506287, + "num_tokens": 278174769.0, + "step": 10745 + }, + { + "epoch": 1.1801010322864045, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3790316581726074, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7131998538970947, + "num_tokens": 278203501.0, + "step": 10746 + }, + { + "epoch": 1.1802108499890183, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.588101387023926, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7192038297653198, + "num_tokens": 278230519.0, + "step": 10747 + }, + { + "epoch": 1.1803206676916318, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.686816453933716, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7183842658996582, + "num_tokens": 278256524.0, + "step": 10748 + }, + { + "epoch": 1.1804304853942456, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.156940221786499, + "learning_rate": 1e-06, + "loss": 0.9262, + "mean_token_accuracy": 0.7234209179878235, + "num_tokens": 278278805.0, + "step": 10749 + }, + { + "epoch": 1.1805403030968593, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.519106149673462, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.7040385007858276, + "num_tokens": 278307389.0, + "step": 10750 + }, + { + "epoch": 1.1806501207994728, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6382241249084473, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6961129903793335, + "num_tokens": 278332455.0, + "step": 10751 + }, + { + "epoch": 1.1807599385020866, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3975436687469482, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7023333311080933, + "num_tokens": 278362515.0, + "step": 10752 + }, + { + "epoch": 1.1808697562047001, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.716801881790161, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7254198789596558, + "num_tokens": 278388616.0, + "step": 10753 + }, + { + "epoch": 1.180979573907314, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9134905338287354, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6966620683670044, + "num_tokens": 278414433.0, + "step": 10754 + }, + { + "epoch": 1.1810893916099274, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.692063808441162, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7190440893173218, + "num_tokens": 278435896.0, + "step": 10755 + }, + { + "epoch": 1.1811992093125412, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7391703128814697, + "learning_rate": 1e-06, + "loss": 0.9265, + "mean_token_accuracy": 0.727454423904419, + "num_tokens": 278457544.0, + "step": 10756 + }, + { + "epoch": 1.1813090270151547, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.572310447692871, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6886699199676514, + "num_tokens": 278483248.0, + "step": 10757 + }, + { + "epoch": 1.1814188447177685, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6786673069000244, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6992238163948059, + "num_tokens": 278506615.0, + "step": 10758 + }, + { + "epoch": 1.1815286624203822, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.55879807472229, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.70055091381073, + "num_tokens": 278533229.0, + "step": 10759 + }, + { + "epoch": 1.1816384801229958, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.542764186859131, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7135753631591797, + "num_tokens": 278559948.0, + "step": 10760 + }, + { + "epoch": 1.1817482978256095, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.699810266494751, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7053940296173096, + "num_tokens": 278583103.0, + "step": 10761 + }, + { + "epoch": 1.181858115528223, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6159019470214844, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7133120894432068, + "num_tokens": 278610819.0, + "step": 10762 + }, + { + "epoch": 1.1819679332308368, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.59560489654541, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.7302312850952148, + "num_tokens": 278639297.0, + "step": 10763 + }, + { + "epoch": 1.1820777509334506, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.41923189163208, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.691606342792511, + "num_tokens": 278668029.0, + "step": 10764 + }, + { + "epoch": 1.1821875686360641, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.3346774578094482, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7055397033691406, + "num_tokens": 278684837.0, + "step": 10765 + }, + { + "epoch": 1.1822973863386779, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4811339378356934, + "learning_rate": 1e-06, + "loss": 1.1631, + "mean_token_accuracy": 0.6677562594413757, + "num_tokens": 278716488.0, + "step": 10766 + }, + { + "epoch": 1.1824072040412914, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7569210529327393, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.701665997505188, + "num_tokens": 278739057.0, + "step": 10767 + }, + { + "epoch": 1.1825170217439052, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8973379135131836, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.7185477018356323, + "num_tokens": 278757143.0, + "step": 10768 + }, + { + "epoch": 1.1826268394465187, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.419796943664551, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7114098072052002, + "num_tokens": 278788023.0, + "step": 10769 + }, + { + "epoch": 1.1827366571491325, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 20.571210861206055, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7062740325927734, + "num_tokens": 278811875.0, + "step": 10770 + }, + { + "epoch": 1.182846474851746, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9865341186523438, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.695807695388794, + "num_tokens": 278837201.0, + "step": 10771 + }, + { + "epoch": 1.1829562925543597, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.2230288982391357, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7242696285247803, + "num_tokens": 278856995.0, + "step": 10772 + }, + { + "epoch": 1.1830661102569735, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.72977876663208, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7014795541763306, + "num_tokens": 278879410.0, + "step": 10773 + }, + { + "epoch": 1.183175927959587, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7113661766052246, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6981980800628662, + "num_tokens": 278902806.0, + "step": 10774 + }, + { + "epoch": 1.1832857456622008, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.703031539916992, + "learning_rate": 1e-06, + "loss": 1.0794, + "mean_token_accuracy": 0.6934207081794739, + "num_tokens": 278928228.0, + "step": 10775 + }, + { + "epoch": 1.1833955633648143, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.506571054458618, + "learning_rate": 1e-06, + "loss": 1.0745, + "mean_token_accuracy": 0.6857735514640808, + "num_tokens": 278956107.0, + "step": 10776 + }, + { + "epoch": 1.183505381067428, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6867411136627197, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7260063290596008, + "num_tokens": 278980976.0, + "step": 10777 + }, + { + "epoch": 1.1836151987700418, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.641273260116577, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7094867825508118, + "num_tokens": 279006438.0, + "step": 10778 + }, + { + "epoch": 1.1837250164726554, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0157952308654785, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7010321021080017, + "num_tokens": 279025046.0, + "step": 10779 + }, + { + "epoch": 1.1838348341752691, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.442763090133667, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7066842913627625, + "num_tokens": 279054000.0, + "step": 10780 + }, + { + "epoch": 1.1839446518778827, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3786089420318604, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.6809679865837097, + "num_tokens": 279086191.0, + "step": 10781 + }, + { + "epoch": 1.1840544695804964, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8503987789154053, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7137466073036194, + "num_tokens": 279105947.0, + "step": 10782 + }, + { + "epoch": 1.18416428728311, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8935275077819824, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7234760522842407, + "num_tokens": 279132323.0, + "step": 10783 + }, + { + "epoch": 1.1842741049857237, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.666753053665161, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6948186159133911, + "num_tokens": 279157452.0, + "step": 10784 + }, + { + "epoch": 1.1843839226883373, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5761172771453857, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.7045848369598389, + "num_tokens": 279185069.0, + "step": 10785 + }, + { + "epoch": 1.184493740390951, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5102133750915527, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7151662111282349, + "num_tokens": 279212124.0, + "step": 10786 + }, + { + "epoch": 1.1846035580935648, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.641988754272461, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7084659337997437, + "num_tokens": 279239591.0, + "step": 10787 + }, + { + "epoch": 1.1847133757961783, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4733974933624268, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7155109643936157, + "num_tokens": 279269866.0, + "step": 10788 + }, + { + "epoch": 1.184823193498792, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.743816614151001, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.6986427307128906, + "num_tokens": 279293521.0, + "step": 10789 + }, + { + "epoch": 1.1849330112014056, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.341643810272217, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7174968719482422, + "num_tokens": 279326470.0, + "step": 10790 + }, + { + "epoch": 1.1850428289040194, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.589155912399292, + "learning_rate": 1e-06, + "loss": 0.9252, + "mean_token_accuracy": 0.7369601726531982, + "num_tokens": 279351307.0, + "step": 10791 + }, + { + "epoch": 1.185152646606633, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4761979579925537, + "learning_rate": 1e-06, + "loss": 1.1079, + "mean_token_accuracy": 0.6864379048347473, + "num_tokens": 279379656.0, + "step": 10792 + }, + { + "epoch": 1.1852624643092466, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.759997844696045, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7175197601318359, + "num_tokens": 279401365.0, + "step": 10793 + }, + { + "epoch": 1.1853722820118604, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4380152225494385, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6902371644973755, + "num_tokens": 279431184.0, + "step": 10794 + }, + { + "epoch": 1.185482099714474, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7290384769439697, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6979094743728638, + "num_tokens": 279455386.0, + "step": 10795 + }, + { + "epoch": 1.1855919174170877, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5196096897125244, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7031735181808472, + "num_tokens": 279480157.0, + "step": 10796 + }, + { + "epoch": 1.1857017351197012, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2366271018981934, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7106540203094482, + "num_tokens": 279511184.0, + "step": 10797 + }, + { + "epoch": 1.185811552822315, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.748660087585449, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7042819261550903, + "num_tokens": 279536340.0, + "step": 10798 + }, + { + "epoch": 1.1859213705249285, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4524662494659424, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.7136125564575195, + "num_tokens": 279568355.0, + "step": 10799 + }, + { + "epoch": 1.1860311882275423, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.660139560699463, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.7041152715682983, + "num_tokens": 279592058.0, + "step": 10800 + }, + { + "epoch": 1.186141005930156, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6810505390167236, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6925727725028992, + "num_tokens": 279616851.0, + "step": 10801 + }, + { + "epoch": 1.1862508236327696, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.512009859085083, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7077941298484802, + "num_tokens": 279643692.0, + "step": 10802 + }, + { + "epoch": 1.1863606413353833, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.549645185470581, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7087355852127075, + "num_tokens": 279668872.0, + "step": 10803 + }, + { + "epoch": 1.1864704590379969, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.629913568496704, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7083083391189575, + "num_tokens": 279695404.0, + "step": 10804 + }, + { + "epoch": 1.1865802767406106, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5823071002960205, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7135450839996338, + "num_tokens": 279718073.0, + "step": 10805 + }, + { + "epoch": 1.1866900944432242, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.528911590576172, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7003337740898132, + "num_tokens": 279747779.0, + "step": 10806 + }, + { + "epoch": 1.186799912145838, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.863755464553833, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6928204894065857, + "num_tokens": 279781480.0, + "step": 10807 + }, + { + "epoch": 1.1869097298484517, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6067185401916504, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7058420777320862, + "num_tokens": 279809082.0, + "step": 10808 + }, + { + "epoch": 1.1870195475510652, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.743643283843994, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7157248258590698, + "num_tokens": 279832111.0, + "step": 10809 + }, + { + "epoch": 1.187129365253679, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.417483329772949, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7005659341812134, + "num_tokens": 279860701.0, + "step": 10810 + }, + { + "epoch": 1.1872391829562925, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3420424461364746, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6944156885147095, + "num_tokens": 279891858.0, + "step": 10811 + }, + { + "epoch": 1.1873490006589063, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8541388511657715, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7117888927459717, + "num_tokens": 279911871.0, + "step": 10812 + }, + { + "epoch": 1.1874588183615198, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5725016593933105, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.7031298875808716, + "num_tokens": 279937887.0, + "step": 10813 + }, + { + "epoch": 1.1875686360641335, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5132596492767334, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.6987476348876953, + "num_tokens": 279965585.0, + "step": 10814 + }, + { + "epoch": 1.1876784537667473, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.033139944076538, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6993544101715088, + "num_tokens": 279985822.0, + "step": 10815 + }, + { + "epoch": 1.1877882714693608, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.486475944519043, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7076132893562317, + "num_tokens": 280016291.0, + "step": 10816 + }, + { + "epoch": 1.1878980891719746, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4749162197113037, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.6945662498474121, + "num_tokens": 280045789.0, + "step": 10817 + }, + { + "epoch": 1.1880079068745881, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.943523406982422, + "learning_rate": 1e-06, + "loss": 0.9314, + "mean_token_accuracy": 0.7276515960693359, + "num_tokens": 280068256.0, + "step": 10818 + }, + { + "epoch": 1.1881177245772019, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6925439834594727, + "learning_rate": 1e-06, + "loss": 0.9059, + "mean_token_accuracy": 0.7390612959861755, + "num_tokens": 280092650.0, + "step": 10819 + }, + { + "epoch": 1.1882275422798154, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6763064861297607, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.703105092048645, + "num_tokens": 280118424.0, + "step": 10820 + }, + { + "epoch": 1.1883373599824292, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5586140155792236, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7034873962402344, + "num_tokens": 280142730.0, + "step": 10821 + }, + { + "epoch": 1.1884471776850427, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7811832427978516, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6948279142379761, + "num_tokens": 280163927.0, + "step": 10822 + }, + { + "epoch": 1.1885569953876565, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8012259006500244, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7089846134185791, + "num_tokens": 280184953.0, + "step": 10823 + }, + { + "epoch": 1.1886668130902702, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.844061851501465, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6971723437309265, + "num_tokens": 280206670.0, + "step": 10824 + }, + { + "epoch": 1.1887766307928838, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.669137954711914, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7122405767440796, + "num_tokens": 280232882.0, + "step": 10825 + }, + { + "epoch": 1.1888864484954975, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.751894474029541, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7167960405349731, + "num_tokens": 280257097.0, + "step": 10826 + }, + { + "epoch": 1.188996266198111, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5273492336273193, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7161686420440674, + "num_tokens": 280283838.0, + "step": 10827 + }, + { + "epoch": 1.1891060839007248, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.812220573425293, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7186194658279419, + "num_tokens": 280305715.0, + "step": 10828 + }, + { + "epoch": 1.1892159016033386, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9855589866638184, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.722273588180542, + "num_tokens": 280323948.0, + "step": 10829 + }, + { + "epoch": 1.189325719305952, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6269690990448, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.7033506035804749, + "num_tokens": 280352256.0, + "step": 10830 + }, + { + "epoch": 1.1894355370085659, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6148672103881836, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6902283430099487, + "num_tokens": 280381460.0, + "step": 10831 + }, + { + "epoch": 1.1895453547111794, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.122751474380493, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7177446484565735, + "num_tokens": 280402295.0, + "step": 10832 + }, + { + "epoch": 1.1896551724137931, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.663921356201172, + "learning_rate": 1e-06, + "loss": 0.8949, + "mean_token_accuracy": 0.741108775138855, + "num_tokens": 280424774.0, + "step": 10833 + }, + { + "epoch": 1.1897649901164067, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.872579336166382, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6885843276977539, + "num_tokens": 280444824.0, + "step": 10834 + }, + { + "epoch": 1.1898748078190204, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6666042804718018, + "learning_rate": 1e-06, + "loss": 1.0418, + "mean_token_accuracy": 0.7001707553863525, + "num_tokens": 280469134.0, + "step": 10835 + }, + { + "epoch": 1.189984625521634, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4066002368927, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.69972163438797, + "num_tokens": 280499041.0, + "step": 10836 + }, + { + "epoch": 1.1900944432242477, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4960994720458984, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6975517868995667, + "num_tokens": 280528314.0, + "step": 10837 + }, + { + "epoch": 1.1902042609268615, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5943448543548584, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6996233463287354, + "num_tokens": 280554192.0, + "step": 10838 + }, + { + "epoch": 1.190314078629475, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5951173305511475, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7122779488563538, + "num_tokens": 280579201.0, + "step": 10839 + }, + { + "epoch": 1.1904238963320888, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4072797298431396, + "learning_rate": 1e-06, + "loss": 1.1248, + "mean_token_accuracy": 0.6833625435829163, + "num_tokens": 280613242.0, + "step": 10840 + }, + { + "epoch": 1.1905337140347023, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.42179012298584, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.710220456123352, + "num_tokens": 280641227.0, + "step": 10841 + }, + { + "epoch": 1.190643531737316, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.386223793029785, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.702671229839325, + "num_tokens": 280671471.0, + "step": 10842 + }, + { + "epoch": 1.1907533494399298, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.645198345184326, + "learning_rate": 1e-06, + "loss": 1.0837, + "mean_token_accuracy": 0.6935186982154846, + "num_tokens": 280695950.0, + "step": 10843 + }, + { + "epoch": 1.1908631671425434, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.940430164337158, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.7300912141799927, + "num_tokens": 280714361.0, + "step": 10844 + }, + { + "epoch": 1.1909729848451571, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9030327796936035, + "learning_rate": 1e-06, + "loss": 0.9131, + "mean_token_accuracy": 0.7330389618873596, + "num_tokens": 280734389.0, + "step": 10845 + }, + { + "epoch": 1.1910828025477707, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.569336414337158, + "learning_rate": 1e-06, + "loss": 0.9358, + "mean_token_accuracy": 0.7200563549995422, + "num_tokens": 280759798.0, + "step": 10846 + }, + { + "epoch": 1.1911926202503844, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8012232780456543, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7174800634384155, + "num_tokens": 280782156.0, + "step": 10847 + }, + { + "epoch": 1.191302437952998, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5670602321624756, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.7005206942558289, + "num_tokens": 280807703.0, + "step": 10848 + }, + { + "epoch": 1.1914122556556117, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.600750207901001, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7010018229484558, + "num_tokens": 280831959.0, + "step": 10849 + }, + { + "epoch": 1.1915220733582252, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2764947414398193, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.706514835357666, + "num_tokens": 280864614.0, + "step": 10850 + }, + { + "epoch": 1.191631891060839, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.616368532180786, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6931672096252441, + "num_tokens": 280891174.0, + "step": 10851 + }, + { + "epoch": 1.1917417087634528, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6271462440490723, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7051997780799866, + "num_tokens": 280919838.0, + "step": 10852 + }, + { + "epoch": 1.1918515264660663, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0697436332702637, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7134119868278503, + "num_tokens": 280938891.0, + "step": 10853 + }, + { + "epoch": 1.19196134416868, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.635035991668701, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7059706449508667, + "num_tokens": 280962915.0, + "step": 10854 + }, + { + "epoch": 1.1920711618712936, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4220383167266846, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.694580078125, + "num_tokens": 280991722.0, + "step": 10855 + }, + { + "epoch": 1.1921809795739073, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4383955001831055, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6941772103309631, + "num_tokens": 281021599.0, + "step": 10856 + }, + { + "epoch": 1.192290797276521, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4246788024902344, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7139692902565002, + "num_tokens": 281051685.0, + "step": 10857 + }, + { + "epoch": 1.1924006149791346, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.813760757446289, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7013676166534424, + "num_tokens": 281074117.0, + "step": 10858 + }, + { + "epoch": 1.1925104326817484, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.516563653945923, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6961309909820557, + "num_tokens": 281102277.0, + "step": 10859 + }, + { + "epoch": 1.192620250384362, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.036118984222412, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.7006292343139648, + "num_tokens": 281128923.0, + "step": 10860 + }, + { + "epoch": 1.1927300680869757, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.451179265975952, + "learning_rate": 1e-06, + "loss": 0.9171, + "mean_token_accuracy": 0.735527515411377, + "num_tokens": 281156396.0, + "step": 10861 + }, + { + "epoch": 1.1928398857895892, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.674720048904419, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7133011817932129, + "num_tokens": 281181795.0, + "step": 10862 + }, + { + "epoch": 1.192949703492203, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7507102489471436, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.6982598900794983, + "num_tokens": 281205956.0, + "step": 10863 + }, + { + "epoch": 1.1930595211948165, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6601722240448, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7007325887680054, + "num_tokens": 281231044.0, + "step": 10864 + }, + { + "epoch": 1.1931693388974303, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 9.70206356048584, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7230710983276367, + "num_tokens": 281256379.0, + "step": 10865 + }, + { + "epoch": 1.193279156600044, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.494032144546509, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7044607400894165, + "num_tokens": 281286506.0, + "step": 10866 + }, + { + "epoch": 1.1933889743026576, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7571260929107666, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.701656699180603, + "num_tokens": 281315045.0, + "step": 10867 + }, + { + "epoch": 1.1934987920052713, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.630748748779297, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7113536596298218, + "num_tokens": 281338268.0, + "step": 10868 + }, + { + "epoch": 1.1936086097078848, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8673131465911865, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7250404953956604, + "num_tokens": 281358237.0, + "step": 10869 + }, + { + "epoch": 1.1937184274104986, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7958712577819824, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7259796261787415, + "num_tokens": 281379501.0, + "step": 10870 + }, + { + "epoch": 1.1938282451131121, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.840822696685791, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7137371301651001, + "num_tokens": 281399907.0, + "step": 10871 + }, + { + "epoch": 1.193938062815726, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5216760635375977, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.707851767539978, + "num_tokens": 281426171.0, + "step": 10872 + }, + { + "epoch": 1.1940478805183394, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.498767614364624, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7016910314559937, + "num_tokens": 281452592.0, + "step": 10873 + }, + { + "epoch": 1.1941576982209532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4680140018463135, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6950457096099854, + "num_tokens": 281479924.0, + "step": 10874 + }, + { + "epoch": 1.194267515923567, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7090108394622803, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7057647705078125, + "num_tokens": 281503398.0, + "step": 10875 + }, + { + "epoch": 1.1943773336261805, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.673820972442627, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7080684900283813, + "num_tokens": 281525676.0, + "step": 10876 + }, + { + "epoch": 1.1944871513287942, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.628868341445923, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.712518572807312, + "num_tokens": 281551631.0, + "step": 10877 + }, + { + "epoch": 1.1945969690314078, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.946363687515259, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7074124813079834, + "num_tokens": 281572796.0, + "step": 10878 + }, + { + "epoch": 1.1947067867340215, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6077699661254883, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7009803652763367, + "num_tokens": 281597170.0, + "step": 10879 + }, + { + "epoch": 1.1948166044366353, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.710207462310791, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7141546607017517, + "num_tokens": 281620614.0, + "step": 10880 + }, + { + "epoch": 1.1949264221392488, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7408928871154785, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7070119380950928, + "num_tokens": 281646288.0, + "step": 10881 + }, + { + "epoch": 1.1950362398418626, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.802656650543213, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7180652618408203, + "num_tokens": 281667782.0, + "step": 10882 + }, + { + "epoch": 1.1951460575444761, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6626813411712646, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.7120899558067322, + "num_tokens": 281694308.0, + "step": 10883 + }, + { + "epoch": 1.1952558752470899, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.742684841156006, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7180312871932983, + "num_tokens": 281716188.0, + "step": 10884 + }, + { + "epoch": 1.1953656929497034, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.692970037460327, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6834633350372314, + "num_tokens": 281743927.0, + "step": 10885 + }, + { + "epoch": 1.1954755106523172, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5011351108551025, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7065777778625488, + "num_tokens": 281770630.0, + "step": 10886 + }, + { + "epoch": 1.1955853283549307, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7778372764587402, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6883413791656494, + "num_tokens": 281793140.0, + "step": 10887 + }, + { + "epoch": 1.1956951460575445, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7882473468780518, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.719132661819458, + "num_tokens": 281814623.0, + "step": 10888 + }, + { + "epoch": 1.1958049637601582, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6767542362213135, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7044327855110168, + "num_tokens": 281838998.0, + "step": 10889 + }, + { + "epoch": 1.1959147814627717, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4708828926086426, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7206430435180664, + "num_tokens": 281867363.0, + "step": 10890 + }, + { + "epoch": 1.1960245991653855, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.649186849594116, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7137429714202881, + "num_tokens": 281889231.0, + "step": 10891 + }, + { + "epoch": 1.196134416867999, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.578508138656616, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.696732759475708, + "num_tokens": 281917109.0, + "step": 10892 + }, + { + "epoch": 1.1962442345706128, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.643047332763672, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6961182355880737, + "num_tokens": 281942238.0, + "step": 10893 + }, + { + "epoch": 1.1963540522732266, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0186829566955566, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7106332778930664, + "num_tokens": 281963200.0, + "step": 10894 + }, + { + "epoch": 1.19646386997584, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.607456922531128, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.6950626969337463, + "num_tokens": 281991008.0, + "step": 10895 + }, + { + "epoch": 1.1965736876784538, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.6873667240142822, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.7186375856399536, + "num_tokens": 282013040.0, + "step": 10896 + }, + { + "epoch": 1.1966835053810674, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.861184597015381, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7173251509666443, + "num_tokens": 282033724.0, + "step": 10897 + }, + { + "epoch": 1.1967933230836811, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4808714389801025, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6903150677680969, + "num_tokens": 282064888.0, + "step": 10898 + }, + { + "epoch": 1.1969031407862947, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.49453067779541, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7103608846664429, + "num_tokens": 282091842.0, + "step": 10899 + }, + { + "epoch": 1.1970129584889084, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3778367042541504, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.7173431515693665, + "num_tokens": 282121276.0, + "step": 10900 + }, + { + "epoch": 1.197122776191522, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1766371726989746, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7169280052185059, + "num_tokens": 282138436.0, + "step": 10901 + }, + { + "epoch": 1.1972325938941357, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.71645188331604, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7134115695953369, + "num_tokens": 282161974.0, + "step": 10902 + }, + { + "epoch": 1.1973424115967495, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6879677772521973, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6826598644256592, + "num_tokens": 282187070.0, + "step": 10903 + }, + { + "epoch": 1.197452229299363, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.943173885345459, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7004914283752441, + "num_tokens": 282210391.0, + "step": 10904 + }, + { + "epoch": 1.1975620470019768, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5098583698272705, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.7190539836883545, + "num_tokens": 282236268.0, + "step": 10905 + }, + { + "epoch": 1.1976718647045903, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3403093814849854, + "learning_rate": 1e-06, + "loss": 1.1234, + "mean_token_accuracy": 0.6812872886657715, + "num_tokens": 282266725.0, + "step": 10906 + }, + { + "epoch": 1.197781682407204, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8042309284210205, + "learning_rate": 1e-06, + "loss": 0.9081, + "mean_token_accuracy": 0.7387120723724365, + "num_tokens": 282286293.0, + "step": 10907 + }, + { + "epoch": 1.1978915001098178, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.37290096282959, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6952469944953918, + "num_tokens": 282317493.0, + "step": 10908 + }, + { + "epoch": 1.1980013178124314, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1342275142669678, + "learning_rate": 1e-06, + "loss": 1.0833, + "mean_token_accuracy": 0.6820629835128784, + "num_tokens": 282342822.0, + "step": 10909 + }, + { + "epoch": 1.1981111355150451, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6498825550079346, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7298473119735718, + "num_tokens": 282367191.0, + "step": 10910 + }, + { + "epoch": 1.1982209532176586, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.44222354888916, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6861758232116699, + "num_tokens": 282397156.0, + "step": 10911 + }, + { + "epoch": 1.1983307709202724, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.561397075653076, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7195962071418762, + "num_tokens": 282422169.0, + "step": 10912 + }, + { + "epoch": 1.198440588622886, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.433980941772461, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7098654508590698, + "num_tokens": 282451878.0, + "step": 10913 + }, + { + "epoch": 1.1985504063254997, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.471668004989624, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6930669546127319, + "num_tokens": 282479141.0, + "step": 10914 + }, + { + "epoch": 1.1986602240281132, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4123177528381348, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6944345235824585, + "num_tokens": 282508341.0, + "step": 10915 + }, + { + "epoch": 1.198770041730727, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4710628986358643, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7165961265563965, + "num_tokens": 282533930.0, + "step": 10916 + }, + { + "epoch": 1.1988798594333407, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7579193115234375, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7089093327522278, + "num_tokens": 282556891.0, + "step": 10917 + }, + { + "epoch": 1.1989896771359543, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7665436267852783, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.711803674697876, + "num_tokens": 282578796.0, + "step": 10918 + }, + { + "epoch": 1.199099494838568, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6448090076446533, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6977465152740479, + "num_tokens": 282603968.0, + "step": 10919 + }, + { + "epoch": 1.1992093125411816, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9749667644500732, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7094358801841736, + "num_tokens": 282623436.0, + "step": 10920 + }, + { + "epoch": 1.1993191302437953, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1243181228637695, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7103335857391357, + "num_tokens": 282644872.0, + "step": 10921 + }, + { + "epoch": 1.199428947946409, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.476092576980591, + "learning_rate": 1e-06, + "loss": 1.1083, + "mean_token_accuracy": 0.6744253635406494, + "num_tokens": 282675436.0, + "step": 10922 + }, + { + "epoch": 1.1995387656490226, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7778356075286865, + "learning_rate": 1e-06, + "loss": 0.9147, + "mean_token_accuracy": 0.728628396987915, + "num_tokens": 282698546.0, + "step": 10923 + }, + { + "epoch": 1.1996485833516364, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.539719820022583, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7002168297767639, + "num_tokens": 282722349.0, + "step": 10924 + }, + { + "epoch": 1.19975840105425, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.686062812805176, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6873332858085632, + "num_tokens": 282745554.0, + "step": 10925 + }, + { + "epoch": 1.1998682187568637, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.780851125717163, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7117124795913696, + "num_tokens": 282770136.0, + "step": 10926 + }, + { + "epoch": 1.1999780364594772, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.735541343688965, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.702774167060852, + "num_tokens": 282793880.0, + "step": 10927 + }, + { + "epoch": 1.200087854162091, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.70599627494812, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7059848308563232, + "num_tokens": 282819050.0, + "step": 10928 + }, + { + "epoch": 1.2001976718647045, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9371798038482666, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7176318764686584, + "num_tokens": 282838645.0, + "step": 10929 + }, + { + "epoch": 1.2003074895673183, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.385671615600586, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7009321451187134, + "num_tokens": 282870167.0, + "step": 10930 + }, + { + "epoch": 1.200417307269932, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.342316150665283, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7071822881698608, + "num_tokens": 282900252.0, + "step": 10931 + }, + { + "epoch": 1.2005271249725455, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.561213970184326, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.7006737589836121, + "num_tokens": 282924793.0, + "step": 10932 + }, + { + "epoch": 1.2006369426751593, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4957168102264404, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.6996159553527832, + "num_tokens": 282951368.0, + "step": 10933 + }, + { + "epoch": 1.2007467603777728, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.32441782951355, + "learning_rate": 1e-06, + "loss": 1.1185, + "mean_token_accuracy": 0.6774884462356567, + "num_tokens": 282982310.0, + "step": 10934 + }, + { + "epoch": 1.2008565780803866, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7317686080932617, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.695756196975708, + "num_tokens": 283005262.0, + "step": 10935 + }, + { + "epoch": 1.2009663957830001, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6538264751434326, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7221201658248901, + "num_tokens": 283029265.0, + "step": 10936 + }, + { + "epoch": 1.2010762134856139, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5316851139068604, + "learning_rate": 1e-06, + "loss": 0.9171, + "mean_token_accuracy": 0.7242294549942017, + "num_tokens": 283055446.0, + "step": 10937 + }, + { + "epoch": 1.2011860311882274, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5054497718811035, + "learning_rate": 1e-06, + "loss": 0.8829, + "mean_token_accuracy": 0.7367247939109802, + "num_tokens": 283080794.0, + "step": 10938 + }, + { + "epoch": 1.2012958488908412, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8334178924560547, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7056134939193726, + "num_tokens": 283103477.0, + "step": 10939 + }, + { + "epoch": 1.201405666593455, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.6407132148742676, + "learning_rate": 1e-06, + "loss": 1.0829, + "mean_token_accuracy": 0.6888096332550049, + "num_tokens": 283128921.0, + "step": 10940 + }, + { + "epoch": 1.2015154842960685, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3741772174835205, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7021890878677368, + "num_tokens": 283159069.0, + "step": 10941 + }, + { + "epoch": 1.2016253019986822, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0134377479553223, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7054487466812134, + "num_tokens": 283182509.0, + "step": 10942 + }, + { + "epoch": 1.2017351197012958, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.644793748855591, + "learning_rate": 1e-06, + "loss": 1.1042, + "mean_token_accuracy": 0.6809889078140259, + "num_tokens": 283211447.0, + "step": 10943 + }, + { + "epoch": 1.2018449374039095, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.467094659805298, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7097570896148682, + "num_tokens": 283239895.0, + "step": 10944 + }, + { + "epoch": 1.2019547551065233, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9777133464813232, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.6996073722839355, + "num_tokens": 283260397.0, + "step": 10945 + }, + { + "epoch": 1.2020645728091368, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.61008620262146, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7203333973884583, + "num_tokens": 283285092.0, + "step": 10946 + }, + { + "epoch": 1.2021743905117506, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.695281744003296, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7181762456893921, + "num_tokens": 283308409.0, + "step": 10947 + }, + { + "epoch": 1.202284208214364, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6427290439605713, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6930754780769348, + "num_tokens": 283337028.0, + "step": 10948 + }, + { + "epoch": 1.2023940259169779, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.529165744781494, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7107048034667969, + "num_tokens": 283362535.0, + "step": 10949 + }, + { + "epoch": 1.2025038436195914, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4586055278778076, + "learning_rate": 1e-06, + "loss": 1.1259, + "mean_token_accuracy": 0.6721312999725342, + "num_tokens": 283392480.0, + "step": 10950 + }, + { + "epoch": 1.2026136613222052, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9232635498046875, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7124651670455933, + "num_tokens": 283414290.0, + "step": 10951 + }, + { + "epoch": 1.2027234790248187, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8829030990600586, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7297564744949341, + "num_tokens": 283434515.0, + "step": 10952 + }, + { + "epoch": 1.2028332967274324, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7383196353912354, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7123878002166748, + "num_tokens": 283459691.0, + "step": 10953 + }, + { + "epoch": 1.2029431144300462, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6356091499328613, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7122858762741089, + "num_tokens": 283483621.0, + "step": 10954 + }, + { + "epoch": 1.2030529321326597, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5791728496551514, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7211418151855469, + "num_tokens": 283507753.0, + "step": 10955 + }, + { + "epoch": 1.2031627498352735, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5612971782684326, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.703738808631897, + "num_tokens": 283532007.0, + "step": 10956 + }, + { + "epoch": 1.203272567537887, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.772305488586426, + "learning_rate": 1e-06, + "loss": 1.1116, + "mean_token_accuracy": 0.6876583695411682, + "num_tokens": 283555781.0, + "step": 10957 + }, + { + "epoch": 1.2033823852405008, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7833993434906006, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7210033535957336, + "num_tokens": 283578730.0, + "step": 10958 + }, + { + "epoch": 1.2034922029431145, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.632399559020996, + "learning_rate": 1e-06, + "loss": 0.9044, + "mean_token_accuracy": 0.7343510985374451, + "num_tokens": 283602621.0, + "step": 10959 + }, + { + "epoch": 1.203602020645728, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.28302264213562, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7072259187698364, + "num_tokens": 283638322.0, + "step": 10960 + }, + { + "epoch": 1.2037118383483418, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.41580867767334, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7014927864074707, + "num_tokens": 283668495.0, + "step": 10961 + }, + { + "epoch": 1.2038216560509554, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.863337993621826, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7155944108963013, + "num_tokens": 283688969.0, + "step": 10962 + }, + { + "epoch": 1.2039314737535691, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5010268688201904, + "learning_rate": 1e-06, + "loss": 1.0857, + "mean_token_accuracy": 0.6882369518280029, + "num_tokens": 283717810.0, + "step": 10963 + }, + { + "epoch": 1.2040412914561827, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.891061544418335, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6915552616119385, + "num_tokens": 283740254.0, + "step": 10964 + }, + { + "epoch": 1.2041511091587964, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.885338544845581, + "learning_rate": 1e-06, + "loss": 0.9418, + "mean_token_accuracy": 0.7140319347381592, + "num_tokens": 283760259.0, + "step": 10965 + }, + { + "epoch": 1.20426092686141, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5472781658172607, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7035788893699646, + "num_tokens": 283784178.0, + "step": 10966 + }, + { + "epoch": 1.2043707445640237, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7805891036987305, + "learning_rate": 1e-06, + "loss": 0.9221, + "mean_token_accuracy": 0.7297665476799011, + "num_tokens": 283806168.0, + "step": 10967 + }, + { + "epoch": 1.2044805622666375, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.726177453994751, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7146173715591431, + "num_tokens": 283830981.0, + "step": 10968 + }, + { + "epoch": 1.204590379969251, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.622366189956665, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.7014399170875549, + "num_tokens": 283856809.0, + "step": 10969 + }, + { + "epoch": 1.2047001976718648, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7274415493011475, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7101564407348633, + "num_tokens": 283880097.0, + "step": 10970 + }, + { + "epoch": 1.2048100153744783, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6360397338867188, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.720461368560791, + "num_tokens": 283904831.0, + "step": 10971 + }, + { + "epoch": 1.204919833077092, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.82432222366333, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.697708249092102, + "num_tokens": 283931547.0, + "step": 10972 + }, + { + "epoch": 1.2050296507797058, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.904893636703491, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7113844752311707, + "num_tokens": 283950698.0, + "step": 10973 + }, + { + "epoch": 1.2051394684823193, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.747795820236206, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7144439816474915, + "num_tokens": 283974470.0, + "step": 10974 + }, + { + "epoch": 1.205249286184933, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6972906589508057, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6926717162132263, + "num_tokens": 284001087.0, + "step": 10975 + }, + { + "epoch": 1.2053591038875466, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8945956230163574, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.728225827217102, + "num_tokens": 284020322.0, + "step": 10976 + }, + { + "epoch": 1.2054689215901604, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9288933277130127, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.686679482460022, + "num_tokens": 284040858.0, + "step": 10977 + }, + { + "epoch": 1.205578739292774, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6047606468200684, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6915349960327148, + "num_tokens": 284067636.0, + "step": 10978 + }, + { + "epoch": 1.2056885569953877, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5230040550231934, + "learning_rate": 1e-06, + "loss": 1.1302, + "mean_token_accuracy": 0.6911237239837646, + "num_tokens": 284096336.0, + "step": 10979 + }, + { + "epoch": 1.2057983746980012, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.714426040649414, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.6983515024185181, + "num_tokens": 284120413.0, + "step": 10980 + }, + { + "epoch": 1.205908192400615, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6139333248138428, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7127653360366821, + "num_tokens": 284145303.0, + "step": 10981 + }, + { + "epoch": 1.2060180101032287, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5430381298065186, + "learning_rate": 1e-06, + "loss": 0.8407, + "mean_token_accuracy": 0.7519052028656006, + "num_tokens": 284167398.0, + "step": 10982 + }, + { + "epoch": 1.2061278278058423, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.678037643432617, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7106239795684814, + "num_tokens": 284191959.0, + "step": 10983 + }, + { + "epoch": 1.206237645508456, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5419058799743652, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7054184079170227, + "num_tokens": 284217135.0, + "step": 10984 + }, + { + "epoch": 1.2063474632110696, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4345614910125732, + "learning_rate": 1e-06, + "loss": 1.0846, + "mean_token_accuracy": 0.6822487115859985, + "num_tokens": 284252636.0, + "step": 10985 + }, + { + "epoch": 1.2064572809136833, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.497749090194702, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.7170758247375488, + "num_tokens": 284277462.0, + "step": 10986 + }, + { + "epoch": 1.2065670986162969, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7770395278930664, + "learning_rate": 1e-06, + "loss": 0.8941, + "mean_token_accuracy": 0.7359024882316589, + "num_tokens": 284298181.0, + "step": 10987 + }, + { + "epoch": 1.2066769163189106, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3729805946350098, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7002130746841431, + "num_tokens": 284326972.0, + "step": 10988 + }, + { + "epoch": 1.2067867340215244, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.742020845413208, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7136131525039673, + "num_tokens": 284352298.0, + "step": 10989 + }, + { + "epoch": 1.206896551724138, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8774468898773193, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7105277180671692, + "num_tokens": 284371893.0, + "step": 10990 + }, + { + "epoch": 1.2070063694267517, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.642477035522461, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6934375762939453, + "num_tokens": 284398281.0, + "step": 10991 + }, + { + "epoch": 1.2071161871293652, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.669492483139038, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7041982412338257, + "num_tokens": 284421089.0, + "step": 10992 + }, + { + "epoch": 1.207226004831979, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8436741828918457, + "learning_rate": 1e-06, + "loss": 0.9453, + "mean_token_accuracy": 0.7184706330299377, + "num_tokens": 284441192.0, + "step": 10993 + }, + { + "epoch": 1.2073358225345925, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.272244691848755, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6860009431838989, + "num_tokens": 284471296.0, + "step": 10994 + }, + { + "epoch": 1.2074456402372062, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.562551259994507, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.6964426636695862, + "num_tokens": 284497238.0, + "step": 10995 + }, + { + "epoch": 1.20755545793982, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.536548614501953, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6885259747505188, + "num_tokens": 284525660.0, + "step": 10996 + }, + { + "epoch": 1.2076652756424335, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.469804048538208, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7129286527633667, + "num_tokens": 284554370.0, + "step": 10997 + }, + { + "epoch": 1.2077750933450473, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6364972591400146, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.713776171207428, + "num_tokens": 284577055.0, + "step": 10998 + }, + { + "epoch": 1.2078849110476608, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6593005657196045, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7203128933906555, + "num_tokens": 284601534.0, + "step": 10999 + }, + { + "epoch": 1.2079947287502746, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7839043140411377, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7063775658607483, + "num_tokens": 284624533.0, + "step": 11000 + }, + { + "epoch": 1.2081045464528881, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4534053802490234, + "learning_rate": 1e-06, + "loss": 1.1385, + "mean_token_accuracy": 0.6686744093894958, + "num_tokens": 284655548.0, + "step": 11001 + }, + { + "epoch": 1.2082143641555019, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.682648181915283, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.703312873840332, + "num_tokens": 284678884.0, + "step": 11002 + }, + { + "epoch": 1.2083241818581154, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8044209480285645, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7028244137763977, + "num_tokens": 284703846.0, + "step": 11003 + }, + { + "epoch": 1.2084339995607292, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.454045057296753, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7114166617393494, + "num_tokens": 284729345.0, + "step": 11004 + }, + { + "epoch": 1.208543817263343, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.398576498031616, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7099430561065674, + "num_tokens": 284759880.0, + "step": 11005 + }, + { + "epoch": 1.2086536349659565, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9102344512939453, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.7228477597236633, + "num_tokens": 284778553.0, + "step": 11006 + }, + { + "epoch": 1.2087634526685702, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5885422229766846, + "learning_rate": 1e-06, + "loss": 0.9423, + "mean_token_accuracy": 0.7239169478416443, + "num_tokens": 284803349.0, + "step": 11007 + }, + { + "epoch": 1.2088732703711838, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4077677726745605, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.680131196975708, + "num_tokens": 284836162.0, + "step": 11008 + }, + { + "epoch": 1.2089830880737975, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8862552642822266, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7157140374183655, + "num_tokens": 284855788.0, + "step": 11009 + }, + { + "epoch": 1.2090929057764113, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8131511211395264, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7066839933395386, + "num_tokens": 284880287.0, + "step": 11010 + }, + { + "epoch": 1.2092027234790248, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.613464117050171, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7146422266960144, + "num_tokens": 284904543.0, + "step": 11011 + }, + { + "epoch": 1.2093125411816386, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.544076681137085, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.6983264684677124, + "num_tokens": 284933876.0, + "step": 11012 + }, + { + "epoch": 1.209422358884252, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6187496185302734, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6967039704322815, + "num_tokens": 284961257.0, + "step": 11013 + }, + { + "epoch": 1.2095321765868658, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6707956790924072, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.6847418546676636, + "num_tokens": 284988656.0, + "step": 11014 + }, + { + "epoch": 1.2096419942894794, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7785756587982178, + "learning_rate": 1e-06, + "loss": 0.921, + "mean_token_accuracy": 0.7209759950637817, + "num_tokens": 285009556.0, + "step": 11015 + }, + { + "epoch": 1.2097518119920931, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5422728061676025, + "learning_rate": 1e-06, + "loss": 0.9131, + "mean_token_accuracy": 0.7275223731994629, + "num_tokens": 285034371.0, + "step": 11016 + }, + { + "epoch": 1.2098616296947067, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6969010829925537, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7126102447509766, + "num_tokens": 285056245.0, + "step": 11017 + }, + { + "epoch": 1.2099714473973204, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6243746280670166, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7072516679763794, + "num_tokens": 285081384.0, + "step": 11018 + }, + { + "epoch": 1.2100812650999342, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.866245985031128, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7157689332962036, + "num_tokens": 285101546.0, + "step": 11019 + }, + { + "epoch": 1.2101910828025477, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5009055137634277, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7227563858032227, + "num_tokens": 285128909.0, + "step": 11020 + }, + { + "epoch": 1.2103009005051615, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.8763527870178223, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.6989866495132446, + "num_tokens": 285149217.0, + "step": 11021 + }, + { + "epoch": 1.210410718207775, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.635923385620117, + "learning_rate": 1e-06, + "loss": 0.8995, + "mean_token_accuracy": 0.7302459478378296, + "num_tokens": 285173057.0, + "step": 11022 + }, + { + "epoch": 1.2105205359103888, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.678415060043335, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7141735553741455, + "num_tokens": 285198193.0, + "step": 11023 + }, + { + "epoch": 1.2106303536130025, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.415471315383911, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6982215642929077, + "num_tokens": 285226208.0, + "step": 11024 + }, + { + "epoch": 1.210740171315616, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6721272468566895, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7024614810943604, + "num_tokens": 285249441.0, + "step": 11025 + }, + { + "epoch": 1.2108499890182298, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.535139799118042, + "learning_rate": 1e-06, + "loss": 0.9097, + "mean_token_accuracy": 0.7304672002792358, + "num_tokens": 285274785.0, + "step": 11026 + }, + { + "epoch": 1.2109598067208434, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3459694385528564, + "learning_rate": 1e-06, + "loss": 1.1224, + "mean_token_accuracy": 0.6809360980987549, + "num_tokens": 285311003.0, + "step": 11027 + }, + { + "epoch": 1.2110696244234571, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6493964195251465, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7118287086486816, + "num_tokens": 285334538.0, + "step": 11028 + }, + { + "epoch": 1.2111794421260706, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5648415088653564, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7057479023933411, + "num_tokens": 285361596.0, + "step": 11029 + }, + { + "epoch": 1.2112892598286844, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4772331714630127, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6933815479278564, + "num_tokens": 285389838.0, + "step": 11030 + }, + { + "epoch": 1.211399077531298, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6399974822998047, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6883482933044434, + "num_tokens": 285417739.0, + "step": 11031 + }, + { + "epoch": 1.2115088952339117, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1187403202056885, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7193187475204468, + "num_tokens": 285435299.0, + "step": 11032 + }, + { + "epoch": 1.2116187129365255, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.832015037536621, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.7023659348487854, + "num_tokens": 285460233.0, + "step": 11033 + }, + { + "epoch": 1.211728530639139, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7908008098602295, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.700151264667511, + "num_tokens": 285485153.0, + "step": 11034 + }, + { + "epoch": 1.2118383483417527, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.524904489517212, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7197113633155823, + "num_tokens": 285510008.0, + "step": 11035 + }, + { + "epoch": 1.2119481660443663, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9629359245300293, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7020488977432251, + "num_tokens": 285535003.0, + "step": 11036 + }, + { + "epoch": 1.21205798374698, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4831693172454834, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.7000784873962402, + "num_tokens": 285565024.0, + "step": 11037 + }, + { + "epoch": 1.2121678014495938, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3194973468780518, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7124948501586914, + "num_tokens": 285598590.0, + "step": 11038 + }, + { + "epoch": 1.2122776191522073, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.706112861633301, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7099079489707947, + "num_tokens": 285619570.0, + "step": 11039 + }, + { + "epoch": 1.212387436854821, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.60821795463562, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6869062185287476, + "num_tokens": 285644388.0, + "step": 11040 + }, + { + "epoch": 1.2124972545574346, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.407027244567871, + "learning_rate": 1e-06, + "loss": 1.1168, + "mean_token_accuracy": 0.6835252046585083, + "num_tokens": 285676890.0, + "step": 11041 + }, + { + "epoch": 1.2126070722600484, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9354772567749023, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7229311466217041, + "num_tokens": 285694848.0, + "step": 11042 + }, + { + "epoch": 1.212716889962662, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0047011375427246, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.7169300317764282, + "num_tokens": 285713171.0, + "step": 11043 + }, + { + "epoch": 1.2128267076652757, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6836183071136475, + "learning_rate": 1e-06, + "loss": 0.8884, + "mean_token_accuracy": 0.7372696399688721, + "num_tokens": 285736481.0, + "step": 11044 + }, + { + "epoch": 1.2129365253678892, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7191219329833984, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.6817290782928467, + "num_tokens": 285766170.0, + "step": 11045 + }, + { + "epoch": 1.213046343070503, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1919806003570557, + "learning_rate": 1e-06, + "loss": 0.9231, + "mean_token_accuracy": 0.7265444993972778, + "num_tokens": 285782962.0, + "step": 11046 + }, + { + "epoch": 1.2131561607731167, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7295238971710205, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7264021635055542, + "num_tokens": 285805223.0, + "step": 11047 + }, + { + "epoch": 1.2132659784757303, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.882521867752075, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7080954313278198, + "num_tokens": 285830478.0, + "step": 11048 + }, + { + "epoch": 1.213375796178344, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0173392295837402, + "learning_rate": 1e-06, + "loss": 0.865, + "mean_token_accuracy": 0.7438505291938782, + "num_tokens": 285848620.0, + "step": 11049 + }, + { + "epoch": 1.2134856138809575, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.500387668609619, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6975874900817871, + "num_tokens": 285876089.0, + "step": 11050 + }, + { + "epoch": 1.2135954315835713, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.711042881011963, + "learning_rate": 1e-06, + "loss": 0.9038, + "mean_token_accuracy": 0.729254961013794, + "num_tokens": 285900376.0, + "step": 11051 + }, + { + "epoch": 1.2137052492861848, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.642738103866577, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7214357852935791, + "num_tokens": 285923703.0, + "step": 11052 + }, + { + "epoch": 1.2138150669887986, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.528745651245117, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6879836320877075, + "num_tokens": 285950717.0, + "step": 11053 + }, + { + "epoch": 1.2139248846914121, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6112968921661377, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.690343976020813, + "num_tokens": 285973567.0, + "step": 11054 + }, + { + "epoch": 1.2140347023940259, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.641512870788574, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.6998661756515503, + "num_tokens": 285997017.0, + "step": 11055 + }, + { + "epoch": 1.2141445200966396, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.3670601844787598, + "learning_rate": 1e-06, + "loss": 1.1492, + "mean_token_accuracy": 0.670636773109436, + "num_tokens": 286028655.0, + "step": 11056 + }, + { + "epoch": 1.2142543377992532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.52243971824646, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.7010658383369446, + "num_tokens": 286054292.0, + "step": 11057 + }, + { + "epoch": 1.214364155501867, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5408082008361816, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7160645723342896, + "num_tokens": 286079944.0, + "step": 11058 + }, + { + "epoch": 1.2144739732044805, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.533406972885132, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6859306693077087, + "num_tokens": 286108619.0, + "step": 11059 + }, + { + "epoch": 1.2145837909070942, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.371890068054199, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7195233106613159, + "num_tokens": 286137641.0, + "step": 11060 + }, + { + "epoch": 1.214693608609708, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.444185495376587, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7100732326507568, + "num_tokens": 286163781.0, + "step": 11061 + }, + { + "epoch": 1.2148034263123215, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4826581478118896, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6982055306434631, + "num_tokens": 286193706.0, + "step": 11062 + }, + { + "epoch": 1.2149132440149353, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.358856201171875, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.696492075920105, + "num_tokens": 286224522.0, + "step": 11063 + }, + { + "epoch": 1.2150230617175488, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3991780281066895, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7237253785133362, + "num_tokens": 286249405.0, + "step": 11064 + }, + { + "epoch": 1.2151328794201626, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.410411834716797, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7102326154708862, + "num_tokens": 286278531.0, + "step": 11065 + }, + { + "epoch": 1.215242697122776, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3723018169403076, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.7014647126197815, + "num_tokens": 286308331.0, + "step": 11066 + }, + { + "epoch": 1.2153525148253899, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4769790172576904, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6940547227859497, + "num_tokens": 286339208.0, + "step": 11067 + }, + { + "epoch": 1.2154623325280034, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6240878105163574, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7204726934432983, + "num_tokens": 286361971.0, + "step": 11068 + }, + { + "epoch": 1.2155721502306172, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.324488878250122, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7057660222053528, + "num_tokens": 286392775.0, + "step": 11069 + }, + { + "epoch": 1.215681967933231, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9364938735961914, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7165994644165039, + "num_tokens": 286411263.0, + "step": 11070 + }, + { + "epoch": 1.2157917856358444, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.566925525665283, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6992487907409668, + "num_tokens": 286437434.0, + "step": 11071 + }, + { + "epoch": 1.2159016033384582, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5710036754608154, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7115089893341064, + "num_tokens": 286461643.0, + "step": 11072 + }, + { + "epoch": 1.2160114210410717, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5395147800445557, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.6961458921432495, + "num_tokens": 286493286.0, + "step": 11073 + }, + { + "epoch": 1.2161212387436855, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7592391967773438, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.7042012214660645, + "num_tokens": 286514586.0, + "step": 11074 + }, + { + "epoch": 1.2162310564462993, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8409671783447266, + "learning_rate": 1e-06, + "loss": 0.9129, + "mean_token_accuracy": 0.7283350825309753, + "num_tokens": 286533142.0, + "step": 11075 + }, + { + "epoch": 1.2163408741489128, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4373605251312256, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7022238969802856, + "num_tokens": 286560108.0, + "step": 11076 + }, + { + "epoch": 1.2164506918515265, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2966461181640625, + "learning_rate": 1e-06, + "loss": 1.1402, + "mean_token_accuracy": 0.6760894060134888, + "num_tokens": 286592101.0, + "step": 11077 + }, + { + "epoch": 1.21656050955414, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.8745322227478027, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.6996004581451416, + "num_tokens": 286614678.0, + "step": 11078 + }, + { + "epoch": 1.2166703272567538, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.755491256713867, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7022184133529663, + "num_tokens": 286636211.0, + "step": 11079 + }, + { + "epoch": 1.2167801449593674, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4891750812530518, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7082512378692627, + "num_tokens": 286662911.0, + "step": 11080 + }, + { + "epoch": 1.2168899626619811, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.38687801361084, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7089323997497559, + "num_tokens": 286690482.0, + "step": 11081 + }, + { + "epoch": 1.2169997803645947, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.623671293258667, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6878423094749451, + "num_tokens": 286716439.0, + "step": 11082 + }, + { + "epoch": 1.2171095980672084, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4657278060913086, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7165386080741882, + "num_tokens": 286745981.0, + "step": 11083 + }, + { + "epoch": 1.2172194157698222, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.460104465484619, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.703782320022583, + "num_tokens": 286774086.0, + "step": 11084 + }, + { + "epoch": 1.2173292334724357, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6579151153564453, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7092466354370117, + "num_tokens": 286797551.0, + "step": 11085 + }, + { + "epoch": 1.2174390511750495, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7095417976379395, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.7061399817466736, + "num_tokens": 286821481.0, + "step": 11086 + }, + { + "epoch": 1.217548868877663, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4872114658355713, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6922104358673096, + "num_tokens": 286851130.0, + "step": 11087 + }, + { + "epoch": 1.2176586865802768, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.021819829940796, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.7149783968925476, + "num_tokens": 286870940.0, + "step": 11088 + }, + { + "epoch": 1.2177685042828905, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7628700733184814, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7160521149635315, + "num_tokens": 286893884.0, + "step": 11089 + }, + { + "epoch": 1.217878321985504, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.753908634185791, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7250820994377136, + "num_tokens": 286916003.0, + "step": 11090 + }, + { + "epoch": 1.2179881396881178, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8855373859405518, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6954261064529419, + "num_tokens": 286938420.0, + "step": 11091 + }, + { + "epoch": 1.2180979573907313, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.083070755004883, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7241857051849365, + "num_tokens": 286958223.0, + "step": 11092 + }, + { + "epoch": 1.218207775093345, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6158864498138428, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7083574533462524, + "num_tokens": 286982199.0, + "step": 11093 + }, + { + "epoch": 1.2183175927959586, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.554094076156616, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7054494619369507, + "num_tokens": 287010430.0, + "step": 11094 + }, + { + "epoch": 1.2184274104985724, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.833695411682129, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.713933527469635, + "num_tokens": 287030970.0, + "step": 11095 + }, + { + "epoch": 1.218537228201186, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5920040607452393, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6934787034988403, + "num_tokens": 287057132.0, + "step": 11096 + }, + { + "epoch": 1.2186470459037997, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6676838397979736, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7204872369766235, + "num_tokens": 287082012.0, + "step": 11097 + }, + { + "epoch": 1.2187568636064134, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.644948720932007, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6904680728912354, + "num_tokens": 287109465.0, + "step": 11098 + }, + { + "epoch": 1.218866681309027, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5845954418182373, + "learning_rate": 1e-06, + "loss": 0.912, + "mean_token_accuracy": 0.7340499758720398, + "num_tokens": 287135634.0, + "step": 11099 + }, + { + "epoch": 1.2189764990116407, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.1621575355529785, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7232295870780945, + "num_tokens": 287151167.0, + "step": 11100 + }, + { + "epoch": 1.2190863167142543, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5659563541412354, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7293989658355713, + "num_tokens": 287176289.0, + "step": 11101 + }, + { + "epoch": 1.219196134416868, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.66756010055542, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7001342177391052, + "num_tokens": 287199749.0, + "step": 11102 + }, + { + "epoch": 1.2193059521194818, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5271854400634766, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6945002675056458, + "num_tokens": 287229177.0, + "step": 11103 + }, + { + "epoch": 1.2194157698220953, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5547831058502197, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6982284784317017, + "num_tokens": 287257362.0, + "step": 11104 + }, + { + "epoch": 1.219525587524709, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4138107299804688, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7112045884132385, + "num_tokens": 287286351.0, + "step": 11105 + }, + { + "epoch": 1.2196354052273226, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.655662775039673, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7079832553863525, + "num_tokens": 287312114.0, + "step": 11106 + }, + { + "epoch": 1.2197452229299364, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.453805685043335, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7122164964675903, + "num_tokens": 287340424.0, + "step": 11107 + }, + { + "epoch": 1.21985504063255, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.614699363708496, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.6939834356307983, + "num_tokens": 287365830.0, + "step": 11108 + }, + { + "epoch": 1.2199648583351637, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8372344970703125, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7167861461639404, + "num_tokens": 287392435.0, + "step": 11109 + }, + { + "epoch": 1.2200746760377772, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.028392791748047, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7154507637023926, + "num_tokens": 287412571.0, + "step": 11110 + }, + { + "epoch": 1.220184493740391, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.49135422706604, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.7015737295150757, + "num_tokens": 287441777.0, + "step": 11111 + }, + { + "epoch": 1.2202943114430047, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.896347761154175, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7008959054946899, + "num_tokens": 287462481.0, + "step": 11112 + }, + { + "epoch": 1.2204041291456182, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4420390129089355, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.7024680376052856, + "num_tokens": 287491128.0, + "step": 11113 + }, + { + "epoch": 1.220513946848232, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2609620094299316, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7103646993637085, + "num_tokens": 287524819.0, + "step": 11114 + }, + { + "epoch": 1.2206237645508455, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.1694955825805664, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7047866582870483, + "num_tokens": 287562937.0, + "step": 11115 + }, + { + "epoch": 1.2207335822534593, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8152568340301514, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7269656658172607, + "num_tokens": 287584404.0, + "step": 11116 + }, + { + "epoch": 1.2208433999560728, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.535156011581421, + "learning_rate": 1e-06, + "loss": 0.8926, + "mean_token_accuracy": 0.7335517406463623, + "num_tokens": 287609866.0, + "step": 11117 + }, + { + "epoch": 1.2209532176586866, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.300424575805664, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6949220895767212, + "num_tokens": 287639847.0, + "step": 11118 + }, + { + "epoch": 1.2210630353613001, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.329320192337036, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7053345441818237, + "num_tokens": 287670981.0, + "step": 11119 + }, + { + "epoch": 1.2211728530639139, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4515035152435303, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6971564292907715, + "num_tokens": 287701178.0, + "step": 11120 + }, + { + "epoch": 1.2212826707665276, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.495086669921875, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6920874118804932, + "num_tokens": 287728371.0, + "step": 11121 + }, + { + "epoch": 1.2213924884691412, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6747236251831055, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7094413638114929, + "num_tokens": 287752359.0, + "step": 11122 + }, + { + "epoch": 1.221502306171755, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7587456703186035, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7200090885162354, + "num_tokens": 287774074.0, + "step": 11123 + }, + { + "epoch": 1.2216121238743685, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5500073432922363, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6877133846282959, + "num_tokens": 287805557.0, + "step": 11124 + }, + { + "epoch": 1.2217219415769822, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6074867248535156, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6929436922073364, + "num_tokens": 287831463.0, + "step": 11125 + }, + { + "epoch": 1.221831759279596, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.690507173538208, + "learning_rate": 1e-06, + "loss": 0.9101, + "mean_token_accuracy": 0.7261103987693787, + "num_tokens": 287852361.0, + "step": 11126 + }, + { + "epoch": 1.2219415769822095, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9719436168670654, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7115988731384277, + "num_tokens": 287871387.0, + "step": 11127 + }, + { + "epoch": 1.2220513946848233, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6491332054138184, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7058959007263184, + "num_tokens": 287897123.0, + "step": 11128 + }, + { + "epoch": 1.2221612123874368, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7217977046966553, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.706707775592804, + "num_tokens": 287920282.0, + "step": 11129 + }, + { + "epoch": 1.2222710300900506, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.697843551635742, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.7043718099594116, + "num_tokens": 287945621.0, + "step": 11130 + }, + { + "epoch": 1.222380847792664, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4168286323547363, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7221179008483887, + "num_tokens": 287972594.0, + "step": 11131 + }, + { + "epoch": 1.2224906654952779, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.205264091491699, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6922624111175537, + "num_tokens": 288005889.0, + "step": 11132 + }, + { + "epoch": 1.2226004831978914, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.708538055419922, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.716475248336792, + "num_tokens": 288026620.0, + "step": 11133 + }, + { + "epoch": 1.2227103009005051, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4931583404541016, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7037011384963989, + "num_tokens": 288055937.0, + "step": 11134 + }, + { + "epoch": 1.222820118603119, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.679189443588257, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7101383805274963, + "num_tokens": 288082276.0, + "step": 11135 + }, + { + "epoch": 1.2229299363057324, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7175376415252686, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.7027791142463684, + "num_tokens": 288105292.0, + "step": 11136 + }, + { + "epoch": 1.2230397540083462, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.641977548599243, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7156985998153687, + "num_tokens": 288131219.0, + "step": 11137 + }, + { + "epoch": 1.2231495717109597, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3044652938842773, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7001993656158447, + "num_tokens": 288164176.0, + "step": 11138 + }, + { + "epoch": 1.2232593894135735, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8942317962646484, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7044225335121155, + "num_tokens": 288188719.0, + "step": 11139 + }, + { + "epoch": 1.2233692071161872, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4752871990203857, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6964132189750671, + "num_tokens": 288216109.0, + "step": 11140 + }, + { + "epoch": 1.2234790248188008, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6559343338012695, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7032136917114258, + "num_tokens": 288240464.0, + "step": 11141 + }, + { + "epoch": 1.2235888425214145, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.586815595626831, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7147355675697327, + "num_tokens": 288266942.0, + "step": 11142 + }, + { + "epoch": 1.223698660224028, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5762956142425537, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.692513108253479, + "num_tokens": 288295463.0, + "step": 11143 + }, + { + "epoch": 1.2238084779266418, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5431644916534424, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7172644138336182, + "num_tokens": 288319423.0, + "step": 11144 + }, + { + "epoch": 1.2239182956292554, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0018863677978516, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7208658456802368, + "num_tokens": 288338142.0, + "step": 11145 + }, + { + "epoch": 1.2240281133318691, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.411968469619751, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7204470038414001, + "num_tokens": 288365441.0, + "step": 11146 + }, + { + "epoch": 1.2241379310344827, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.797877073287964, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6877594590187073, + "num_tokens": 288391090.0, + "step": 11147 + }, + { + "epoch": 1.2242477487370964, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3065402507781982, + "learning_rate": 1e-06, + "loss": 1.1105, + "mean_token_accuracy": 0.6837174892425537, + "num_tokens": 288423660.0, + "step": 11148 + }, + { + "epoch": 1.2243575664397102, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4939870834350586, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7104636430740356, + "num_tokens": 288452342.0, + "step": 11149 + }, + { + "epoch": 1.2244673841423237, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.411785364151001, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6970016956329346, + "num_tokens": 288480932.0, + "step": 11150 + }, + { + "epoch": 1.2245772018449375, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.211904764175415, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.734499990940094, + "num_tokens": 288498077.0, + "step": 11151 + }, + { + "epoch": 1.224687019547551, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5604357719421387, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6951804161071777, + "num_tokens": 288524421.0, + "step": 11152 + }, + { + "epoch": 1.2247968372501647, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6061108112335205, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7108111381530762, + "num_tokens": 288551768.0, + "step": 11153 + }, + { + "epoch": 1.2249066549527785, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7914984226226807, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7162100672721863, + "num_tokens": 288575893.0, + "step": 11154 + }, + { + "epoch": 1.225016472655392, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.684215784072876, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7143696546554565, + "num_tokens": 288598787.0, + "step": 11155 + }, + { + "epoch": 1.2251262903580058, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.8021392822265625, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.704318106174469, + "num_tokens": 288620878.0, + "step": 11156 + }, + { + "epoch": 1.2252361080606193, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6298530101776123, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7088952660560608, + "num_tokens": 288648869.0, + "step": 11157 + }, + { + "epoch": 1.225345925763233, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9064276218414307, + "learning_rate": 1e-06, + "loss": 0.8964, + "mean_token_accuracy": 0.7351093888282776, + "num_tokens": 288668783.0, + "step": 11158 + }, + { + "epoch": 1.2254557434658466, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6492645740509033, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7070142030715942, + "num_tokens": 288696825.0, + "step": 11159 + }, + { + "epoch": 1.2255655611684604, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.301284074783325, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7138696908950806, + "num_tokens": 288729299.0, + "step": 11160 + }, + { + "epoch": 1.225675378871074, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.627295732498169, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7092930674552917, + "num_tokens": 288754789.0, + "step": 11161 + }, + { + "epoch": 1.2257851965736877, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6972248554229736, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6877428293228149, + "num_tokens": 288780435.0, + "step": 11162 + }, + { + "epoch": 1.2258950142763014, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8151705265045166, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.701935887336731, + "num_tokens": 288803995.0, + "step": 11163 + }, + { + "epoch": 1.226004831978915, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4424519538879395, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6935960650444031, + "num_tokens": 288832003.0, + "step": 11164 + }, + { + "epoch": 1.2261146496815287, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.43993878364563, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.6937707662582397, + "num_tokens": 288862135.0, + "step": 11165 + }, + { + "epoch": 1.2262244673841423, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4595441818237305, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6919398307800293, + "num_tokens": 288890024.0, + "step": 11166 + }, + { + "epoch": 1.226334285086756, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5706732273101807, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7063050270080566, + "num_tokens": 288913336.0, + "step": 11167 + }, + { + "epoch": 1.2264441027893696, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4598500728607178, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7087029814720154, + "num_tokens": 288944772.0, + "step": 11168 + }, + { + "epoch": 1.2265539204919833, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.685885429382324, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7054047584533691, + "num_tokens": 288972703.0, + "step": 11169 + }, + { + "epoch": 1.226663738194597, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.583458662033081, + "learning_rate": 1e-06, + "loss": 0.9302, + "mean_token_accuracy": 0.7273731231689453, + "num_tokens": 288997926.0, + "step": 11170 + }, + { + "epoch": 1.2267735558972106, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.677382230758667, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6918250322341919, + "num_tokens": 289020373.0, + "step": 11171 + }, + { + "epoch": 1.2268833735998244, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.554180860519409, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6826575994491577, + "num_tokens": 289048954.0, + "step": 11172 + }, + { + "epoch": 1.226993191302438, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.743898630142212, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7074247002601624, + "num_tokens": 289073845.0, + "step": 11173 + }, + { + "epoch": 1.2271030090050516, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9031572341918945, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7139520049095154, + "num_tokens": 289094320.0, + "step": 11174 + }, + { + "epoch": 1.2272128267076652, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6042964458465576, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7311426401138306, + "num_tokens": 289123281.0, + "step": 11175 + }, + { + "epoch": 1.227322644410279, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.852400779724121, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6934987306594849, + "num_tokens": 289146395.0, + "step": 11176 + }, + { + "epoch": 1.2274324621128927, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.776855945587158, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7115885019302368, + "num_tokens": 289166789.0, + "step": 11177 + }, + { + "epoch": 1.2275422798155062, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.609449625015259, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.6908508539199829, + "num_tokens": 289190811.0, + "step": 11178 + }, + { + "epoch": 1.22765209751812, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7407453060150146, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.709461510181427, + "num_tokens": 289216576.0, + "step": 11179 + }, + { + "epoch": 1.2277619152207335, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5713047981262207, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7079818844795227, + "num_tokens": 289241926.0, + "step": 11180 + }, + { + "epoch": 1.2278717329233473, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.831319570541382, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.7287771105766296, + "num_tokens": 289261528.0, + "step": 11181 + }, + { + "epoch": 1.2279815506259608, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7773280143737793, + "learning_rate": 1e-06, + "loss": 1.1396, + "mean_token_accuracy": 0.6812934279441833, + "num_tokens": 289284892.0, + "step": 11182 + }, + { + "epoch": 1.2280913683285746, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4556527137756348, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6970876455307007, + "num_tokens": 289311750.0, + "step": 11183 + }, + { + "epoch": 1.228201186031188, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6305904388427734, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7041118741035461, + "num_tokens": 289336750.0, + "step": 11184 + }, + { + "epoch": 1.2283110037338019, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.1716561317443848, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.6985244750976562, + "num_tokens": 289372893.0, + "step": 11185 + }, + { + "epoch": 1.2284208214364156, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4619195461273193, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6919535398483276, + "num_tokens": 289400788.0, + "step": 11186 + }, + { + "epoch": 1.2285306391390292, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5120675563812256, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.7056354284286499, + "num_tokens": 289428204.0, + "step": 11187 + }, + { + "epoch": 1.228640456841643, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7945029735565186, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7277767062187195, + "num_tokens": 289448773.0, + "step": 11188 + }, + { + "epoch": 1.2287502745442564, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7144665718078613, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7133442163467407, + "num_tokens": 289472201.0, + "step": 11189 + }, + { + "epoch": 1.2288600922468702, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.735224485397339, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.7277340292930603, + "num_tokens": 289493334.0, + "step": 11190 + }, + { + "epoch": 1.228969909949484, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4864864349365234, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.686428964138031, + "num_tokens": 289522313.0, + "step": 11191 + }, + { + "epoch": 1.2290797276520975, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 16.70289421081543, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6768842935562134, + "num_tokens": 289555995.0, + "step": 11192 + }, + { + "epoch": 1.2291895453547113, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.57528018951416, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6917270421981812, + "num_tokens": 289588907.0, + "step": 11193 + }, + { + "epoch": 1.2292993630573248, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.835489511489868, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.7034393548965454, + "num_tokens": 289613880.0, + "step": 11194 + }, + { + "epoch": 1.2294091807599385, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7599291801452637, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7103111743927002, + "num_tokens": 289635850.0, + "step": 11195 + }, + { + "epoch": 1.229518998462552, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.576730728149414, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7068823575973511, + "num_tokens": 289661786.0, + "step": 11196 + }, + { + "epoch": 1.2296288161651658, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.448448896408081, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7185865640640259, + "num_tokens": 289690312.0, + "step": 11197 + }, + { + "epoch": 1.2297386338677794, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5803961753845215, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6939113140106201, + "num_tokens": 289716247.0, + "step": 11198 + }, + { + "epoch": 1.2298484515703931, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.78991961479187, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6929224729537964, + "num_tokens": 289737054.0, + "step": 11199 + }, + { + "epoch": 1.2299582692730069, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4301278591156006, + "learning_rate": 1e-06, + "loss": 1.1237, + "mean_token_accuracy": 0.6798630356788635, + "num_tokens": 289765541.0, + "step": 11200 + }, + { + "epoch": 1.2300680869756204, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.565854072570801, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7086045145988464, + "num_tokens": 289793721.0, + "step": 11201 + }, + { + "epoch": 1.2301779046782342, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7160191535949707, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7115500569343567, + "num_tokens": 289818886.0, + "step": 11202 + }, + { + "epoch": 1.2302877223808477, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.943460464477539, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7119070291519165, + "num_tokens": 289839342.0, + "step": 11203 + }, + { + "epoch": 1.2303975400834615, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0253262519836426, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.697260856628418, + "num_tokens": 289859828.0, + "step": 11204 + }, + { + "epoch": 1.2305073577860752, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9190285205841064, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.7298751473426819, + "num_tokens": 289878027.0, + "step": 11205 + }, + { + "epoch": 1.2306171754886888, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.686485767364502, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6927903890609741, + "num_tokens": 289902423.0, + "step": 11206 + }, + { + "epoch": 1.2307269931913025, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6978397369384766, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6965621709823608, + "num_tokens": 289926283.0, + "step": 11207 + }, + { + "epoch": 1.230836810893916, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.824965238571167, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7100691795349121, + "num_tokens": 289947044.0, + "step": 11208 + }, + { + "epoch": 1.2309466285965298, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6441497802734375, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6943084597587585, + "num_tokens": 289973967.0, + "step": 11209 + }, + { + "epoch": 1.2310564462991433, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.814809799194336, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7035677433013916, + "num_tokens": 289997008.0, + "step": 11210 + }, + { + "epoch": 1.231166264001757, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.539175033569336, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.6843634843826294, + "num_tokens": 290025815.0, + "step": 11211 + }, + { + "epoch": 1.2312760817043706, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 16.72422981262207, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.6974688768386841, + "num_tokens": 290049001.0, + "step": 11212 + }, + { + "epoch": 1.2313858994069844, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.967602491378784, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7037487030029297, + "num_tokens": 290072534.0, + "step": 11213 + }, + { + "epoch": 1.2314957171095982, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8177006244659424, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6891416907310486, + "num_tokens": 290098182.0, + "step": 11214 + }, + { + "epoch": 1.2316055348122117, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.769386053085327, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7155493497848511, + "num_tokens": 290121546.0, + "step": 11215 + }, + { + "epoch": 1.2317153525148254, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6377193927764893, + "learning_rate": 1e-06, + "loss": 0.9146, + "mean_token_accuracy": 0.7287108302116394, + "num_tokens": 290144929.0, + "step": 11216 + }, + { + "epoch": 1.231825170217439, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5549449920654297, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7085742950439453, + "num_tokens": 290169041.0, + "step": 11217 + }, + { + "epoch": 1.2319349879200527, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.442089080810547, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7184185981750488, + "num_tokens": 290196061.0, + "step": 11218 + }, + { + "epoch": 1.2320448056226665, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4292619228363037, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.6995339393615723, + "num_tokens": 290222130.0, + "step": 11219 + }, + { + "epoch": 1.23215462332528, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6872384548187256, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7166829109191895, + "num_tokens": 290245566.0, + "step": 11220 + }, + { + "epoch": 1.2322644410278938, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7317073345184326, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.6971249580383301, + "num_tokens": 290268633.0, + "step": 11221 + }, + { + "epoch": 1.2323742587305073, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3051130771636963, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7279257774353027, + "num_tokens": 290298052.0, + "step": 11222 + }, + { + "epoch": 1.232484076433121, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.70963978767395, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.7201235890388489, + "num_tokens": 290323306.0, + "step": 11223 + }, + { + "epoch": 1.2325938941357346, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4151182174682617, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7083771228790283, + "num_tokens": 290351222.0, + "step": 11224 + }, + { + "epoch": 1.2327037118383484, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6292829513549805, + "learning_rate": 1e-06, + "loss": 1.1187, + "mean_token_accuracy": 0.6783876419067383, + "num_tokens": 290380297.0, + "step": 11225 + }, + { + "epoch": 1.232813529540962, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7391388416290283, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.72110915184021, + "num_tokens": 290410462.0, + "step": 11226 + }, + { + "epoch": 1.2329233472435757, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.628873109817505, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.7023230195045471, + "num_tokens": 290438048.0, + "step": 11227 + }, + { + "epoch": 1.2330331649461894, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3151819705963135, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6940316557884216, + "num_tokens": 290469821.0, + "step": 11228 + }, + { + "epoch": 1.233142982648803, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.638050079345703, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7092017531394958, + "num_tokens": 290495476.0, + "step": 11229 + }, + { + "epoch": 1.2332528003514167, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6576967239379883, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6885007619857788, + "num_tokens": 290519504.0, + "step": 11230 + }, + { + "epoch": 1.2333626180540302, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.523376226425171, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.691752016544342, + "num_tokens": 290547917.0, + "step": 11231 + }, + { + "epoch": 1.233472435756644, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4660913944244385, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6957578659057617, + "num_tokens": 290575889.0, + "step": 11232 + }, + { + "epoch": 1.2335822534592575, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7264835834503174, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7207638621330261, + "num_tokens": 290597692.0, + "step": 11233 + }, + { + "epoch": 1.2336920711618713, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7396605014801025, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6887068152427673, + "num_tokens": 290624743.0, + "step": 11234 + }, + { + "epoch": 1.2338018888644848, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.633065700531006, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6983629465103149, + "num_tokens": 290651588.0, + "step": 11235 + }, + { + "epoch": 1.2339117065670986, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.557739496231079, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.6963683366775513, + "num_tokens": 290679796.0, + "step": 11236 + }, + { + "epoch": 1.2340215242697123, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.0897231101989746, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7115226984024048, + "num_tokens": 290698190.0, + "step": 11237 + }, + { + "epoch": 1.2341313419723259, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.489617347717285, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7148033976554871, + "num_tokens": 290722165.0, + "step": 11238 + }, + { + "epoch": 1.2342411596749396, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3262505531311035, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.683866024017334, + "num_tokens": 290755251.0, + "step": 11239 + }, + { + "epoch": 1.2343509773775532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.532059669494629, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.7054867744445801, + "num_tokens": 290782938.0, + "step": 11240 + }, + { + "epoch": 1.234460795080167, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3432037830352783, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7204722762107849, + "num_tokens": 290814105.0, + "step": 11241 + }, + { + "epoch": 1.2345706127827807, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.446873188018799, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.7234622836112976, + "num_tokens": 290841887.0, + "step": 11242 + }, + { + "epoch": 1.2346804304853942, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.387450695037842, + "learning_rate": 1e-06, + "loss": 1.1727, + "mean_token_accuracy": 0.66147381067276, + "num_tokens": 290872908.0, + "step": 11243 + }, + { + "epoch": 1.234790248188008, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.469489812850952, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.703420877456665, + "num_tokens": 290902229.0, + "step": 11244 + }, + { + "epoch": 1.2349000658906215, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9306833744049072, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6835094690322876, + "num_tokens": 290930274.0, + "step": 11245 + }, + { + "epoch": 1.2350098835932353, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7643420696258545, + "learning_rate": 1e-06, + "loss": 1.1333, + "mean_token_accuracy": 0.6732361316680908, + "num_tokens": 290957237.0, + "step": 11246 + }, + { + "epoch": 1.2351197012958488, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.599393606185913, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7140783071517944, + "num_tokens": 290982019.0, + "step": 11247 + }, + { + "epoch": 1.2352295189984626, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7181215286254883, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7210065722465515, + "num_tokens": 291004914.0, + "step": 11248 + }, + { + "epoch": 1.235339336701076, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.378821611404419, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.7010454535484314, + "num_tokens": 291035370.0, + "step": 11249 + }, + { + "epoch": 1.2354491544036899, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6552462577819824, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.6854040026664734, + "num_tokens": 291064938.0, + "step": 11250 + }, + { + "epoch": 1.2355589721063036, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5868606567382812, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6999397277832031, + "num_tokens": 291090672.0, + "step": 11251 + }, + { + "epoch": 1.2356687898089171, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.746089220046997, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7057584524154663, + "num_tokens": 291113233.0, + "step": 11252 + }, + { + "epoch": 1.235778607511531, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.610017776489258, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7086030840873718, + "num_tokens": 291142475.0, + "step": 11253 + }, + { + "epoch": 1.2358884252141444, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4577860832214355, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7041659951210022, + "num_tokens": 291169504.0, + "step": 11254 + }, + { + "epoch": 1.2359982429167582, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.374201774597168, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7002490758895874, + "num_tokens": 291202195.0, + "step": 11255 + }, + { + "epoch": 1.236108060619372, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.750685214996338, + "learning_rate": 1e-06, + "loss": 1.1356, + "mean_token_accuracy": 0.6787567138671875, + "num_tokens": 291228469.0, + "step": 11256 + }, + { + "epoch": 1.2362178783219855, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.342527389526367, + "learning_rate": 1e-06, + "loss": 1.1037, + "mean_token_accuracy": 0.6776968836784363, + "num_tokens": 291259022.0, + "step": 11257 + }, + { + "epoch": 1.2363276960245992, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5190279483795166, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7157109379768372, + "num_tokens": 291283767.0, + "step": 11258 + }, + { + "epoch": 1.2364375137272128, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7682416439056396, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7163199782371521, + "num_tokens": 291306139.0, + "step": 11259 + }, + { + "epoch": 1.2365473314298265, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3694815635681152, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7052343487739563, + "num_tokens": 291336887.0, + "step": 11260 + }, + { + "epoch": 1.23665714913244, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7986156940460205, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.6900753974914551, + "num_tokens": 291363600.0, + "step": 11261 + }, + { + "epoch": 1.2367669668350538, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.542070150375366, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7118421196937561, + "num_tokens": 291389067.0, + "step": 11262 + }, + { + "epoch": 1.2368767845376674, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9821054935455322, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6826521158218384, + "num_tokens": 291413585.0, + "step": 11263 + }, + { + "epoch": 1.2369866022402811, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.5976312160491943, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7240485548973083, + "num_tokens": 291434867.0, + "step": 11264 + }, + { + "epoch": 1.2370964199428949, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5626564025878906, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7103465795516968, + "num_tokens": 291461764.0, + "step": 11265 + }, + { + "epoch": 1.2372062376455084, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4528918266296387, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.6969387531280518, + "num_tokens": 291488950.0, + "step": 11266 + }, + { + "epoch": 1.2373160553481222, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.485701084136963, + "learning_rate": 1e-06, + "loss": 1.1406, + "mean_token_accuracy": 0.6700173020362854, + "num_tokens": 291516738.0, + "step": 11267 + }, + { + "epoch": 1.2374258730507357, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5048158168792725, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7010529041290283, + "num_tokens": 291548428.0, + "step": 11268 + }, + { + "epoch": 1.2375356907533495, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6878175735473633, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7133851647377014, + "num_tokens": 291572552.0, + "step": 11269 + }, + { + "epoch": 1.2376455084559632, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.48624324798584, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7289589047431946, + "num_tokens": 291597888.0, + "step": 11270 + }, + { + "epoch": 1.2377553261585768, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6145386695861816, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7058573961257935, + "num_tokens": 291624458.0, + "step": 11271 + }, + { + "epoch": 1.2378651438611905, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.592740058898926, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.6989163160324097, + "num_tokens": 291649258.0, + "step": 11272 + }, + { + "epoch": 1.237974961563804, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.54439377784729, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7267237901687622, + "num_tokens": 291673698.0, + "step": 11273 + }, + { + "epoch": 1.2380847792664178, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7493085861206055, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7119799852371216, + "num_tokens": 291697675.0, + "step": 11274 + }, + { + "epoch": 1.2381945969690313, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.603986978530884, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6974921226501465, + "num_tokens": 291724573.0, + "step": 11275 + }, + { + "epoch": 1.238304414671645, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.030738592147827, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7187533378601074, + "num_tokens": 291743913.0, + "step": 11276 + }, + { + "epoch": 1.2384142323742586, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.530435800552368, + "learning_rate": 1e-06, + "loss": 0.9552, + "mean_token_accuracy": 0.719420313835144, + "num_tokens": 291768326.0, + "step": 11277 + }, + { + "epoch": 1.2385240500768724, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.604806900024414, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6955505013465881, + "num_tokens": 291796062.0, + "step": 11278 + }, + { + "epoch": 1.2386338677794861, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6325459480285645, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7059473991394043, + "num_tokens": 291825094.0, + "step": 11279 + }, + { + "epoch": 1.2387436854820997, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.640740156173706, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7075604796409607, + "num_tokens": 291851180.0, + "step": 11280 + }, + { + "epoch": 1.2388535031847134, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.556828498840332, + "learning_rate": 1e-06, + "loss": 0.9565, + "mean_token_accuracy": 0.717639148235321, + "num_tokens": 291875864.0, + "step": 11281 + }, + { + "epoch": 1.238963320887327, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4713430404663086, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7090554237365723, + "num_tokens": 291901757.0, + "step": 11282 + }, + { + "epoch": 1.2390731385899407, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.71627140045166, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7236586809158325, + "num_tokens": 291924418.0, + "step": 11283 + }, + { + "epoch": 1.2391829562925545, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5395126342773438, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6996970176696777, + "num_tokens": 291953676.0, + "step": 11284 + }, + { + "epoch": 1.239292773995168, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.566098928451538, + "learning_rate": 1e-06, + "loss": 1.1305, + "mean_token_accuracy": 0.6757187843322754, + "num_tokens": 291982067.0, + "step": 11285 + }, + { + "epoch": 1.2394025916977818, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.815613269805908, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7138004302978516, + "num_tokens": 292002505.0, + "step": 11286 + }, + { + "epoch": 1.2395124094003953, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.440962076187134, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.7023934125900269, + "num_tokens": 292030540.0, + "step": 11287 + }, + { + "epoch": 1.239622227103009, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.545811176300049, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6906702518463135, + "num_tokens": 292055186.0, + "step": 11288 + }, + { + "epoch": 1.2397320448056226, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8690848350524902, + "learning_rate": 1e-06, + "loss": 0.9146, + "mean_token_accuracy": 0.7285290360450745, + "num_tokens": 292075507.0, + "step": 11289 + }, + { + "epoch": 1.2398418625082364, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7472918033599854, + "learning_rate": 1e-06, + "loss": 1.1068, + "mean_token_accuracy": 0.6871718168258667, + "num_tokens": 292103532.0, + "step": 11290 + }, + { + "epoch": 1.23995168021085, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5141961574554443, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7193179130554199, + "num_tokens": 292131001.0, + "step": 11291 + }, + { + "epoch": 1.2400614979134637, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3681347370147705, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7054638862609863, + "num_tokens": 292159966.0, + "step": 11292 + }, + { + "epoch": 1.2401713156160774, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.569593906402588, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7163927555084229, + "num_tokens": 292185311.0, + "step": 11293 + }, + { + "epoch": 1.240281133318691, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.477499008178711, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7136753797531128, + "num_tokens": 292210456.0, + "step": 11294 + }, + { + "epoch": 1.2403909510213047, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.444410562515259, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6958552002906799, + "num_tokens": 292238954.0, + "step": 11295 + }, + { + "epoch": 1.2405007687239182, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.341587543487549, + "learning_rate": 1e-06, + "loss": 0.8866, + "mean_token_accuracy": 0.7402283549308777, + "num_tokens": 292268336.0, + "step": 11296 + }, + { + "epoch": 1.240610586426532, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7142858505249023, + "learning_rate": 1e-06, + "loss": 0.9361, + "mean_token_accuracy": 0.7224790453910828, + "num_tokens": 292290510.0, + "step": 11297 + }, + { + "epoch": 1.2407204041291455, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 17.07057762145996, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7188924551010132, + "num_tokens": 292319469.0, + "step": 11298 + }, + { + "epoch": 1.2408302218317593, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8767189979553223, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.6998164057731628, + "num_tokens": 292345632.0, + "step": 11299 + }, + { + "epoch": 1.2409400395343728, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6756982803344727, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6984623670578003, + "num_tokens": 292372304.0, + "step": 11300 + }, + { + "epoch": 1.2410498572369866, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.677140712738037, + "learning_rate": 1e-06, + "loss": 0.9053, + "mean_token_accuracy": 0.737785279750824, + "num_tokens": 292395329.0, + "step": 11301 + }, + { + "epoch": 1.2411596749396003, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.588000535964966, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7100937366485596, + "num_tokens": 292420030.0, + "step": 11302 + }, + { + "epoch": 1.2412694926422139, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4811134338378906, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6996952891349792, + "num_tokens": 292444679.0, + "step": 11303 + }, + { + "epoch": 1.2413793103448276, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.414280414581299, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7079989910125732, + "num_tokens": 292471191.0, + "step": 11304 + }, + { + "epoch": 1.2414891280474412, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6928796768188477, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7098470330238342, + "num_tokens": 292493761.0, + "step": 11305 + }, + { + "epoch": 1.241598945750055, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7545485496520996, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7052832841873169, + "num_tokens": 292516844.0, + "step": 11306 + }, + { + "epoch": 1.2417087634526687, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8738670349121094, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6872267723083496, + "num_tokens": 292540030.0, + "step": 11307 + }, + { + "epoch": 1.2418185811552822, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.371575355529785, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7077279686927795, + "num_tokens": 292568982.0, + "step": 11308 + }, + { + "epoch": 1.241928398857896, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.896411657333374, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7072851657867432, + "num_tokens": 292589148.0, + "step": 11309 + }, + { + "epoch": 1.2420382165605095, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.959336519241333, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6945544481277466, + "num_tokens": 292617921.0, + "step": 11310 + }, + { + "epoch": 1.2421480342631233, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.503042221069336, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7266827821731567, + "num_tokens": 292644301.0, + "step": 11311 + }, + { + "epoch": 1.2422578519657368, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.394409656524658, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7106344699859619, + "num_tokens": 292674415.0, + "step": 11312 + }, + { + "epoch": 1.2423676696683505, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6123414039611816, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7075008153915405, + "num_tokens": 292698202.0, + "step": 11313 + }, + { + "epoch": 1.242477487370964, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.624544382095337, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7035065293312073, + "num_tokens": 292727895.0, + "step": 11314 + }, + { + "epoch": 1.2425873050735778, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8525779247283936, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6849763989448547, + "num_tokens": 292751726.0, + "step": 11315 + }, + { + "epoch": 1.2426971227761916, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 16.81394386291504, + "learning_rate": 1e-06, + "loss": 1.1593, + "mean_token_accuracy": 0.6681005954742432, + "num_tokens": 292782025.0, + "step": 11316 + }, + { + "epoch": 1.2428069404788051, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6695351600646973, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7041104435920715, + "num_tokens": 292809852.0, + "step": 11317 + }, + { + "epoch": 1.242916758181419, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3844311237335205, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6965570449829102, + "num_tokens": 292840054.0, + "step": 11318 + }, + { + "epoch": 1.2430265758840324, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.9694175720214844, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.7336598038673401, + "num_tokens": 292858529.0, + "step": 11319 + }, + { + "epoch": 1.2431363935866462, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.548449993133545, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6945600509643555, + "num_tokens": 292888213.0, + "step": 11320 + }, + { + "epoch": 1.24324621128926, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.409419059753418, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7056588530540466, + "num_tokens": 292921633.0, + "step": 11321 + }, + { + "epoch": 1.2433560289918735, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4734103679656982, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7029220461845398, + "num_tokens": 292950474.0, + "step": 11322 + }, + { + "epoch": 1.2434658466944872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5603809356689453, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6901125907897949, + "num_tokens": 292975140.0, + "step": 11323 + }, + { + "epoch": 1.2435756643971008, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.4336037635803223, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7117317318916321, + "num_tokens": 293005180.0, + "step": 11324 + }, + { + "epoch": 1.2436854820997145, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.2784714698791504, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7153415679931641, + "num_tokens": 293028606.0, + "step": 11325 + }, + { + "epoch": 1.243795299802328, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7721519470214844, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.720104455947876, + "num_tokens": 293049332.0, + "step": 11326 + }, + { + "epoch": 1.2439051175049418, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.358165740966797, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7061131000518799, + "num_tokens": 293079125.0, + "step": 11327 + }, + { + "epoch": 1.2440149352075554, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.687166929244995, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6872471570968628, + "num_tokens": 293106284.0, + "step": 11328 + }, + { + "epoch": 1.244124752910169, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.2704918384552, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7123546004295349, + "num_tokens": 293137777.0, + "step": 11329 + }, + { + "epoch": 1.2442345706127829, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.945431709289551, + "learning_rate": 1e-06, + "loss": 0.9168, + "mean_token_accuracy": 0.7293556928634644, + "num_tokens": 293157064.0, + "step": 11330 + }, + { + "epoch": 1.2443443883153964, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.7264225482940674, + "learning_rate": 1e-06, + "loss": 1.0917, + "mean_token_accuracy": 0.6842290759086609, + "num_tokens": 293181353.0, + "step": 11331 + }, + { + "epoch": 1.2444542060180102, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.377511501312256, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6925373673439026, + "num_tokens": 293211082.0, + "step": 11332 + }, + { + "epoch": 1.2445640237206237, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.443587064743042, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7170298099517822, + "num_tokens": 293238078.0, + "step": 11333 + }, + { + "epoch": 1.2446738414232374, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.107828378677368, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.714417576789856, + "num_tokens": 293255608.0, + "step": 11334 + }, + { + "epoch": 1.2447836591258512, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.58384370803833, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.6970661878585815, + "num_tokens": 293285103.0, + "step": 11335 + }, + { + "epoch": 1.2448934768284647, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.658935546875, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.6997826099395752, + "num_tokens": 293309605.0, + "step": 11336 + }, + { + "epoch": 1.2450032945310785, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.574665069580078, + "learning_rate": 1e-06, + "loss": 1.1185, + "mean_token_accuracy": 0.677275538444519, + "num_tokens": 293336245.0, + "step": 11337 + }, + { + "epoch": 1.245113112233692, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.619235038757324, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6927202939987183, + "num_tokens": 293361734.0, + "step": 11338 + }, + { + "epoch": 1.2452229299363058, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.3632559776306152, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7129179239273071, + "num_tokens": 293389235.0, + "step": 11339 + }, + { + "epoch": 1.2453327476389193, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6616079807281494, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7071829438209534, + "num_tokens": 293410452.0, + "step": 11340 + }, + { + "epoch": 1.245442565341533, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.6242427825927734, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7230103015899658, + "num_tokens": 293435361.0, + "step": 11341 + }, + { + "epoch": 1.2455523830441466, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.666086196899414, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7239567041397095, + "num_tokens": 293457915.0, + "step": 11342 + }, + { + "epoch": 1.2456622007467604, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.8112175464630127, + "learning_rate": 1e-06, + "loss": 0.9108, + "mean_token_accuracy": 0.7341703772544861, + "num_tokens": 293478810.0, + "step": 11343 + }, + { + "epoch": 1.2457720184493741, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.576967239379883, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.721045732498169, + "num_tokens": 293502615.0, + "step": 11344 + }, + { + "epoch": 1.2458818361519877, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 3.386395215988159, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.709565281867981, + "num_tokens": 293527291.0, + "step": 11345 + }, + { + "epoch": 1.2459916538546014, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.620018482208252, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7226168513298035, + "num_tokens": 293554129.0, + "step": 11346 + }, + { + "epoch": 1.246101471557215, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6091771125793457, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7136541604995728, + "num_tokens": 293579967.0, + "step": 11347 + }, + { + "epoch": 1.2462112892598287, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.57981014251709, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7206107378005981, + "num_tokens": 293606901.0, + "step": 11348 + }, + { + "epoch": 1.2463211069624422, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.5596959590911865, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6915692090988159, + "num_tokens": 293634345.0, + "step": 11349 + }, + { + "epoch": 1.246430924665056, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.535226345062256, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.6979087591171265, + "num_tokens": 293660439.0, + "step": 11350 + }, + { + "epoch": 1.2465407423676698, + "ewc_loss": 9.238719940185547e-06, + "grad_norm": 2.823488712310791, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.7240610718727112, + "num_tokens": 293685799.0, + "step": 11351 + }, + { + "epoch": 1.2466505600702833, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.592444896697998, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7183191776275635, + "num_tokens": 293710985.0, + "step": 11352 + }, + { + "epoch": 1.246760377772897, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4726948738098145, + "learning_rate": 1e-06, + "loss": 1.1297, + "mean_token_accuracy": 0.6761770844459534, + "num_tokens": 293742124.0, + "step": 11353 + }, + { + "epoch": 1.2468701954755106, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7790656089782715, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7021561861038208, + "num_tokens": 293764646.0, + "step": 11354 + }, + { + "epoch": 1.2469800131781243, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.592871904373169, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7138839960098267, + "num_tokens": 293788042.0, + "step": 11355 + }, + { + "epoch": 1.2470898308807379, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.841360569000244, + "learning_rate": 1e-06, + "loss": 0.9532, + "mean_token_accuracy": 0.7200875282287598, + "num_tokens": 293811110.0, + "step": 11356 + }, + { + "epoch": 1.2471996485833516, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6026265621185303, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6879862546920776, + "num_tokens": 293839656.0, + "step": 11357 + }, + { + "epoch": 1.2473094662859654, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7257959842681885, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.683319091796875, + "num_tokens": 293864921.0, + "step": 11358 + }, + { + "epoch": 1.247419283988579, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7179856300354004, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.7001123428344727, + "num_tokens": 293891997.0, + "step": 11359 + }, + { + "epoch": 1.2475291016911927, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 4.1895575523376465, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6856215000152588, + "num_tokens": 293928415.0, + "step": 11360 + }, + { + "epoch": 1.2476389193938062, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2684285640716553, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.7013629674911499, + "num_tokens": 293964945.0, + "step": 11361 + }, + { + "epoch": 1.24774873709642, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.453052282333374, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6956455707550049, + "num_tokens": 293991970.0, + "step": 11362 + }, + { + "epoch": 1.2478585547990335, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.51519775390625, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7057540416717529, + "num_tokens": 294018661.0, + "step": 11363 + }, + { + "epoch": 1.2479683725016473, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5185203552246094, + "learning_rate": 1e-06, + "loss": 0.903, + "mean_token_accuracy": 0.7373136281967163, + "num_tokens": 294046716.0, + "step": 11364 + }, + { + "epoch": 1.2480781902042608, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.543226957321167, + "learning_rate": 1e-06, + "loss": 0.8951, + "mean_token_accuracy": 0.7341527938842773, + "num_tokens": 294071987.0, + "step": 11365 + }, + { + "epoch": 1.2481880079068746, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5938446521759033, + "learning_rate": 1e-06, + "loss": 0.9158, + "mean_token_accuracy": 0.729544460773468, + "num_tokens": 294095503.0, + "step": 11366 + }, + { + "epoch": 1.2482978256094883, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.566183567047119, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.6857271790504456, + "num_tokens": 294122424.0, + "step": 11367 + }, + { + "epoch": 1.2484076433121019, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.550835132598877, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.7045966386795044, + "num_tokens": 294150376.0, + "step": 11368 + }, + { + "epoch": 1.2485174610147156, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4494800567626953, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7102622985839844, + "num_tokens": 294177393.0, + "step": 11369 + }, + { + "epoch": 1.2486272787173291, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.722142457962036, + "learning_rate": 1e-06, + "loss": 1.0992, + "mean_token_accuracy": 0.6962548494338989, + "num_tokens": 294203511.0, + "step": 11370 + }, + { + "epoch": 1.248737096419943, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4927265644073486, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7042416334152222, + "num_tokens": 294232230.0, + "step": 11371 + }, + { + "epoch": 1.2488469141225567, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.394456148147583, + "learning_rate": 1e-06, + "loss": 0.9202, + "mean_token_accuracy": 0.73097825050354, + "num_tokens": 294261107.0, + "step": 11372 + }, + { + "epoch": 1.2489567318251702, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.872918128967285, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7175543308258057, + "num_tokens": 294283492.0, + "step": 11373 + }, + { + "epoch": 1.249066549527784, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6787147521972656, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7133712768554688, + "num_tokens": 294308887.0, + "step": 11374 + }, + { + "epoch": 1.2491763672303975, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.883225202560425, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6972084045410156, + "num_tokens": 294336422.0, + "step": 11375 + }, + { + "epoch": 1.2492861849330112, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6706132888793945, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.6873993277549744, + "num_tokens": 294363697.0, + "step": 11376 + }, + { + "epoch": 1.2493960026356248, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.641342878341675, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.72056645154953, + "num_tokens": 294387983.0, + "step": 11377 + }, + { + "epoch": 1.2495058203382385, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.495615005493164, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6967417001724243, + "num_tokens": 294414661.0, + "step": 11378 + }, + { + "epoch": 1.249615638040852, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.602977991104126, + "learning_rate": 1e-06, + "loss": 0.9606, + "mean_token_accuracy": 0.7167565822601318, + "num_tokens": 294440692.0, + "step": 11379 + }, + { + "epoch": 1.2497254557434658, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6474852561950684, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7056528925895691, + "num_tokens": 294468741.0, + "step": 11380 + }, + { + "epoch": 1.2498352734460796, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.782672882080078, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7150647640228271, + "num_tokens": 294489488.0, + "step": 11381 + }, + { + "epoch": 1.2499450911486931, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.874677896499634, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7022144794464111, + "num_tokens": 294513430.0, + "step": 11382 + }, + { + "epoch": 1.2500549088513069, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.428318500518799, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7073098421096802, + "num_tokens": 294542188.0, + "step": 11383 + }, + { + "epoch": 1.2501647265539204, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 16.721214294433594, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7023432850837708, + "num_tokens": 294566002.0, + "step": 11384 + }, + { + "epoch": 1.2502745442565342, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 7.981588840484619, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7012884616851807, + "num_tokens": 294591308.0, + "step": 11385 + }, + { + "epoch": 1.250384361959148, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.995457887649536, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6906105279922485, + "num_tokens": 294615500.0, + "step": 11386 + }, + { + "epoch": 1.2504941796617615, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8552346229553223, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7078880667686462, + "num_tokens": 294640524.0, + "step": 11387 + }, + { + "epoch": 1.2506039973643752, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4968433380126953, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.700779914855957, + "num_tokens": 294670595.0, + "step": 11388 + }, + { + "epoch": 1.2507138150669888, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.533905267715454, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6921778917312622, + "num_tokens": 294696095.0, + "step": 11389 + }, + { + "epoch": 1.2508236327696025, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4263179302215576, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7144562602043152, + "num_tokens": 294722676.0, + "step": 11390 + }, + { + "epoch": 1.250933450472216, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.331569194793701, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.704637348651886, + "num_tokens": 294752360.0, + "step": 11391 + }, + { + "epoch": 1.2510432681748298, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4784929752349854, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6974982619285583, + "num_tokens": 294777603.0, + "step": 11392 + }, + { + "epoch": 1.2511530858774433, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4737048149108887, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7195897102355957, + "num_tokens": 294802529.0, + "step": 11393 + }, + { + "epoch": 1.251262903580057, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 8.046642303466797, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6866484880447388, + "num_tokens": 294827373.0, + "step": 11394 + }, + { + "epoch": 1.2513727212826709, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6696395874023438, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7076889276504517, + "num_tokens": 294855380.0, + "step": 11395 + }, + { + "epoch": 1.2514825389852844, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7458667755126953, + "learning_rate": 1e-06, + "loss": 0.889, + "mean_token_accuracy": 0.738311767578125, + "num_tokens": 294877037.0, + "step": 11396 + }, + { + "epoch": 1.2515923566878981, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.636013984680176, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.7011963725090027, + "num_tokens": 294901232.0, + "step": 11397 + }, + { + "epoch": 1.2517021743905117, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.688917398452759, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7194780111312866, + "num_tokens": 294921628.0, + "step": 11398 + }, + { + "epoch": 1.2518119920931254, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.638532876968384, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7151353359222412, + "num_tokens": 294947376.0, + "step": 11399 + }, + { + "epoch": 1.2519218097957392, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.619804859161377, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.718941867351532, + "num_tokens": 294976345.0, + "step": 11400 + }, + { + "epoch": 1.2520316274983527, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.747443914413452, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7100235223770142, + "num_tokens": 295000527.0, + "step": 11401 + }, + { + "epoch": 1.2521414452009663, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.859459161758423, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6846873164176941, + "num_tokens": 295021857.0, + "step": 11402 + }, + { + "epoch": 1.25225126290358, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.733612298965454, + "learning_rate": 1e-06, + "loss": 0.911, + "mean_token_accuracy": 0.7353712916374207, + "num_tokens": 295045127.0, + "step": 11403 + }, + { + "epoch": 1.2523610806061938, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7276551723480225, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7251126766204834, + "num_tokens": 295070148.0, + "step": 11404 + }, + { + "epoch": 1.2524708983088073, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8376758098602295, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7051788568496704, + "num_tokens": 295089819.0, + "step": 11405 + }, + { + "epoch": 1.252580716011421, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0078659057617188, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7071821689605713, + "num_tokens": 295107843.0, + "step": 11406 + }, + { + "epoch": 1.2526905337140346, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.8948886394500732, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7179257869720459, + "num_tokens": 295133757.0, + "step": 11407 + }, + { + "epoch": 1.2528003514166484, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.607663631439209, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.695388674736023, + "num_tokens": 295161640.0, + "step": 11408 + }, + { + "epoch": 1.2529101691192621, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.408566474914551, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7203702330589294, + "num_tokens": 295189968.0, + "step": 11409 + }, + { + "epoch": 1.2530199868218757, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6022109985351562, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7235502600669861, + "num_tokens": 295214500.0, + "step": 11410 + }, + { + "epoch": 1.2531298045244894, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7751965522766113, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.6953913569450378, + "num_tokens": 295237895.0, + "step": 11411 + }, + { + "epoch": 1.253239622227103, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7058684825897217, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7171832323074341, + "num_tokens": 295260666.0, + "step": 11412 + }, + { + "epoch": 1.2533494399297167, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.234973907470703, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6927629709243774, + "num_tokens": 295293423.0, + "step": 11413 + }, + { + "epoch": 1.2534592576323305, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6610424518585205, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7117764949798584, + "num_tokens": 295318031.0, + "step": 11414 + }, + { + "epoch": 1.253569075334944, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5780913829803467, + "learning_rate": 1e-06, + "loss": 1.095, + "mean_token_accuracy": 0.6790008544921875, + "num_tokens": 295345764.0, + "step": 11415 + }, + { + "epoch": 1.2536788930375575, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.670840263366699, + "learning_rate": 1e-06, + "loss": 1.1244, + "mean_token_accuracy": 0.6764311790466309, + "num_tokens": 295369222.0, + "step": 11416 + }, + { + "epoch": 1.2537887107401713, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7918789386749268, + "learning_rate": 1e-06, + "loss": 0.89, + "mean_token_accuracy": 0.7313267588615417, + "num_tokens": 295390231.0, + "step": 11417 + }, + { + "epoch": 1.253898528442785, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.0637640953063965, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7115439176559448, + "num_tokens": 295411044.0, + "step": 11418 + }, + { + "epoch": 1.2540083461453986, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.0710721015930176, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.720722496509552, + "num_tokens": 295432250.0, + "step": 11419 + }, + { + "epoch": 1.2541181638480123, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.698340654373169, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7075868844985962, + "num_tokens": 295462432.0, + "step": 11420 + }, + { + "epoch": 1.2542279815506259, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3922622203826904, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6904072165489197, + "num_tokens": 295490547.0, + "step": 11421 + }, + { + "epoch": 1.2543377992532396, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7927615642547607, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.7059197425842285, + "num_tokens": 295513531.0, + "step": 11422 + }, + { + "epoch": 1.2544476169558534, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7575457096099854, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7249101400375366, + "num_tokens": 295535425.0, + "step": 11423 + }, + { + "epoch": 1.254557434658467, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6040914058685303, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7138441801071167, + "num_tokens": 295561283.0, + "step": 11424 + }, + { + "epoch": 1.2546672523610807, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3779938220977783, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7108330130577087, + "num_tokens": 295589515.0, + "step": 11425 + }, + { + "epoch": 1.2547770700636942, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6903505325317383, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6974678635597229, + "num_tokens": 295615953.0, + "step": 11426 + }, + { + "epoch": 1.254886887766308, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.561983585357666, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.7049210667610168, + "num_tokens": 295644770.0, + "step": 11427 + }, + { + "epoch": 1.2549967054689217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7810540199279785, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7186601161956787, + "num_tokens": 295665904.0, + "step": 11428 + }, + { + "epoch": 1.2551065231715353, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3158254623413086, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.698101818561554, + "num_tokens": 295698262.0, + "step": 11429 + }, + { + "epoch": 1.2552163408741488, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.67117977142334, + "learning_rate": 1e-06, + "loss": 1.1336, + "mean_token_accuracy": 0.6761388778686523, + "num_tokens": 295723447.0, + "step": 11430 + }, + { + "epoch": 1.2553261585767626, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8609867095947266, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.719428539276123, + "num_tokens": 295744698.0, + "step": 11431 + }, + { + "epoch": 1.2554359762793763, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.980501651763916, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7191885709762573, + "num_tokens": 295764393.0, + "step": 11432 + }, + { + "epoch": 1.2555457939819898, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.550447940826416, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7005428075790405, + "num_tokens": 295790134.0, + "step": 11433 + }, + { + "epoch": 1.2556556116846036, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8722686767578125, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7081232070922852, + "num_tokens": 295811072.0, + "step": 11434 + }, + { + "epoch": 1.2557654293872171, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.514085292816162, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7332987785339355, + "num_tokens": 295837017.0, + "step": 11435 + }, + { + "epoch": 1.255875247089831, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.893972158432007, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7060084342956543, + "num_tokens": 295859040.0, + "step": 11436 + }, + { + "epoch": 1.2559850647924446, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.999755620956421, + "learning_rate": 1e-06, + "loss": 0.8928, + "mean_token_accuracy": 0.7334465980529785, + "num_tokens": 295875491.0, + "step": 11437 + }, + { + "epoch": 1.2560948824950582, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.722865104675293, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.7229905128479004, + "num_tokens": 295897898.0, + "step": 11438 + }, + { + "epoch": 1.256204700197672, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5517630577087402, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6873449087142944, + "num_tokens": 295928789.0, + "step": 11439 + }, + { + "epoch": 1.2563145179002855, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.958559989929199, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7099037766456604, + "num_tokens": 295953425.0, + "step": 11440 + }, + { + "epoch": 1.2564243356028992, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.378811836242676, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.686434805393219, + "num_tokens": 295982609.0, + "step": 11441 + }, + { + "epoch": 1.2565341533055128, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.639650821685791, + "learning_rate": 1e-06, + "loss": 1.1115, + "mean_token_accuracy": 0.6791259050369263, + "num_tokens": 296008120.0, + "step": 11442 + }, + { + "epoch": 1.2566439710081265, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.0710813999176025, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.720525860786438, + "num_tokens": 296025613.0, + "step": 11443 + }, + { + "epoch": 1.25675378871074, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4370977878570557, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7018308639526367, + "num_tokens": 296053874.0, + "step": 11444 + }, + { + "epoch": 1.2568636064133538, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.09944748878479, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.7192884087562561, + "num_tokens": 296074215.0, + "step": 11445 + }, + { + "epoch": 1.2569734241159676, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6211941242218018, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6961122751235962, + "num_tokens": 296102474.0, + "step": 11446 + }, + { + "epoch": 1.257083241818581, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.442824602127075, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.6849503517150879, + "num_tokens": 296131694.0, + "step": 11447 + }, + { + "epoch": 1.2571930595211949, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7315680980682373, + "learning_rate": 1e-06, + "loss": 0.9082, + "mean_token_accuracy": 0.7264062166213989, + "num_tokens": 296153126.0, + "step": 11448 + }, + { + "epoch": 1.2573028772238084, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7650578022003174, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6882445812225342, + "num_tokens": 296179671.0, + "step": 11449 + }, + { + "epoch": 1.2574126949264222, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.678509473800659, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6921922564506531, + "num_tokens": 296204581.0, + "step": 11450 + }, + { + "epoch": 1.257522512629036, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.0069918632507324, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7067664861679077, + "num_tokens": 296224412.0, + "step": 11451 + }, + { + "epoch": 1.2576323303316495, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8599002361297607, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.7036317586898804, + "num_tokens": 296246127.0, + "step": 11452 + }, + { + "epoch": 1.257742148034263, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.605970859527588, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6975687742233276, + "num_tokens": 296269949.0, + "step": 11453 + }, + { + "epoch": 1.2578519657368767, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7139086723327637, + "learning_rate": 1e-06, + "loss": 1.1381, + "mean_token_accuracy": 0.6818791031837463, + "num_tokens": 296291955.0, + "step": 11454 + }, + { + "epoch": 1.2579617834394905, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.567445755004883, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7101322412490845, + "num_tokens": 296319512.0, + "step": 11455 + }, + { + "epoch": 1.258071601142104, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.838204860687256, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7005255222320557, + "num_tokens": 296344515.0, + "step": 11456 + }, + { + "epoch": 1.2581814188447178, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6122047901153564, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.698652982711792, + "num_tokens": 296371510.0, + "step": 11457 + }, + { + "epoch": 1.2582912365473313, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6325035095214844, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7124775052070618, + "num_tokens": 296395037.0, + "step": 11458 + }, + { + "epoch": 1.258401054249945, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.581698417663574, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7075663208961487, + "num_tokens": 296420556.0, + "step": 11459 + }, + { + "epoch": 1.2585108719525588, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4637625217437744, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6887460350990295, + "num_tokens": 296449227.0, + "step": 11460 + }, + { + "epoch": 1.2586206896551724, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.534437656402588, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.7088915705680847, + "num_tokens": 296474535.0, + "step": 11461 + }, + { + "epoch": 1.2587305073577861, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.892033815383911, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6907625794410706, + "num_tokens": 296498255.0, + "step": 11462 + }, + { + "epoch": 1.2588403250603997, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5443387031555176, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7164371609687805, + "num_tokens": 296527472.0, + "step": 11463 + }, + { + "epoch": 1.2589501427630134, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.673245906829834, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6986215114593506, + "num_tokens": 296551741.0, + "step": 11464 + }, + { + "epoch": 1.2590599604656272, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3906397819519043, + "learning_rate": 1e-06, + "loss": 0.8906, + "mean_token_accuracy": 0.7313373684883118, + "num_tokens": 296577777.0, + "step": 11465 + }, + { + "epoch": 1.2591697781682407, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.300908088684082, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7053602337837219, + "num_tokens": 296611448.0, + "step": 11466 + }, + { + "epoch": 1.2592795958708543, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.68784761428833, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.70377117395401, + "num_tokens": 296638612.0, + "step": 11467 + }, + { + "epoch": 1.259389413573468, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.522116184234619, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7087720036506653, + "num_tokens": 296663290.0, + "step": 11468 + }, + { + "epoch": 1.2594992312760818, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.473036050796509, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7201192378997803, + "num_tokens": 296688612.0, + "step": 11469 + }, + { + "epoch": 1.2596090489786953, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.578935146331787, + "learning_rate": 1e-06, + "loss": 1.1014, + "mean_token_accuracy": 0.6798486709594727, + "num_tokens": 296718311.0, + "step": 11470 + }, + { + "epoch": 1.259718866681309, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6376426219940186, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7011880874633789, + "num_tokens": 296741384.0, + "step": 11471 + }, + { + "epoch": 1.2598286843839226, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6381726264953613, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7227790355682373, + "num_tokens": 296764291.0, + "step": 11472 + }, + { + "epoch": 1.2599385020865363, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7129199504852295, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.7013363838195801, + "num_tokens": 296790654.0, + "step": 11473 + }, + { + "epoch": 1.26004831978915, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.349895477294922, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.7179768085479736, + "num_tokens": 296821205.0, + "step": 11474 + }, + { + "epoch": 1.2601581374917636, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3073313236236572, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7193325757980347, + "num_tokens": 296849946.0, + "step": 11475 + }, + { + "epoch": 1.2602679551943774, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.612729072570801, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7030949592590332, + "num_tokens": 296879116.0, + "step": 11476 + }, + { + "epoch": 1.260377772896991, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5949409008026123, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6857733130455017, + "num_tokens": 296905714.0, + "step": 11477 + }, + { + "epoch": 1.2604875905996047, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6837873458862305, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.7031322121620178, + "num_tokens": 296932215.0, + "step": 11478 + }, + { + "epoch": 1.2605974083022184, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6859631538391113, + "learning_rate": 1e-06, + "loss": 0.8917, + "mean_token_accuracy": 0.7511112689971924, + "num_tokens": 296959360.0, + "step": 11479 + }, + { + "epoch": 1.260707226004832, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.902172327041626, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6915464997291565, + "num_tokens": 296981937.0, + "step": 11480 + }, + { + "epoch": 1.2608170437074455, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.358454704284668, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.7003360390663147, + "num_tokens": 297013783.0, + "step": 11481 + }, + { + "epoch": 1.2609268614100593, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3859243392944336, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.687565803527832, + "num_tokens": 297045639.0, + "step": 11482 + }, + { + "epoch": 1.261036679112673, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.470496416091919, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6922580003738403, + "num_tokens": 297074810.0, + "step": 11483 + }, + { + "epoch": 1.2611464968152866, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.064162492752075, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7271376252174377, + "num_tokens": 297100410.0, + "step": 11484 + }, + { + "epoch": 1.2612563145179003, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3774397373199463, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7130767107009888, + "num_tokens": 297132999.0, + "step": 11485 + }, + { + "epoch": 1.2613661322205139, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6655731201171875, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7089824080467224, + "num_tokens": 297157221.0, + "step": 11486 + }, + { + "epoch": 1.2614759499231276, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6308491230010986, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.6921564340591431, + "num_tokens": 297183062.0, + "step": 11487 + }, + { + "epoch": 1.2615857676257414, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9720852375030518, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7214922904968262, + "num_tokens": 297206154.0, + "step": 11488 + }, + { + "epoch": 1.261695585328355, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8647518157958984, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7216607332229614, + "num_tokens": 297228651.0, + "step": 11489 + }, + { + "epoch": 1.2618054030309687, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8861944675445557, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7025394439697266, + "num_tokens": 297259028.0, + "step": 11490 + }, + { + "epoch": 1.2619152207335822, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.592154026031494, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7085240483283997, + "num_tokens": 297285857.0, + "step": 11491 + }, + { + "epoch": 1.262025038436196, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7065412998199463, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6837774515151978, + "num_tokens": 297312496.0, + "step": 11492 + }, + { + "epoch": 1.2621348561388097, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.656494379043579, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.68822181224823, + "num_tokens": 297338471.0, + "step": 11493 + }, + { + "epoch": 1.2622446738414232, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.657456398010254, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.7148400545120239, + "num_tokens": 297363122.0, + "step": 11494 + }, + { + "epoch": 1.2623544915440368, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3807156085968018, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.6992917656898499, + "num_tokens": 297393466.0, + "step": 11495 + }, + { + "epoch": 1.2624643092466505, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4154934883117676, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7084509134292603, + "num_tokens": 297422391.0, + "step": 11496 + }, + { + "epoch": 1.2625741269492643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5190887451171875, + "learning_rate": 1e-06, + "loss": 1.1087, + "mean_token_accuracy": 0.6816102266311646, + "num_tokens": 297451099.0, + "step": 11497 + }, + { + "epoch": 1.2626839446518778, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7092387676239014, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.695826530456543, + "num_tokens": 297476115.0, + "step": 11498 + }, + { + "epoch": 1.2627937623544916, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7577316761016846, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6892420649528503, + "num_tokens": 297498409.0, + "step": 11499 + }, + { + "epoch": 1.2629035800571051, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6320056915283203, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.719713568687439, + "num_tokens": 297521611.0, + "step": 11500 + }, + { + "epoch": 1.2630133977597189, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.709637403488159, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7210673689842224, + "num_tokens": 297548205.0, + "step": 11501 + }, + { + "epoch": 1.2631232154623326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.121678113937378, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6860724687576294, + "num_tokens": 297573850.0, + "step": 11502 + }, + { + "epoch": 1.2632330331649462, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1087682247161865, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7116618752479553, + "num_tokens": 297593424.0, + "step": 11503 + }, + { + "epoch": 1.26334285086756, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.701869249343872, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6885377764701843, + "num_tokens": 297618308.0, + "step": 11504 + }, + { + "epoch": 1.2634526685701735, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7982611656188965, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7064963579177856, + "num_tokens": 297642742.0, + "step": 11505 + }, + { + "epoch": 1.2635624862727872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5432379245758057, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7148902416229248, + "num_tokens": 297667547.0, + "step": 11506 + }, + { + "epoch": 1.2636723039754008, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5942866802215576, + "learning_rate": 1e-06, + "loss": 0.8822, + "mean_token_accuracy": 0.7366342544555664, + "num_tokens": 297691845.0, + "step": 11507 + }, + { + "epoch": 1.2637821216780145, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8074090480804443, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7136093378067017, + "num_tokens": 297716930.0, + "step": 11508 + }, + { + "epoch": 1.263891939380628, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4992358684539795, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6978344321250916, + "num_tokens": 297747549.0, + "step": 11509 + }, + { + "epoch": 1.2640017570832418, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9670166969299316, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.7210272550582886, + "num_tokens": 297766898.0, + "step": 11510 + }, + { + "epoch": 1.2641115747858556, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8275818824768066, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6936211585998535, + "num_tokens": 297791316.0, + "step": 11511 + }, + { + "epoch": 1.264221392488469, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9071199893951416, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7038034200668335, + "num_tokens": 297811952.0, + "step": 11512 + }, + { + "epoch": 1.2643312101910829, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8211874961853027, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7107410430908203, + "num_tokens": 297833174.0, + "step": 11513 + }, + { + "epoch": 1.2644410278936964, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.998307466506958, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.6991446018218994, + "num_tokens": 297853684.0, + "step": 11514 + }, + { + "epoch": 1.2645508455963101, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.923574924468994, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6887273788452148, + "num_tokens": 297880080.0, + "step": 11515 + }, + { + "epoch": 1.264660663298924, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.684094190597534, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.690065860748291, + "num_tokens": 297909560.0, + "step": 11516 + }, + { + "epoch": 1.2647704810015374, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.791165590286255, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7035003304481506, + "num_tokens": 297932492.0, + "step": 11517 + }, + { + "epoch": 1.264880298704151, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4619503021240234, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.6875211596488953, + "num_tokens": 297965319.0, + "step": 11518 + }, + { + "epoch": 1.2649901164067647, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.61562180519104, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6893798112869263, + "num_tokens": 297993903.0, + "step": 11519 + }, + { + "epoch": 1.2650999341093785, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5407772064208984, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.693834662437439, + "num_tokens": 298022800.0, + "step": 11520 + }, + { + "epoch": 1.265209751811992, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.739121437072754, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7197995781898499, + "num_tokens": 298052026.0, + "step": 11521 + }, + { + "epoch": 1.2653195695146058, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.68176007270813, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7159379720687866, + "num_tokens": 298076619.0, + "step": 11522 + }, + { + "epoch": 1.2654293872172193, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7190866470336914, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.7126712799072266, + "num_tokens": 298100965.0, + "step": 11523 + }, + { + "epoch": 1.265539204919833, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.202810525894165, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6805251836776733, + "num_tokens": 298124809.0, + "step": 11524 + }, + { + "epoch": 1.2656490226224468, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3858559131622314, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.6929042339324951, + "num_tokens": 298155550.0, + "step": 11525 + }, + { + "epoch": 1.2657588403250604, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6119556427001953, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6986982226371765, + "num_tokens": 298181998.0, + "step": 11526 + }, + { + "epoch": 1.2658686580276741, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.46929931640625, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7160833477973938, + "num_tokens": 298211803.0, + "step": 11527 + }, + { + "epoch": 1.2659784757302877, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9133739471435547, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.695044219493866, + "num_tokens": 298234110.0, + "step": 11528 + }, + { + "epoch": 1.2660882934329014, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.344985008239746, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.684209942817688, + "num_tokens": 298265890.0, + "step": 11529 + }, + { + "epoch": 1.2661981111355152, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6961169242858887, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7100749015808105, + "num_tokens": 298290478.0, + "step": 11530 + }, + { + "epoch": 1.2663079288381287, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5015668869018555, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.718748927116394, + "num_tokens": 298318961.0, + "step": 11531 + }, + { + "epoch": 1.2664177465407422, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.807648181915283, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.7232016921043396, + "num_tokens": 298341171.0, + "step": 11532 + }, + { + "epoch": 1.266527564243356, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5870652198791504, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7171404957771301, + "num_tokens": 298365083.0, + "step": 11533 + }, + { + "epoch": 1.2666373819459698, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.31878399848938, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6997570991516113, + "num_tokens": 298396270.0, + "step": 11534 + }, + { + "epoch": 1.2667471996485833, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.667325258255005, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.6987664699554443, + "num_tokens": 298420344.0, + "step": 11535 + }, + { + "epoch": 1.266857017351197, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4207851886749268, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7038307189941406, + "num_tokens": 298451386.0, + "step": 11536 + }, + { + "epoch": 1.2669668350538106, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.776937961578369, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6953248977661133, + "num_tokens": 298473767.0, + "step": 11537 + }, + { + "epoch": 1.2670766527564243, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7525529861450195, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7172228097915649, + "num_tokens": 298495330.0, + "step": 11538 + }, + { + "epoch": 1.267186470459038, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4809184074401855, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6933173537254333, + "num_tokens": 298524502.0, + "step": 11539 + }, + { + "epoch": 1.2672962881616516, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5185599327087402, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7020078897476196, + "num_tokens": 298552364.0, + "step": 11540 + }, + { + "epoch": 1.2674061058642654, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6589977741241455, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.701739490032196, + "num_tokens": 298578078.0, + "step": 11541 + }, + { + "epoch": 1.267515923566879, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5542705059051514, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7070178985595703, + "num_tokens": 298605107.0, + "step": 11542 + }, + { + "epoch": 1.2676257412694927, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8930535316467285, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7180467844009399, + "num_tokens": 298625913.0, + "step": 11543 + }, + { + "epoch": 1.2677355589721064, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4301910400390625, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6794354915618896, + "num_tokens": 298656895.0, + "step": 11544 + }, + { + "epoch": 1.26784537667472, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7855312824249268, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7028626799583435, + "num_tokens": 298680044.0, + "step": 11545 + }, + { + "epoch": 1.2679551943773335, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6598246097564697, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.695368766784668, + "num_tokens": 298707118.0, + "step": 11546 + }, + { + "epoch": 1.2680650120799473, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.556792974472046, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.692711591720581, + "num_tokens": 298731089.0, + "step": 11547 + }, + { + "epoch": 1.268174829782561, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.576678514480591, + "learning_rate": 1e-06, + "loss": 0.9314, + "mean_token_accuracy": 0.7255550622940063, + "num_tokens": 298757668.0, + "step": 11548 + }, + { + "epoch": 1.2682846474851746, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4067118167877197, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7175592184066772, + "num_tokens": 298785147.0, + "step": 11549 + }, + { + "epoch": 1.2683944651877883, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.501798152923584, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6954610347747803, + "num_tokens": 298812858.0, + "step": 11550 + }, + { + "epoch": 1.2685042828904018, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.810025930404663, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7168872356414795, + "num_tokens": 298835848.0, + "step": 11551 + }, + { + "epoch": 1.2686141005930156, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7353057861328125, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7006402611732483, + "num_tokens": 298856536.0, + "step": 11552 + }, + { + "epoch": 1.2687239182956294, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6992523670196533, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.7155186533927917, + "num_tokens": 298883334.0, + "step": 11553 + }, + { + "epoch": 1.268833735998243, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9514589309692383, + "learning_rate": 1e-06, + "loss": 0.9311, + "mean_token_accuracy": 0.7285809516906738, + "num_tokens": 298902690.0, + "step": 11554 + }, + { + "epoch": 1.2689435537008567, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.447694778442383, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6968448758125305, + "num_tokens": 298935723.0, + "step": 11555 + }, + { + "epoch": 1.2690533714034702, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.636061191558838, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7036161422729492, + "num_tokens": 298961129.0, + "step": 11556 + }, + { + "epoch": 1.269163189106084, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3726677894592285, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7171725034713745, + "num_tokens": 298989359.0, + "step": 11557 + }, + { + "epoch": 1.2692730068086975, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.493678569793701, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.716898500919342, + "num_tokens": 299018271.0, + "step": 11558 + }, + { + "epoch": 1.2693828245113112, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7676150798797607, + "learning_rate": 1e-06, + "loss": 1.1186, + "mean_token_accuracy": 0.6846033334732056, + "num_tokens": 299041732.0, + "step": 11559 + }, + { + "epoch": 1.2694926422139248, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5306105613708496, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7132934331893921, + "num_tokens": 299066729.0, + "step": 11560 + }, + { + "epoch": 1.2696024599165385, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.824023723602295, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6937694549560547, + "num_tokens": 299088342.0, + "step": 11561 + }, + { + "epoch": 1.2697122776191523, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.764368772506714, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.685973048210144, + "num_tokens": 299113587.0, + "step": 11562 + }, + { + "epoch": 1.2698220953217658, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6896767616271973, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6993131637573242, + "num_tokens": 299140214.0, + "step": 11563 + }, + { + "epoch": 1.2699319130243796, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2782437801361084, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6936419010162354, + "num_tokens": 299174699.0, + "step": 11564 + }, + { + "epoch": 1.2700417307269931, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5388259887695312, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7179914116859436, + "num_tokens": 299206309.0, + "step": 11565 + }, + { + "epoch": 1.2701515484296069, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8717525005340576, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7131248116493225, + "num_tokens": 299225998.0, + "step": 11566 + }, + { + "epoch": 1.2702613661322206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4586925506591797, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7021893262863159, + "num_tokens": 299258633.0, + "step": 11567 + }, + { + "epoch": 1.2703711838348342, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7595722675323486, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6957851648330688, + "num_tokens": 299280740.0, + "step": 11568 + }, + { + "epoch": 1.270481001537448, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.67226243019104, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7042333483695984, + "num_tokens": 299304369.0, + "step": 11569 + }, + { + "epoch": 1.2705908192400615, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.486863613128662, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6893770694732666, + "num_tokens": 299332152.0, + "step": 11570 + }, + { + "epoch": 1.2707006369426752, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5799667835235596, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6845575571060181, + "num_tokens": 299361670.0, + "step": 11571 + }, + { + "epoch": 1.2708104546452887, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.693624496459961, + "learning_rate": 1e-06, + "loss": 1.1218, + "mean_token_accuracy": 0.6835572719573975, + "num_tokens": 299389803.0, + "step": 11572 + }, + { + "epoch": 1.2709202723479025, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.386364221572876, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7031102180480957, + "num_tokens": 299420332.0, + "step": 11573 + }, + { + "epoch": 1.271030090050516, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7006163597106934, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6847424507141113, + "num_tokens": 299444266.0, + "step": 11574 + }, + { + "epoch": 1.2711399077531298, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3429315090179443, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7003334760665894, + "num_tokens": 299476084.0, + "step": 11575 + }, + { + "epoch": 1.2712497254557436, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5734024047851562, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.708059549331665, + "num_tokens": 299501733.0, + "step": 11576 + }, + { + "epoch": 1.271359543158357, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6706788539886475, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6980520486831665, + "num_tokens": 299524569.0, + "step": 11577 + }, + { + "epoch": 1.2714693608609708, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.705871105194092, + "learning_rate": 1e-06, + "loss": 0.893, + "mean_token_accuracy": 0.7325330972671509, + "num_tokens": 299546818.0, + "step": 11578 + }, + { + "epoch": 1.2715791785635844, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8419127464294434, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7038583755493164, + "num_tokens": 299568630.0, + "step": 11579 + }, + { + "epoch": 1.2716889962661981, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.618778705596924, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6931905746459961, + "num_tokens": 299595091.0, + "step": 11580 + }, + { + "epoch": 1.271798813968812, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.470306158065796, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7074359059333801, + "num_tokens": 299624124.0, + "step": 11581 + }, + { + "epoch": 1.2719086316714254, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.60488224029541, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6893254518508911, + "num_tokens": 299649454.0, + "step": 11582 + }, + { + "epoch": 1.272018449374039, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.757892608642578, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6952935457229614, + "num_tokens": 299675046.0, + "step": 11583 + }, + { + "epoch": 1.2721282670766527, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5530107021331787, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.685844898223877, + "num_tokens": 299704160.0, + "step": 11584 + }, + { + "epoch": 1.2722380847792665, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.726830005645752, + "learning_rate": 1e-06, + "loss": 1.1176, + "mean_token_accuracy": 0.6753562092781067, + "num_tokens": 299730361.0, + "step": 11585 + }, + { + "epoch": 1.27234790248188, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.908963918685913, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7060441374778748, + "num_tokens": 299757251.0, + "step": 11586 + }, + { + "epoch": 1.2724577201844938, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9373676776885986, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7209527492523193, + "num_tokens": 299777812.0, + "step": 11587 + }, + { + "epoch": 1.2725675378871073, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7664694786071777, + "learning_rate": 1e-06, + "loss": 0.9125, + "mean_token_accuracy": 0.7337837219238281, + "num_tokens": 299799763.0, + "step": 11588 + }, + { + "epoch": 1.272677355589721, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6434011459350586, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.7017030715942383, + "num_tokens": 299823944.0, + "step": 11589 + }, + { + "epoch": 1.2727871732923348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7625975608825684, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.7259234189987183, + "num_tokens": 299846384.0, + "step": 11590 + }, + { + "epoch": 1.2728969909949484, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.537903070449829, + "learning_rate": 1e-06, + "loss": 1.1502, + "mean_token_accuracy": 0.6668771505355835, + "num_tokens": 299877965.0, + "step": 11591 + }, + { + "epoch": 1.273006808697562, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.532122850418091, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7177196741104126, + "num_tokens": 299907103.0, + "step": 11592 + }, + { + "epoch": 1.2731166264001756, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.374828577041626, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.694555401802063, + "num_tokens": 299939433.0, + "step": 11593 + }, + { + "epoch": 1.2732264441027894, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.434140205383301, + "learning_rate": 1e-06, + "loss": 0.9192, + "mean_token_accuracy": 0.730150580406189, + "num_tokens": 299965674.0, + "step": 11594 + }, + { + "epoch": 1.2733362618054032, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.840184211730957, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.696870744228363, + "num_tokens": 299988125.0, + "step": 11595 + }, + { + "epoch": 1.2734460795080167, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.33746075630188, + "learning_rate": 1e-06, + "loss": 0.9046, + "mean_token_accuracy": 0.7279751300811768, + "num_tokens": 300007609.0, + "step": 11596 + }, + { + "epoch": 1.2735558972106302, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4110171794891357, + "learning_rate": 1e-06, + "loss": 1.1839, + "mean_token_accuracy": 0.6622295379638672, + "num_tokens": 300042810.0, + "step": 11597 + }, + { + "epoch": 1.273665714913244, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8061773777008057, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.7008213996887207, + "num_tokens": 300065000.0, + "step": 11598 + }, + { + "epoch": 1.2737755326158577, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5482277870178223, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7122702598571777, + "num_tokens": 300093292.0, + "step": 11599 + }, + { + "epoch": 1.2738853503184713, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5840306282043457, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7186293601989746, + "num_tokens": 300116894.0, + "step": 11600 + }, + { + "epoch": 1.273995168021085, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.050337791442871, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6922268271446228, + "num_tokens": 300138074.0, + "step": 11601 + }, + { + "epoch": 1.2741049857236986, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.489776134490967, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7142040133476257, + "num_tokens": 300167547.0, + "step": 11602 + }, + { + "epoch": 1.2742148034263123, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4816300868988037, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6959471702575684, + "num_tokens": 300197864.0, + "step": 11603 + }, + { + "epoch": 1.274324621128926, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6159749031066895, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6935958862304688, + "num_tokens": 300223913.0, + "step": 11604 + }, + { + "epoch": 1.2744344388315396, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.401689052581787, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7113432288169861, + "num_tokens": 300253707.0, + "step": 11605 + }, + { + "epoch": 1.2745442565341534, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.692725419998169, + "learning_rate": 1e-06, + "loss": 0.9367, + "mean_token_accuracy": 0.7215415239334106, + "num_tokens": 300277603.0, + "step": 11606 + }, + { + "epoch": 1.274654074236767, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4276909828186035, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7021529674530029, + "num_tokens": 300308542.0, + "step": 11607 + }, + { + "epoch": 1.2747638919393807, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1275882720947266, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7062548398971558, + "num_tokens": 300326984.0, + "step": 11608 + }, + { + "epoch": 1.2748737096419944, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.494138240814209, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7050668001174927, + "num_tokens": 300353479.0, + "step": 11609 + }, + { + "epoch": 1.274983527344608, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.59203839302063, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.687113881111145, + "num_tokens": 300377842.0, + "step": 11610 + }, + { + "epoch": 1.2750933450472215, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.828239679336548, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7065515518188477, + "num_tokens": 300399800.0, + "step": 11611 + }, + { + "epoch": 1.2752031627498353, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.95528507232666, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6880959868431091, + "num_tokens": 300420136.0, + "step": 11612 + }, + { + "epoch": 1.275312980452449, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8785390853881836, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7220351099967957, + "num_tokens": 300443520.0, + "step": 11613 + }, + { + "epoch": 1.2754227981550625, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7392923831939697, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7195549011230469, + "num_tokens": 300468267.0, + "step": 11614 + }, + { + "epoch": 1.2755326158576763, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.498248338699341, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7000977993011475, + "num_tokens": 300498103.0, + "step": 11615 + }, + { + "epoch": 1.2756424335602898, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.777329921722412, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7184792757034302, + "num_tokens": 300521745.0, + "step": 11616 + }, + { + "epoch": 1.2757522512629036, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.686608076095581, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7126423120498657, + "num_tokens": 300550409.0, + "step": 11617 + }, + { + "epoch": 1.2758620689655173, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8737564086914062, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7143038511276245, + "num_tokens": 300573697.0, + "step": 11618 + }, + { + "epoch": 1.2759718866681309, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7070419788360596, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7095503807067871, + "num_tokens": 300595692.0, + "step": 11619 + }, + { + "epoch": 1.2760817043707446, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4919421672821045, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7087926864624023, + "num_tokens": 300621855.0, + "step": 11620 + }, + { + "epoch": 1.2761915220733582, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.738797187805176, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7307291626930237, + "num_tokens": 300648529.0, + "step": 11621 + }, + { + "epoch": 1.276301339775972, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3987083435058594, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6940569281578064, + "num_tokens": 300680264.0, + "step": 11622 + }, + { + "epoch": 1.2764111574785855, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7120604515075684, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7014909982681274, + "num_tokens": 300703359.0, + "step": 11623 + }, + { + "epoch": 1.2765209751811992, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0363264083862305, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.696793258190155, + "num_tokens": 300722624.0, + "step": 11624 + }, + { + "epoch": 1.2766307928838128, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.519648790359497, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7139521837234497, + "num_tokens": 300749852.0, + "step": 11625 + }, + { + "epoch": 1.2767406105864265, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7363908290863037, + "learning_rate": 1e-06, + "loss": 0.9297, + "mean_token_accuracy": 0.7276855111122131, + "num_tokens": 300772727.0, + "step": 11626 + }, + { + "epoch": 1.2768504282890403, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.01530385017395, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7046701908111572, + "num_tokens": 300796476.0, + "step": 11627 + }, + { + "epoch": 1.2769602459916538, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.893911838531494, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.702701210975647, + "num_tokens": 300821612.0, + "step": 11628 + }, + { + "epoch": 1.2770700636942676, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5840651988983154, + "learning_rate": 1e-06, + "loss": 0.9111, + "mean_token_accuracy": 0.7302635312080383, + "num_tokens": 300847114.0, + "step": 11629 + }, + { + "epoch": 1.277179881396881, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9955763816833496, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7046054601669312, + "num_tokens": 300867258.0, + "step": 11630 + }, + { + "epoch": 1.2772896990994949, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4804773330688477, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.6780259013175964, + "num_tokens": 300895077.0, + "step": 11631 + }, + { + "epoch": 1.2773995168021086, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.468209743499756, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6939454078674316, + "num_tokens": 300926950.0, + "step": 11632 + }, + { + "epoch": 1.2775093345047221, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.520437717437744, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7094907760620117, + "num_tokens": 300954676.0, + "step": 11633 + }, + { + "epoch": 1.2776191522073357, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6525890827178955, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7078843116760254, + "num_tokens": 300979637.0, + "step": 11634 + }, + { + "epoch": 1.2777289699099494, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5134639739990234, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6945639252662659, + "num_tokens": 301008969.0, + "step": 11635 + }, + { + "epoch": 1.2778387876125632, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.387683153152466, + "learning_rate": 1e-06, + "loss": 1.1088, + "mean_token_accuracy": 0.6903655529022217, + "num_tokens": 301039515.0, + "step": 11636 + }, + { + "epoch": 1.2779486053151767, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4723122119903564, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7108692526817322, + "num_tokens": 301066854.0, + "step": 11637 + }, + { + "epoch": 1.2780584230177905, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5988337993621826, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.6925327777862549, + "num_tokens": 301097544.0, + "step": 11638 + }, + { + "epoch": 1.278168240720404, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2839853763580322, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7080490589141846, + "num_tokens": 301128404.0, + "step": 11639 + }, + { + "epoch": 1.2782780584230178, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.504417896270752, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.703255295753479, + "num_tokens": 301154182.0, + "step": 11640 + }, + { + "epoch": 1.2783878761256315, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9017176628112793, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7165701389312744, + "num_tokens": 301174498.0, + "step": 11641 + }, + { + "epoch": 1.278497693828245, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6201765537261963, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7072527408599854, + "num_tokens": 301198593.0, + "step": 11642 + }, + { + "epoch": 1.2786075115308588, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.743665933609009, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7216875553131104, + "num_tokens": 301219877.0, + "step": 11643 + }, + { + "epoch": 1.2787173292334724, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.052314519882202, + "learning_rate": 1e-06, + "loss": 0.9185, + "mean_token_accuracy": 0.726892352104187, + "num_tokens": 301237174.0, + "step": 11644 + }, + { + "epoch": 1.2788271469360861, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.630668878555298, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7108633518218994, + "num_tokens": 301264961.0, + "step": 11645 + }, + { + "epoch": 1.2789369646386999, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5586671829223633, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7095456123352051, + "num_tokens": 301288854.0, + "step": 11646 + }, + { + "epoch": 1.2790467823413134, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.527817726135254, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7079055309295654, + "num_tokens": 301317681.0, + "step": 11647 + }, + { + "epoch": 1.279156600043927, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.567450761795044, + "learning_rate": 1e-06, + "loss": 0.9395, + "mean_token_accuracy": 0.720728874206543, + "num_tokens": 301342976.0, + "step": 11648 + }, + { + "epoch": 1.2792664177465407, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.399031639099121, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.7004359364509583, + "num_tokens": 301371968.0, + "step": 11649 + }, + { + "epoch": 1.2793762354491545, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.441993236541748, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6975452899932861, + "num_tokens": 301399863.0, + "step": 11650 + }, + { + "epoch": 1.279486053151768, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.808823823928833, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7033071517944336, + "num_tokens": 301421566.0, + "step": 11651 + }, + { + "epoch": 1.2795958708543818, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9072515964508057, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7173774242401123, + "num_tokens": 301448732.0, + "step": 11652 + }, + { + "epoch": 1.2797056885569953, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6159451007843018, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6821681261062622, + "num_tokens": 301472964.0, + "step": 11653 + }, + { + "epoch": 1.279815506259609, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7058517932891846, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7073572874069214, + "num_tokens": 301499186.0, + "step": 11654 + }, + { + "epoch": 1.2799253239622228, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5433437824249268, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.7077416181564331, + "num_tokens": 301526014.0, + "step": 11655 + }, + { + "epoch": 1.2800351416648363, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7125139236450195, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.695715069770813, + "num_tokens": 301551649.0, + "step": 11656 + }, + { + "epoch": 1.28014495936745, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4508056640625, + "learning_rate": 1e-06, + "loss": 1.1217, + "mean_token_accuracy": 0.6818199157714844, + "num_tokens": 301580729.0, + "step": 11657 + }, + { + "epoch": 1.2802547770700636, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3804659843444824, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.685060977935791, + "num_tokens": 301610803.0, + "step": 11658 + }, + { + "epoch": 1.2803645947726774, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9007370471954346, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7127496600151062, + "num_tokens": 301634563.0, + "step": 11659 + }, + { + "epoch": 1.2804744124752911, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.906616449356079, + "learning_rate": 1e-06, + "loss": 0.9353, + "mean_token_accuracy": 0.7185150384902954, + "num_tokens": 301655487.0, + "step": 11660 + }, + { + "epoch": 1.2805842301779047, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.2842183113098145, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.6926549673080444, + "num_tokens": 301692870.0, + "step": 11661 + }, + { + "epoch": 1.2806940478805182, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6559970378875732, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.690680980682373, + "num_tokens": 301718774.0, + "step": 11662 + }, + { + "epoch": 1.280803865583132, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4582130908966064, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.726459264755249, + "num_tokens": 301745583.0, + "step": 11663 + }, + { + "epoch": 1.2809136832857457, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8542258739471436, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7050082683563232, + "num_tokens": 301766400.0, + "step": 11664 + }, + { + "epoch": 1.2810235009883593, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7115213871002197, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7196555137634277, + "num_tokens": 301788337.0, + "step": 11665 + }, + { + "epoch": 1.281133318690973, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7049062252044678, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.6979434490203857, + "num_tokens": 301812213.0, + "step": 11666 + }, + { + "epoch": 1.2812431363935866, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0721018314361572, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.7274879217147827, + "num_tokens": 301834284.0, + "step": 11667 + }, + { + "epoch": 1.2813529540962003, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5326619148254395, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6972467303276062, + "num_tokens": 301866714.0, + "step": 11668 + }, + { + "epoch": 1.281462771798814, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5923547744750977, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7055137753486633, + "num_tokens": 301895602.0, + "step": 11669 + }, + { + "epoch": 1.2815725895014276, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5923104286193848, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.697668194770813, + "num_tokens": 301922347.0, + "step": 11670 + }, + { + "epoch": 1.2816824072040414, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7070112228393555, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.712763786315918, + "num_tokens": 301948200.0, + "step": 11671 + }, + { + "epoch": 1.281792224906655, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6322479248046875, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7050765752792358, + "num_tokens": 301972825.0, + "step": 11672 + }, + { + "epoch": 1.2819020426092687, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8917837142944336, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7172896862030029, + "num_tokens": 301992609.0, + "step": 11673 + }, + { + "epoch": 1.2820118603118824, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4851903915405273, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.694875180721283, + "num_tokens": 302019194.0, + "step": 11674 + }, + { + "epoch": 1.282121678014496, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6139540672302246, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6945295333862305, + "num_tokens": 302045526.0, + "step": 11675 + }, + { + "epoch": 1.2822314957171095, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.819511651992798, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6959894895553589, + "num_tokens": 302071242.0, + "step": 11676 + }, + { + "epoch": 1.2823413134197232, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5294313430786133, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6934068202972412, + "num_tokens": 302099235.0, + "step": 11677 + }, + { + "epoch": 1.282451131122337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5843167304992676, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7044761180877686, + "num_tokens": 302125416.0, + "step": 11678 + }, + { + "epoch": 1.2825609488249505, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4499661922454834, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7026423215866089, + "num_tokens": 302155002.0, + "step": 11679 + }, + { + "epoch": 1.2826707665275643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8398828506469727, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7037762403488159, + "num_tokens": 302177355.0, + "step": 11680 + }, + { + "epoch": 1.2827805842301778, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7646291255950928, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7154409885406494, + "num_tokens": 302199821.0, + "step": 11681 + }, + { + "epoch": 1.2828904019327916, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.646362543106079, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.6983178853988647, + "num_tokens": 302228048.0, + "step": 11682 + }, + { + "epoch": 1.2830002196354053, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.749938488006592, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7203338742256165, + "num_tokens": 302250998.0, + "step": 11683 + }, + { + "epoch": 1.2831100373380189, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6883580684661865, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7217810153961182, + "num_tokens": 302275876.0, + "step": 11684 + }, + { + "epoch": 1.2832198550406326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5436956882476807, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7107357382774353, + "num_tokens": 302303707.0, + "step": 11685 + }, + { + "epoch": 1.2833296727432462, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.1177124977111816, + "learning_rate": 1e-06, + "loss": 0.8876, + "mean_token_accuracy": 0.729559600353241, + "num_tokens": 302329954.0, + "step": 11686 + }, + { + "epoch": 1.28343949044586, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0034003257751465, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7115238904953003, + "num_tokens": 302352435.0, + "step": 11687 + }, + { + "epoch": 1.2835493081484735, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3013603687286377, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7129483222961426, + "num_tokens": 302383080.0, + "step": 11688 + }, + { + "epoch": 1.2836591258510872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5776007175445557, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6867813467979431, + "num_tokens": 302407828.0, + "step": 11689 + }, + { + "epoch": 1.2837689435537007, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.603876829147339, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7051122784614563, + "num_tokens": 302434324.0, + "step": 11690 + }, + { + "epoch": 1.2838787612563145, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.307968854904175, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7044833302497864, + "num_tokens": 302466784.0, + "step": 11691 + }, + { + "epoch": 1.2839885789589283, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6120245456695557, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.725398063659668, + "num_tokens": 302490422.0, + "step": 11692 + }, + { + "epoch": 1.2840983966615418, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.364306926727295, + "learning_rate": 1e-06, + "loss": 1.1373, + "mean_token_accuracy": 0.6810410022735596, + "num_tokens": 302521864.0, + "step": 11693 + }, + { + "epoch": 1.2842082143641556, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5044384002685547, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7044296264648438, + "num_tokens": 302550059.0, + "step": 11694 + }, + { + "epoch": 1.284318032066769, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7181344032287598, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7132338285446167, + "num_tokens": 302572257.0, + "step": 11695 + }, + { + "epoch": 1.2844278497693828, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6358401775360107, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7076892852783203, + "num_tokens": 302597184.0, + "step": 11696 + }, + { + "epoch": 1.2845376674719966, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.5170059204101562, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7123329043388367, + "num_tokens": 302625370.0, + "step": 11697 + }, + { + "epoch": 1.2846474851746101, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8961985111236572, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7011483907699585, + "num_tokens": 302647698.0, + "step": 11698 + }, + { + "epoch": 1.2847573028772237, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5824074745178223, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.7123766541481018, + "num_tokens": 302675217.0, + "step": 11699 + }, + { + "epoch": 1.2848671205798374, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.664344072341919, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.7007938623428345, + "num_tokens": 302701118.0, + "step": 11700 + }, + { + "epoch": 1.2849769382824512, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9337852001190186, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.6928179264068604, + "num_tokens": 302721586.0, + "step": 11701 + }, + { + "epoch": 1.2850867559850647, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6008849143981934, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7050660848617554, + "num_tokens": 302749143.0, + "step": 11702 + }, + { + "epoch": 1.2851965736876785, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.63875150680542, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7087498903274536, + "num_tokens": 302771892.0, + "step": 11703 + }, + { + "epoch": 1.285306391390292, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.594856023788452, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7050188779830933, + "num_tokens": 302795399.0, + "step": 11704 + }, + { + "epoch": 1.2854162090929058, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.587834358215332, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7154687643051147, + "num_tokens": 302820330.0, + "step": 11705 + }, + { + "epoch": 1.2855260267955195, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5574088096618652, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6887592077255249, + "num_tokens": 302844470.0, + "step": 11706 + }, + { + "epoch": 1.285635844498133, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.637589693069458, + "learning_rate": 1e-06, + "loss": 1.1015, + "mean_token_accuracy": 0.6795256733894348, + "num_tokens": 302875201.0, + "step": 11707 + }, + { + "epoch": 1.2857456622007468, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.579078435897827, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7181058526039124, + "num_tokens": 302900116.0, + "step": 11708 + }, + { + "epoch": 1.2858554799033604, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5886456966400146, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6887147426605225, + "num_tokens": 302926066.0, + "step": 11709 + }, + { + "epoch": 1.2859652976059741, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5400094985961914, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.7064135074615479, + "num_tokens": 302951977.0, + "step": 11710 + }, + { + "epoch": 1.2860751153085879, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.6701385974884033, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7026224136352539, + "num_tokens": 302976941.0, + "step": 11711 + }, + { + "epoch": 1.2861849330112014, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.815385103225708, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7102633714675903, + "num_tokens": 302999808.0, + "step": 11712 + }, + { + "epoch": 1.286294750713815, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3388311862945557, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7323416471481323, + "num_tokens": 303030124.0, + "step": 11713 + }, + { + "epoch": 1.2864045684164287, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4640417098999023, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7175737619400024, + "num_tokens": 303062520.0, + "step": 11714 + }, + { + "epoch": 1.2865143861190425, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6045494079589844, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.706513524055481, + "num_tokens": 303087211.0, + "step": 11715 + }, + { + "epoch": 1.286624203821656, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5080533027648926, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7109584808349609, + "num_tokens": 303114010.0, + "step": 11716 + }, + { + "epoch": 1.2867340215242697, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.0324888229370117, + "learning_rate": 1e-06, + "loss": 0.9351, + "mean_token_accuracy": 0.723369836807251, + "num_tokens": 303132223.0, + "step": 11717 + }, + { + "epoch": 1.2868438392268833, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7140440940856934, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7121951580047607, + "num_tokens": 303154590.0, + "step": 11718 + }, + { + "epoch": 1.286953656929497, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.479280710220337, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6943169832229614, + "num_tokens": 303181959.0, + "step": 11719 + }, + { + "epoch": 1.2870634746321108, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.513563394546509, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.7271617650985718, + "num_tokens": 303206582.0, + "step": 11720 + }, + { + "epoch": 1.2871732923347243, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.771681308746338, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6941078901290894, + "num_tokens": 303230746.0, + "step": 11721 + }, + { + "epoch": 1.287283110037338, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4989776611328125, + "learning_rate": 1e-06, + "loss": 0.8982, + "mean_token_accuracy": 0.7333763241767883, + "num_tokens": 303256952.0, + "step": 11722 + }, + { + "epoch": 1.2873929277399516, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.890353202819824, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7231338620185852, + "num_tokens": 303276667.0, + "step": 11723 + }, + { + "epoch": 1.2875027454425654, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3789610862731934, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7212984561920166, + "num_tokens": 303302982.0, + "step": 11724 + }, + { + "epoch": 1.2876125631451791, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.882087230682373, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6969294548034668, + "num_tokens": 303330334.0, + "step": 11725 + }, + { + "epoch": 1.2877223808477927, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.569045305252075, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6825774908065796, + "num_tokens": 303358251.0, + "step": 11726 + }, + { + "epoch": 1.2878321985504062, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7041563987731934, + "learning_rate": 1e-06, + "loss": 0.908, + "mean_token_accuracy": 0.7331243753433228, + "num_tokens": 303381982.0, + "step": 11727 + }, + { + "epoch": 1.28794201625302, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8553740978240967, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7206835746765137, + "num_tokens": 303401346.0, + "step": 11728 + }, + { + "epoch": 1.2880518339556337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.540604829788208, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6979693174362183, + "num_tokens": 303426493.0, + "step": 11729 + }, + { + "epoch": 1.2881616516582473, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3789761066436768, + "learning_rate": 1e-06, + "loss": 1.154, + "mean_token_accuracy": 0.6712177395820618, + "num_tokens": 303458971.0, + "step": 11730 + }, + { + "epoch": 1.288271469360861, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6392343044281006, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.732910692691803, + "num_tokens": 303482654.0, + "step": 11731 + }, + { + "epoch": 1.2883812870634745, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2241933345794678, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7128803730010986, + "num_tokens": 303515731.0, + "step": 11732 + }, + { + "epoch": 1.2884911047660883, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.469446897506714, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7235268354415894, + "num_tokens": 303540606.0, + "step": 11733 + }, + { + "epoch": 1.288600922468702, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6130268573760986, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7132440805435181, + "num_tokens": 303562877.0, + "step": 11734 + }, + { + "epoch": 1.2887107401713156, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3515448570251465, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6929271817207336, + "num_tokens": 303590329.0, + "step": 11735 + }, + { + "epoch": 1.2888205578739294, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2936806678771973, + "learning_rate": 1e-06, + "loss": 1.153, + "mean_token_accuracy": 0.6651701927185059, + "num_tokens": 303622275.0, + "step": 11736 + }, + { + "epoch": 1.2889303755765429, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3648788928985596, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6938316226005554, + "num_tokens": 303652835.0, + "step": 11737 + }, + { + "epoch": 1.2890401932791566, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.664990186691284, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7068682909011841, + "num_tokens": 303677543.0, + "step": 11738 + }, + { + "epoch": 1.2891500109817704, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8623063564300537, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.7020663022994995, + "num_tokens": 303705579.0, + "step": 11739 + }, + { + "epoch": 1.289259828684384, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6935248374938965, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7200731039047241, + "num_tokens": 303727515.0, + "step": 11740 + }, + { + "epoch": 1.2893696463869975, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.428128719329834, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7062420845031738, + "num_tokens": 303757209.0, + "step": 11741 + }, + { + "epoch": 1.2894794640896112, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5838820934295654, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6885925531387329, + "num_tokens": 303785195.0, + "step": 11742 + }, + { + "epoch": 1.289589281792225, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.109790563583374, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.7037742137908936, + "num_tokens": 303804918.0, + "step": 11743 + }, + { + "epoch": 1.2896990994948385, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6317026615142822, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.697929859161377, + "num_tokens": 303829718.0, + "step": 11744 + }, + { + "epoch": 1.2898089171974523, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5179483890533447, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.697359025478363, + "num_tokens": 303858466.0, + "step": 11745 + }, + { + "epoch": 1.2899187349000658, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.554614782333374, + "learning_rate": 1e-06, + "loss": 0.8851, + "mean_token_accuracy": 0.7328321933746338, + "num_tokens": 303881926.0, + "step": 11746 + }, + { + "epoch": 1.2900285526026796, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5048866271972656, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7162879705429077, + "num_tokens": 303908711.0, + "step": 11747 + }, + { + "epoch": 1.2901383703052933, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.456693410873413, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6954162120819092, + "num_tokens": 303934977.0, + "step": 11748 + }, + { + "epoch": 1.2902481880079069, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.361973524093628, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7097986936569214, + "num_tokens": 303962166.0, + "step": 11749 + }, + { + "epoch": 1.2903580057105206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8647103309631348, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7001399993896484, + "num_tokens": 303993221.0, + "step": 11750 + }, + { + "epoch": 1.2904678234131342, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.768296003341675, + "learning_rate": 1e-06, + "loss": 1.1263, + "mean_token_accuracy": 0.6846194863319397, + "num_tokens": 304018005.0, + "step": 11751 + }, + { + "epoch": 1.290577641115748, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6347460746765137, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7118303775787354, + "num_tokens": 304042860.0, + "step": 11752 + }, + { + "epoch": 1.2906874588183614, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3422887325286865, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6985350251197815, + "num_tokens": 304072245.0, + "step": 11753 + }, + { + "epoch": 1.2907972765209752, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.53525447845459, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7059259414672852, + "num_tokens": 304096862.0, + "step": 11754 + }, + { + "epoch": 1.2909070942235887, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8392529487609863, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7085743546485901, + "num_tokens": 304124158.0, + "step": 11755 + }, + { + "epoch": 1.2910169119262025, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.613365650177002, + "learning_rate": 1e-06, + "loss": 1.1125, + "mean_token_accuracy": 0.6778567433357239, + "num_tokens": 304152468.0, + "step": 11756 + }, + { + "epoch": 1.2911267296288162, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7765371799468994, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7037692070007324, + "num_tokens": 304175812.0, + "step": 11757 + }, + { + "epoch": 1.2912365473314298, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.59761118888855, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7240003347396851, + "num_tokens": 304200028.0, + "step": 11758 + }, + { + "epoch": 1.2913463650340435, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.800299644470215, + "learning_rate": 1e-06, + "loss": 0.9443, + "mean_token_accuracy": 0.7200407981872559, + "num_tokens": 304224531.0, + "step": 11759 + }, + { + "epoch": 1.291456182736657, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.424379587173462, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6821695566177368, + "num_tokens": 304256347.0, + "step": 11760 + }, + { + "epoch": 1.2915660004392708, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5157854557037354, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6922119855880737, + "num_tokens": 304281763.0, + "step": 11761 + }, + { + "epoch": 1.2916758181418846, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.891474962234497, + "learning_rate": 1e-06, + "loss": 0.9154, + "mean_token_accuracy": 0.7427189350128174, + "num_tokens": 304304235.0, + "step": 11762 + }, + { + "epoch": 1.2917856358444981, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6767828464508057, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7138881683349609, + "num_tokens": 304329706.0, + "step": 11763 + }, + { + "epoch": 1.2918954535471117, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5724949836730957, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7210342884063721, + "num_tokens": 304360223.0, + "step": 11764 + }, + { + "epoch": 1.2920052712497254, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.660845994949341, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7165807485580444, + "num_tokens": 304382198.0, + "step": 11765 + }, + { + "epoch": 1.2921150889523392, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.279144525527954, + "learning_rate": 1e-06, + "loss": 0.9566, + "mean_token_accuracy": 0.7184287309646606, + "num_tokens": 304415819.0, + "step": 11766 + }, + { + "epoch": 1.2922249066549527, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4994125366210938, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.6954265236854553, + "num_tokens": 304443269.0, + "step": 11767 + }, + { + "epoch": 1.2923347243575665, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.696934223175049, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6896913051605225, + "num_tokens": 304467737.0, + "step": 11768 + }, + { + "epoch": 1.29244454206018, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.52201771736145, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6948586702346802, + "num_tokens": 304498891.0, + "step": 11769 + }, + { + "epoch": 1.2925543597627938, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.0283989906311035, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7053937911987305, + "num_tokens": 304521902.0, + "step": 11770 + }, + { + "epoch": 1.2926641774654075, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7395269870758057, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7089489698410034, + "num_tokens": 304546070.0, + "step": 11771 + }, + { + "epoch": 1.292773995168021, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.818005084991455, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7023438215255737, + "num_tokens": 304567669.0, + "step": 11772 + }, + { + "epoch": 1.2928838128706348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5452654361724854, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7036740779876709, + "num_tokens": 304594245.0, + "step": 11773 + }, + { + "epoch": 1.2929936305732483, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5466504096984863, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7154223918914795, + "num_tokens": 304622361.0, + "step": 11774 + }, + { + "epoch": 1.293103448275862, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.878483533859253, + "learning_rate": 1e-06, + "loss": 0.8532, + "mean_token_accuracy": 0.7454129457473755, + "num_tokens": 304641523.0, + "step": 11775 + }, + { + "epoch": 1.2932132659784759, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.620195150375366, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7183417677879333, + "num_tokens": 304666763.0, + "step": 11776 + }, + { + "epoch": 1.2933230836810894, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6847918033599854, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.700799822807312, + "num_tokens": 304691674.0, + "step": 11777 + }, + { + "epoch": 1.293432901383703, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.752091884613037, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.710385799407959, + "num_tokens": 304713565.0, + "step": 11778 + }, + { + "epoch": 1.2935427190863167, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6927881240844727, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.700225830078125, + "num_tokens": 304738147.0, + "step": 11779 + }, + { + "epoch": 1.2936525367889304, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.649935483932495, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.6952387094497681, + "num_tokens": 304763544.0, + "step": 11780 + }, + { + "epoch": 1.293762354491544, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.45658278465271, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6883131265640259, + "num_tokens": 304793019.0, + "step": 11781 + }, + { + "epoch": 1.2938721721941577, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3742258548736572, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6936973333358765, + "num_tokens": 304821249.0, + "step": 11782 + }, + { + "epoch": 1.2939819898967713, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3996570110321045, + "learning_rate": 1e-06, + "loss": 1.1249, + "mean_token_accuracy": 0.6848218441009521, + "num_tokens": 304849658.0, + "step": 11783 + }, + { + "epoch": 1.294091807599385, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.614962577819824, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7057903409004211, + "num_tokens": 304872325.0, + "step": 11784 + }, + { + "epoch": 1.2942016253019988, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.598984479904175, + "learning_rate": 1e-06, + "loss": 0.92, + "mean_token_accuracy": 0.7321195006370544, + "num_tokens": 304896426.0, + "step": 11785 + }, + { + "epoch": 1.2943114430046123, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6959495544433594, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6891912221908569, + "num_tokens": 304920851.0, + "step": 11786 + }, + { + "epoch": 1.294421260707226, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.613044261932373, + "learning_rate": 1e-06, + "loss": 0.8556, + "mean_token_accuracy": 0.7518901824951172, + "num_tokens": 304943334.0, + "step": 11787 + }, + { + "epoch": 1.2945310784098396, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.963909864425659, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6856735944747925, + "num_tokens": 304969500.0, + "step": 11788 + }, + { + "epoch": 1.2946408961124534, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7501535415649414, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7091408967971802, + "num_tokens": 304991110.0, + "step": 11789 + }, + { + "epoch": 1.2947507138150671, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.684178590774536, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7077261209487915, + "num_tokens": 305015955.0, + "step": 11790 + }, + { + "epoch": 1.2948605315176807, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7798590660095215, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7035452723503113, + "num_tokens": 305042520.0, + "step": 11791 + }, + { + "epoch": 1.2949703492202942, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.640026330947876, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6959830522537231, + "num_tokens": 305071471.0, + "step": 11792 + }, + { + "epoch": 1.295080166922908, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6115951538085938, + "learning_rate": 1e-06, + "loss": 0.897, + "mean_token_accuracy": 0.7371882200241089, + "num_tokens": 305095514.0, + "step": 11793 + }, + { + "epoch": 1.2951899846255217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.787121295928955, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6948850154876709, + "num_tokens": 305117920.0, + "step": 11794 + }, + { + "epoch": 1.2952998023281352, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7827813625335693, + "learning_rate": 1e-06, + "loss": 0.8994, + "mean_token_accuracy": 0.7348973751068115, + "num_tokens": 305139672.0, + "step": 11795 + }, + { + "epoch": 1.295409620030749, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6726620197296143, + "learning_rate": 1e-06, + "loss": 0.977, + "mean_token_accuracy": 0.71409672498703, + "num_tokens": 305164213.0, + "step": 11796 + }, + { + "epoch": 1.2955194377333625, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.65207576751709, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6960995197296143, + "num_tokens": 305188419.0, + "step": 11797 + }, + { + "epoch": 1.2956292554359763, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.454829454421997, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.695911169052124, + "num_tokens": 305217696.0, + "step": 11798 + }, + { + "epoch": 1.29573907313859, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5428988933563232, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.707741379737854, + "num_tokens": 305244888.0, + "step": 11799 + }, + { + "epoch": 1.2958488908412036, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4662463665008545, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7385634183883667, + "num_tokens": 305271407.0, + "step": 11800 + }, + { + "epoch": 1.2959587085438173, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8205599784851074, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.706972599029541, + "num_tokens": 305293901.0, + "step": 11801 + }, + { + "epoch": 1.2960685262464309, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.535423994064331, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7037360668182373, + "num_tokens": 305319982.0, + "step": 11802 + }, + { + "epoch": 1.2961783439490446, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7038514614105225, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.707866907119751, + "num_tokens": 305341512.0, + "step": 11803 + }, + { + "epoch": 1.2962881616516582, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5712497234344482, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6921297311782837, + "num_tokens": 305369414.0, + "step": 11804 + }, + { + "epoch": 1.296397979354272, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.717017889022827, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7235434055328369, + "num_tokens": 305392887.0, + "step": 11805 + }, + { + "epoch": 1.2965077970568855, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3417508602142334, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6825776100158691, + "num_tokens": 305423865.0, + "step": 11806 + }, + { + "epoch": 1.2966176147594992, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.413205146789551, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.694996178150177, + "num_tokens": 305454469.0, + "step": 11807 + }, + { + "epoch": 1.296727432462113, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7438080310821533, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7067918181419373, + "num_tokens": 305483821.0, + "step": 11808 + }, + { + "epoch": 1.2968372501647265, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.71635103225708, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7207567095756531, + "num_tokens": 305506685.0, + "step": 11809 + }, + { + "epoch": 1.2969470678673403, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6064000129699707, + "learning_rate": 1e-06, + "loss": 0.9597, + "mean_token_accuracy": 0.7217411994934082, + "num_tokens": 305530489.0, + "step": 11810 + }, + { + "epoch": 1.2970568855699538, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6746890544891357, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7186119556427002, + "num_tokens": 305552683.0, + "step": 11811 + }, + { + "epoch": 1.2971667032725676, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6061532497406006, + "learning_rate": 1e-06, + "loss": 0.8252, + "mean_token_accuracy": 0.7538607716560364, + "num_tokens": 305575005.0, + "step": 11812 + }, + { + "epoch": 1.2972765209751813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.583217144012451, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7231510877609253, + "num_tokens": 305601169.0, + "step": 11813 + }, + { + "epoch": 1.2973863386777948, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.491722345352173, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7205354571342468, + "num_tokens": 305628213.0, + "step": 11814 + }, + { + "epoch": 1.2974961563804086, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.502540111541748, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.6980969309806824, + "num_tokens": 305655998.0, + "step": 11815 + }, + { + "epoch": 1.2976059740830221, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.626481056213379, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.6946529150009155, + "num_tokens": 305682178.0, + "step": 11816 + }, + { + "epoch": 1.297715791785636, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.709766149520874, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6951217651367188, + "num_tokens": 305706363.0, + "step": 11817 + }, + { + "epoch": 1.2978256094882494, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.284764051437378, + "learning_rate": 1e-06, + "loss": 0.9555, + "mean_token_accuracy": 0.7150084972381592, + "num_tokens": 305736478.0, + "step": 11818 + }, + { + "epoch": 1.2979354271908632, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.611170768737793, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.6993836164474487, + "num_tokens": 305765312.0, + "step": 11819 + }, + { + "epoch": 1.2980452448934767, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.619948625564575, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7260557413101196, + "num_tokens": 305788731.0, + "step": 11820 + }, + { + "epoch": 1.2981550625960905, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5052077770233154, + "learning_rate": 1e-06, + "loss": 1.1254, + "mean_token_accuracy": 0.6823197603225708, + "num_tokens": 305814487.0, + "step": 11821 + }, + { + "epoch": 1.2982648802987042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.522784948348999, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7267568111419678, + "num_tokens": 305840110.0, + "step": 11822 + }, + { + "epoch": 1.2983746980013178, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3556230068206787, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7099828720092773, + "num_tokens": 305870635.0, + "step": 11823 + }, + { + "epoch": 1.2984845157039315, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.418691873550415, + "learning_rate": 1e-06, + "loss": 1.0915, + "mean_token_accuracy": 0.693523645401001, + "num_tokens": 305900461.0, + "step": 11824 + }, + { + "epoch": 1.298594333406545, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7493021488189697, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7117166519165039, + "num_tokens": 305920842.0, + "step": 11825 + }, + { + "epoch": 1.2987041511091588, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.52634859085083, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6938819885253906, + "num_tokens": 305947118.0, + "step": 11826 + }, + { + "epoch": 1.2988139688117726, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.583714246749878, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7108317613601685, + "num_tokens": 305973592.0, + "step": 11827 + }, + { + "epoch": 1.2989237865143861, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4910051822662354, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7071340084075928, + "num_tokens": 306000339.0, + "step": 11828 + }, + { + "epoch": 1.2990336042169996, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1753337383270264, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7004258036613464, + "num_tokens": 306018889.0, + "step": 11829 + }, + { + "epoch": 1.2991434219196134, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8427281379699707, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7114360928535461, + "num_tokens": 306041685.0, + "step": 11830 + }, + { + "epoch": 1.2992532396222272, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.421180486679077, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6984325647354126, + "num_tokens": 306072866.0, + "step": 11831 + }, + { + "epoch": 1.2993630573248407, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.550950527191162, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7033129930496216, + "num_tokens": 306100711.0, + "step": 11832 + }, + { + "epoch": 1.2994728750274545, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8107478618621826, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6870999932289124, + "num_tokens": 306122537.0, + "step": 11833 + }, + { + "epoch": 1.299582692730068, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6407923698425293, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7066183686256409, + "num_tokens": 306145470.0, + "step": 11834 + }, + { + "epoch": 1.2996925104326817, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6088449954986572, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6968625783920288, + "num_tokens": 306171805.0, + "step": 11835 + }, + { + "epoch": 1.2998023281352955, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5446889400482178, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7181017398834229, + "num_tokens": 306196865.0, + "step": 11836 + }, + { + "epoch": 1.299912145837909, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4485199451446533, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7090033292770386, + "num_tokens": 306224739.0, + "step": 11837 + }, + { + "epoch": 1.3000219635405228, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9678146839141846, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6923278570175171, + "num_tokens": 306248787.0, + "step": 11838 + }, + { + "epoch": 1.3001317812431363, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.688272476196289, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7011300325393677, + "num_tokens": 306274373.0, + "step": 11839 + }, + { + "epoch": 1.30024159894575, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7110180854797363, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7068405151367188, + "num_tokens": 306299504.0, + "step": 11840 + }, + { + "epoch": 1.3003514166483638, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4260847568511963, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6961679458618164, + "num_tokens": 306328299.0, + "step": 11841 + }, + { + "epoch": 1.3004612343509774, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7765660285949707, + "learning_rate": 1e-06, + "loss": 0.9185, + "mean_token_accuracy": 0.7239254713058472, + "num_tokens": 306351122.0, + "step": 11842 + }, + { + "epoch": 1.300571052053591, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0671653747558594, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7193244099617004, + "num_tokens": 306371039.0, + "step": 11843 + }, + { + "epoch": 1.3006808697562047, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.577141523361206, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7017977237701416, + "num_tokens": 306398171.0, + "step": 11844 + }, + { + "epoch": 1.3007906874588184, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4522581100463867, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7051520347595215, + "num_tokens": 306426726.0, + "step": 11845 + }, + { + "epoch": 1.300900505161432, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3051843643188477, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6854325532913208, + "num_tokens": 306459056.0, + "step": 11846 + }, + { + "epoch": 1.3010103228640457, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3995492458343506, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6943441033363342, + "num_tokens": 306488203.0, + "step": 11847 + }, + { + "epoch": 1.3011201405666593, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.561089515686035, + "learning_rate": 1e-06, + "loss": 0.9238, + "mean_token_accuracy": 0.7241275906562805, + "num_tokens": 306514586.0, + "step": 11848 + }, + { + "epoch": 1.301229958269273, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6797075271606445, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7203474044799805, + "num_tokens": 306541307.0, + "step": 11849 + }, + { + "epoch": 1.3013397759718868, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.505610227584839, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7146939039230347, + "num_tokens": 306568190.0, + "step": 11850 + }, + { + "epoch": 1.3014495936745003, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4131228923797607, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7099327445030212, + "num_tokens": 306598005.0, + "step": 11851 + }, + { + "epoch": 1.301559411377114, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.415588855743408, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7111603021621704, + "num_tokens": 306626348.0, + "step": 11852 + }, + { + "epoch": 1.3016692290797276, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.783256769180298, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7016828060150146, + "num_tokens": 306648604.0, + "step": 11853 + }, + { + "epoch": 1.3017790467823414, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4213881492614746, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.69493567943573, + "num_tokens": 306677033.0, + "step": 11854 + }, + { + "epoch": 1.301888864484955, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5435547828674316, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.714026689529419, + "num_tokens": 306704591.0, + "step": 11855 + }, + { + "epoch": 1.3019986821875686, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9119930267333984, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.716991662979126, + "num_tokens": 306722762.0, + "step": 11856 + }, + { + "epoch": 1.3021084998901822, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4144961833953857, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6918623447418213, + "num_tokens": 306755451.0, + "step": 11857 + }, + { + "epoch": 1.302218317592796, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.223008871078491, + "learning_rate": 1e-06, + "loss": 0.9095, + "mean_token_accuracy": 0.7327659130096436, + "num_tokens": 306777213.0, + "step": 11858 + }, + { + "epoch": 1.3023281352954097, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5318567752838135, + "learning_rate": 1e-06, + "loss": 1.1338, + "mean_token_accuracy": 0.6783642172813416, + "num_tokens": 306811483.0, + "step": 11859 + }, + { + "epoch": 1.3024379529980232, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6498196125030518, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.7058935761451721, + "num_tokens": 306835767.0, + "step": 11860 + }, + { + "epoch": 1.302547770700637, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7197649478912354, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7206283807754517, + "num_tokens": 306858057.0, + "step": 11861 + }, + { + "epoch": 1.3026575884032505, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7071406841278076, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7218708992004395, + "num_tokens": 306881002.0, + "step": 11862 + }, + { + "epoch": 1.3027674061058643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5084965229034424, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.6999533176422119, + "num_tokens": 306907122.0, + "step": 11863 + }, + { + "epoch": 1.302877223808478, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.588649272918701, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7114637494087219, + "num_tokens": 306931095.0, + "step": 11864 + }, + { + "epoch": 1.3029870415110916, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.602752685546875, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7152888178825378, + "num_tokens": 306953333.0, + "step": 11865 + }, + { + "epoch": 1.3030968592137053, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8917975425720215, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.7017940282821655, + "num_tokens": 306979310.0, + "step": 11866 + }, + { + "epoch": 1.3032066769163189, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.696381092071533, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7108676433563232, + "num_tokens": 307005402.0, + "step": 11867 + }, + { + "epoch": 1.3033164946189326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5847179889678955, + "learning_rate": 1e-06, + "loss": 0.9995, + "mean_token_accuracy": 0.7062820196151733, + "num_tokens": 307031279.0, + "step": 11868 + }, + { + "epoch": 1.3034263123215462, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.2066972255706787, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7153884172439575, + "num_tokens": 307054553.0, + "step": 11869 + }, + { + "epoch": 1.30353613002416, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.552443027496338, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6951150894165039, + "num_tokens": 307081607.0, + "step": 11870 + }, + { + "epoch": 1.3036459477267734, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4572606086730957, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7078437805175781, + "num_tokens": 307108952.0, + "step": 11871 + }, + { + "epoch": 1.3037557654293872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.950148344039917, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6987733244895935, + "num_tokens": 307129054.0, + "step": 11872 + }, + { + "epoch": 1.303865583132001, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.789968729019165, + "learning_rate": 1e-06, + "loss": 0.9173, + "mean_token_accuracy": 0.7313396334648132, + "num_tokens": 307155838.0, + "step": 11873 + }, + { + "epoch": 1.3039754008346145, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7124595642089844, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6902245879173279, + "num_tokens": 307182441.0, + "step": 11874 + }, + { + "epoch": 1.3040852185372283, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.974275827407837, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7157531976699829, + "num_tokens": 307204790.0, + "step": 11875 + }, + { + "epoch": 1.3041950362398418, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6056110858917236, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.6978809237480164, + "num_tokens": 307230438.0, + "step": 11876 + }, + { + "epoch": 1.3043048539424555, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.622498035430908, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6956750154495239, + "num_tokens": 307262556.0, + "step": 11877 + }, + { + "epoch": 1.3044146716450693, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8277294635772705, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6904534101486206, + "num_tokens": 307283816.0, + "step": 11878 + }, + { + "epoch": 1.3045244893476828, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.842555522918701, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6980422735214233, + "num_tokens": 307304357.0, + "step": 11879 + }, + { + "epoch": 1.3046343070502964, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.606363534927368, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.70097815990448, + "num_tokens": 307329642.0, + "step": 11880 + }, + { + "epoch": 1.3047441247529101, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3987317085266113, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7140177488327026, + "num_tokens": 307357631.0, + "step": 11881 + }, + { + "epoch": 1.3048539424555239, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.52260422706604, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7043169736862183, + "num_tokens": 307384204.0, + "step": 11882 + }, + { + "epoch": 1.3049637601581374, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.966578722000122, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6881839036941528, + "num_tokens": 307411603.0, + "step": 11883 + }, + { + "epoch": 1.3050735778607512, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4896061420440674, + "learning_rate": 1e-06, + "loss": 1.1199, + "mean_token_accuracy": 0.6731846332550049, + "num_tokens": 307440959.0, + "step": 11884 + }, + { + "epoch": 1.3051833955633647, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.416733503341675, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6867309212684631, + "num_tokens": 307472194.0, + "step": 11885 + }, + { + "epoch": 1.3052932132659785, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.39078426361084, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.7080774307250977, + "num_tokens": 307502651.0, + "step": 11886 + }, + { + "epoch": 1.3054030309685922, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8606975078582764, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7154849171638489, + "num_tokens": 307525708.0, + "step": 11887 + }, + { + "epoch": 1.3055128486712058, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5843329429626465, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6897714138031006, + "num_tokens": 307551774.0, + "step": 11888 + }, + { + "epoch": 1.3056226663738195, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.407918691635132, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7157373428344727, + "num_tokens": 307577389.0, + "step": 11889 + }, + { + "epoch": 1.305732484076433, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4275741577148438, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.700124204158783, + "num_tokens": 307605984.0, + "step": 11890 + }, + { + "epoch": 1.3058423017790468, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.409520149230957, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7104282975196838, + "num_tokens": 307627950.0, + "step": 11891 + }, + { + "epoch": 1.3059521194816606, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8168816566467285, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7036402225494385, + "num_tokens": 307653276.0, + "step": 11892 + }, + { + "epoch": 1.306061937184274, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5052473545074463, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.6944023370742798, + "num_tokens": 307681643.0, + "step": 11893 + }, + { + "epoch": 1.3061717548868876, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.52581787109375, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7205910682678223, + "num_tokens": 307707993.0, + "step": 11894 + }, + { + "epoch": 1.3062815725895014, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.692032814025879, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.698235809803009, + "num_tokens": 307730284.0, + "step": 11895 + }, + { + "epoch": 1.3063913902921152, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.813194513320923, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7201534509658813, + "num_tokens": 307750876.0, + "step": 11896 + }, + { + "epoch": 1.3065012079947287, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8063530921936035, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7262597680091858, + "num_tokens": 307771352.0, + "step": 11897 + }, + { + "epoch": 1.3066110256973424, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6112422943115234, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.686120331287384, + "num_tokens": 307799490.0, + "step": 11898 + }, + { + "epoch": 1.306720843399956, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7468793392181396, + "learning_rate": 1e-06, + "loss": 0.907, + "mean_token_accuracy": 0.7288469672203064, + "num_tokens": 307822092.0, + "step": 11899 + }, + { + "epoch": 1.3068306611025697, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.729602098464966, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.707588255405426, + "num_tokens": 307847578.0, + "step": 11900 + }, + { + "epoch": 1.3069404788051835, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.391697645187378, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.6993273496627808, + "num_tokens": 307879184.0, + "step": 11901 + }, + { + "epoch": 1.307050296507797, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9033262729644775, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6925740242004395, + "num_tokens": 307902946.0, + "step": 11902 + }, + { + "epoch": 1.3071601142104108, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.802858829498291, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6868990659713745, + "num_tokens": 307925745.0, + "step": 11903 + }, + { + "epoch": 1.3072699319130243, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.68298077583313, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6934521198272705, + "num_tokens": 307950738.0, + "step": 11904 + }, + { + "epoch": 1.307379749615638, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.433004379272461, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6974514126777649, + "num_tokens": 307980167.0, + "step": 11905 + }, + { + "epoch": 1.3074895673182518, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.794222593307495, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7072087526321411, + "num_tokens": 308003106.0, + "step": 11906 + }, + { + "epoch": 1.3075993850208654, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4631869792938232, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7131680250167847, + "num_tokens": 308029924.0, + "step": 11907 + }, + { + "epoch": 1.307709202723479, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.642240047454834, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6930252313613892, + "num_tokens": 308053551.0, + "step": 11908 + }, + { + "epoch": 1.3078190204260927, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6984975337982178, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7135002613067627, + "num_tokens": 308075839.0, + "step": 11909 + }, + { + "epoch": 1.3079288381287064, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.657085657119751, + "learning_rate": 1e-06, + "loss": 0.9081, + "mean_token_accuracy": 0.730233371257782, + "num_tokens": 308098003.0, + "step": 11910 + }, + { + "epoch": 1.30803865583132, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3521344661712646, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6940163969993591, + "num_tokens": 308131484.0, + "step": 11911 + }, + { + "epoch": 1.3081484735339337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.678708791732788, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7094696164131165, + "num_tokens": 308152707.0, + "step": 11912 + }, + { + "epoch": 1.3082582912365472, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.339858055114746, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.7059363126754761, + "num_tokens": 308183132.0, + "step": 11913 + }, + { + "epoch": 1.308368108939161, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0091962814331055, + "learning_rate": 1e-06, + "loss": 0.8792, + "mean_token_accuracy": 0.7363442182540894, + "num_tokens": 308201717.0, + "step": 11914 + }, + { + "epoch": 1.3084779266417748, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1554629802703857, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7091547250747681, + "num_tokens": 308221777.0, + "step": 11915 + }, + { + "epoch": 1.3085877443443883, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.463641405105591, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7000380754470825, + "num_tokens": 308252221.0, + "step": 11916 + }, + { + "epoch": 1.308697562047002, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.617720603942871, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7127572894096375, + "num_tokens": 308274420.0, + "step": 11917 + }, + { + "epoch": 1.3088073797496156, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.438990592956543, + "learning_rate": 1e-06, + "loss": 1.1007, + "mean_token_accuracy": 0.68043452501297, + "num_tokens": 308309469.0, + "step": 11918 + }, + { + "epoch": 1.3089171974522293, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.633958578109741, + "learning_rate": 1e-06, + "loss": 0.9402, + "mean_token_accuracy": 0.7230616807937622, + "num_tokens": 308334493.0, + "step": 11919 + }, + { + "epoch": 1.309027015154843, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.56817889213562, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.6996695399284363, + "num_tokens": 308362169.0, + "step": 11920 + }, + { + "epoch": 1.3091368328574566, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.804851770401001, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7187197208404541, + "num_tokens": 308383792.0, + "step": 11921 + }, + { + "epoch": 1.3092466505600702, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6317155361175537, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7076044082641602, + "num_tokens": 308408941.0, + "step": 11922 + }, + { + "epoch": 1.309356468262684, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5859155654907227, + "learning_rate": 1e-06, + "loss": 1.0943, + "mean_token_accuracy": 0.6908015012741089, + "num_tokens": 308437407.0, + "step": 11923 + }, + { + "epoch": 1.3094662859652977, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6567609310150146, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6964809894561768, + "num_tokens": 308464422.0, + "step": 11924 + }, + { + "epoch": 1.3095761036679112, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.614448308944702, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.681262731552124, + "num_tokens": 308493901.0, + "step": 11925 + }, + { + "epoch": 1.309685921370525, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 16.72126579284668, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7089471817016602, + "num_tokens": 308527056.0, + "step": 11926 + }, + { + "epoch": 1.3097957390731385, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.76275897026062, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7041783332824707, + "num_tokens": 308559454.0, + "step": 11927 + }, + { + "epoch": 1.3099055567757523, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7021336555480957, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7122641801834106, + "num_tokens": 308582588.0, + "step": 11928 + }, + { + "epoch": 1.310015374478366, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6384942531585693, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7014847993850708, + "num_tokens": 308607198.0, + "step": 11929 + }, + { + "epoch": 1.3101251921809796, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5063016414642334, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7007487416267395, + "num_tokens": 308633119.0, + "step": 11930 + }, + { + "epoch": 1.3102350098835933, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.532257080078125, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6991857290267944, + "num_tokens": 308658039.0, + "step": 11931 + }, + { + "epoch": 1.3103448275862069, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6851823329925537, + "learning_rate": 1e-06, + "loss": 0.891, + "mean_token_accuracy": 0.7283509969711304, + "num_tokens": 308678716.0, + "step": 11932 + }, + { + "epoch": 1.3104546452888206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.451605796813965, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.7021827101707458, + "num_tokens": 308708883.0, + "step": 11933 + }, + { + "epoch": 1.3105644629914341, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.515110969543457, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7151156663894653, + "num_tokens": 308733671.0, + "step": 11934 + }, + { + "epoch": 1.310674280694048, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.486025810241699, + "learning_rate": 1e-06, + "loss": 1.0854, + "mean_token_accuracy": 0.698091447353363, + "num_tokens": 308760280.0, + "step": 11935 + }, + { + "epoch": 1.3107840983966614, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8344879150390625, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7211250066757202, + "num_tokens": 308780609.0, + "step": 11936 + }, + { + "epoch": 1.3108939160992752, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6703968048095703, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6896342039108276, + "num_tokens": 308804795.0, + "step": 11937 + }, + { + "epoch": 1.311003733801889, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8460440635681152, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7154251933097839, + "num_tokens": 308827372.0, + "step": 11938 + }, + { + "epoch": 1.3111135515045025, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 6.701202392578125, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7172532081604004, + "num_tokens": 308851157.0, + "step": 11939 + }, + { + "epoch": 1.3112233692071162, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.093991994857788, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7080575823783875, + "num_tokens": 308876515.0, + "step": 11940 + }, + { + "epoch": 1.3113331869097298, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9343557357788086, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6877841949462891, + "num_tokens": 308898561.0, + "step": 11941 + }, + { + "epoch": 1.3114430046123435, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.404900550842285, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7037505507469177, + "num_tokens": 308928792.0, + "step": 11942 + }, + { + "epoch": 1.3115528223149573, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.415046453475952, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7253713011741638, + "num_tokens": 308955868.0, + "step": 11943 + }, + { + "epoch": 1.3116626400175708, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5297346115112305, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.704910933971405, + "num_tokens": 308979800.0, + "step": 11944 + }, + { + "epoch": 1.3117724577201844, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4217658042907715, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.6970480680465698, + "num_tokens": 309009514.0, + "step": 11945 + }, + { + "epoch": 1.3118822754227981, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8374388217926025, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7072793841362, + "num_tokens": 309030237.0, + "step": 11946 + }, + { + "epoch": 1.3119920931254119, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.518662929534912, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7189793586730957, + "num_tokens": 309058352.0, + "step": 11947 + }, + { + "epoch": 1.3121019108280254, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8291444778442383, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7068158388137817, + "num_tokens": 309078411.0, + "step": 11948 + }, + { + "epoch": 1.3122117285306392, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.707204580307007, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7259057760238647, + "num_tokens": 309099362.0, + "step": 11949 + }, + { + "epoch": 1.3123215462332527, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9448294639587402, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7257131338119507, + "num_tokens": 309120392.0, + "step": 11950 + }, + { + "epoch": 1.3124313639358665, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4042208194732666, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7148052453994751, + "num_tokens": 309149883.0, + "step": 11951 + }, + { + "epoch": 1.3125411816384802, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.504451036453247, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7029342651367188, + "num_tokens": 309176404.0, + "step": 11952 + }, + { + "epoch": 1.3126509993410937, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.332535982131958, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6916836500167847, + "num_tokens": 309204974.0, + "step": 11953 + }, + { + "epoch": 1.3127608170437075, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3846967220306396, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7085933685302734, + "num_tokens": 309233538.0, + "step": 11954 + }, + { + "epoch": 1.312870634746321, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.336355686187744, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6965007185935974, + "num_tokens": 309265874.0, + "step": 11955 + }, + { + "epoch": 1.3129804524489348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.558351755142212, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6817400455474854, + "num_tokens": 309292134.0, + "step": 11956 + }, + { + "epoch": 1.3130902701515486, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.70624041557312, + "learning_rate": 1e-06, + "loss": 0.9244, + "mean_token_accuracy": 0.727452278137207, + "num_tokens": 309314202.0, + "step": 11957 + }, + { + "epoch": 1.313200087854162, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4489541053771973, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7072919607162476, + "num_tokens": 309345545.0, + "step": 11958 + }, + { + "epoch": 1.3133099055567756, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7956790924072266, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6919050812721252, + "num_tokens": 309368933.0, + "step": 11959 + }, + { + "epoch": 1.3134197232593894, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6838910579681396, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.705376386642456, + "num_tokens": 309395945.0, + "step": 11960 + }, + { + "epoch": 1.3135295409620031, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6928536891937256, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.691205620765686, + "num_tokens": 309421344.0, + "step": 11961 + }, + { + "epoch": 1.3136393586646167, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7049906253814697, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7037245035171509, + "num_tokens": 309444483.0, + "step": 11962 + }, + { + "epoch": 1.3137491763672304, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8217461109161377, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7066141366958618, + "num_tokens": 309468167.0, + "step": 11963 + }, + { + "epoch": 1.313858994069844, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3287971019744873, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6922783851623535, + "num_tokens": 309503149.0, + "step": 11964 + }, + { + "epoch": 1.3139688117724577, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5953431129455566, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7226397395133972, + "num_tokens": 309528160.0, + "step": 11965 + }, + { + "epoch": 1.3140786294750715, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.5024445056915283, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7140884399414062, + "num_tokens": 309556572.0, + "step": 11966 + }, + { + "epoch": 1.314188447177685, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.620936393737793, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7112689018249512, + "num_tokens": 309582611.0, + "step": 11967 + }, + { + "epoch": 1.3142982648802988, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4469287395477295, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.7259416580200195, + "num_tokens": 309609761.0, + "step": 11968 + }, + { + "epoch": 1.3144080825829123, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.745699644088745, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7043847441673279, + "num_tokens": 309635992.0, + "step": 11969 + }, + { + "epoch": 1.314517900285526, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.878636360168457, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.705219030380249, + "num_tokens": 309659002.0, + "step": 11970 + }, + { + "epoch": 1.3146277179881398, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5941905975341797, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6974152326583862, + "num_tokens": 309689015.0, + "step": 11971 + }, + { + "epoch": 1.3147375356907534, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6380536556243896, + "learning_rate": 1e-06, + "loss": 1.0983, + "mean_token_accuracy": 0.6829816102981567, + "num_tokens": 309715038.0, + "step": 11972 + }, + { + "epoch": 1.314847353393367, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7316105365753174, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.6981281042098999, + "num_tokens": 309739071.0, + "step": 11973 + }, + { + "epoch": 1.3149571710959806, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4319684505462646, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.6803369522094727, + "num_tokens": 309769186.0, + "step": 11974 + }, + { + "epoch": 1.3150669887985944, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.344296932220459, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7073795795440674, + "num_tokens": 309800157.0, + "step": 11975 + }, + { + "epoch": 1.315176806501208, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5093843936920166, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6817017793655396, + "num_tokens": 309827522.0, + "step": 11976 + }, + { + "epoch": 1.3152866242038217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4198098182678223, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.681138277053833, + "num_tokens": 309855405.0, + "step": 11977 + }, + { + "epoch": 1.3153964419064352, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.368380308151245, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7036442160606384, + "num_tokens": 309885220.0, + "step": 11978 + }, + { + "epoch": 1.315506259609049, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.555238723754883, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7162885665893555, + "num_tokens": 309912524.0, + "step": 11979 + }, + { + "epoch": 1.3156160773116627, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.438779592514038, + "learning_rate": 1e-06, + "loss": 1.1318, + "mean_token_accuracy": 0.6776027679443359, + "num_tokens": 309944880.0, + "step": 11980 + }, + { + "epoch": 1.3157258950142763, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6203439235687256, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.6958651542663574, + "num_tokens": 309971327.0, + "step": 11981 + }, + { + "epoch": 1.31583571271689, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7248897552490234, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6911646127700806, + "num_tokens": 309996491.0, + "step": 11982 + }, + { + "epoch": 1.3159455304195036, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.974980115890503, + "learning_rate": 1e-06, + "loss": 0.9365, + "mean_token_accuracy": 0.7237554788589478, + "num_tokens": 310016335.0, + "step": 11983 + }, + { + "epoch": 1.3160553481221173, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4026153087615967, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6935085654258728, + "num_tokens": 310051258.0, + "step": 11984 + }, + { + "epoch": 1.3161651658247309, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.548614263534546, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6995916962623596, + "num_tokens": 310084879.0, + "step": 11985 + }, + { + "epoch": 1.3162749835273446, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.682389974594116, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7263680696487427, + "num_tokens": 310110181.0, + "step": 11986 + }, + { + "epoch": 1.3163848012299582, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 3.5258426666259766, + "learning_rate": 1e-06, + "loss": 0.8836, + "mean_token_accuracy": 0.7354822158813477, + "num_tokens": 310132279.0, + "step": 11987 + }, + { + "epoch": 1.316494618932572, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.592637062072754, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.7218091487884521, + "num_tokens": 310154905.0, + "step": 11988 + }, + { + "epoch": 1.3166044366351857, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4907734394073486, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7084774971008301, + "num_tokens": 310182554.0, + "step": 11989 + }, + { + "epoch": 1.3167142543377992, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.322218418121338, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7079452276229858, + "num_tokens": 310216138.0, + "step": 11990 + }, + { + "epoch": 1.316824072040413, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.533334493637085, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7065552473068237, + "num_tokens": 310243173.0, + "step": 11991 + }, + { + "epoch": 1.3169338897430265, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.735248565673828, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7087448835372925, + "num_tokens": 310266769.0, + "step": 11992 + }, + { + "epoch": 1.3170437074456403, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.3776183128356934, + "learning_rate": 1e-06, + "loss": 1.1264, + "mean_token_accuracy": 0.682338535785675, + "num_tokens": 310299363.0, + "step": 11993 + }, + { + "epoch": 1.317153525148254, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.426788806915283, + "learning_rate": 1e-06, + "loss": 1.109, + "mean_token_accuracy": 0.681178092956543, + "num_tokens": 310327603.0, + "step": 11994 + }, + { + "epoch": 1.3172633428508675, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6925883293151855, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7091412544250488, + "num_tokens": 310352373.0, + "step": 11995 + }, + { + "epoch": 1.3173731605534813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.430767774581909, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7182127237319946, + "num_tokens": 310378023.0, + "step": 11996 + }, + { + "epoch": 1.3174829782560948, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.428377389907837, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6898089051246643, + "num_tokens": 310408772.0, + "step": 11997 + }, + { + "epoch": 1.3175927959587086, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5510103702545166, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.6990427374839783, + "num_tokens": 310437185.0, + "step": 11998 + }, + { + "epoch": 1.3177026136613221, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.173464775085449, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7103303670883179, + "num_tokens": 310472318.0, + "step": 11999 + }, + { + "epoch": 1.3178124313639359, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.397818088531494, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7230156660079956, + "num_tokens": 310501902.0, + "step": 12000 + }, + { + "epoch": 1.3179222490665494, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.859243631362915, + "learning_rate": 1e-06, + "loss": 0.891, + "mean_token_accuracy": 0.7378835678100586, + "num_tokens": 310523070.0, + "step": 12001 + }, + { + "epoch": 1.3180320667691632, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.492094039916992, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7067378759384155, + "num_tokens": 310548975.0, + "step": 12002 + }, + { + "epoch": 1.318141884471777, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.318711519241333, + "learning_rate": 1e-06, + "loss": 0.8872, + "mean_token_accuracy": 0.7423942685127258, + "num_tokens": 310578440.0, + "step": 12003 + }, + { + "epoch": 1.3182517021743905, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.44006085395813, + "learning_rate": 1e-06, + "loss": 1.0842, + "mean_token_accuracy": 0.6865665912628174, + "num_tokens": 310608050.0, + "step": 12004 + }, + { + "epoch": 1.3183615198770042, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6939170360565186, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7059879302978516, + "num_tokens": 310634990.0, + "step": 12005 + }, + { + "epoch": 1.3184713375796178, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.478008985519409, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.704980194568634, + "num_tokens": 310663671.0, + "step": 12006 + }, + { + "epoch": 1.3185811552822315, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.711073637008667, + "learning_rate": 1e-06, + "loss": 0.8894, + "mean_token_accuracy": 0.7275874614715576, + "num_tokens": 310686032.0, + "step": 12007 + }, + { + "epoch": 1.3186909729848453, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9101052284240723, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7014204263687134, + "num_tokens": 310715174.0, + "step": 12008 + }, + { + "epoch": 1.3188007906874588, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5937836170196533, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6846143007278442, + "num_tokens": 310741269.0, + "step": 12009 + }, + { + "epoch": 1.3189106083900723, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9671831130981445, + "learning_rate": 1e-06, + "loss": 0.9046, + "mean_token_accuracy": 0.7206922769546509, + "num_tokens": 310761641.0, + "step": 12010 + }, + { + "epoch": 1.319020426092686, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.648191452026367, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7166103720664978, + "num_tokens": 310784991.0, + "step": 12011 + }, + { + "epoch": 1.3191302437952999, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.874128580093384, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6847007274627686, + "num_tokens": 310807560.0, + "step": 12012 + }, + { + "epoch": 1.3192400614979134, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.465200185775757, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7097412943840027, + "num_tokens": 310833251.0, + "step": 12013 + }, + { + "epoch": 1.3193498792005272, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.138962984085083, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7036365270614624, + "num_tokens": 310865625.0, + "step": 12014 + }, + { + "epoch": 1.3194596969031407, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.014571189880371, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.7054574489593506, + "num_tokens": 310889811.0, + "step": 12015 + }, + { + "epoch": 1.3195695146057544, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.266853094100952, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7072498798370361, + "num_tokens": 310908529.0, + "step": 12016 + }, + { + "epoch": 1.3196793323083682, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5304951667785645, + "learning_rate": 1e-06, + "loss": 1.1607, + "mean_token_accuracy": 0.669853687286377, + "num_tokens": 310940143.0, + "step": 12017 + }, + { + "epoch": 1.3197891500109817, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7897133827209473, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7090888023376465, + "num_tokens": 310960261.0, + "step": 12018 + }, + { + "epoch": 1.3198989677135955, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.214653253555298, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7033863067626953, + "num_tokens": 310983025.0, + "step": 12019 + }, + { + "epoch": 1.320008785416209, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6719841957092285, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6992393732070923, + "num_tokens": 311008978.0, + "step": 12020 + }, + { + "epoch": 1.3201186031188228, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.559112071990967, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7149791717529297, + "num_tokens": 311037576.0, + "step": 12021 + }, + { + "epoch": 1.3202284208214365, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.754897356033325, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.687725841999054, + "num_tokens": 311061893.0, + "step": 12022 + }, + { + "epoch": 1.32033823852405, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.849090337753296, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7255613803863525, + "num_tokens": 311084726.0, + "step": 12023 + }, + { + "epoch": 1.3204480562266636, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.473621368408203, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7128220796585083, + "num_tokens": 311110099.0, + "step": 12024 + }, + { + "epoch": 1.3205578739292774, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.047794818878174, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7224102020263672, + "num_tokens": 311130690.0, + "step": 12025 + }, + { + "epoch": 1.3206676916318911, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9461240768432617, + "learning_rate": 1e-06, + "loss": 1.0796, + "mean_token_accuracy": 0.6828826665878296, + "num_tokens": 311157366.0, + "step": 12026 + }, + { + "epoch": 1.3207775093345047, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.80088210105896, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.6979199647903442, + "num_tokens": 311179686.0, + "step": 12027 + }, + { + "epoch": 1.3208873270371184, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.846560001373291, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6921252012252808, + "num_tokens": 311203656.0, + "step": 12028 + }, + { + "epoch": 1.320997144739732, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.47552752494812, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.702666163444519, + "num_tokens": 311230369.0, + "step": 12029 + }, + { + "epoch": 1.3211069624423457, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.331207513809204, + "learning_rate": 1e-06, + "loss": 0.9454, + "mean_token_accuracy": 0.7270394563674927, + "num_tokens": 311260256.0, + "step": 12030 + }, + { + "epoch": 1.3212167801449595, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.813133955001831, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6900360584259033, + "num_tokens": 311282552.0, + "step": 12031 + }, + { + "epoch": 1.321326597847573, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5880234241485596, + "learning_rate": 1e-06, + "loss": 0.9073, + "mean_token_accuracy": 0.729195237159729, + "num_tokens": 311304267.0, + "step": 12032 + }, + { + "epoch": 1.3214364155501868, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5437114238739014, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6896792650222778, + "num_tokens": 311330001.0, + "step": 12033 + }, + { + "epoch": 1.3215462332528003, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.310755968093872, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7085307836532593, + "num_tokens": 311355827.0, + "step": 12034 + }, + { + "epoch": 1.321656050955414, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7475838661193848, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7087751626968384, + "num_tokens": 311379536.0, + "step": 12035 + }, + { + "epoch": 1.3217658686580278, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5967607498168945, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7130171060562134, + "num_tokens": 311404203.0, + "step": 12036 + }, + { + "epoch": 1.3218756863606413, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.481325626373291, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7161813378334045, + "num_tokens": 311429686.0, + "step": 12037 + }, + { + "epoch": 1.3219855040632549, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2907021045684814, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6990745067596436, + "num_tokens": 311459298.0, + "step": 12038 + }, + { + "epoch": 1.3220953217658686, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7972750663757324, + "learning_rate": 1e-06, + "loss": 0.9026, + "mean_token_accuracy": 0.7312700748443604, + "num_tokens": 311479998.0, + "step": 12039 + }, + { + "epoch": 1.3222051394684824, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.612943649291992, + "learning_rate": 1e-06, + "loss": 1.0679, + "mean_token_accuracy": 0.6878488063812256, + "num_tokens": 311508414.0, + "step": 12040 + }, + { + "epoch": 1.322314957171096, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5938446521759033, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7189061641693115, + "num_tokens": 311531813.0, + "step": 12041 + }, + { + "epoch": 1.3224247748737097, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.340228319168091, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7082566022872925, + "num_tokens": 311565499.0, + "step": 12042 + }, + { + "epoch": 1.3225345925763232, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.9514546394348145, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7061893939971924, + "num_tokens": 311585634.0, + "step": 12043 + }, + { + "epoch": 1.322644410278937, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.589198589324951, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7054035067558289, + "num_tokens": 311612145.0, + "step": 12044 + }, + { + "epoch": 1.3227542279815507, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4760751724243164, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6874840259552002, + "num_tokens": 311637952.0, + "step": 12045 + }, + { + "epoch": 1.3228640456841643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.349654197692871, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7177774310112, + "num_tokens": 311667932.0, + "step": 12046 + }, + { + "epoch": 1.322973863386778, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.2355918884277344, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7163753509521484, + "num_tokens": 311691471.0, + "step": 12047 + }, + { + "epoch": 1.3230836810893916, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9192023277282715, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7151354551315308, + "num_tokens": 311710910.0, + "step": 12048 + }, + { + "epoch": 1.3231934987920053, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6480486392974854, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7261621356010437, + "num_tokens": 311733474.0, + "step": 12049 + }, + { + "epoch": 1.3233033164946189, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8620095252990723, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.703742265701294, + "num_tokens": 311754050.0, + "step": 12050 + }, + { + "epoch": 1.3234131341972326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7087512016296387, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6915076971054077, + "num_tokens": 311782576.0, + "step": 12051 + }, + { + "epoch": 1.3235229518998461, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4232096672058105, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7179627418518066, + "num_tokens": 311810613.0, + "step": 12052 + }, + { + "epoch": 1.32363276960246, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.303121566772461, + "learning_rate": 1e-06, + "loss": 1.1176, + "mean_token_accuracy": 0.6799439191818237, + "num_tokens": 311845004.0, + "step": 12053 + }, + { + "epoch": 1.3237425873050737, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.897697687149048, + "learning_rate": 1e-06, + "loss": 0.9211, + "mean_token_accuracy": 0.7306722402572632, + "num_tokens": 311865498.0, + "step": 12054 + }, + { + "epoch": 1.3238524050076872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.357329845428467, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6973499655723572, + "num_tokens": 311895993.0, + "step": 12055 + }, + { + "epoch": 1.323962222710301, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5268871784210205, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7093988656997681, + "num_tokens": 311922674.0, + "step": 12056 + }, + { + "epoch": 1.3240720404129145, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.757612943649292, + "learning_rate": 1e-06, + "loss": 0.848, + "mean_token_accuracy": 0.7472878694534302, + "num_tokens": 311944228.0, + "step": 12057 + }, + { + "epoch": 1.3241818581155282, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.674747943878174, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7014943361282349, + "num_tokens": 311967283.0, + "step": 12058 + }, + { + "epoch": 1.324291675818142, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9203829765319824, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7129706144332886, + "num_tokens": 311986322.0, + "step": 12059 + }, + { + "epoch": 1.3244014935207555, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5981616973876953, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7054161429405212, + "num_tokens": 312011426.0, + "step": 12060 + }, + { + "epoch": 1.324511311223369, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.466303586959839, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6828653812408447, + "num_tokens": 312043022.0, + "step": 12061 + }, + { + "epoch": 1.3246211289259828, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5807478427886963, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6814514398574829, + "num_tokens": 312073029.0, + "step": 12062 + }, + { + "epoch": 1.3247309466285966, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.602916717529297, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7178936004638672, + "num_tokens": 312098938.0, + "step": 12063 + }, + { + "epoch": 1.3248407643312101, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6294474601745605, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7047186493873596, + "num_tokens": 312122336.0, + "step": 12064 + }, + { + "epoch": 1.3249505820338239, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6628496646881104, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6980499029159546, + "num_tokens": 312145917.0, + "step": 12065 + }, + { + "epoch": 1.3250603997364374, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7731399536132812, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6935266852378845, + "num_tokens": 312167339.0, + "step": 12066 + }, + { + "epoch": 1.3251702174390512, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7137153148651123, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7150002121925354, + "num_tokens": 312189994.0, + "step": 12067 + }, + { + "epoch": 1.325280035141665, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6497340202331543, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7039806246757507, + "num_tokens": 312220602.0, + "step": 12068 + }, + { + "epoch": 1.3253898528442785, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.559684991836548, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7001940608024597, + "num_tokens": 312244048.0, + "step": 12069 + }, + { + "epoch": 1.3254996705468922, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.474557399749756, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7002732157707214, + "num_tokens": 312272462.0, + "step": 12070 + }, + { + "epoch": 1.3256094882495058, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.474989891052246, + "learning_rate": 1e-06, + "loss": 1.1299, + "mean_token_accuracy": 0.6829977035522461, + "num_tokens": 312304011.0, + "step": 12071 + }, + { + "epoch": 1.3257193059521195, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2454068660736084, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.6960240602493286, + "num_tokens": 312339018.0, + "step": 12072 + }, + { + "epoch": 1.3258291236547333, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.578256607055664, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6992576122283936, + "num_tokens": 312365181.0, + "step": 12073 + }, + { + "epoch": 1.3259389413573468, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1404592990875244, + "learning_rate": 1e-06, + "loss": 0.9073, + "mean_token_accuracy": 0.7300946712493896, + "num_tokens": 312380926.0, + "step": 12074 + }, + { + "epoch": 1.3260487590599603, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.534853458404541, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7135723233222961, + "num_tokens": 312408282.0, + "step": 12075 + }, + { + "epoch": 1.326158576762574, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.574895143508911, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7171703577041626, + "num_tokens": 312432842.0, + "step": 12076 + }, + { + "epoch": 1.3262683944651878, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.969935655593872, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7203245162963867, + "num_tokens": 312457932.0, + "step": 12077 + }, + { + "epoch": 1.3263782121678014, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5639805793762207, + "learning_rate": 1e-06, + "loss": 0.8849, + "mean_token_accuracy": 0.7381249666213989, + "num_tokens": 312482510.0, + "step": 12078 + }, + { + "epoch": 1.3264880298704151, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3406553268432617, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6834532022476196, + "num_tokens": 312522526.0, + "step": 12079 + }, + { + "epoch": 1.3265978475730287, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5454940795898438, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7009086608886719, + "num_tokens": 312547872.0, + "step": 12080 + }, + { + "epoch": 1.3267076652756424, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7537338733673096, + "learning_rate": 1e-06, + "loss": 0.8836, + "mean_token_accuracy": 0.7352256774902344, + "num_tokens": 312567048.0, + "step": 12081 + }, + { + "epoch": 1.3268174829782562, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6348068714141846, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7062934637069702, + "num_tokens": 312590450.0, + "step": 12082 + }, + { + "epoch": 1.3269273006808697, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3925588130950928, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6909387111663818, + "num_tokens": 312621839.0, + "step": 12083 + }, + { + "epoch": 1.3270371183834835, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4676177501678467, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6877481937408447, + "num_tokens": 312649493.0, + "step": 12084 + }, + { + "epoch": 1.327146936086097, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5754926204681396, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6969701051712036, + "num_tokens": 312678801.0, + "step": 12085 + }, + { + "epoch": 1.3272567537887108, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5257253646850586, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7189013361930847, + "num_tokens": 312702894.0, + "step": 12086 + }, + { + "epoch": 1.3273665714913245, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.62038516998291, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7054746150970459, + "num_tokens": 312726396.0, + "step": 12087 + }, + { + "epoch": 1.327476389193938, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3457119464874268, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6933494210243225, + "num_tokens": 312755864.0, + "step": 12088 + }, + { + "epoch": 1.3275862068965516, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.805263042449951, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.68336021900177, + "num_tokens": 312780319.0, + "step": 12089 + }, + { + "epoch": 1.3276960245991654, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8215391635894775, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7026302218437195, + "num_tokens": 312808674.0, + "step": 12090 + }, + { + "epoch": 1.3278058423017791, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.547374963760376, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6915889978408813, + "num_tokens": 312834242.0, + "step": 12091 + }, + { + "epoch": 1.3279156600043927, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.7641921043395996, + "learning_rate": 1e-06, + "loss": 0.935, + "mean_token_accuracy": 0.7195098996162415, + "num_tokens": 312856847.0, + "step": 12092 + }, + { + "epoch": 1.3280254777070064, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.482900619506836, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6913875341415405, + "num_tokens": 312885152.0, + "step": 12093 + }, + { + "epoch": 1.32813529540962, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.78482985496521, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6999104022979736, + "num_tokens": 312907134.0, + "step": 12094 + }, + { + "epoch": 1.3282451131122337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8443055152893066, + "learning_rate": 1e-06, + "loss": 0.9048, + "mean_token_accuracy": 0.7277601957321167, + "num_tokens": 312929375.0, + "step": 12095 + }, + { + "epoch": 1.3283549308148475, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.466806411743164, + "learning_rate": 1e-06, + "loss": 1.1097, + "mean_token_accuracy": 0.6758790612220764, + "num_tokens": 312959592.0, + "step": 12096 + }, + { + "epoch": 1.328464748517461, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5068118572235107, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7178798913955688, + "num_tokens": 312984791.0, + "step": 12097 + }, + { + "epoch": 1.3285745662200747, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5693366527557373, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7236737012863159, + "num_tokens": 313008554.0, + "step": 12098 + }, + { + "epoch": 1.3286843839226883, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.519289493560791, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.6995906829833984, + "num_tokens": 313036100.0, + "step": 12099 + }, + { + "epoch": 1.328794201625302, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.598342180252075, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7105448842048645, + "num_tokens": 313060742.0, + "step": 12100 + }, + { + "epoch": 1.3289040193279158, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.57383394241333, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.715922474861145, + "num_tokens": 313085469.0, + "step": 12101 + }, + { + "epoch": 1.3290138370305293, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.573723554611206, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6957362294197083, + "num_tokens": 313112793.0, + "step": 12102 + }, + { + "epoch": 1.3291236547331429, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.781987428665161, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.7242801785469055, + "num_tokens": 313134516.0, + "step": 12103 + }, + { + "epoch": 1.3292334724357566, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.450181007385254, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7025984525680542, + "num_tokens": 313163023.0, + "step": 12104 + }, + { + "epoch": 1.3293432901383704, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.370668411254883, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7104179859161377, + "num_tokens": 313196724.0, + "step": 12105 + }, + { + "epoch": 1.329453107840984, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.395627021789551, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.7037474513053894, + "num_tokens": 313227346.0, + "step": 12106 + }, + { + "epoch": 1.3295629255435977, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 80.5580825805664, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6965843439102173, + "num_tokens": 313255908.0, + "step": 12107 + }, + { + "epoch": 1.3296727432462112, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.1593780517578125, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7035152912139893, + "num_tokens": 313282227.0, + "step": 12108 + }, + { + "epoch": 1.329782560948825, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7881970405578613, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.718602180480957, + "num_tokens": 313303207.0, + "step": 12109 + }, + { + "epoch": 1.3298923786514387, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8265485763549805, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6930724382400513, + "num_tokens": 313325047.0, + "step": 12110 + }, + { + "epoch": 1.3300021963540523, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4615800380706787, + "learning_rate": 1e-06, + "loss": 1.131, + "mean_token_accuracy": 0.6783101558685303, + "num_tokens": 313356491.0, + "step": 12111 + }, + { + "epoch": 1.330112014056666, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2566192150115967, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6899036765098572, + "num_tokens": 313389447.0, + "step": 12112 + }, + { + "epoch": 1.3302218317592795, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7316298484802246, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7015095949172974, + "num_tokens": 313415681.0, + "step": 12113 + }, + { + "epoch": 1.3303316494618933, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.744072437286377, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7106919288635254, + "num_tokens": 313440534.0, + "step": 12114 + }, + { + "epoch": 1.3304414671645068, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.453343391418457, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7038564682006836, + "num_tokens": 313472238.0, + "step": 12115 + }, + { + "epoch": 1.3305512848671206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.743227481842041, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.6979889869689941, + "num_tokens": 313496901.0, + "step": 12116 + }, + { + "epoch": 1.3306611025697341, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6227972507476807, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.686270534992218, + "num_tokens": 313521690.0, + "step": 12117 + }, + { + "epoch": 1.330770920272348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.611565113067627, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.708800733089447, + "num_tokens": 313546813.0, + "step": 12118 + }, + { + "epoch": 1.3308807379749616, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.449848175048828, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7111630439758301, + "num_tokens": 313574997.0, + "step": 12119 + }, + { + "epoch": 1.3309905556775752, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3012537956237793, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6941866874694824, + "num_tokens": 313607098.0, + "step": 12120 + }, + { + "epoch": 1.331100373380189, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6880173683166504, + "learning_rate": 1e-06, + "loss": 0.9077, + "mean_token_accuracy": 0.7340059280395508, + "num_tokens": 313630093.0, + "step": 12121 + }, + { + "epoch": 1.3312101910828025, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.411450147628784, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7058209180831909, + "num_tokens": 313657207.0, + "step": 12122 + }, + { + "epoch": 1.3313200087854162, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.490462303161621, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.6991124153137207, + "num_tokens": 313686686.0, + "step": 12123 + }, + { + "epoch": 1.33142982648803, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3785271644592285, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6981047987937927, + "num_tokens": 313718020.0, + "step": 12124 + }, + { + "epoch": 1.3315396441906435, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.469578266143799, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7146351337432861, + "num_tokens": 313744954.0, + "step": 12125 + }, + { + "epoch": 1.331649461893257, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5985352993011475, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6907198429107666, + "num_tokens": 313767651.0, + "step": 12126 + }, + { + "epoch": 1.3317592795958708, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.539412498474121, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7083072662353516, + "num_tokens": 313793030.0, + "step": 12127 + }, + { + "epoch": 1.3318690972984846, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7106661796569824, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7029284834861755, + "num_tokens": 313819612.0, + "step": 12128 + }, + { + "epoch": 1.331978915001098, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.517282247543335, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7159171104431152, + "num_tokens": 313847069.0, + "step": 12129 + }, + { + "epoch": 1.3320887327037119, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.583065986633301, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7194288969039917, + "num_tokens": 313875458.0, + "step": 12130 + }, + { + "epoch": 1.3321985504063254, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8866374492645264, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7189497351646423, + "num_tokens": 313896527.0, + "step": 12131 + }, + { + "epoch": 1.3323083681089392, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.784095287322998, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.7032526135444641, + "num_tokens": 313918697.0, + "step": 12132 + }, + { + "epoch": 1.332418185811553, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4451749324798584, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7049371600151062, + "num_tokens": 313946077.0, + "step": 12133 + }, + { + "epoch": 1.3325280035141664, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.4758100509643555, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.6976215839385986, + "num_tokens": 313972542.0, + "step": 12134 + }, + { + "epoch": 1.3326378212167802, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4940195083618164, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7125997543334961, + "num_tokens": 314003809.0, + "step": 12135 + }, + { + "epoch": 1.3327476389193937, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.670318126678467, + "learning_rate": 1e-06, + "loss": 1.0943, + "mean_token_accuracy": 0.6818435192108154, + "num_tokens": 314030491.0, + "step": 12136 + }, + { + "epoch": 1.3328574566220075, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3088254928588867, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7119635343551636, + "num_tokens": 314062339.0, + "step": 12137 + }, + { + "epoch": 1.3329672743246213, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3384897708892822, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7199486494064331, + "num_tokens": 314089700.0, + "step": 12138 + }, + { + "epoch": 1.3330770920272348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4886703491210938, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6935703158378601, + "num_tokens": 314120020.0, + "step": 12139 + }, + { + "epoch": 1.3331869097298483, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.071014881134033, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7244613766670227, + "num_tokens": 314143549.0, + "step": 12140 + }, + { + "epoch": 1.333296727432462, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.414336681365967, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7082013487815857, + "num_tokens": 314173193.0, + "step": 12141 + }, + { + "epoch": 1.3334065451350758, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.9079442024230957, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7020633816719055, + "num_tokens": 314200701.0, + "step": 12142 + }, + { + "epoch": 1.3335163628376894, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.539299726486206, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7181089520454407, + "num_tokens": 314224618.0, + "step": 12143 + }, + { + "epoch": 1.3336261805403031, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4584813117980957, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6895349621772766, + "num_tokens": 314254988.0, + "step": 12144 + }, + { + "epoch": 1.3337359982429167, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4670016765594482, + "learning_rate": 1e-06, + "loss": 1.0796, + "mean_token_accuracy": 0.692912220954895, + "num_tokens": 314282456.0, + "step": 12145 + }, + { + "epoch": 1.3338458159455304, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.682039260864258, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6875383257865906, + "num_tokens": 314305152.0, + "step": 12146 + }, + { + "epoch": 1.3339556336481442, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.534492254257202, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7009989023208618, + "num_tokens": 314333641.0, + "step": 12147 + }, + { + "epoch": 1.3340654513507577, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.40226411819458, + "learning_rate": 1e-06, + "loss": 1.0779, + "mean_token_accuracy": 0.6889487504959106, + "num_tokens": 314368978.0, + "step": 12148 + }, + { + "epoch": 1.3341752690533715, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.441288471221924, + "learning_rate": 1e-06, + "loss": 0.9308, + "mean_token_accuracy": 0.7187580466270447, + "num_tokens": 314394530.0, + "step": 12149 + }, + { + "epoch": 1.334285086755985, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.66599178314209, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7035333514213562, + "num_tokens": 314417140.0, + "step": 12150 + }, + { + "epoch": 1.3343949044585988, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5540101528167725, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6994321346282959, + "num_tokens": 314440596.0, + "step": 12151 + }, + { + "epoch": 1.3345047221612125, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6102912425994873, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7160179615020752, + "num_tokens": 314462690.0, + "step": 12152 + }, + { + "epoch": 1.334614539863826, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.89762806892395, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6963040232658386, + "num_tokens": 314485297.0, + "step": 12153 + }, + { + "epoch": 1.3347243575664396, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8669302463531494, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7054020762443542, + "num_tokens": 314511086.0, + "step": 12154 + }, + { + "epoch": 1.3348341752690533, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.631481885910034, + "learning_rate": 1e-06, + "loss": 0.9291, + "mean_token_accuracy": 0.7300461530685425, + "num_tokens": 314534220.0, + "step": 12155 + }, + { + "epoch": 1.334943992971667, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5360047817230225, + "learning_rate": 1e-06, + "loss": 1.082, + "mean_token_accuracy": 0.6931279897689819, + "num_tokens": 314559951.0, + "step": 12156 + }, + { + "epoch": 1.3350538106742806, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.619612216949463, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7079409956932068, + "num_tokens": 314584385.0, + "step": 12157 + }, + { + "epoch": 1.3351636283768944, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.668940782546997, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7201204299926758, + "num_tokens": 314608107.0, + "step": 12158 + }, + { + "epoch": 1.335273446079508, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.478954553604126, + "learning_rate": 1e-06, + "loss": 1.1261, + "mean_token_accuracy": 0.6741916537284851, + "num_tokens": 314636606.0, + "step": 12159 + }, + { + "epoch": 1.3353832637821217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.661036729812622, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7141298055648804, + "num_tokens": 314660635.0, + "step": 12160 + }, + { + "epoch": 1.3354930814847354, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7561495304107666, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7145565748214722, + "num_tokens": 314687780.0, + "step": 12161 + }, + { + "epoch": 1.335602899187349, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.646308183670044, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.719756007194519, + "num_tokens": 314713069.0, + "step": 12162 + }, + { + "epoch": 1.3357127168899627, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.521731376647949, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7035216093063354, + "num_tokens": 314737852.0, + "step": 12163 + }, + { + "epoch": 1.3358225345925763, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4559688568115234, + "learning_rate": 1e-06, + "loss": 1.1264, + "mean_token_accuracy": 0.6782745718955994, + "num_tokens": 314764199.0, + "step": 12164 + }, + { + "epoch": 1.33593235229519, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.635870933532715, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7146269083023071, + "num_tokens": 314788466.0, + "step": 12165 + }, + { + "epoch": 1.3360421699978036, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.554590940475464, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6954302787780762, + "num_tokens": 314814228.0, + "step": 12166 + }, + { + "epoch": 1.3361519877004173, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3171279430389404, + "learning_rate": 1e-06, + "loss": 1.1324, + "mean_token_accuracy": 0.6740641593933105, + "num_tokens": 314847990.0, + "step": 12167 + }, + { + "epoch": 1.3362618054030309, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6704092025756836, + "learning_rate": 1e-06, + "loss": 0.9122, + "mean_token_accuracy": 0.7261365652084351, + "num_tokens": 314871778.0, + "step": 12168 + }, + { + "epoch": 1.3363716231056446, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7823832035064697, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7045576572418213, + "num_tokens": 314894790.0, + "step": 12169 + }, + { + "epoch": 1.3364814408082584, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.619342803955078, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7033404111862183, + "num_tokens": 314918890.0, + "step": 12170 + }, + { + "epoch": 1.336591258510872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6381280422210693, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7164751887321472, + "num_tokens": 314944008.0, + "step": 12171 + }, + { + "epoch": 1.3367010762134857, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.285527467727661, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6905015707015991, + "num_tokens": 314979013.0, + "step": 12172 + }, + { + "epoch": 1.3368108939160992, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4926881790161133, + "learning_rate": 1e-06, + "loss": 0.9162, + "mean_token_accuracy": 0.7308337688446045, + "num_tokens": 315003209.0, + "step": 12173 + }, + { + "epoch": 1.336920711618713, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4830334186553955, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7243874073028564, + "num_tokens": 315029803.0, + "step": 12174 + }, + { + "epoch": 1.3370305293213267, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.688281536102295, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6990800499916077, + "num_tokens": 315058104.0, + "step": 12175 + }, + { + "epoch": 1.3371403470239402, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6620705127716064, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.720928430557251, + "num_tokens": 315080546.0, + "step": 12176 + }, + { + "epoch": 1.337250164726554, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6501171588897705, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7135375738143921, + "num_tokens": 315105139.0, + "step": 12177 + }, + { + "epoch": 1.3373599824291675, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.683788776397705, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6970531940460205, + "num_tokens": 315128254.0, + "step": 12178 + }, + { + "epoch": 1.3374698001317813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.499565362930298, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.703583836555481, + "num_tokens": 315151826.0, + "step": 12179 + }, + { + "epoch": 1.3375796178343948, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.424067735671997, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6995863318443298, + "num_tokens": 315180659.0, + "step": 12180 + }, + { + "epoch": 1.3376894355370086, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4001033306121826, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7063939571380615, + "num_tokens": 315208419.0, + "step": 12181 + }, + { + "epoch": 1.3377992532396221, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5095391273498535, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7136670351028442, + "num_tokens": 315233736.0, + "step": 12182 + }, + { + "epoch": 1.3379090709422359, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5944933891296387, + "learning_rate": 1e-06, + "loss": 0.9013, + "mean_token_accuracy": 0.7349931001663208, + "num_tokens": 315258558.0, + "step": 12183 + }, + { + "epoch": 1.3380188886448496, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.590787649154663, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7009778022766113, + "num_tokens": 315284503.0, + "step": 12184 + }, + { + "epoch": 1.3381287063474632, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4740989208221436, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7005033493041992, + "num_tokens": 315314227.0, + "step": 12185 + }, + { + "epoch": 1.338238524050077, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.416184902191162, + "learning_rate": 1e-06, + "loss": 0.9358, + "mean_token_accuracy": 0.7319068908691406, + "num_tokens": 315341861.0, + "step": 12186 + }, + { + "epoch": 1.3383483417526905, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8565618991851807, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.722977876663208, + "num_tokens": 315362198.0, + "step": 12187 + }, + { + "epoch": 1.3384581594553042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7314236164093018, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7008854150772095, + "num_tokens": 315384800.0, + "step": 12188 + }, + { + "epoch": 1.338567977157918, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6714272499084473, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.709237813949585, + "num_tokens": 315412593.0, + "step": 12189 + }, + { + "epoch": 1.3386777948605315, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.80210280418396, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7059054374694824, + "num_tokens": 315437974.0, + "step": 12190 + }, + { + "epoch": 1.338787612563145, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5141303539276123, + "learning_rate": 1e-06, + "loss": 1.0999, + "mean_token_accuracy": 0.678032398223877, + "num_tokens": 315466659.0, + "step": 12191 + }, + { + "epoch": 1.3388974302657588, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3135221004486084, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7177013158798218, + "num_tokens": 315498402.0, + "step": 12192 + }, + { + "epoch": 1.3390072479683726, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7848780155181885, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7144933342933655, + "num_tokens": 315521133.0, + "step": 12193 + }, + { + "epoch": 1.339117065670986, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.568718671798706, + "learning_rate": 1e-06, + "loss": 0.9319, + "mean_token_accuracy": 0.7268550992012024, + "num_tokens": 315545156.0, + "step": 12194 + }, + { + "epoch": 1.3392268833735999, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4888317584991455, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7050151228904724, + "num_tokens": 315571491.0, + "step": 12195 + }, + { + "epoch": 1.3393367010762134, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.477973222732544, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7155934572219849, + "num_tokens": 315597388.0, + "step": 12196 + }, + { + "epoch": 1.3394465187788271, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9438064098358154, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7136648893356323, + "num_tokens": 315618018.0, + "step": 12197 + }, + { + "epoch": 1.339556336481441, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.550215244293213, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7163540124893188, + "num_tokens": 315644214.0, + "step": 12198 + }, + { + "epoch": 1.3396661541840544, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.473078727722168, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7142853736877441, + "num_tokens": 315670775.0, + "step": 12199 + }, + { + "epoch": 1.3397759718866682, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.769777536392212, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7112372517585754, + "num_tokens": 315692463.0, + "step": 12200 + }, + { + "epoch": 1.3398857895892817, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5491204261779785, + "learning_rate": 1e-06, + "loss": 1.1138, + "mean_token_accuracy": 0.6797956824302673, + "num_tokens": 315719338.0, + "step": 12201 + }, + { + "epoch": 1.3399956072918955, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5464107990264893, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6819623112678528, + "num_tokens": 315748717.0, + "step": 12202 + }, + { + "epoch": 1.3401054249945092, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6210227012634277, + "learning_rate": 1e-06, + "loss": 0.873, + "mean_token_accuracy": 0.7364907264709473, + "num_tokens": 315771304.0, + "step": 12203 + }, + { + "epoch": 1.3402152426971228, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.540423631668091, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.6997432112693787, + "num_tokens": 315801733.0, + "step": 12204 + }, + { + "epoch": 1.3403250603997363, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7081453800201416, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7006685733795166, + "num_tokens": 315825516.0, + "step": 12205 + }, + { + "epoch": 1.34043487810235, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4920411109924316, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6952176094055176, + "num_tokens": 315854523.0, + "step": 12206 + }, + { + "epoch": 1.3405446958049638, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 4.067249774932861, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7092810869216919, + "num_tokens": 315881088.0, + "step": 12207 + }, + { + "epoch": 1.3406545135075774, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3453369140625, + "learning_rate": 1e-06, + "loss": 1.1574, + "mean_token_accuracy": 0.6712191700935364, + "num_tokens": 315913033.0, + "step": 12208 + }, + { + "epoch": 1.3407643312101911, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5001060962677, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6958820223808289, + "num_tokens": 315939853.0, + "step": 12209 + }, + { + "epoch": 1.3408741489128047, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.499884605407715, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.7000635862350464, + "num_tokens": 315964538.0, + "step": 12210 + }, + { + "epoch": 1.3409839666154184, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.36289381980896, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6958749890327454, + "num_tokens": 316001696.0, + "step": 12211 + }, + { + "epoch": 1.3410937843180322, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3992629051208496, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7021993398666382, + "num_tokens": 316027487.0, + "step": 12212 + }, + { + "epoch": 1.3412036020206457, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.438650131225586, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.714405357837677, + "num_tokens": 316052502.0, + "step": 12213 + }, + { + "epoch": 1.3413134197232595, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6083076000213623, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7267656326293945, + "num_tokens": 316074290.0, + "step": 12214 + }, + { + "epoch": 1.341423237425873, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.870664596557617, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7044309973716736, + "num_tokens": 316095713.0, + "step": 12215 + }, + { + "epoch": 1.3415330551284868, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.867321491241455, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6952863931655884, + "num_tokens": 316119090.0, + "step": 12216 + }, + { + "epoch": 1.3416428728311005, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.407618284225464, + "learning_rate": 1e-06, + "loss": 1.1634, + "mean_token_accuracy": 0.6651601791381836, + "num_tokens": 316151908.0, + "step": 12217 + }, + { + "epoch": 1.341752690533714, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.847026824951172, + "learning_rate": 1e-06, + "loss": 0.9196, + "mean_token_accuracy": 0.7274736166000366, + "num_tokens": 316173112.0, + "step": 12218 + }, + { + "epoch": 1.3418625082363276, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5867576599121094, + "learning_rate": 1e-06, + "loss": 0.9312, + "mean_token_accuracy": 0.7226578593254089, + "num_tokens": 316195721.0, + "step": 12219 + }, + { + "epoch": 1.3419723259389413, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5335354804992676, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7004826068878174, + "num_tokens": 316224989.0, + "step": 12220 + }, + { + "epoch": 1.342082143641555, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4281208515167236, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6961624622344971, + "num_tokens": 316259046.0, + "step": 12221 + }, + { + "epoch": 1.3421919613441686, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7225639820098877, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7216334342956543, + "num_tokens": 316281973.0, + "step": 12222 + }, + { + "epoch": 1.3423017790467824, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6823599338531494, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6985271573066711, + "num_tokens": 316307353.0, + "step": 12223 + }, + { + "epoch": 1.342411596749396, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.669236898422241, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7031884789466858, + "num_tokens": 316332042.0, + "step": 12224 + }, + { + "epoch": 1.3425214144520097, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.11785626411438, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7131137251853943, + "num_tokens": 316361949.0, + "step": 12225 + }, + { + "epoch": 1.3426312321546234, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5317907333374023, + "learning_rate": 1e-06, + "loss": 0.9439, + "mean_token_accuracy": 0.730195164680481, + "num_tokens": 316389388.0, + "step": 12226 + }, + { + "epoch": 1.342741049857237, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.845426082611084, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7137715220451355, + "num_tokens": 316411616.0, + "step": 12227 + }, + { + "epoch": 1.3428508675598507, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.656680107116699, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7121453881263733, + "num_tokens": 316436210.0, + "step": 12228 + }, + { + "epoch": 1.3429606852624643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6529531478881836, + "learning_rate": 1e-06, + "loss": 1.1035, + "mean_token_accuracy": 0.687415599822998, + "num_tokens": 316460338.0, + "step": 12229 + }, + { + "epoch": 1.343070502965078, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.226335287094116, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7100096344947815, + "num_tokens": 316492222.0, + "step": 12230 + }, + { + "epoch": 1.3431803206676916, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.443817138671875, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7179701328277588, + "num_tokens": 316520060.0, + "step": 12231 + }, + { + "epoch": 1.3432901383703053, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.1481471061706543, + "learning_rate": 1e-06, + "loss": 1.1081, + "mean_token_accuracy": 0.6818249225616455, + "num_tokens": 316557495.0, + "step": 12232 + }, + { + "epoch": 1.3433999560729188, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.68208646774292, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6989870071411133, + "num_tokens": 316580479.0, + "step": 12233 + }, + { + "epoch": 1.3435097737755326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6071934700012207, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7116168737411499, + "num_tokens": 316604763.0, + "step": 12234 + }, + { + "epoch": 1.3436195914781464, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.548100471496582, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7274596691131592, + "num_tokens": 316628819.0, + "step": 12235 + }, + { + "epoch": 1.34372940918076, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.776573657989502, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7119203209877014, + "num_tokens": 316649717.0, + "step": 12236 + }, + { + "epoch": 1.3438392268833736, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5499091148376465, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.7041639089584351, + "num_tokens": 316673641.0, + "step": 12237 + }, + { + "epoch": 1.3439490445859872, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.5464468002319336, + "learning_rate": 1e-06, + "loss": 1.0466, + "mean_token_accuracy": 0.7005571722984314, + "num_tokens": 316703302.0, + "step": 12238 + }, + { + "epoch": 1.344058862288601, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5041141510009766, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7073752880096436, + "num_tokens": 316731974.0, + "step": 12239 + }, + { + "epoch": 1.3441686799912147, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.714724063873291, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7136715054512024, + "num_tokens": 316753735.0, + "step": 12240 + }, + { + "epoch": 1.3442784976938282, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.719346284866333, + "learning_rate": 1e-06, + "loss": 0.937, + "mean_token_accuracy": 0.7210757732391357, + "num_tokens": 316777376.0, + "step": 12241 + }, + { + "epoch": 1.3443883153964418, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.54109263420105, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7261308431625366, + "num_tokens": 316803200.0, + "step": 12242 + }, + { + "epoch": 1.3444981330990555, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.621861457824707, + "learning_rate": 1e-06, + "loss": 1.1574, + "mean_token_accuracy": 0.6619400978088379, + "num_tokens": 316830502.0, + "step": 12243 + }, + { + "epoch": 1.3446079508016693, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4548423290252686, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7151565551757812, + "num_tokens": 316858731.0, + "step": 12244 + }, + { + "epoch": 1.3447177685042828, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2881240844726562, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.6916124820709229, + "num_tokens": 316890304.0, + "step": 12245 + }, + { + "epoch": 1.3448275862068966, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5178163051605225, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7113451361656189, + "num_tokens": 316914171.0, + "step": 12246 + }, + { + "epoch": 1.34493740390951, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.395383834838867, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7151938676834106, + "num_tokens": 316945920.0, + "step": 12247 + }, + { + "epoch": 1.3450472216121239, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9142842292785645, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7183847427368164, + "num_tokens": 316967233.0, + "step": 12248 + }, + { + "epoch": 1.3451570393147376, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4035656452178955, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6972114443778992, + "num_tokens": 316996454.0, + "step": 12249 + }, + { + "epoch": 1.3452668570173512, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.6371877193450928, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7022610306739807, + "num_tokens": 317021126.0, + "step": 12250 + }, + { + "epoch": 1.345376674719965, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7181060314178467, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6912674903869629, + "num_tokens": 317049288.0, + "step": 12251 + }, + { + "epoch": 1.3454864924225785, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6554572582244873, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6939850449562073, + "num_tokens": 317075763.0, + "step": 12252 + }, + { + "epoch": 1.3455963101251922, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.849609375, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.7134801149368286, + "num_tokens": 317105490.0, + "step": 12253 + }, + { + "epoch": 1.345706127827806, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7697083950042725, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7107950448989868, + "num_tokens": 317128532.0, + "step": 12254 + }, + { + "epoch": 1.3458159455304195, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.787119150161743, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7131010293960571, + "num_tokens": 317149149.0, + "step": 12255 + }, + { + "epoch": 1.345925763233033, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.342097520828247, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.715936541557312, + "num_tokens": 317179995.0, + "step": 12256 + }, + { + "epoch": 1.3460355809356468, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4500527381896973, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7089322209358215, + "num_tokens": 317208166.0, + "step": 12257 + }, + { + "epoch": 1.3461453986382605, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5456948280334473, + "learning_rate": 1e-06, + "loss": 1.1104, + "mean_token_accuracy": 0.678749680519104, + "num_tokens": 317235731.0, + "step": 12258 + }, + { + "epoch": 1.346255216340874, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6291749477386475, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7192445397377014, + "num_tokens": 317259161.0, + "step": 12259 + }, + { + "epoch": 1.3463650340434878, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9411630630493164, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7015882730484009, + "num_tokens": 317280911.0, + "step": 12260 + }, + { + "epoch": 1.3464748517461014, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6583120822906494, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7061542272567749, + "num_tokens": 317307150.0, + "step": 12261 + }, + { + "epoch": 1.3465846694487151, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.395303726196289, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7063103914260864, + "num_tokens": 317336466.0, + "step": 12262 + }, + { + "epoch": 1.346694487151329, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8039979934692383, + "learning_rate": 1e-06, + "loss": 0.9314, + "mean_token_accuracy": 0.7303255796432495, + "num_tokens": 317358547.0, + "step": 12263 + }, + { + "epoch": 1.3468043048539424, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.1507718563079834, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7157118320465088, + "num_tokens": 317395401.0, + "step": 12264 + }, + { + "epoch": 1.3469141225565562, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5534939765930176, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7059472799301147, + "num_tokens": 317423197.0, + "step": 12265 + }, + { + "epoch": 1.3470239402591697, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.766585111618042, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6959853172302246, + "num_tokens": 317446824.0, + "step": 12266 + }, + { + "epoch": 1.3471337579617835, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7194783687591553, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7076235413551331, + "num_tokens": 317471013.0, + "step": 12267 + }, + { + "epoch": 1.3472435756643972, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.571423053741455, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7233148813247681, + "num_tokens": 317494213.0, + "step": 12268 + }, + { + "epoch": 1.3473533933670108, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.421408176422119, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7109360694885254, + "num_tokens": 317521586.0, + "step": 12269 + }, + { + "epoch": 1.3474632110696243, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4840619564056396, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7118065357208252, + "num_tokens": 317550203.0, + "step": 12270 + }, + { + "epoch": 1.347573028772238, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4216978549957275, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7139031887054443, + "num_tokens": 317577786.0, + "step": 12271 + }, + { + "epoch": 1.3476828464748518, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5141069889068604, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6912570595741272, + "num_tokens": 317605820.0, + "step": 12272 + }, + { + "epoch": 1.3477926641774653, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4085731506347656, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6814234852790833, + "num_tokens": 317632019.0, + "step": 12273 + }, + { + "epoch": 1.347902481880079, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4198694229125977, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7018032073974609, + "num_tokens": 317660099.0, + "step": 12274 + }, + { + "epoch": 1.3480122995826926, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5750572681427, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.6965539455413818, + "num_tokens": 317686365.0, + "step": 12275 + }, + { + "epoch": 1.3481221172853064, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.765979766845703, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.709132194519043, + "num_tokens": 317707757.0, + "step": 12276 + }, + { + "epoch": 1.3482319349879202, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5908892154693604, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7287566065788269, + "num_tokens": 317731141.0, + "step": 12277 + }, + { + "epoch": 1.3483417526905337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5491108894348145, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7240428924560547, + "num_tokens": 317758618.0, + "step": 12278 + }, + { + "epoch": 1.3484515703931474, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7521557807922363, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7043890953063965, + "num_tokens": 317780188.0, + "step": 12279 + }, + { + "epoch": 1.348561388095761, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5556929111480713, + "learning_rate": 1e-06, + "loss": 0.9205, + "mean_token_accuracy": 0.7313023805618286, + "num_tokens": 317803852.0, + "step": 12280 + }, + { + "epoch": 1.3486712057983747, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5014398097991943, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6998672485351562, + "num_tokens": 317834475.0, + "step": 12281 + }, + { + "epoch": 1.3487810235009885, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6164588928222656, + "learning_rate": 1e-06, + "loss": 0.9177, + "mean_token_accuracy": 0.7287353277206421, + "num_tokens": 317860962.0, + "step": 12282 + }, + { + "epoch": 1.348890841203602, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.604738473892212, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7084729671478271, + "num_tokens": 317887167.0, + "step": 12283 + }, + { + "epoch": 1.3490006589062156, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.981363534927368, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7247396111488342, + "num_tokens": 317907342.0, + "step": 12284 + }, + { + "epoch": 1.3491104766088293, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.928602933883667, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7039012908935547, + "num_tokens": 317929542.0, + "step": 12285 + }, + { + "epoch": 1.349220294311443, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.88273549079895, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7016388177871704, + "num_tokens": 317950297.0, + "step": 12286 + }, + { + "epoch": 1.3493301120140566, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5337557792663574, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.6920228004455566, + "num_tokens": 317978347.0, + "step": 12287 + }, + { + "epoch": 1.3494399297166704, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9352269172668457, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7229785919189453, + "num_tokens": 318003315.0, + "step": 12288 + }, + { + "epoch": 1.349549747419284, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.729189395904541, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7166445851325989, + "num_tokens": 318027089.0, + "step": 12289 + }, + { + "epoch": 1.3496595651218977, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8255934715270996, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7113295793533325, + "num_tokens": 318049983.0, + "step": 12290 + }, + { + "epoch": 1.3497693828245114, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5502052307128906, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.7010918855667114, + "num_tokens": 318076446.0, + "step": 12291 + }, + { + "epoch": 1.349879200527125, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.807753562927246, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7119027376174927, + "num_tokens": 318099282.0, + "step": 12292 + }, + { + "epoch": 1.3499890182297387, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5730738639831543, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7014899253845215, + "num_tokens": 318125169.0, + "step": 12293 + }, + { + "epoch": 1.3500988359323522, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3527462482452393, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7293065190315247, + "num_tokens": 318157779.0, + "step": 12294 + }, + { + "epoch": 1.350208653634966, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.314023017883301, + "learning_rate": 1e-06, + "loss": 0.9117, + "mean_token_accuracy": 0.7343086004257202, + "num_tokens": 318187301.0, + "step": 12295 + }, + { + "epoch": 1.3503184713375795, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5307722091674805, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.6992155313491821, + "num_tokens": 318212475.0, + "step": 12296 + }, + { + "epoch": 1.3504282890401933, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4498684406280518, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6940242052078247, + "num_tokens": 318245545.0, + "step": 12297 + }, + { + "epoch": 1.3505381067428068, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5832674503326416, + "learning_rate": 1e-06, + "loss": 0.9227, + "mean_token_accuracy": 0.7190085649490356, + "num_tokens": 318268347.0, + "step": 12298 + }, + { + "epoch": 1.3506479244454206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3223912715911865, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7155966758728027, + "num_tokens": 318296588.0, + "step": 12299 + }, + { + "epoch": 1.3507577421480343, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.655331611633301, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7135838270187378, + "num_tokens": 318318387.0, + "step": 12300 + }, + { + "epoch": 1.3508675598506479, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5675201416015625, + "learning_rate": 1e-06, + "loss": 0.8767, + "mean_token_accuracy": 0.7431224584579468, + "num_tokens": 318340938.0, + "step": 12301 + }, + { + "epoch": 1.3509773775532616, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.905073881149292, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7111228108406067, + "num_tokens": 318362842.0, + "step": 12302 + }, + { + "epoch": 1.3510871952558752, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6900672912597656, + "learning_rate": 1e-06, + "loss": 0.9353, + "mean_token_accuracy": 0.7204974293708801, + "num_tokens": 318388176.0, + "step": 12303 + }, + { + "epoch": 1.351197012958489, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.743812084197998, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.7198022603988647, + "num_tokens": 318411027.0, + "step": 12304 + }, + { + "epoch": 1.3513068306611027, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7715888023376465, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7178868651390076, + "num_tokens": 318431454.0, + "step": 12305 + }, + { + "epoch": 1.3514166483637162, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.685177803039551, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6958662271499634, + "num_tokens": 318456081.0, + "step": 12306 + }, + { + "epoch": 1.3515264660663298, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.421375274658203, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7050610780715942, + "num_tokens": 318484595.0, + "step": 12307 + }, + { + "epoch": 1.3516362837689435, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7626469135284424, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.693657398223877, + "num_tokens": 318507267.0, + "step": 12308 + }, + { + "epoch": 1.3517461014715573, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.65761399269104, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7226104736328125, + "num_tokens": 318530460.0, + "step": 12309 + }, + { + "epoch": 1.3518559191741708, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5105371475219727, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.6999257802963257, + "num_tokens": 318557169.0, + "step": 12310 + }, + { + "epoch": 1.3519657368767846, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4285712242126465, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6945936679840088, + "num_tokens": 318585845.0, + "step": 12311 + }, + { + "epoch": 1.352075554579398, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7342422008514404, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6906265616416931, + "num_tokens": 318607694.0, + "step": 12312 + }, + { + "epoch": 1.3521853722820119, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6892309188842773, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7020730376243591, + "num_tokens": 318630886.0, + "step": 12313 + }, + { + "epoch": 1.3522951899846256, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8271913528442383, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7231553792953491, + "num_tokens": 318651318.0, + "step": 12314 + }, + { + "epoch": 1.3524050076872391, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4583799839019775, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6984209418296814, + "num_tokens": 318681369.0, + "step": 12315 + }, + { + "epoch": 1.352514825389853, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5587382316589355, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6972413063049316, + "num_tokens": 318708443.0, + "step": 12316 + }, + { + "epoch": 1.3526246430924664, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8428547382354736, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7102971076965332, + "num_tokens": 318728753.0, + "step": 12317 + }, + { + "epoch": 1.3527344607950802, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.718433141708374, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6987133026123047, + "num_tokens": 318752398.0, + "step": 12318 + }, + { + "epoch": 1.352844278497694, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7050609588623047, + "learning_rate": 1e-06, + "loss": 1.092, + "mean_token_accuracy": 0.6873412132263184, + "num_tokens": 318777523.0, + "step": 12319 + }, + { + "epoch": 1.3529540962003075, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.524507761001587, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7035070657730103, + "num_tokens": 318804092.0, + "step": 12320 + }, + { + "epoch": 1.353063913902921, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.582977533340454, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7156116962432861, + "num_tokens": 318830073.0, + "step": 12321 + }, + { + "epoch": 1.3531737316055348, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5315070152282715, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7048189640045166, + "num_tokens": 318855429.0, + "step": 12322 + }, + { + "epoch": 1.3532835493081485, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.724944829940796, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7214586734771729, + "num_tokens": 318878567.0, + "step": 12323 + }, + { + "epoch": 1.353393367010762, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5685884952545166, + "learning_rate": 1e-06, + "loss": 0.932, + "mean_token_accuracy": 0.7273346185684204, + "num_tokens": 318905612.0, + "step": 12324 + }, + { + "epoch": 1.3535031847133758, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.605273723602295, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.702217698097229, + "num_tokens": 318932808.0, + "step": 12325 + }, + { + "epoch": 1.3536130024159894, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6473281383514404, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.7001482248306274, + "num_tokens": 318958431.0, + "step": 12326 + }, + { + "epoch": 1.3537228201186031, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5866403579711914, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7217637300491333, + "num_tokens": 318981995.0, + "step": 12327 + }, + { + "epoch": 1.3538326378212169, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5002546310424805, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6843183040618896, + "num_tokens": 319011387.0, + "step": 12328 + }, + { + "epoch": 1.3539424555238304, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4071662425994873, + "learning_rate": 1e-06, + "loss": 1.1103, + "mean_token_accuracy": 0.6781555414199829, + "num_tokens": 319044796.0, + "step": 12329 + }, + { + "epoch": 1.3540522732264442, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.408334970474243, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7042150497436523, + "num_tokens": 319074717.0, + "step": 12330 + }, + { + "epoch": 1.3541620909290577, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6671032905578613, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6848926544189453, + "num_tokens": 319099557.0, + "step": 12331 + }, + { + "epoch": 1.3542719086316715, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8576345443725586, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7161282300949097, + "num_tokens": 319123101.0, + "step": 12332 + }, + { + "epoch": 1.3543817263342852, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1616225242614746, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7062791585922241, + "num_tokens": 319145466.0, + "step": 12333 + }, + { + "epoch": 1.3544915440368988, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2006382942199707, + "learning_rate": 1e-06, + "loss": 1.1349, + "mean_token_accuracy": 0.6770578622817993, + "num_tokens": 319184665.0, + "step": 12334 + }, + { + "epoch": 1.3546013617395123, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8062925338745117, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6884127855300903, + "num_tokens": 319205907.0, + "step": 12335 + }, + { + "epoch": 1.354711179442126, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8264307975769043, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7104924917221069, + "num_tokens": 319227066.0, + "step": 12336 + }, + { + "epoch": 1.3548209971447398, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6587350368499756, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6904462575912476, + "num_tokens": 319252101.0, + "step": 12337 + }, + { + "epoch": 1.3549308148473533, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7474617958068848, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7063251733779907, + "num_tokens": 319275052.0, + "step": 12338 + }, + { + "epoch": 1.355040632549967, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 16.71407127380371, + "learning_rate": 1e-06, + "loss": 0.8842, + "mean_token_accuracy": 0.7373062968254089, + "num_tokens": 319295313.0, + "step": 12339 + }, + { + "epoch": 1.3551504502525806, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5734829902648926, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6956665515899658, + "num_tokens": 319321658.0, + "step": 12340 + }, + { + "epoch": 1.3552602679551944, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7158212661743164, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.7078997492790222, + "num_tokens": 319344828.0, + "step": 12341 + }, + { + "epoch": 1.3553700856578081, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.653343915939331, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7048108577728271, + "num_tokens": 319372964.0, + "step": 12342 + }, + { + "epoch": 1.3554799033604217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.752584218978882, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7094627618789673, + "num_tokens": 319397019.0, + "step": 12343 + }, + { + "epoch": 1.3555897210630354, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.772972822189331, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6830581426620483, + "num_tokens": 319423150.0, + "step": 12344 + }, + { + "epoch": 1.355699538765649, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7517592906951904, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7290924787521362, + "num_tokens": 319444028.0, + "step": 12345 + }, + { + "epoch": 1.3558093564682627, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7238333225250244, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.6935561895370483, + "num_tokens": 319471726.0, + "step": 12346 + }, + { + "epoch": 1.3559191741708763, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.792872428894043, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6984206438064575, + "num_tokens": 319496920.0, + "step": 12347 + }, + { + "epoch": 1.35602899187349, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5176899433135986, + "learning_rate": 1e-06, + "loss": 0.9094, + "mean_token_accuracy": 0.7303909063339233, + "num_tokens": 319522394.0, + "step": 12348 + }, + { + "epoch": 1.3561388095761036, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.400193214416504, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7233213186264038, + "num_tokens": 319549366.0, + "step": 12349 + }, + { + "epoch": 1.3562486272787173, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5745503902435303, + "learning_rate": 1e-06, + "loss": 0.9552, + "mean_token_accuracy": 0.7278184294700623, + "num_tokens": 319573807.0, + "step": 12350 + }, + { + "epoch": 1.356358444981331, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.785175085067749, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7036249041557312, + "num_tokens": 319597163.0, + "step": 12351 + }, + { + "epoch": 1.3564682626839446, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.619572639465332, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7192544937133789, + "num_tokens": 319624851.0, + "step": 12352 + }, + { + "epoch": 1.3565780803865584, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.696782112121582, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6946245431900024, + "num_tokens": 319648659.0, + "step": 12353 + }, + { + "epoch": 1.356687898089172, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.946134567260742, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.701112687587738, + "num_tokens": 319668633.0, + "step": 12354 + }, + { + "epoch": 1.3567977157917857, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.44441819190979, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7029542326927185, + "num_tokens": 319697321.0, + "step": 12355 + }, + { + "epoch": 1.3569075334943994, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3977742195129395, + "learning_rate": 1e-06, + "loss": 1.1107, + "mean_token_accuracy": 0.6797198057174683, + "num_tokens": 319727386.0, + "step": 12356 + }, + { + "epoch": 1.357017351197013, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5139684677124023, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7074119448661804, + "num_tokens": 319751705.0, + "step": 12357 + }, + { + "epoch": 1.3571271688996267, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6077592372894287, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.716528058052063, + "num_tokens": 319777094.0, + "step": 12358 + }, + { + "epoch": 1.3572369866022402, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5083489418029785, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7037370800971985, + "num_tokens": 319806021.0, + "step": 12359 + }, + { + "epoch": 1.357346804304854, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.63291335105896, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6977387070655823, + "num_tokens": 319834341.0, + "step": 12360 + }, + { + "epoch": 1.3574566220074675, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6013903617858887, + "learning_rate": 1e-06, + "loss": 1.1205, + "mean_token_accuracy": 0.6731483936309814, + "num_tokens": 319863895.0, + "step": 12361 + }, + { + "epoch": 1.3575664397100813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.488797187805176, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6927371025085449, + "num_tokens": 319893533.0, + "step": 12362 + }, + { + "epoch": 1.3576762574126948, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.392955780029297, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.686755359172821, + "num_tokens": 319925237.0, + "step": 12363 + }, + { + "epoch": 1.3577860751153086, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4627747535705566, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.7002646327018738, + "num_tokens": 319953740.0, + "step": 12364 + }, + { + "epoch": 1.3578958928179223, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.791933536529541, + "learning_rate": 1e-06, + "loss": 0.8789, + "mean_token_accuracy": 0.7262855768203735, + "num_tokens": 319975003.0, + "step": 12365 + }, + { + "epoch": 1.3580057105205359, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3928661346435547, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7038857936859131, + "num_tokens": 320005624.0, + "step": 12366 + }, + { + "epoch": 1.3581155282231496, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6543052196502686, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7056896686553955, + "num_tokens": 320030292.0, + "step": 12367 + }, + { + "epoch": 1.3582253459257632, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.614697217941284, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6958726048469543, + "num_tokens": 320054812.0, + "step": 12368 + }, + { + "epoch": 1.358335163628377, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5288567543029785, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7044044733047485, + "num_tokens": 320083315.0, + "step": 12369 + }, + { + "epoch": 1.3584449813309907, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6044366359710693, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6994614005088806, + "num_tokens": 320107574.0, + "step": 12370 + }, + { + "epoch": 1.3585547990336042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.667633533477783, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6900047063827515, + "num_tokens": 320137930.0, + "step": 12371 + }, + { + "epoch": 1.3586646167362177, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8093438148498535, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.6998805999755859, + "num_tokens": 320160452.0, + "step": 12372 + }, + { + "epoch": 1.3587744344388315, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5088045597076416, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6939659118652344, + "num_tokens": 320186376.0, + "step": 12373 + }, + { + "epoch": 1.3588842521414453, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.616417407989502, + "learning_rate": 1e-06, + "loss": 1.1254, + "mean_token_accuracy": 0.6941487789154053, + "num_tokens": 320212668.0, + "step": 12374 + }, + { + "epoch": 1.3589940698440588, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.44120717048645, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7086585760116577, + "num_tokens": 320242462.0, + "step": 12375 + }, + { + "epoch": 1.3591038875466726, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4137797355651855, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7218140363693237, + "num_tokens": 320270761.0, + "step": 12376 + }, + { + "epoch": 1.359213705249286, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3991806507110596, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7184332609176636, + "num_tokens": 320299388.0, + "step": 12377 + }, + { + "epoch": 1.3593235229518998, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3602185249328613, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7117834091186523, + "num_tokens": 320329246.0, + "step": 12378 + }, + { + "epoch": 1.3594333406545136, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.429936408996582, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7039839029312134, + "num_tokens": 320357784.0, + "step": 12379 + }, + { + "epoch": 1.3595431583571271, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4006495475769043, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7077318429946899, + "num_tokens": 320388559.0, + "step": 12380 + }, + { + "epoch": 1.359652976059741, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.538547992706299, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.6950777769088745, + "num_tokens": 320413902.0, + "step": 12381 + }, + { + "epoch": 1.3597627937623544, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0444040298461914, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7057540416717529, + "num_tokens": 320433528.0, + "step": 12382 + }, + { + "epoch": 1.3598726114649682, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5559322834014893, + "learning_rate": 1e-06, + "loss": 1.0857, + "mean_token_accuracy": 0.6860798001289368, + "num_tokens": 320461910.0, + "step": 12383 + }, + { + "epoch": 1.359982429167582, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6242117881774902, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.7006499171257019, + "num_tokens": 320488243.0, + "step": 12384 + }, + { + "epoch": 1.3600922468701955, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.689605236053467, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6886522769927979, + "num_tokens": 320513492.0, + "step": 12385 + }, + { + "epoch": 1.360202064572809, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4081056118011475, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.691259503364563, + "num_tokens": 320541935.0, + "step": 12386 + }, + { + "epoch": 1.3603118822754228, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 16.780384063720703, + "learning_rate": 1e-06, + "loss": 0.8995, + "mean_token_accuracy": 0.7371568083763123, + "num_tokens": 320559125.0, + "step": 12387 + }, + { + "epoch": 1.3604216999780365, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7735323905944824, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7042537927627563, + "num_tokens": 320583098.0, + "step": 12388 + }, + { + "epoch": 1.36053151768065, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9275496006011963, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7059513926506042, + "num_tokens": 320603576.0, + "step": 12389 + }, + { + "epoch": 1.3606413353832638, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.0588483810424805, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6849559545516968, + "num_tokens": 320644237.0, + "step": 12390 + }, + { + "epoch": 1.3607511530858774, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3782835006713867, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.689014732837677, + "num_tokens": 320673042.0, + "step": 12391 + }, + { + "epoch": 1.360860970788491, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6174538135528564, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.7086886167526245, + "num_tokens": 320696520.0, + "step": 12392 + }, + { + "epoch": 1.3609707884911049, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.582228660583496, + "learning_rate": 1e-06, + "loss": 0.9141, + "mean_token_accuracy": 0.7269699573516846, + "num_tokens": 320719902.0, + "step": 12393 + }, + { + "epoch": 1.3610806061937184, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.388806104660034, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.7077581286430359, + "num_tokens": 320753425.0, + "step": 12394 + }, + { + "epoch": 1.3611904238963322, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.504096269607544, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7037582397460938, + "num_tokens": 320780438.0, + "step": 12395 + }, + { + "epoch": 1.3613002415989457, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.700056791305542, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7185994386672974, + "num_tokens": 320801567.0, + "step": 12396 + }, + { + "epoch": 1.3614100593015594, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.878877878189087, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7124010324478149, + "num_tokens": 320825226.0, + "step": 12397 + }, + { + "epoch": 1.3615198770041732, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.020395517349243, + "learning_rate": 1e-06, + "loss": 0.9295, + "mean_token_accuracy": 0.723666250705719, + "num_tokens": 320845317.0, + "step": 12398 + }, + { + "epoch": 1.3616296947067867, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.140096664428711, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7142187356948853, + "num_tokens": 320863226.0, + "step": 12399 + }, + { + "epoch": 1.3617395124094003, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.759481906890869, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7156959772109985, + "num_tokens": 320884355.0, + "step": 12400 + }, + { + "epoch": 1.361849330112014, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6581835746765137, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7072744369506836, + "num_tokens": 320909532.0, + "step": 12401 + }, + { + "epoch": 1.3619591478146278, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6882786750793457, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6982568502426147, + "num_tokens": 320937048.0, + "step": 12402 + }, + { + "epoch": 1.3620689655172413, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5690062046051025, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7095640301704407, + "num_tokens": 320960287.0, + "step": 12403 + }, + { + "epoch": 1.362178783219855, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.897167444229126, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6930322647094727, + "num_tokens": 320979821.0, + "step": 12404 + }, + { + "epoch": 1.3622886009224686, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.812701463699341, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6878707408905029, + "num_tokens": 321002289.0, + "step": 12405 + }, + { + "epoch": 1.3623984186250824, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.492097854614258, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6839993596076965, + "num_tokens": 321031301.0, + "step": 12406 + }, + { + "epoch": 1.3625082363276961, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.765031576156616, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6930366158485413, + "num_tokens": 321054918.0, + "step": 12407 + }, + { + "epoch": 1.3626180540303097, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5591323375701904, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7188771963119507, + "num_tokens": 321081261.0, + "step": 12408 + }, + { + "epoch": 1.3627278717329234, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.418332815170288, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7210370302200317, + "num_tokens": 321109810.0, + "step": 12409 + }, + { + "epoch": 1.362837689435537, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7685468196868896, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7095879316329956, + "num_tokens": 321132982.0, + "step": 12410 + }, + { + "epoch": 1.3629475071381507, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4700655937194824, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7135316133499146, + "num_tokens": 321162779.0, + "step": 12411 + }, + { + "epoch": 1.3630573248407643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8840725421905518, + "learning_rate": 1e-06, + "loss": 0.866, + "mean_token_accuracy": 0.743789553642273, + "num_tokens": 321181944.0, + "step": 12412 + }, + { + "epoch": 1.363167142543378, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.473677396774292, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6947047710418701, + "num_tokens": 321210380.0, + "step": 12413 + }, + { + "epoch": 1.3632769602459915, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5843615531921387, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7048225402832031, + "num_tokens": 321235617.0, + "step": 12414 + }, + { + "epoch": 1.3633867779486053, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.663522720336914, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7101714611053467, + "num_tokens": 321260493.0, + "step": 12415 + }, + { + "epoch": 1.363496595651219, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.917109489440918, + "learning_rate": 1e-06, + "loss": 0.9286, + "mean_token_accuracy": 0.7247034311294556, + "num_tokens": 321282094.0, + "step": 12416 + }, + { + "epoch": 1.3636064133538326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4665706157684326, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.698732852935791, + "num_tokens": 321313841.0, + "step": 12417 + }, + { + "epoch": 1.3637162310564463, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.818063735961914, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7099770307540894, + "num_tokens": 321336344.0, + "step": 12418 + }, + { + "epoch": 1.3638260487590599, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2260541915893555, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6896743178367615, + "num_tokens": 321369387.0, + "step": 12419 + }, + { + "epoch": 1.3639358664616736, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6533427238464355, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7147079110145569, + "num_tokens": 321395461.0, + "step": 12420 + }, + { + "epoch": 1.3640456841642874, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5943589210510254, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7179400324821472, + "num_tokens": 321420665.0, + "step": 12421 + }, + { + "epoch": 1.364155501866901, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7323720455169678, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.715178370475769, + "num_tokens": 321446602.0, + "step": 12422 + }, + { + "epoch": 1.3642653195695145, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8122661113739014, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6989343166351318, + "num_tokens": 321472037.0, + "step": 12423 + }, + { + "epoch": 1.3643751372721282, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.496276378631592, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7084293961524963, + "num_tokens": 321498650.0, + "step": 12424 + }, + { + "epoch": 1.364484954974742, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.657663106918335, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.6860394477844238, + "num_tokens": 321524056.0, + "step": 12425 + }, + { + "epoch": 1.3645947726773555, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.627105236053467, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6880083680152893, + "num_tokens": 321548443.0, + "step": 12426 + }, + { + "epoch": 1.3647045903799693, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7495415210723877, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6941677927970886, + "num_tokens": 321578386.0, + "step": 12427 + }, + { + "epoch": 1.3648144080825828, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3691887855529785, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7053804397583008, + "num_tokens": 321607798.0, + "step": 12428 + }, + { + "epoch": 1.3649242257851966, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.44230580329895, + "learning_rate": 1e-06, + "loss": 1.1301, + "mean_token_accuracy": 0.6766365766525269, + "num_tokens": 321637720.0, + "step": 12429 + }, + { + "epoch": 1.3650340434878103, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.873199462890625, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6849136352539062, + "num_tokens": 321665072.0, + "step": 12430 + }, + { + "epoch": 1.3651438611904239, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.546781301498413, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7146823406219482, + "num_tokens": 321690054.0, + "step": 12431 + }, + { + "epoch": 1.3652536788930376, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.274223804473877, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7151144742965698, + "num_tokens": 321722282.0, + "step": 12432 + }, + { + "epoch": 1.3653634965956511, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3653063774108887, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6844009160995483, + "num_tokens": 321754184.0, + "step": 12433 + }, + { + "epoch": 1.365473314298265, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9560673236846924, + "learning_rate": 1e-06, + "loss": 0.8527, + "mean_token_accuracy": 0.7466111183166504, + "num_tokens": 321774303.0, + "step": 12434 + }, + { + "epoch": 1.3655831320008787, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.537912368774414, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7063679695129395, + "num_tokens": 321799656.0, + "step": 12435 + }, + { + "epoch": 1.3656929497034922, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.072693109512329, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7039840817451477, + "num_tokens": 321818801.0, + "step": 12436 + }, + { + "epoch": 1.3658027674061057, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7309517860412598, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6942011117935181, + "num_tokens": 321844992.0, + "step": 12437 + }, + { + "epoch": 1.3659125851087195, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.2384192943573, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7090232372283936, + "num_tokens": 321866970.0, + "step": 12438 + }, + { + "epoch": 1.3660224028113332, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5514490604400635, + "learning_rate": 1e-06, + "loss": 0.8909, + "mean_token_accuracy": 0.7450801134109497, + "num_tokens": 321891675.0, + "step": 12439 + }, + { + "epoch": 1.3661322205139468, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3002052307128906, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7115511298179626, + "num_tokens": 321926398.0, + "step": 12440 + }, + { + "epoch": 1.3662420382165605, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.731459379196167, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7209652662277222, + "num_tokens": 321950293.0, + "step": 12441 + }, + { + "epoch": 1.366351855919174, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3986425399780273, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7105783820152283, + "num_tokens": 321979849.0, + "step": 12442 + }, + { + "epoch": 1.3664616736217878, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.735384702682495, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7080954909324646, + "num_tokens": 322001928.0, + "step": 12443 + }, + { + "epoch": 1.3665714913244016, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5019404888153076, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7122076153755188, + "num_tokens": 322029905.0, + "step": 12444 + }, + { + "epoch": 1.3666813090270151, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.649698495864868, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6994318962097168, + "num_tokens": 322055755.0, + "step": 12445 + }, + { + "epoch": 1.3667911267296289, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.744616746902466, + "learning_rate": 1e-06, + "loss": 1.0927, + "mean_token_accuracy": 0.6867490410804749, + "num_tokens": 322081185.0, + "step": 12446 + }, + { + "epoch": 1.3669009444322424, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.90432071685791, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7017898559570312, + "num_tokens": 322107051.0, + "step": 12447 + }, + { + "epoch": 1.3670107621348562, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7097482681274414, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7200392484664917, + "num_tokens": 322130687.0, + "step": 12448 + }, + { + "epoch": 1.36712057983747, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.380208730697632, + "learning_rate": 1e-06, + "loss": 1.1634, + "mean_token_accuracy": 0.6660994291305542, + "num_tokens": 322164725.0, + "step": 12449 + }, + { + "epoch": 1.3672303975400835, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.296085834503174, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7093083262443542, + "num_tokens": 322193166.0, + "step": 12450 + }, + { + "epoch": 1.367340215242697, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7791757583618164, + "learning_rate": 1e-06, + "loss": 0.8934, + "mean_token_accuracy": 0.7321522235870361, + "num_tokens": 322212005.0, + "step": 12451 + }, + { + "epoch": 1.3674500329453108, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5996906757354736, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6877684593200684, + "num_tokens": 322239553.0, + "step": 12452 + }, + { + "epoch": 1.3675598506479245, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.44925594329834, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7072569727897644, + "num_tokens": 322269202.0, + "step": 12453 + }, + { + "epoch": 1.367669668350538, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.560089349746704, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.691848635673523, + "num_tokens": 322297129.0, + "step": 12454 + }, + { + "epoch": 1.3677794860531518, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2158079147338867, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6923484206199646, + "num_tokens": 322331302.0, + "step": 12455 + }, + { + "epoch": 1.3678893037557653, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.529940605163574, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.7018049955368042, + "num_tokens": 322356109.0, + "step": 12456 + }, + { + "epoch": 1.367999121458379, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.568484306335449, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6838506460189819, + "num_tokens": 322382372.0, + "step": 12457 + }, + { + "epoch": 1.3681089391609929, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6015536785125732, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.7271007895469666, + "num_tokens": 322408429.0, + "step": 12458 + }, + { + "epoch": 1.3682187568636064, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6620442867279053, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7110174894332886, + "num_tokens": 322433048.0, + "step": 12459 + }, + { + "epoch": 1.3683285745662201, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5890181064605713, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7167301177978516, + "num_tokens": 322458545.0, + "step": 12460 + }, + { + "epoch": 1.3684383922688337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.5459115505218506, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7281116247177124, + "num_tokens": 322478183.0, + "step": 12461 + }, + { + "epoch": 1.3685482099714474, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9031567573547363, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7020870447158813, + "num_tokens": 322499421.0, + "step": 12462 + }, + { + "epoch": 1.3686580276740612, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.482024908065796, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7126668691635132, + "num_tokens": 322528571.0, + "step": 12463 + }, + { + "epoch": 1.3687678453766747, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6543850898742676, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6938062906265259, + "num_tokens": 322552969.0, + "step": 12464 + }, + { + "epoch": 1.3688776630792883, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6808278560638428, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7047092914581299, + "num_tokens": 322576081.0, + "step": 12465 + }, + { + "epoch": 1.368987480781902, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.498199462890625, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7049216032028198, + "num_tokens": 322605259.0, + "step": 12466 + }, + { + "epoch": 1.3690972984845158, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5298163890838623, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7157776355743408, + "num_tokens": 322632366.0, + "step": 12467 + }, + { + "epoch": 1.3692071161871293, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6812000274658203, + "learning_rate": 1e-06, + "loss": 1.114, + "mean_token_accuracy": 0.6734291315078735, + "num_tokens": 322658014.0, + "step": 12468 + }, + { + "epoch": 1.369316933889743, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.580472946166992, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.7079169154167175, + "num_tokens": 322685064.0, + "step": 12469 + }, + { + "epoch": 1.3694267515923566, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.515383720397949, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7155710458755493, + "num_tokens": 322711501.0, + "step": 12470 + }, + { + "epoch": 1.3695365692949704, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6197309494018555, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7201066017150879, + "num_tokens": 322738897.0, + "step": 12471 + }, + { + "epoch": 1.3696463869975841, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8728296756744385, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7153263092041016, + "num_tokens": 322762890.0, + "step": 12472 + }, + { + "epoch": 1.3697562047001977, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.563563108444214, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6951707005500793, + "num_tokens": 322788541.0, + "step": 12473 + }, + { + "epoch": 1.3698660224028114, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.803234577178955, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7011833190917969, + "num_tokens": 322812278.0, + "step": 12474 + }, + { + "epoch": 1.369975840105425, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3921525478363037, + "learning_rate": 1e-06, + "loss": 1.1493, + "mean_token_accuracy": 0.6758548617362976, + "num_tokens": 322844266.0, + "step": 12475 + }, + { + "epoch": 1.3700856578080387, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9227375984191895, + "learning_rate": 1e-06, + "loss": 1.1109, + "mean_token_accuracy": 0.6849514245986938, + "num_tokens": 322867518.0, + "step": 12476 + }, + { + "epoch": 1.3701954755106522, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.712247610092163, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7110240459442139, + "num_tokens": 322891856.0, + "step": 12477 + }, + { + "epoch": 1.370305293213266, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7752223014831543, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6862574219703674, + "num_tokens": 322916582.0, + "step": 12478 + }, + { + "epoch": 1.3704151109158795, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 16.745304107666016, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.703906774520874, + "num_tokens": 322938281.0, + "step": 12479 + }, + { + "epoch": 1.3705249286184933, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.833674907684326, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.7056012749671936, + "num_tokens": 322961516.0, + "step": 12480 + }, + { + "epoch": 1.370634746321107, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1025350093841553, + "learning_rate": 1e-06, + "loss": 0.8931, + "mean_token_accuracy": 0.7338336110115051, + "num_tokens": 322979819.0, + "step": 12481 + }, + { + "epoch": 1.3707445640237206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.305291175842285, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6950475573539734, + "num_tokens": 323013915.0, + "step": 12482 + }, + { + "epoch": 1.3708543817263343, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.716710329055786, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7252175807952881, + "num_tokens": 323036826.0, + "step": 12483 + }, + { + "epoch": 1.3709641994289479, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2671055793762207, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6947982311248779, + "num_tokens": 323067041.0, + "step": 12484 + }, + { + "epoch": 1.3710740171315616, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7089126110076904, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.70456463098526, + "num_tokens": 323090266.0, + "step": 12485 + }, + { + "epoch": 1.3711838348341754, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6046526432037354, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6899682879447937, + "num_tokens": 323116084.0, + "step": 12486 + }, + { + "epoch": 1.371293652536789, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3335933685302734, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7091660499572754, + "num_tokens": 323145061.0, + "step": 12487 + }, + { + "epoch": 1.3714034702394025, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8383188247680664, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7092896699905396, + "num_tokens": 323167555.0, + "step": 12488 + }, + { + "epoch": 1.3715132879420162, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8302438259124756, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7175481915473938, + "num_tokens": 323189204.0, + "step": 12489 + }, + { + "epoch": 1.37162310564463, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.546664237976074, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7187706232070923, + "num_tokens": 323212801.0, + "step": 12490 + }, + { + "epoch": 1.3717329233472435, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.450521945953369, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7181274890899658, + "num_tokens": 323239307.0, + "step": 12491 + }, + { + "epoch": 1.3718427410498573, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.59659481048584, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6981017589569092, + "num_tokens": 323264180.0, + "step": 12492 + }, + { + "epoch": 1.3719525587524708, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6060903072357178, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6940649151802063, + "num_tokens": 323292103.0, + "step": 12493 + }, + { + "epoch": 1.3720623764550846, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6298258304595947, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.702674150466919, + "num_tokens": 323315127.0, + "step": 12494 + }, + { + "epoch": 1.3721721941576983, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6533923149108887, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6847989559173584, + "num_tokens": 323340495.0, + "step": 12495 + }, + { + "epoch": 1.3722820118603118, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.709035873413086, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6854156255722046, + "num_tokens": 323366720.0, + "step": 12496 + }, + { + "epoch": 1.3723918295629256, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7330820560455322, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7156621217727661, + "num_tokens": 323390323.0, + "step": 12497 + }, + { + "epoch": 1.3725016472655391, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6587674617767334, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7154862880706787, + "num_tokens": 323412899.0, + "step": 12498 + }, + { + "epoch": 1.372611464968153, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0927655696868896, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7199347019195557, + "num_tokens": 323430627.0, + "step": 12499 + }, + { + "epoch": 1.3727212826707667, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.866568088531494, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.7220916152000427, + "num_tokens": 323451115.0, + "step": 12500 + }, + { + "epoch": 1.3728311003733802, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6065585613250732, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7012244462966919, + "num_tokens": 323477932.0, + "step": 12501 + }, + { + "epoch": 1.3729409180759937, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5939908027648926, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.6961162090301514, + "num_tokens": 323507554.0, + "step": 12502 + }, + { + "epoch": 1.3730507357786075, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4855751991271973, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6929545402526855, + "num_tokens": 323536096.0, + "step": 12503 + }, + { + "epoch": 1.3731605534812212, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.322120189666748, + "learning_rate": 1e-06, + "loss": 1.1182, + "mean_token_accuracy": 0.674393892288208, + "num_tokens": 323568211.0, + "step": 12504 + }, + { + "epoch": 1.3732703711838348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4246926307678223, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6906625032424927, + "num_tokens": 323597033.0, + "step": 12505 + }, + { + "epoch": 1.3733801888864485, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5161776542663574, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7209160327911377, + "num_tokens": 323623987.0, + "step": 12506 + }, + { + "epoch": 1.373490006589062, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4989612102508545, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7083864808082581, + "num_tokens": 323651896.0, + "step": 12507 + }, + { + "epoch": 1.3735998242916758, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4994633197784424, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.6901599764823914, + "num_tokens": 323679655.0, + "step": 12508 + }, + { + "epoch": 1.3737096419942896, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.560269594192505, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6994163990020752, + "num_tokens": 323704418.0, + "step": 12509 + }, + { + "epoch": 1.3738194596969031, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9023516178131104, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.701799750328064, + "num_tokens": 323727290.0, + "step": 12510 + }, + { + "epoch": 1.3739292773995169, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.505136728286743, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.68751460313797, + "num_tokens": 323756261.0, + "step": 12511 + }, + { + "epoch": 1.3740390951021304, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4308383464813232, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6901564002037048, + "num_tokens": 323785724.0, + "step": 12512 + }, + { + "epoch": 1.3741489128047442, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5478014945983887, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6898261308670044, + "num_tokens": 323812965.0, + "step": 12513 + }, + { + "epoch": 1.374258730507358, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6415205001831055, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7260525226593018, + "num_tokens": 323838244.0, + "step": 12514 + }, + { + "epoch": 1.3743685482099715, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.930264711380005, + "learning_rate": 1e-06, + "loss": 0.883, + "mean_token_accuracy": 0.7345490455627441, + "num_tokens": 323858223.0, + "step": 12515 + }, + { + "epoch": 1.374478365912585, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.337985038757324, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.694277286529541, + "num_tokens": 323890650.0, + "step": 12516 + }, + { + "epoch": 1.3745881836151987, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5317933559417725, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7040308713912964, + "num_tokens": 323915138.0, + "step": 12517 + }, + { + "epoch": 1.3746980013178125, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6663498878479004, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6862041354179382, + "num_tokens": 323939118.0, + "step": 12518 + }, + { + "epoch": 1.374807819020426, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3308353424072266, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7041604518890381, + "num_tokens": 323967959.0, + "step": 12519 + }, + { + "epoch": 1.3749176367230398, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7105655670166016, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7088313102722168, + "num_tokens": 323989323.0, + "step": 12520 + }, + { + "epoch": 1.3750274544256533, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7031867504119873, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7159112691879272, + "num_tokens": 324010929.0, + "step": 12521 + }, + { + "epoch": 1.375137272128267, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0593976974487305, + "learning_rate": 1e-06, + "loss": 0.8901, + "mean_token_accuracy": 0.735626220703125, + "num_tokens": 324029764.0, + "step": 12522 + }, + { + "epoch": 1.3752470898308808, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8083481788635254, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7255143523216248, + "num_tokens": 324051254.0, + "step": 12523 + }, + { + "epoch": 1.3753569075334944, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.997952938079834, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7156952619552612, + "num_tokens": 324073161.0, + "step": 12524 + }, + { + "epoch": 1.3754667252361081, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7461438179016113, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7080734968185425, + "num_tokens": 324096249.0, + "step": 12525 + }, + { + "epoch": 1.3755765429387217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.729015827178955, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7043434977531433, + "num_tokens": 324119618.0, + "step": 12526 + }, + { + "epoch": 1.3756863606413354, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4477925300598145, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7034858465194702, + "num_tokens": 324149285.0, + "step": 12527 + }, + { + "epoch": 1.3757961783439492, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8438243865966797, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7051112651824951, + "num_tokens": 324170406.0, + "step": 12528 + }, + { + "epoch": 1.3759059960465627, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.546330690383911, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6866475343704224, + "num_tokens": 324198163.0, + "step": 12529 + }, + { + "epoch": 1.3760158137491763, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.64746356010437, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.7031017541885376, + "num_tokens": 324226223.0, + "step": 12530 + }, + { + "epoch": 1.37612563145179, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4961471557617188, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.690668523311615, + "num_tokens": 324252547.0, + "step": 12531 + }, + { + "epoch": 1.3762354491544038, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.272953987121582, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7160236239433289, + "num_tokens": 324269247.0, + "step": 12532 + }, + { + "epoch": 1.3763452668570173, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5659186840057373, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6959631443023682, + "num_tokens": 324294598.0, + "step": 12533 + }, + { + "epoch": 1.376455084559631, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.447453737258911, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.6928812265396118, + "num_tokens": 324322960.0, + "step": 12534 + }, + { + "epoch": 1.3765649022622446, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.83451247215271, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7327381372451782, + "num_tokens": 324345051.0, + "step": 12535 + }, + { + "epoch": 1.3766747199648584, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6989030838012695, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6908119916915894, + "num_tokens": 324370134.0, + "step": 12536 + }, + { + "epoch": 1.376784537667472, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.784801721572876, + "learning_rate": 1e-06, + "loss": 1.1036, + "mean_token_accuracy": 0.6805842518806458, + "num_tokens": 324396827.0, + "step": 12537 + }, + { + "epoch": 1.3768943553700856, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.798888921737671, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7029407620429993, + "num_tokens": 324419711.0, + "step": 12538 + }, + { + "epoch": 1.3770041730726994, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.755763530731201, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7047305703163147, + "num_tokens": 324442608.0, + "step": 12539 + }, + { + "epoch": 1.377113990775313, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.753779411315918, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7048866152763367, + "num_tokens": 324463862.0, + "step": 12540 + }, + { + "epoch": 1.3772238084779267, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.674532890319824, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7093439102172852, + "num_tokens": 324485779.0, + "step": 12541 + }, + { + "epoch": 1.3773336261805402, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.515800714492798, + "learning_rate": 1e-06, + "loss": 1.1152, + "mean_token_accuracy": 0.6876153945922852, + "num_tokens": 324513413.0, + "step": 12542 + }, + { + "epoch": 1.377443443883154, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.329465866088867, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7155852317810059, + "num_tokens": 324543901.0, + "step": 12543 + }, + { + "epoch": 1.3775532615857675, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9296116828918457, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6899666786193848, + "num_tokens": 324566134.0, + "step": 12544 + }, + { + "epoch": 1.3776630792883813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3966150283813477, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6838312745094299, + "num_tokens": 324594770.0, + "step": 12545 + }, + { + "epoch": 1.377772896990995, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5841474533081055, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6950831413269043, + "num_tokens": 324619443.0, + "step": 12546 + }, + { + "epoch": 1.3778827146936086, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.881957769393921, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7130388021469116, + "num_tokens": 324639903.0, + "step": 12547 + }, + { + "epoch": 1.3779925323962223, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.629622220993042, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7170772552490234, + "num_tokens": 324661797.0, + "step": 12548 + }, + { + "epoch": 1.3781023500988359, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4790515899658203, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7236474752426147, + "num_tokens": 324686076.0, + "step": 12549 + }, + { + "epoch": 1.3782121678014496, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9155287742614746, + "learning_rate": 1e-06, + "loss": 0.9454, + "mean_token_accuracy": 0.7233955264091492, + "num_tokens": 324709315.0, + "step": 12550 + }, + { + "epoch": 1.3783219855040634, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6246402263641357, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7202492952346802, + "num_tokens": 324733142.0, + "step": 12551 + }, + { + "epoch": 1.378431803206677, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.732639789581299, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7260158061981201, + "num_tokens": 324756808.0, + "step": 12552 + }, + { + "epoch": 1.3785416209092904, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.834047555923462, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7037073969841003, + "num_tokens": 324777993.0, + "step": 12553 + }, + { + "epoch": 1.3786514386119042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.465259552001953, + "learning_rate": 1e-06, + "loss": 0.9105, + "mean_token_accuracy": 0.7288748025894165, + "num_tokens": 324806322.0, + "step": 12554 + }, + { + "epoch": 1.378761256314518, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3488845825195312, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6881067752838135, + "num_tokens": 324837557.0, + "step": 12555 + }, + { + "epoch": 1.3788710740171315, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.938990831375122, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.6879811882972717, + "num_tokens": 324857980.0, + "step": 12556 + }, + { + "epoch": 1.3789808917197452, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.634882688522339, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6966990828514099, + "num_tokens": 324887487.0, + "step": 12557 + }, + { + "epoch": 1.3790907094223588, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8486392498016357, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.731597900390625, + "num_tokens": 324908476.0, + "step": 12558 + }, + { + "epoch": 1.3792005271249725, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.613696575164795, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7119340300559998, + "num_tokens": 324934794.0, + "step": 12559 + }, + { + "epoch": 1.3793103448275863, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.559826374053955, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7218659520149231, + "num_tokens": 324965560.0, + "step": 12560 + }, + { + "epoch": 1.3794201625301998, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4817402362823486, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7078622579574585, + "num_tokens": 324992511.0, + "step": 12561 + }, + { + "epoch": 1.3795299802328136, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5225014686584473, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7161746025085449, + "num_tokens": 325019487.0, + "step": 12562 + }, + { + "epoch": 1.3796397979354271, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.558446168899536, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7085914015769958, + "num_tokens": 325043901.0, + "step": 12563 + }, + { + "epoch": 1.3797496156380409, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.517928123474121, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7185781598091125, + "num_tokens": 325068013.0, + "step": 12564 + }, + { + "epoch": 1.3798594333406546, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.702056646347046, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6963776350021362, + "num_tokens": 325094329.0, + "step": 12565 + }, + { + "epoch": 1.3799692510432682, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.521759510040283, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7081959247589111, + "num_tokens": 325121449.0, + "step": 12566 + }, + { + "epoch": 1.3800790687458817, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.524427652359009, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7127223014831543, + "num_tokens": 325146118.0, + "step": 12567 + }, + { + "epoch": 1.3801888864484955, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4481189250946045, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6974834203720093, + "num_tokens": 325177571.0, + "step": 12568 + }, + { + "epoch": 1.3802987041511092, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5019776821136475, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.6822574138641357, + "num_tokens": 325207045.0, + "step": 12569 + }, + { + "epoch": 1.3804085218537228, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.613422155380249, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7027225494384766, + "num_tokens": 325233060.0, + "step": 12570 + }, + { + "epoch": 1.3805183395563365, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7498373985290527, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7191532850265503, + "num_tokens": 325255279.0, + "step": 12571 + }, + { + "epoch": 1.38062815725895, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6874513626098633, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7037777900695801, + "num_tokens": 325279494.0, + "step": 12572 + }, + { + "epoch": 1.3807379749615638, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5402185916900635, + "learning_rate": 1e-06, + "loss": 0.8947, + "mean_token_accuracy": 0.7375325560569763, + "num_tokens": 325303432.0, + "step": 12573 + }, + { + "epoch": 1.3808477926641776, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5391550064086914, + "learning_rate": 1e-06, + "loss": 1.0802, + "mean_token_accuracy": 0.69232177734375, + "num_tokens": 325329242.0, + "step": 12574 + }, + { + "epoch": 1.380957610366791, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5643653869628906, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.7016446590423584, + "num_tokens": 325355446.0, + "step": 12575 + }, + { + "epoch": 1.3810674280694049, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4098780155181885, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.6999456882476807, + "num_tokens": 325382954.0, + "step": 12576 + }, + { + "epoch": 1.3811772457720184, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7124383449554443, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6903985738754272, + "num_tokens": 325407603.0, + "step": 12577 + }, + { + "epoch": 1.3812870634746321, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7502853870391846, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6899659633636475, + "num_tokens": 325434185.0, + "step": 12578 + }, + { + "epoch": 1.381396881177246, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.437098264694214, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7094690799713135, + "num_tokens": 325461343.0, + "step": 12579 + }, + { + "epoch": 1.3815066988798594, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6593735218048096, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.6970767378807068, + "num_tokens": 325483556.0, + "step": 12580 + }, + { + "epoch": 1.381616516582473, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5643672943115234, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6966347098350525, + "num_tokens": 325510162.0, + "step": 12581 + }, + { + "epoch": 1.3817263342850867, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6350278854370117, + "learning_rate": 1e-06, + "loss": 0.8836, + "mean_token_accuracy": 0.7385572195053101, + "num_tokens": 325533463.0, + "step": 12582 + }, + { + "epoch": 1.3818361519877005, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4985766410827637, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.7330639958381653, + "num_tokens": 325558235.0, + "step": 12583 + }, + { + "epoch": 1.381945969690314, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.753735065460205, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7124007940292358, + "num_tokens": 325579757.0, + "step": 12584 + }, + { + "epoch": 1.3820557873929278, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.614971160888672, + "learning_rate": 1e-06, + "loss": 1.1054, + "mean_token_accuracy": 0.6845649480819702, + "num_tokens": 325605573.0, + "step": 12585 + }, + { + "epoch": 1.3821656050955413, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5752370357513428, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7066537141799927, + "num_tokens": 325629148.0, + "step": 12586 + }, + { + "epoch": 1.382275422798155, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7176198959350586, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7064324021339417, + "num_tokens": 325655848.0, + "step": 12587 + }, + { + "epoch": 1.3823852405007688, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.724822998046875, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7068836688995361, + "num_tokens": 325679168.0, + "step": 12588 + }, + { + "epoch": 1.3824950582033824, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.490583658218384, + "learning_rate": 1e-06, + "loss": 1.1062, + "mean_token_accuracy": 0.6878392100334167, + "num_tokens": 325711489.0, + "step": 12589 + }, + { + "epoch": 1.3826048759059961, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.4671504497528076, + "learning_rate": 1e-06, + "loss": 1.0956, + "mean_token_accuracy": 0.6808043718338013, + "num_tokens": 325740922.0, + "step": 12590 + }, + { + "epoch": 1.3827146936086097, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.688044548034668, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6929612755775452, + "num_tokens": 325770207.0, + "step": 12591 + }, + { + "epoch": 1.3828245113112234, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.387737989425659, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7278617024421692, + "num_tokens": 325798524.0, + "step": 12592 + }, + { + "epoch": 1.382934329013837, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7895987033843994, + "learning_rate": 1e-06, + "loss": 1.0937, + "mean_token_accuracy": 0.6849681735038757, + "num_tokens": 325825687.0, + "step": 12593 + }, + { + "epoch": 1.3830441467164507, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.514253616333008, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7049312591552734, + "num_tokens": 325854967.0, + "step": 12594 + }, + { + "epoch": 1.3831539644190642, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.396794080734253, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6817923784255981, + "num_tokens": 325887490.0, + "step": 12595 + }, + { + "epoch": 1.383263782121678, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.8480703830718994, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6905970573425293, + "num_tokens": 325908570.0, + "step": 12596 + }, + { + "epoch": 1.3833735998242918, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9539942741394043, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7076968550682068, + "num_tokens": 325928528.0, + "step": 12597 + }, + { + "epoch": 1.3834834175269053, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5289454460144043, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6923040151596069, + "num_tokens": 325956713.0, + "step": 12598 + }, + { + "epoch": 1.383593235229519, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.600992202758789, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7011533975601196, + "num_tokens": 325984839.0, + "step": 12599 + }, + { + "epoch": 1.3837030529321326, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.476724624633789, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7158312201499939, + "num_tokens": 326011963.0, + "step": 12600 + }, + { + "epoch": 1.3838128706347463, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.605064630508423, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7103067636489868, + "num_tokens": 326037627.0, + "step": 12601 + }, + { + "epoch": 1.38392268833736, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.3514277935028076, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7052602767944336, + "num_tokens": 326068689.0, + "step": 12602 + }, + { + "epoch": 1.3840325060399736, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.5888359546661377, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6986953020095825, + "num_tokens": 326099103.0, + "step": 12603 + }, + { + "epoch": 1.3841423237425874, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.650183916091919, + "learning_rate": 1e-06, + "loss": 1.1227, + "mean_token_accuracy": 0.6773874759674072, + "num_tokens": 326125592.0, + "step": 12604 + }, + { + "epoch": 1.384252141445201, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.543989658355713, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7047241926193237, + "num_tokens": 326154108.0, + "step": 12605 + }, + { + "epoch": 1.3843619591478147, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 7.945913314819336, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7130235433578491, + "num_tokens": 326178614.0, + "step": 12606 + }, + { + "epoch": 1.3844717768504282, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.604294776916504, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6901863813400269, + "num_tokens": 326205864.0, + "step": 12607 + }, + { + "epoch": 1.384581594553042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.796360969543457, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6988669633865356, + "num_tokens": 326229546.0, + "step": 12608 + }, + { + "epoch": 1.3846914122556555, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.514660120010376, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6922423839569092, + "num_tokens": 326257447.0, + "step": 12609 + }, + { + "epoch": 1.3848012299582693, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7697455883026123, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7119112610816956, + "num_tokens": 326279636.0, + "step": 12610 + }, + { + "epoch": 1.384911047660883, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4889564514160156, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.693274736404419, + "num_tokens": 326307847.0, + "step": 12611 + }, + { + "epoch": 1.3850208653634966, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.549201488494873, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6988674998283386, + "num_tokens": 326334739.0, + "step": 12612 + }, + { + "epoch": 1.3851306830661103, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.969611883163452, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.731294572353363, + "num_tokens": 326354685.0, + "step": 12613 + }, + { + "epoch": 1.3852405007687238, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.472409725189209, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6866850256919861, + "num_tokens": 326384519.0, + "step": 12614 + }, + { + "epoch": 1.3853503184713376, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.775071382522583, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7258662581443787, + "num_tokens": 326407163.0, + "step": 12615 + }, + { + "epoch": 1.3854601361739514, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 7.929542064666748, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.699549674987793, + "num_tokens": 326432569.0, + "step": 12616 + }, + { + "epoch": 1.385569953876565, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.672959804534912, + "learning_rate": 1e-06, + "loss": 1.0883, + "mean_token_accuracy": 0.6888606548309326, + "num_tokens": 326462060.0, + "step": 12617 + }, + { + "epoch": 1.3856797715791784, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.627855062484741, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7181788682937622, + "num_tokens": 326487727.0, + "step": 12618 + }, + { + "epoch": 1.3857895892817922, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.853119373321533, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.7208123803138733, + "num_tokens": 326510858.0, + "step": 12619 + }, + { + "epoch": 1.385899406984406, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8042831420898438, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7152005434036255, + "num_tokens": 326532694.0, + "step": 12620 + }, + { + "epoch": 1.3860092246870195, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0911865234375, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7187823057174683, + "num_tokens": 326560597.0, + "step": 12621 + }, + { + "epoch": 1.3861190423896332, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7685670852661133, + "learning_rate": 1e-06, + "loss": 1.1136, + "mean_token_accuracy": 0.675121545791626, + "num_tokens": 326582964.0, + "step": 12622 + }, + { + "epoch": 1.3862288600922468, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5491631031036377, + "learning_rate": 1e-06, + "loss": 1.1192, + "mean_token_accuracy": 0.6902906894683838, + "num_tokens": 326609534.0, + "step": 12623 + }, + { + "epoch": 1.3863386777948605, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5221192836761475, + "learning_rate": 1e-06, + "loss": 0.9204, + "mean_token_accuracy": 0.7214009761810303, + "num_tokens": 326634252.0, + "step": 12624 + }, + { + "epoch": 1.3864484954974743, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6179492473602295, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7195857763290405, + "num_tokens": 326658361.0, + "step": 12625 + }, + { + "epoch": 1.3865583132000878, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9051783084869385, + "learning_rate": 1e-06, + "loss": 0.9298, + "mean_token_accuracy": 0.7228696346282959, + "num_tokens": 326678597.0, + "step": 12626 + }, + { + "epoch": 1.3866681309027016, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4830312728881836, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6939857602119446, + "num_tokens": 326707463.0, + "step": 12627 + }, + { + "epoch": 1.3867779486053151, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4446845054626465, + "learning_rate": 1e-06, + "loss": 1.1389, + "mean_token_accuracy": 0.6737822890281677, + "num_tokens": 326740808.0, + "step": 12628 + }, + { + "epoch": 1.3868877663079289, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.4642512798309326, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.6955221891403198, + "num_tokens": 326770351.0, + "step": 12629 + }, + { + "epoch": 1.3869975840105426, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.2349112033843994, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7170583009719849, + "num_tokens": 326795628.0, + "step": 12630 + }, + { + "epoch": 1.3871074017131562, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2787179946899414, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.698638916015625, + "num_tokens": 326828612.0, + "step": 12631 + }, + { + "epoch": 1.3872172194157697, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5859248638153076, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.716362476348877, + "num_tokens": 326852073.0, + "step": 12632 + }, + { + "epoch": 1.3873270371183835, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.592823028564453, + "learning_rate": 1e-06, + "loss": 0.9252, + "mean_token_accuracy": 0.7296162843704224, + "num_tokens": 326877643.0, + "step": 12633 + }, + { + "epoch": 1.3874368548209972, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.708462953567505, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6975885033607483, + "num_tokens": 326902042.0, + "step": 12634 + }, + { + "epoch": 1.3875466725236107, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4889161586761475, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.7040128707885742, + "num_tokens": 326927492.0, + "step": 12635 + }, + { + "epoch": 1.3876564902262245, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.482306718826294, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7073560953140259, + "num_tokens": 326955640.0, + "step": 12636 + }, + { + "epoch": 1.387766307928838, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.495500326156616, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7021028995513916, + "num_tokens": 326982934.0, + "step": 12637 + }, + { + "epoch": 1.3878761256314518, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.928542375564575, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.6969795227050781, + "num_tokens": 327003284.0, + "step": 12638 + }, + { + "epoch": 1.3879859433340656, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1895687580108643, + "learning_rate": 1e-06, + "loss": 1.1346, + "mean_token_accuracy": 0.6720666289329529, + "num_tokens": 327039297.0, + "step": 12639 + }, + { + "epoch": 1.388095761036679, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4283952713012695, + "learning_rate": 1e-06, + "loss": 1.1162, + "mean_token_accuracy": 0.6806215047836304, + "num_tokens": 327068485.0, + "step": 12640 + }, + { + "epoch": 1.3882055787392928, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5304481983184814, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7020410895347595, + "num_tokens": 327094322.0, + "step": 12641 + }, + { + "epoch": 1.3883153964419064, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.593341588973999, + "learning_rate": 1e-06, + "loss": 0.891, + "mean_token_accuracy": 0.7355396151542664, + "num_tokens": 327118567.0, + "step": 12642 + }, + { + "epoch": 1.3884252141445201, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.495382785797119, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7083722949028015, + "num_tokens": 327144865.0, + "step": 12643 + }, + { + "epoch": 1.388535031847134, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5052599906921387, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6963202953338623, + "num_tokens": 327174998.0, + "step": 12644 + }, + { + "epoch": 1.3886448495497474, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4794883728027344, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6896215081214905, + "num_tokens": 327200509.0, + "step": 12645 + }, + { + "epoch": 1.388754667252361, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.456372022628784, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.7164240479469299, + "num_tokens": 327225745.0, + "step": 12646 + }, + { + "epoch": 1.3888644849549747, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 5.405942916870117, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6831083297729492, + "num_tokens": 327254984.0, + "step": 12647 + }, + { + "epoch": 1.3889743026575885, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6685335636138916, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7132527828216553, + "num_tokens": 327279223.0, + "step": 12648 + }, + { + "epoch": 1.389084120360202, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7698235511779785, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7087516188621521, + "num_tokens": 327303319.0, + "step": 12649 + }, + { + "epoch": 1.3891939380628158, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3889670372009277, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6978482604026794, + "num_tokens": 327331675.0, + "step": 12650 + }, + { + "epoch": 1.3893037557654293, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6542553901672363, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7073521018028259, + "num_tokens": 327355990.0, + "step": 12651 + }, + { + "epoch": 1.389413573468043, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.6061127185821533, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7165974974632263, + "num_tokens": 327379324.0, + "step": 12652 + }, + { + "epoch": 1.3895233911706568, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.82361102104187, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.6955697536468506, + "num_tokens": 327402498.0, + "step": 12653 + }, + { + "epoch": 1.3896332088732704, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.66191029548645, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6988359689712524, + "num_tokens": 327426181.0, + "step": 12654 + }, + { + "epoch": 1.389743026575884, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4271609783172607, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7012319564819336, + "num_tokens": 327455846.0, + "step": 12655 + }, + { + "epoch": 1.3898528442784976, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5357248783111572, + "learning_rate": 1e-06, + "loss": 1.1215, + "mean_token_accuracy": 0.675818145275116, + "num_tokens": 327486203.0, + "step": 12656 + }, + { + "epoch": 1.3899626619811114, + "ewc_loss": 9.298324584960938e-06, + "grad_norm": 2.571903705596924, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7282408475875854, + "num_tokens": 327510687.0, + "step": 12657 + }, + { + "epoch": 1.390072479683725, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7172844409942627, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7051721811294556, + "num_tokens": 327536849.0, + "step": 12658 + }, + { + "epoch": 1.3901822973863387, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9131975173950195, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.6994173526763916, + "num_tokens": 327559433.0, + "step": 12659 + }, + { + "epoch": 1.3902921150889522, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8467392921447754, + "learning_rate": 1e-06, + "loss": 0.8317, + "mean_token_accuracy": 0.7515170574188232, + "num_tokens": 327580154.0, + "step": 12660 + }, + { + "epoch": 1.390401932791566, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.781315803527832, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7203356027603149, + "num_tokens": 327602813.0, + "step": 12661 + }, + { + "epoch": 1.3905117504941797, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.594454765319824, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7190630435943604, + "num_tokens": 327629603.0, + "step": 12662 + }, + { + "epoch": 1.3906215681967933, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.795825719833374, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6998019218444824, + "num_tokens": 327654648.0, + "step": 12663 + }, + { + "epoch": 1.390731385899407, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3535096645355225, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6904816627502441, + "num_tokens": 327684120.0, + "step": 12664 + }, + { + "epoch": 1.3908412036020206, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.740262985229492, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6978477239608765, + "num_tokens": 327707176.0, + "step": 12665 + }, + { + "epoch": 1.3909510213046343, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6249208450317383, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6855787634849548, + "num_tokens": 327734720.0, + "step": 12666 + }, + { + "epoch": 1.391060839007248, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.633361339569092, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7107905745506287, + "num_tokens": 327760150.0, + "step": 12667 + }, + { + "epoch": 1.3911706567098616, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6022849082946777, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7080775499343872, + "num_tokens": 327784149.0, + "step": 12668 + }, + { + "epoch": 1.3912804744124752, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5959060192108154, + "learning_rate": 1e-06, + "loss": 0.9373, + "mean_token_accuracy": 0.7188007831573486, + "num_tokens": 327808125.0, + "step": 12669 + }, + { + "epoch": 1.391390292115089, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8004910945892334, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7208784818649292, + "num_tokens": 327830077.0, + "step": 12670 + }, + { + "epoch": 1.3915001098177027, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4889473915100098, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7052952647209167, + "num_tokens": 327857461.0, + "step": 12671 + }, + { + "epoch": 1.3916099275203162, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.959336757659912, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7118837833404541, + "num_tokens": 327877176.0, + "step": 12672 + }, + { + "epoch": 1.39171974522293, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1665334701538086, + "learning_rate": 1e-06, + "loss": 0.9169, + "mean_token_accuracy": 0.7300647497177124, + "num_tokens": 327895395.0, + "step": 12673 + }, + { + "epoch": 1.3918295629255435, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4546406269073486, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7129433751106262, + "num_tokens": 327927128.0, + "step": 12674 + }, + { + "epoch": 1.3919393806281573, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.601893424987793, + "learning_rate": 1e-06, + "loss": 0.9309, + "mean_token_accuracy": 0.728028416633606, + "num_tokens": 327951863.0, + "step": 12675 + }, + { + "epoch": 1.392049198330771, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.581702947616577, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7205204963684082, + "num_tokens": 327975457.0, + "step": 12676 + }, + { + "epoch": 1.3921590160333845, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.710991859436035, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6968872547149658, + "num_tokens": 328001139.0, + "step": 12677 + }, + { + "epoch": 1.3922688337359983, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5904762744903564, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6956942677497864, + "num_tokens": 328027431.0, + "step": 12678 + }, + { + "epoch": 1.3923786514386118, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5779612064361572, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6876876354217529, + "num_tokens": 328054469.0, + "step": 12679 + }, + { + "epoch": 1.3924884691412256, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2816591262817383, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6977568864822388, + "num_tokens": 328086474.0, + "step": 12680 + }, + { + "epoch": 1.3925982868438394, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6342289447784424, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6960530877113342, + "num_tokens": 328114272.0, + "step": 12681 + }, + { + "epoch": 1.3927081045464529, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5203442573547363, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7019881010055542, + "num_tokens": 328139879.0, + "step": 12682 + }, + { + "epoch": 1.3928179222490664, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.73009991645813, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.6920570135116577, + "num_tokens": 328163458.0, + "step": 12683 + }, + { + "epoch": 1.3929277399516802, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.577423334121704, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7064675092697144, + "num_tokens": 328189557.0, + "step": 12684 + }, + { + "epoch": 1.393037557654294, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5229790210723877, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6973208785057068, + "num_tokens": 328216781.0, + "step": 12685 + }, + { + "epoch": 1.3931473753569075, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4878838062286377, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7163367867469788, + "num_tokens": 328248094.0, + "step": 12686 + }, + { + "epoch": 1.3932571930595212, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3049569129943848, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6859171390533447, + "num_tokens": 328282358.0, + "step": 12687 + }, + { + "epoch": 1.3933670107621348, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2673420906066895, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6871892809867859, + "num_tokens": 328318452.0, + "step": 12688 + }, + { + "epoch": 1.3934768284647485, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5306599140167236, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6893370747566223, + "num_tokens": 328344819.0, + "step": 12689 + }, + { + "epoch": 1.3935866461673623, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8814384937286377, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6868302822113037, + "num_tokens": 328366894.0, + "step": 12690 + }, + { + "epoch": 1.3936964638699758, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5228335857391357, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6979939937591553, + "num_tokens": 328393601.0, + "step": 12691 + }, + { + "epoch": 1.3938062815725896, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5087666511535645, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7120332717895508, + "num_tokens": 328419230.0, + "step": 12692 + }, + { + "epoch": 1.393916099275203, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6439871788024902, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.6972058415412903, + "num_tokens": 328443372.0, + "step": 12693 + }, + { + "epoch": 1.3940259169778169, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.747145652770996, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7155272960662842, + "num_tokens": 328465375.0, + "step": 12694 + }, + { + "epoch": 1.3941357346804306, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4052464962005615, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.703626275062561, + "num_tokens": 328497099.0, + "step": 12695 + }, + { + "epoch": 1.3942455523830442, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.676811695098877, + "learning_rate": 1e-06, + "loss": 0.8822, + "mean_token_accuracy": 0.7346138954162598, + "num_tokens": 328517455.0, + "step": 12696 + }, + { + "epoch": 1.3943553700856577, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5313103199005127, + "learning_rate": 1e-06, + "loss": 1.0842, + "mean_token_accuracy": 0.6930557489395142, + "num_tokens": 328542932.0, + "step": 12697 + }, + { + "epoch": 1.3944651877882714, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6286864280700684, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7158355116844177, + "num_tokens": 328566209.0, + "step": 12698 + }, + { + "epoch": 1.3945750054908852, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7074568271636963, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7062730193138123, + "num_tokens": 328591339.0, + "step": 12699 + }, + { + "epoch": 1.3946848231934987, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 4.2353901863098145, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6819387078285217, + "num_tokens": 328624959.0, + "step": 12700 + }, + { + "epoch": 1.3947946408961125, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6157941818237305, + "learning_rate": 1e-06, + "loss": 1.1108, + "mean_token_accuracy": 0.6786043643951416, + "num_tokens": 328653922.0, + "step": 12701 + }, + { + "epoch": 1.394904458598726, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.931521415710449, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7163723111152649, + "num_tokens": 328674331.0, + "step": 12702 + }, + { + "epoch": 1.3950142763013398, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.569486141204834, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7186431884765625, + "num_tokens": 328698819.0, + "step": 12703 + }, + { + "epoch": 1.3951240940039535, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3711600303649902, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7023038864135742, + "num_tokens": 328727751.0, + "step": 12704 + }, + { + "epoch": 1.395233911706567, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.703359842300415, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.7232410907745361, + "num_tokens": 328749952.0, + "step": 12705 + }, + { + "epoch": 1.3953437294091808, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7210371494293213, + "learning_rate": 1e-06, + "loss": 0.8973, + "mean_token_accuracy": 0.7301204204559326, + "num_tokens": 328773941.0, + "step": 12706 + }, + { + "epoch": 1.3954535471117944, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.679434061050415, + "learning_rate": 1e-06, + "loss": 1.0855, + "mean_token_accuracy": 0.6843185424804688, + "num_tokens": 328799129.0, + "step": 12707 + }, + { + "epoch": 1.3955633648144081, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.551381826400757, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7036470174789429, + "num_tokens": 328826607.0, + "step": 12708 + }, + { + "epoch": 1.3956731825170219, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.728827714920044, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7286402583122253, + "num_tokens": 328851609.0, + "step": 12709 + }, + { + "epoch": 1.3957830002196354, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4039015769958496, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7033876776695251, + "num_tokens": 328878492.0, + "step": 12710 + }, + { + "epoch": 1.395892817922249, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4988162517547607, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7096178531646729, + "num_tokens": 328903482.0, + "step": 12711 + }, + { + "epoch": 1.3960026356248627, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1128640174865723, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.713079035282135, + "num_tokens": 328925219.0, + "step": 12712 + }, + { + "epoch": 1.3961124533274765, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7539055347442627, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7009556293487549, + "num_tokens": 328949992.0, + "step": 12713 + }, + { + "epoch": 1.39622227103009, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.904078245162964, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7165160179138184, + "num_tokens": 328972712.0, + "step": 12714 + }, + { + "epoch": 1.3963320887327038, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.540851354598999, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6893210411071777, + "num_tokens": 329000792.0, + "step": 12715 + }, + { + "epoch": 1.3964419064353173, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.66249680519104, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7139570713043213, + "num_tokens": 329028808.0, + "step": 12716 + }, + { + "epoch": 1.396551724137931, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5317606925964355, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7112314701080322, + "num_tokens": 329056023.0, + "step": 12717 + }, + { + "epoch": 1.3966615418405448, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.756150722503662, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6900203227996826, + "num_tokens": 329087466.0, + "step": 12718 + }, + { + "epoch": 1.3967713595431583, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5247890949249268, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7056891918182373, + "num_tokens": 329113202.0, + "step": 12719 + }, + { + "epoch": 1.396881177245772, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.820114850997925, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6911807060241699, + "num_tokens": 329138724.0, + "step": 12720 + }, + { + "epoch": 1.3969909949483856, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5159127712249756, + "learning_rate": 1e-06, + "loss": 1.1012, + "mean_token_accuracy": 0.6886926889419556, + "num_tokens": 329165168.0, + "step": 12721 + }, + { + "epoch": 1.3971008126509994, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6336050033569336, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.6952298879623413, + "num_tokens": 329189652.0, + "step": 12722 + }, + { + "epoch": 1.397210630353613, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.789470911026001, + "learning_rate": 1e-06, + "loss": 0.9064, + "mean_token_accuracy": 0.7393189072608948, + "num_tokens": 329211107.0, + "step": 12723 + }, + { + "epoch": 1.3973204480562267, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.822538137435913, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6980834007263184, + "num_tokens": 329233763.0, + "step": 12724 + }, + { + "epoch": 1.3974302657588402, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.801539421081543, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7040063738822937, + "num_tokens": 329256232.0, + "step": 12725 + }, + { + "epoch": 1.397540083461454, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.686237096786499, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7051881551742554, + "num_tokens": 329278118.0, + "step": 12726 + }, + { + "epoch": 1.3976499011640677, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5094645023345947, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7314027547836304, + "num_tokens": 329306441.0, + "step": 12727 + }, + { + "epoch": 1.3977597188666813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4671645164489746, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6909184455871582, + "num_tokens": 329338047.0, + "step": 12728 + }, + { + "epoch": 1.397869536569295, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7671117782592773, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7196715474128723, + "num_tokens": 329361361.0, + "step": 12729 + }, + { + "epoch": 1.3979793542719086, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.844609022140503, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7146825194358826, + "num_tokens": 329382056.0, + "step": 12730 + }, + { + "epoch": 1.3980891719745223, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9723894596099854, + "learning_rate": 1e-06, + "loss": 0.9761, + "mean_token_accuracy": 0.7171741127967834, + "num_tokens": 329404022.0, + "step": 12731 + }, + { + "epoch": 1.398198989677136, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5047802925109863, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.695919930934906, + "num_tokens": 329430905.0, + "step": 12732 + }, + { + "epoch": 1.3983088073797496, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.965010404586792, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7095097899436951, + "num_tokens": 329453502.0, + "step": 12733 + }, + { + "epoch": 1.3984186250823631, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4327545166015625, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6935573220252991, + "num_tokens": 329482453.0, + "step": 12734 + }, + { + "epoch": 1.398528442784977, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.336665630340576, + "learning_rate": 1e-06, + "loss": 1.0986, + "mean_token_accuracy": 0.6821281313896179, + "num_tokens": 329514440.0, + "step": 12735 + }, + { + "epoch": 1.3986382604875907, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.361992359161377, + "learning_rate": 1e-06, + "loss": 0.8752, + "mean_token_accuracy": 0.7455393075942993, + "num_tokens": 329534295.0, + "step": 12736 + }, + { + "epoch": 1.3987480781902042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0260825157165527, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6962080001831055, + "num_tokens": 329555527.0, + "step": 12737 + }, + { + "epoch": 1.398857895892818, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4066855907440186, + "learning_rate": 1e-06, + "loss": 1.131, + "mean_token_accuracy": 0.6699658632278442, + "num_tokens": 329586322.0, + "step": 12738 + }, + { + "epoch": 1.3989677135954315, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6672730445861816, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7072872519493103, + "num_tokens": 329611497.0, + "step": 12739 + }, + { + "epoch": 1.3990775312980452, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.646024703979492, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7096623182296753, + "num_tokens": 329636836.0, + "step": 12740 + }, + { + "epoch": 1.399187349000659, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6784000396728516, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7051066160202026, + "num_tokens": 329661010.0, + "step": 12741 + }, + { + "epoch": 1.3992971667032725, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.733438491821289, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7085238695144653, + "num_tokens": 329682644.0, + "step": 12742 + }, + { + "epoch": 1.3994069844058863, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.546936511993408, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7115026712417603, + "num_tokens": 329707774.0, + "step": 12743 + }, + { + "epoch": 1.3995168021084998, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5663394927978516, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7079175710678101, + "num_tokens": 329735999.0, + "step": 12744 + }, + { + "epoch": 1.3996266198111136, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4192466735839844, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6832368969917297, + "num_tokens": 329765151.0, + "step": 12745 + }, + { + "epoch": 1.3997364375137273, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.428091049194336, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7254728078842163, + "num_tokens": 329791205.0, + "step": 12746 + }, + { + "epoch": 1.3998462552163409, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6428074836730957, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.6829856038093567, + "num_tokens": 329815303.0, + "step": 12747 + }, + { + "epoch": 1.3999560729189544, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.3751447200775146, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7202349305152893, + "num_tokens": 329838161.0, + "step": 12748 + }, + { + "epoch": 1.4000658906215682, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4361281394958496, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.685164213180542, + "num_tokens": 329869874.0, + "step": 12749 + }, + { + "epoch": 1.400175708324182, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7063724994659424, + "learning_rate": 1e-06, + "loss": 0.8807, + "mean_token_accuracy": 0.7298545241355896, + "num_tokens": 329890223.0, + "step": 12750 + }, + { + "epoch": 1.4002855260267955, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.482279062271118, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7019457221031189, + "num_tokens": 329918639.0, + "step": 12751 + }, + { + "epoch": 1.4003953437294092, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6192831993103027, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.6990869045257568, + "num_tokens": 329943018.0, + "step": 12752 + }, + { + "epoch": 1.4005051614320227, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7521560192108154, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7107276916503906, + "num_tokens": 329968118.0, + "step": 12753 + }, + { + "epoch": 1.4006149791346365, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.39223313331604, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7158356308937073, + "num_tokens": 329996407.0, + "step": 12754 + }, + { + "epoch": 1.4007247968372503, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.683391571044922, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.6905921697616577, + "num_tokens": 330020797.0, + "step": 12755 + }, + { + "epoch": 1.4008346145398638, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5523905754089355, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.6989073753356934, + "num_tokens": 330046151.0, + "step": 12756 + }, + { + "epoch": 1.4009444322424776, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7824041843414307, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.724799633026123, + "num_tokens": 330067459.0, + "step": 12757 + }, + { + "epoch": 1.401054249945091, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.629601240158081, + "learning_rate": 1e-06, + "loss": 0.9368, + "mean_token_accuracy": 0.7224581241607666, + "num_tokens": 330090836.0, + "step": 12758 + }, + { + "epoch": 1.4011640676477048, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.005375862121582, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6834372878074646, + "num_tokens": 330112581.0, + "step": 12759 + }, + { + "epoch": 1.4012738853503186, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8395614624023438, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7018270492553711, + "num_tokens": 330134770.0, + "step": 12760 + }, + { + "epoch": 1.4013837030529321, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.594817638397217, + "learning_rate": 1e-06, + "loss": 1.1447, + "mean_token_accuracy": 0.677903413772583, + "num_tokens": 330162305.0, + "step": 12761 + }, + { + "epoch": 1.4014935207555457, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.375171661376953, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6818885207176208, + "num_tokens": 330195513.0, + "step": 12762 + }, + { + "epoch": 1.4016033384581594, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.690147876739502, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.688955545425415, + "num_tokens": 330224612.0, + "step": 12763 + }, + { + "epoch": 1.4017131561607732, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9609365463256836, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.6905274391174316, + "num_tokens": 330245872.0, + "step": 12764 + }, + { + "epoch": 1.4018229738633867, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.054647445678711, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7166706919670105, + "num_tokens": 330263478.0, + "step": 12765 + }, + { + "epoch": 1.4019327915660005, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.510122776031494, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.7077359557151794, + "num_tokens": 330292579.0, + "step": 12766 + }, + { + "epoch": 1.402042609268614, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7203783988952637, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7024292945861816, + "num_tokens": 330315212.0, + "step": 12767 + }, + { + "epoch": 1.4021524269712278, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.454371213912964, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6994122266769409, + "num_tokens": 330343511.0, + "step": 12768 + }, + { + "epoch": 1.4022622446738415, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.288851022720337, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7122647762298584, + "num_tokens": 330371013.0, + "step": 12769 + }, + { + "epoch": 1.402372062376455, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.866999626159668, + "learning_rate": 1e-06, + "loss": 0.8805, + "mean_token_accuracy": 0.7365671992301941, + "num_tokens": 330390565.0, + "step": 12770 + }, + { + "epoch": 1.4024818800790688, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.676007032394409, + "learning_rate": 1e-06, + "loss": 0.8911, + "mean_token_accuracy": 0.7272249460220337, + "num_tokens": 330414047.0, + "step": 12771 + }, + { + "epoch": 1.4025916977816824, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.622265100479126, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7099635601043701, + "num_tokens": 330440831.0, + "step": 12772 + }, + { + "epoch": 1.4027015154842961, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.774627923965454, + "learning_rate": 1e-06, + "loss": 0.9103, + "mean_token_accuracy": 0.7328253984451294, + "num_tokens": 330460607.0, + "step": 12773 + }, + { + "epoch": 1.4028113331869096, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.414202928543091, + "learning_rate": 1e-06, + "loss": 1.1008, + "mean_token_accuracy": 0.6825047731399536, + "num_tokens": 330493053.0, + "step": 12774 + }, + { + "epoch": 1.4029211508895234, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.243155002593994, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7074692249298096, + "num_tokens": 330525499.0, + "step": 12775 + }, + { + "epoch": 1.403030968592137, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5231926441192627, + "learning_rate": 1e-06, + "loss": 1.125, + "mean_token_accuracy": 0.6750653386116028, + "num_tokens": 330552529.0, + "step": 12776 + }, + { + "epoch": 1.4031407862947507, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2862040996551514, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6823866367340088, + "num_tokens": 330584224.0, + "step": 12777 + }, + { + "epoch": 1.4032506039973645, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.569606065750122, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7020339965820312, + "num_tokens": 330608743.0, + "step": 12778 + }, + { + "epoch": 1.403360421699978, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.468007802963257, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.7039580345153809, + "num_tokens": 330637403.0, + "step": 12779 + }, + { + "epoch": 1.4034702394025917, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4126274585723877, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.7016574144363403, + "num_tokens": 330669501.0, + "step": 12780 + }, + { + "epoch": 1.4035800571052053, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.294041156768799, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.695510745048523, + "num_tokens": 330701549.0, + "step": 12781 + }, + { + "epoch": 1.403689874807819, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4159646034240723, + "learning_rate": 1e-06, + "loss": 1.0989, + "mean_token_accuracy": 0.6897419095039368, + "num_tokens": 330729181.0, + "step": 12782 + }, + { + "epoch": 1.4037996925104328, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.450676679611206, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.6838915348052979, + "num_tokens": 330757862.0, + "step": 12783 + }, + { + "epoch": 1.4039095102130463, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6406891345977783, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7019924521446228, + "num_tokens": 330783695.0, + "step": 12784 + }, + { + "epoch": 1.40401932791566, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.196251153945923, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6941326260566711, + "num_tokens": 330817814.0, + "step": 12785 + }, + { + "epoch": 1.4041291456182736, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4690029621124268, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7128484845161438, + "num_tokens": 330844176.0, + "step": 12786 + }, + { + "epoch": 1.4042389633208874, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.461824655532837, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6921348571777344, + "num_tokens": 330870796.0, + "step": 12787 + }, + { + "epoch": 1.404348781023501, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5554325580596924, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7025191783905029, + "num_tokens": 330894438.0, + "step": 12788 + }, + { + "epoch": 1.4044585987261147, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5868194103240967, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6946470737457275, + "num_tokens": 330921480.0, + "step": 12789 + }, + { + "epoch": 1.4045684164287282, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8688509464263916, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6966695189476013, + "num_tokens": 330946772.0, + "step": 12790 + }, + { + "epoch": 1.404678234131342, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3518640995025635, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7041422724723816, + "num_tokens": 330979370.0, + "step": 12791 + }, + { + "epoch": 1.4047880518339557, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.614996910095215, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6905379891395569, + "num_tokens": 331002954.0, + "step": 12792 + }, + { + "epoch": 1.4048978695365693, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.325148582458496, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6843195557594299, + "num_tokens": 331035129.0, + "step": 12793 + }, + { + "epoch": 1.405007687239183, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7424063682556152, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.6984535455703735, + "num_tokens": 331059434.0, + "step": 12794 + }, + { + "epoch": 1.4051175049417965, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.884373903274536, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6907953023910522, + "num_tokens": 331080140.0, + "step": 12795 + }, + { + "epoch": 1.4052273226444103, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.571455240249634, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7135894894599915, + "num_tokens": 331106893.0, + "step": 12796 + }, + { + "epoch": 1.405337140347024, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.861361503601074, + "learning_rate": 1e-06, + "loss": 0.9269, + "mean_token_accuracy": 0.7338472604751587, + "num_tokens": 331126529.0, + "step": 12797 + }, + { + "epoch": 1.4054469580496376, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.332089900970459, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6928832530975342, + "num_tokens": 331158566.0, + "step": 12798 + }, + { + "epoch": 1.4055567757522511, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.46939754486084, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7049276828765869, + "num_tokens": 331186311.0, + "step": 12799 + }, + { + "epoch": 1.4056665934548649, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.466444730758667, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7272231578826904, + "num_tokens": 331215102.0, + "step": 12800 + }, + { + "epoch": 1.4057764111574786, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8010964393615723, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.6984223127365112, + "num_tokens": 331236980.0, + "step": 12801 + }, + { + "epoch": 1.4058862288600922, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6937434673309326, + "learning_rate": 1e-06, + "loss": 1.0842, + "mean_token_accuracy": 0.6868517398834229, + "num_tokens": 331261948.0, + "step": 12802 + }, + { + "epoch": 1.405996046562706, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8018722534179688, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.6991791725158691, + "num_tokens": 331288192.0, + "step": 12803 + }, + { + "epoch": 1.4061058642653195, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.390991687774658, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7060858011245728, + "num_tokens": 331317084.0, + "step": 12804 + }, + { + "epoch": 1.4062156819679332, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.266697406768799, + "learning_rate": 1e-06, + "loss": 0.9301, + "mean_token_accuracy": 0.7241952419281006, + "num_tokens": 331345302.0, + "step": 12805 + }, + { + "epoch": 1.406325499670547, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5517187118530273, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7085566520690918, + "num_tokens": 331372262.0, + "step": 12806 + }, + { + "epoch": 1.4064353173731605, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4939024448394775, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7149385213851929, + "num_tokens": 331398490.0, + "step": 12807 + }, + { + "epoch": 1.4065451350757743, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5779001712799072, + "learning_rate": 1e-06, + "loss": 1.1041, + "mean_token_accuracy": 0.6817682385444641, + "num_tokens": 331424470.0, + "step": 12808 + }, + { + "epoch": 1.4066549527783878, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.685638427734375, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6916475892066956, + "num_tokens": 331447838.0, + "step": 12809 + }, + { + "epoch": 1.4067647704810016, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.80318546295166, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7002882957458496, + "num_tokens": 331470010.0, + "step": 12810 + }, + { + "epoch": 1.4068745881836153, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5555689334869385, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6904691457748413, + "num_tokens": 331496230.0, + "step": 12811 + }, + { + "epoch": 1.4069844058862289, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7751543521881104, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6787453889846802, + "num_tokens": 331525604.0, + "step": 12812 + }, + { + "epoch": 1.4070942235888424, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.456049919128418, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7022066116333008, + "num_tokens": 331553471.0, + "step": 12813 + }, + { + "epoch": 1.4072040412914562, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6892457008361816, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7222810983657837, + "num_tokens": 331575373.0, + "step": 12814 + }, + { + "epoch": 1.40731385899407, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.405480146408081, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7036334872245789, + "num_tokens": 331604945.0, + "step": 12815 + }, + { + "epoch": 1.4074236766966834, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.523733377456665, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6872223019599915, + "num_tokens": 331632112.0, + "step": 12816 + }, + { + "epoch": 1.4075334943992972, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4801504611968994, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.708296537399292, + "num_tokens": 331659631.0, + "step": 12817 + }, + { + "epoch": 1.4076433121019107, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7698135375976562, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7258319854736328, + "num_tokens": 331681471.0, + "step": 12818 + }, + { + "epoch": 1.4077531298045245, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8179256916046143, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7225600481033325, + "num_tokens": 331704106.0, + "step": 12819 + }, + { + "epoch": 1.4078629475071383, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.30080509185791, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6833619475364685, + "num_tokens": 331736036.0, + "step": 12820 + }, + { + "epoch": 1.4079727652097518, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.625401496887207, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7009851932525635, + "num_tokens": 331761772.0, + "step": 12821 + }, + { + "epoch": 1.4080825829123655, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7815535068511963, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7134963870048523, + "num_tokens": 331785723.0, + "step": 12822 + }, + { + "epoch": 1.408192400614979, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.605095386505127, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7100450992584229, + "num_tokens": 331809006.0, + "step": 12823 + }, + { + "epoch": 1.4083022183175928, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.1915669441223145, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.7262871861457825, + "num_tokens": 331830095.0, + "step": 12824 + }, + { + "epoch": 1.4084120360202066, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7609946727752686, + "learning_rate": 1e-06, + "loss": 0.8818, + "mean_token_accuracy": 0.7310088872909546, + "num_tokens": 331852970.0, + "step": 12825 + }, + { + "epoch": 1.4085218537228201, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4159622192382812, + "learning_rate": 1e-06, + "loss": 0.9504, + "mean_token_accuracy": 0.7197061777114868, + "num_tokens": 331880021.0, + "step": 12826 + }, + { + "epoch": 1.4086316714254337, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.523402452468872, + "learning_rate": 1e-06, + "loss": 1.1199, + "mean_token_accuracy": 0.6806192994117737, + "num_tokens": 331909539.0, + "step": 12827 + }, + { + "epoch": 1.4087414891280474, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.411400079727173, + "learning_rate": 1e-06, + "loss": 1.1288, + "mean_token_accuracy": 0.6773239970207214, + "num_tokens": 331938496.0, + "step": 12828 + }, + { + "epoch": 1.4088513068306612, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3779687881469727, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7211689949035645, + "num_tokens": 331970350.0, + "step": 12829 + }, + { + "epoch": 1.4089611245332747, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.733802556991577, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7044222950935364, + "num_tokens": 331991869.0, + "step": 12830 + }, + { + "epoch": 1.4090709422358885, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7054176330566406, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7109609246253967, + "num_tokens": 332013233.0, + "step": 12831 + }, + { + "epoch": 1.409180759938502, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.494187355041504, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.698610782623291, + "num_tokens": 332039007.0, + "step": 12832 + }, + { + "epoch": 1.4092905776411158, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.505631923675537, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7027896642684937, + "num_tokens": 332066044.0, + "step": 12833 + }, + { + "epoch": 1.4094003953437295, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4609408378601074, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.7335996627807617, + "num_tokens": 332089319.0, + "step": 12834 + }, + { + "epoch": 1.409510213046343, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8326406478881836, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.706346869468689, + "num_tokens": 332109374.0, + "step": 12835 + }, + { + "epoch": 1.4096200307489568, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.546678066253662, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.6966450810432434, + "num_tokens": 332137601.0, + "step": 12836 + }, + { + "epoch": 1.4097298484515703, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.650909185409546, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7006435394287109, + "num_tokens": 332162374.0, + "step": 12837 + }, + { + "epoch": 1.409839666154184, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.482361316680908, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7007368803024292, + "num_tokens": 332191028.0, + "step": 12838 + }, + { + "epoch": 1.4099494838567976, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.799395799636841, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7095657587051392, + "num_tokens": 332212579.0, + "step": 12839 + }, + { + "epoch": 1.4100593015594114, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.6715426445007324, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7099184989929199, + "num_tokens": 332240660.0, + "step": 12840 + }, + { + "epoch": 1.410169119262025, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4617676734924316, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6964365243911743, + "num_tokens": 332271878.0, + "step": 12841 + }, + { + "epoch": 1.4102789369646387, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4147725105285645, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7113496661186218, + "num_tokens": 332300385.0, + "step": 12842 + }, + { + "epoch": 1.4103887546672524, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6492514610290527, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.709396481513977, + "num_tokens": 332324927.0, + "step": 12843 + }, + { + "epoch": 1.410498572369866, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6961464881896973, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7065542340278625, + "num_tokens": 332347441.0, + "step": 12844 + }, + { + "epoch": 1.4106083900724797, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8805601596832275, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7080442905426025, + "num_tokens": 332367849.0, + "step": 12845 + }, + { + "epoch": 1.4107182077750933, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5622270107269287, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.696510910987854, + "num_tokens": 332394262.0, + "step": 12846 + }, + { + "epoch": 1.410828025477707, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6727194786071777, + "learning_rate": 1e-06, + "loss": 1.0914, + "mean_token_accuracy": 0.6861050128936768, + "num_tokens": 332417902.0, + "step": 12847 + }, + { + "epoch": 1.4109378431803208, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4023828506469727, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6934811472892761, + "num_tokens": 332450973.0, + "step": 12848 + }, + { + "epoch": 1.4110476608829343, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.815124750137329, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7049319744110107, + "num_tokens": 332474891.0, + "step": 12849 + }, + { + "epoch": 1.4111574785855479, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.602233409881592, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7041360139846802, + "num_tokens": 332501341.0, + "step": 12850 + }, + { + "epoch": 1.4112672962881616, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5881500244140625, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6965317726135254, + "num_tokens": 332529082.0, + "step": 12851 + }, + { + "epoch": 1.4113771139907754, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.591019630432129, + "learning_rate": 1e-06, + "loss": 1.1412, + "mean_token_accuracy": 0.6766202449798584, + "num_tokens": 332556574.0, + "step": 12852 + }, + { + "epoch": 1.411486931693389, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.727519989013672, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.701012372970581, + "num_tokens": 332579569.0, + "step": 12853 + }, + { + "epoch": 1.4115967493960027, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.407285213470459, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6916603446006775, + "num_tokens": 332611878.0, + "step": 12854 + }, + { + "epoch": 1.4117065670986162, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.548872947692871, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6934770941734314, + "num_tokens": 332638058.0, + "step": 12855 + }, + { + "epoch": 1.41181638480123, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.848557710647583, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6855788826942444, + "num_tokens": 332666000.0, + "step": 12856 + }, + { + "epoch": 1.4119262025038437, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.51992130279541, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7118538618087769, + "num_tokens": 332694149.0, + "step": 12857 + }, + { + "epoch": 1.4120360202064572, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.606037139892578, + "learning_rate": 1e-06, + "loss": 1.1145, + "mean_token_accuracy": 0.6779630780220032, + "num_tokens": 332724188.0, + "step": 12858 + }, + { + "epoch": 1.412145837909071, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4761509895324707, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6873669624328613, + "num_tokens": 332752306.0, + "step": 12859 + }, + { + "epoch": 1.4122556556116845, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5423378944396973, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.695378303527832, + "num_tokens": 332780115.0, + "step": 12860 + }, + { + "epoch": 1.4123654733142983, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9383416175842285, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7013075351715088, + "num_tokens": 332800506.0, + "step": 12861 + }, + { + "epoch": 1.412475291016912, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.432962417602539, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6994346380233765, + "num_tokens": 332830867.0, + "step": 12862 + }, + { + "epoch": 1.4125851087195256, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7089927196502686, + "learning_rate": 1e-06, + "loss": 1.1201, + "mean_token_accuracy": 0.6724776029586792, + "num_tokens": 332860462.0, + "step": 12863 + }, + { + "epoch": 1.4126949264221391, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8163950443267822, + "learning_rate": 1e-06, + "loss": 0.9349, + "mean_token_accuracy": 0.7148007154464722, + "num_tokens": 332880760.0, + "step": 12864 + }, + { + "epoch": 1.4128047441247529, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.408578395843506, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6916379332542419, + "num_tokens": 332910473.0, + "step": 12865 + }, + { + "epoch": 1.4129145618273666, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.694822072982788, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.684383749961853, + "num_tokens": 332936802.0, + "step": 12866 + }, + { + "epoch": 1.4130243795299802, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.068498134613037, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7248046398162842, + "num_tokens": 332959432.0, + "step": 12867 + }, + { + "epoch": 1.413134197232594, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6588268280029297, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6838364005088806, + "num_tokens": 332987546.0, + "step": 12868 + }, + { + "epoch": 1.4132440149352075, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6546382904052734, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6844276785850525, + "num_tokens": 333014776.0, + "step": 12869 + }, + { + "epoch": 1.4133538326378212, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 4.6259236335754395, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7011187076568604, + "num_tokens": 333040984.0, + "step": 12870 + }, + { + "epoch": 1.413463650340435, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.689070463180542, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.716346263885498, + "num_tokens": 333064260.0, + "step": 12871 + }, + { + "epoch": 1.4135734680430485, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9679412841796875, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7167489528656006, + "num_tokens": 333085230.0, + "step": 12872 + }, + { + "epoch": 1.4136832857456623, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4709558486938477, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6932050585746765, + "num_tokens": 333111202.0, + "step": 12873 + }, + { + "epoch": 1.4137931034482758, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4699819087982178, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6987820863723755, + "num_tokens": 333135562.0, + "step": 12874 + }, + { + "epoch": 1.4139029211508896, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6151247024536133, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7025727033615112, + "num_tokens": 333162233.0, + "step": 12875 + }, + { + "epoch": 1.4140127388535033, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5723345279693604, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7047816514968872, + "num_tokens": 333187389.0, + "step": 12876 + }, + { + "epoch": 1.4141225565561168, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 9.772024154663086, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7066378593444824, + "num_tokens": 333216193.0, + "step": 12877 + }, + { + "epoch": 1.4142323742587304, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8503775596618652, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7146527171134949, + "num_tokens": 333237036.0, + "step": 12878 + }, + { + "epoch": 1.4143421919613441, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.566622257232666, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7137342691421509, + "num_tokens": 333261687.0, + "step": 12879 + }, + { + "epoch": 1.414452009663958, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7416183948516846, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.7012566328048706, + "num_tokens": 333285475.0, + "step": 12880 + }, + { + "epoch": 1.4145618273665714, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6619677543640137, + "learning_rate": 1e-06, + "loss": 1.1163, + "mean_token_accuracy": 0.6938817501068115, + "num_tokens": 333308813.0, + "step": 12881 + }, + { + "epoch": 1.4146716450691852, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5280163288116455, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7038917541503906, + "num_tokens": 333335279.0, + "step": 12882 + }, + { + "epoch": 1.4147814627717987, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5341904163360596, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6959521770477295, + "num_tokens": 333364811.0, + "step": 12883 + }, + { + "epoch": 1.4148912804744125, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.9292120933532715, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7122572660446167, + "num_tokens": 333385365.0, + "step": 12884 + }, + { + "epoch": 1.4150010981770262, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.0674867630004883, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7170456647872925, + "num_tokens": 333405477.0, + "step": 12885 + }, + { + "epoch": 1.4151109158796398, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8099710941314697, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.704718291759491, + "num_tokens": 333429113.0, + "step": 12886 + }, + { + "epoch": 1.4152207335822535, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6190717220306396, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7157069444656372, + "num_tokens": 333451628.0, + "step": 12887 + }, + { + "epoch": 1.415330551284867, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.921534299850464, + "learning_rate": 1e-06, + "loss": 0.9022, + "mean_token_accuracy": 0.726211428642273, + "num_tokens": 333469094.0, + "step": 12888 + }, + { + "epoch": 1.4154403689874808, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.751539945602417, + "learning_rate": 1e-06, + "loss": 0.9081, + "mean_token_accuracy": 0.7315553426742554, + "num_tokens": 333492280.0, + "step": 12889 + }, + { + "epoch": 1.4155501866900946, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6497836112976074, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6911419630050659, + "num_tokens": 333518127.0, + "step": 12890 + }, + { + "epoch": 1.4156600043927081, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.559262990951538, + "learning_rate": 1e-06, + "loss": 1.1305, + "mean_token_accuracy": 0.6694085597991943, + "num_tokens": 333546384.0, + "step": 12891 + }, + { + "epoch": 1.4157698220953217, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6980292797088623, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7023264169692993, + "num_tokens": 333571273.0, + "step": 12892 + }, + { + "epoch": 1.4158796397979354, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6108920574188232, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7126411199569702, + "num_tokens": 333594588.0, + "step": 12893 + }, + { + "epoch": 1.4159894575005492, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4449546337127686, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7266266345977783, + "num_tokens": 333622666.0, + "step": 12894 + }, + { + "epoch": 1.4160992752031627, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7060415744781494, + "learning_rate": 1e-06, + "loss": 0.9389, + "mean_token_accuracy": 0.7206175923347473, + "num_tokens": 333644976.0, + "step": 12895 + }, + { + "epoch": 1.4162090929057765, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.579035520553589, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.7027561068534851, + "num_tokens": 333673260.0, + "step": 12896 + }, + { + "epoch": 1.41631891060839, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.511467933654785, + "learning_rate": 1e-06, + "loss": 1.113, + "mean_token_accuracy": 0.6784018278121948, + "num_tokens": 333701124.0, + "step": 12897 + }, + { + "epoch": 1.4164287283110037, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5638651847839355, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7017390727996826, + "num_tokens": 333728031.0, + "step": 12898 + }, + { + "epoch": 1.4165385460136175, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8046813011169434, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7186450958251953, + "num_tokens": 333753218.0, + "step": 12899 + }, + { + "epoch": 1.416648363716231, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8615026473999023, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.7020676136016846, + "num_tokens": 333780133.0, + "step": 12900 + }, + { + "epoch": 1.4167581814188448, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6743686199188232, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6942772269248962, + "num_tokens": 333805585.0, + "step": 12901 + }, + { + "epoch": 1.4168679991214583, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5338337421417236, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6954135894775391, + "num_tokens": 333832788.0, + "step": 12902 + }, + { + "epoch": 1.416977816824072, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4607810974121094, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7124211192131042, + "num_tokens": 333861074.0, + "step": 12903 + }, + { + "epoch": 1.4170876345266856, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.799367666244507, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7032372951507568, + "num_tokens": 333883421.0, + "step": 12904 + }, + { + "epoch": 1.4171974522292994, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8809988498687744, + "learning_rate": 1e-06, + "loss": 0.8515, + "mean_token_accuracy": 0.7444696426391602, + "num_tokens": 333902876.0, + "step": 12905 + }, + { + "epoch": 1.417307269931913, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.409230947494507, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7018249034881592, + "num_tokens": 333932878.0, + "step": 12906 + }, + { + "epoch": 1.4174170876345267, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3935599327087402, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7016726136207581, + "num_tokens": 333964610.0, + "step": 12907 + }, + { + "epoch": 1.4175269053371404, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5941574573516846, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6926512718200684, + "num_tokens": 333990179.0, + "step": 12908 + }, + { + "epoch": 1.417636723039754, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6714608669281006, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7133655548095703, + "num_tokens": 334014098.0, + "step": 12909 + }, + { + "epoch": 1.4177465407423677, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7990963459014893, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.709154486656189, + "num_tokens": 334035745.0, + "step": 12910 + }, + { + "epoch": 1.4178563584449813, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.446239948272705, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7062039375305176, + "num_tokens": 334063259.0, + "step": 12911 + }, + { + "epoch": 1.417966176147595, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3412697315216064, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7047906517982483, + "num_tokens": 334092798.0, + "step": 12912 + }, + { + "epoch": 1.4180759938502088, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3921520709991455, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.714889407157898, + "num_tokens": 334121983.0, + "step": 12913 + }, + { + "epoch": 1.4181858115528223, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4741194248199463, + "learning_rate": 1e-06, + "loss": 1.1049, + "mean_token_accuracy": 0.6850795149803162, + "num_tokens": 334150366.0, + "step": 12914 + }, + { + "epoch": 1.4182956292554358, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5756795406341553, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7145802974700928, + "num_tokens": 334173217.0, + "step": 12915 + }, + { + "epoch": 1.4184054469580496, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3760855197906494, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.6813144683837891, + "num_tokens": 334203261.0, + "step": 12916 + }, + { + "epoch": 1.4185152646606634, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6474595069885254, + "learning_rate": 1e-06, + "loss": 1.0727, + "mean_token_accuracy": 0.6862711906433105, + "num_tokens": 334229166.0, + "step": 12917 + }, + { + "epoch": 1.418625082363277, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.458064079284668, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.6807407140731812, + "num_tokens": 334258769.0, + "step": 12918 + }, + { + "epoch": 1.4187349000658906, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.402761220932007, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6994674801826477, + "num_tokens": 334287469.0, + "step": 12919 + }, + { + "epoch": 1.4188447177685042, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.529101848602295, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.703270435333252, + "num_tokens": 334314416.0, + "step": 12920 + }, + { + "epoch": 1.418954535471118, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3849565982818604, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7151467800140381, + "num_tokens": 334339914.0, + "step": 12921 + }, + { + "epoch": 1.4190643531737317, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.694420576095581, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7064903974533081, + "num_tokens": 334362130.0, + "step": 12922 + }, + { + "epoch": 1.4191741708763452, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2529635429382324, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6916824579238892, + "num_tokens": 334394745.0, + "step": 12923 + }, + { + "epoch": 1.419283988578959, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5753977298736572, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.711087703704834, + "num_tokens": 334422706.0, + "step": 12924 + }, + { + "epoch": 1.4193938062815725, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4727399349212646, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6957657337188721, + "num_tokens": 334448475.0, + "step": 12925 + }, + { + "epoch": 1.4195036239841863, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4725918769836426, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7143344879150391, + "num_tokens": 334473179.0, + "step": 12926 + }, + { + "epoch": 1.4196134416868, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.649745464324951, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7048462629318237, + "num_tokens": 334499427.0, + "step": 12927 + }, + { + "epoch": 1.4197232593894136, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4888999462127686, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6905896663665771, + "num_tokens": 334531301.0, + "step": 12928 + }, + { + "epoch": 1.419833077092027, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6576907634735107, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7035945057868958, + "num_tokens": 334555507.0, + "step": 12929 + }, + { + "epoch": 1.4199428947946409, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.276010036468506, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.6997356414794922, + "num_tokens": 334587652.0, + "step": 12930 + }, + { + "epoch": 1.4200527124972546, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8228392601013184, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7048101425170898, + "num_tokens": 334610022.0, + "step": 12931 + }, + { + "epoch": 1.4201625301998682, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5642824172973633, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6901350021362305, + "num_tokens": 334639176.0, + "step": 12932 + }, + { + "epoch": 1.420272347902482, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.883568525314331, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.720614492893219, + "num_tokens": 334658933.0, + "step": 12933 + }, + { + "epoch": 1.4203821656050954, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.579411506652832, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6865373253822327, + "num_tokens": 334687502.0, + "step": 12934 + }, + { + "epoch": 1.4204919833077092, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.2035984992980957, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6946719884872437, + "num_tokens": 334719409.0, + "step": 12935 + }, + { + "epoch": 1.420601801010323, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4812729358673096, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.692744791507721, + "num_tokens": 334749363.0, + "step": 12936 + }, + { + "epoch": 1.4207116187129365, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4709603786468506, + "learning_rate": 1e-06, + "loss": 1.1084, + "mean_token_accuracy": 0.6808338165283203, + "num_tokens": 334780552.0, + "step": 12937 + }, + { + "epoch": 1.4208214364155503, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.504568099975586, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6888970136642456, + "num_tokens": 334808206.0, + "step": 12938 + }, + { + "epoch": 1.4209312541181638, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6519038677215576, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7037153840065002, + "num_tokens": 334831561.0, + "step": 12939 + }, + { + "epoch": 1.4210410718207775, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4962058067321777, + "learning_rate": 1e-06, + "loss": 1.1511, + "mean_token_accuracy": 0.6727021336555481, + "num_tokens": 334859188.0, + "step": 12940 + }, + { + "epoch": 1.4211508895233913, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.579986333847046, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7135710120201111, + "num_tokens": 334884809.0, + "step": 12941 + }, + { + "epoch": 1.4212607072260048, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5961925983428955, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6992465257644653, + "num_tokens": 334910764.0, + "step": 12942 + }, + { + "epoch": 1.4213705249286184, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.63492488861084, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.7088834047317505, + "num_tokens": 334935069.0, + "step": 12943 + }, + { + "epoch": 1.4214803426312321, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.8183505535125732, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7021527886390686, + "num_tokens": 334961762.0, + "step": 12944 + }, + { + "epoch": 1.4215901603338459, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6869847774505615, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.710241436958313, + "num_tokens": 334988095.0, + "step": 12945 + }, + { + "epoch": 1.4216999780364594, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3847849369049072, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7188960313796997, + "num_tokens": 335015639.0, + "step": 12946 + }, + { + "epoch": 1.4218097957390732, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6907780170440674, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7149383425712585, + "num_tokens": 335038331.0, + "step": 12947 + }, + { + "epoch": 1.4219196134416867, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5866897106170654, + "learning_rate": 1e-06, + "loss": 1.1232, + "mean_token_accuracy": 0.684029221534729, + "num_tokens": 335064612.0, + "step": 12948 + }, + { + "epoch": 1.4220294311443005, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5740857124328613, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7107788324356079, + "num_tokens": 335088936.0, + "step": 12949 + }, + { + "epoch": 1.4221392488469142, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5252814292907715, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.7249352931976318, + "num_tokens": 335111531.0, + "step": 12950 + }, + { + "epoch": 1.4222490665495278, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5547003746032715, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7015636563301086, + "num_tokens": 335138758.0, + "step": 12951 + }, + { + "epoch": 1.4223588842521415, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5786759853363037, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7095009088516235, + "num_tokens": 335165080.0, + "step": 12952 + }, + { + "epoch": 1.422468701954755, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6870739459991455, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6933237910270691, + "num_tokens": 335189641.0, + "step": 12953 + }, + { + "epoch": 1.4225785196573688, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.344562530517578, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6936244368553162, + "num_tokens": 335219361.0, + "step": 12954 + }, + { + "epoch": 1.4226883373599823, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7836127281188965, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7045700550079346, + "num_tokens": 335242980.0, + "step": 12955 + }, + { + "epoch": 1.422798155062596, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.818572998046875, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6918728351593018, + "num_tokens": 335266402.0, + "step": 12956 + }, + { + "epoch": 1.4229079727652096, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.812105417251587, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7180466651916504, + "num_tokens": 335287588.0, + "step": 12957 + }, + { + "epoch": 1.4230177904678234, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6721863746643066, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7141018509864807, + "num_tokens": 335312230.0, + "step": 12958 + }, + { + "epoch": 1.4231276081704372, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.324719190597534, + "learning_rate": 1e-06, + "loss": 0.8735, + "mean_token_accuracy": 0.7360728979110718, + "num_tokens": 335339674.0, + "step": 12959 + }, + { + "epoch": 1.4232374258730507, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6460275650024414, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7010413408279419, + "num_tokens": 335366323.0, + "step": 12960 + }, + { + "epoch": 1.4233472435756644, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8476481437683105, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7231021523475647, + "num_tokens": 335388086.0, + "step": 12961 + }, + { + "epoch": 1.423457061278278, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.552574872970581, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7183910012245178, + "num_tokens": 335416717.0, + "step": 12962 + }, + { + "epoch": 1.4235668789808917, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4352142810821533, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.7006040215492249, + "num_tokens": 335448787.0, + "step": 12963 + }, + { + "epoch": 1.4236766966835055, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.038809299468994, + "learning_rate": 1e-06, + "loss": 0.85, + "mean_token_accuracy": 0.7476582527160645, + "num_tokens": 335471561.0, + "step": 12964 + }, + { + "epoch": 1.423786514386119, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5817205905914307, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7074211239814758, + "num_tokens": 335498861.0, + "step": 12965 + }, + { + "epoch": 1.4238963320887328, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.958944320678711, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7146905660629272, + "num_tokens": 335518377.0, + "step": 12966 + }, + { + "epoch": 1.4240061497913463, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.782885789871216, + "learning_rate": 1e-06, + "loss": 0.941, + "mean_token_accuracy": 0.7188838720321655, + "num_tokens": 335539438.0, + "step": 12967 + }, + { + "epoch": 1.42411596749396, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4983208179473877, + "learning_rate": 1e-06, + "loss": 0.8603, + "mean_token_accuracy": 0.7457190155982971, + "num_tokens": 335563992.0, + "step": 12968 + }, + { + "epoch": 1.4242257851965736, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5066020488739014, + "learning_rate": 1e-06, + "loss": 1.1074, + "mean_token_accuracy": 0.6849528551101685, + "num_tokens": 335595342.0, + "step": 12969 + }, + { + "epoch": 1.4243356028991874, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.533548355102539, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6854131817817688, + "num_tokens": 335623167.0, + "step": 12970 + }, + { + "epoch": 1.424445420601801, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6776087284088135, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7217162847518921, + "num_tokens": 335645301.0, + "step": 12971 + }, + { + "epoch": 1.4245552383044147, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.1844334602355957, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7181102633476257, + "num_tokens": 335668265.0, + "step": 12972 + }, + { + "epoch": 1.4246650560070284, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.413438320159912, + "learning_rate": 1e-06, + "loss": 1.1468, + "mean_token_accuracy": 0.6716832518577576, + "num_tokens": 335702380.0, + "step": 12973 + }, + { + "epoch": 1.424774873709642, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.610152244567871, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6989243030548096, + "num_tokens": 335728746.0, + "step": 12974 + }, + { + "epoch": 1.4248846914122557, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.798731803894043, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6993923187255859, + "num_tokens": 335760911.0, + "step": 12975 + }, + { + "epoch": 1.4249945091148692, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8503293991088867, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7185702323913574, + "num_tokens": 335787964.0, + "step": 12976 + }, + { + "epoch": 1.425104326817483, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 8.055075645446777, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.707428514957428, + "num_tokens": 335808830.0, + "step": 12977 + }, + { + "epoch": 1.4252141445200968, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8841392993927, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.694393515586853, + "num_tokens": 335838235.0, + "step": 12978 + }, + { + "epoch": 1.4253239622227103, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7934155464172363, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7064424157142639, + "num_tokens": 335861638.0, + "step": 12979 + }, + { + "epoch": 1.4254337799253238, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.63494610786438, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6986868381500244, + "num_tokens": 335887560.0, + "step": 12980 + }, + { + "epoch": 1.4255435976279376, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4115259647369385, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6980624198913574, + "num_tokens": 335916147.0, + "step": 12981 + }, + { + "epoch": 1.4256534153305513, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3804783821105957, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6829395890235901, + "num_tokens": 335945961.0, + "step": 12982 + }, + { + "epoch": 1.4257632330331649, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5882351398468018, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7133929133415222, + "num_tokens": 335968366.0, + "step": 12983 + }, + { + "epoch": 1.4258730507357786, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.666384696960449, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7242935299873352, + "num_tokens": 335989642.0, + "step": 12984 + }, + { + "epoch": 1.4259828684383922, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.384178400039673, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7009619474411011, + "num_tokens": 336017457.0, + "step": 12985 + }, + { + "epoch": 1.426092686141006, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5565459728240967, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7036089301109314, + "num_tokens": 336043602.0, + "step": 12986 + }, + { + "epoch": 1.4262025038436197, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5838751792907715, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7135255336761475, + "num_tokens": 336070210.0, + "step": 12987 + }, + { + "epoch": 1.4263123215462332, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.531214714050293, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7095726132392883, + "num_tokens": 336097055.0, + "step": 12988 + }, + { + "epoch": 1.426422139248847, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4775710105895996, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6958253383636475, + "num_tokens": 336124255.0, + "step": 12989 + }, + { + "epoch": 1.4265319569514605, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5274691581726074, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7039533853530884, + "num_tokens": 336151343.0, + "step": 12990 + }, + { + "epoch": 1.4266417746540743, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3708431720733643, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7046798467636108, + "num_tokens": 336183647.0, + "step": 12991 + }, + { + "epoch": 1.426751592356688, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0820791721343994, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7120586037635803, + "num_tokens": 336204677.0, + "step": 12992 + }, + { + "epoch": 1.4268614100593016, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6950600147247314, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.696448028087616, + "num_tokens": 336232008.0, + "step": 12993 + }, + { + "epoch": 1.426971227761915, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.294334888458252, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6854150295257568, + "num_tokens": 336265650.0, + "step": 12994 + }, + { + "epoch": 1.4270810454645289, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5256597995758057, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6861748695373535, + "num_tokens": 336293078.0, + "step": 12995 + }, + { + "epoch": 1.4271908631671426, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3559484481811523, + "learning_rate": 1e-06, + "loss": 1.0909, + "mean_token_accuracy": 0.6895260810852051, + "num_tokens": 336325403.0, + "step": 12996 + }, + { + "epoch": 1.4273006808697561, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6621105670928955, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7018447518348694, + "num_tokens": 336350773.0, + "step": 12997 + }, + { + "epoch": 1.42741049857237, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 8.053731918334961, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7093570232391357, + "num_tokens": 336375432.0, + "step": 12998 + }, + { + "epoch": 1.4275203162749834, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6292383670806885, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7112918496131897, + "num_tokens": 336402467.0, + "step": 12999 + }, + { + "epoch": 1.4276301339775972, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.683253049850464, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.710157573223114, + "num_tokens": 336426745.0, + "step": 13000 + }, + { + "epoch": 1.427739951680211, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.400265693664551, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.705362856388092, + "num_tokens": 336456691.0, + "step": 13001 + }, + { + "epoch": 1.4278497693828245, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.4604685306549072, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7252320051193237, + "num_tokens": 336481532.0, + "step": 13002 + }, + { + "epoch": 1.4279595870854382, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.703047275543213, + "learning_rate": 1e-06, + "loss": 0.9041, + "mean_token_accuracy": 0.7337969541549683, + "num_tokens": 336502282.0, + "step": 13003 + }, + { + "epoch": 1.4280694047880518, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.264594316482544, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.692689061164856, + "num_tokens": 336536836.0, + "step": 13004 + }, + { + "epoch": 1.4281792224906655, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.896169424057007, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.718967080116272, + "num_tokens": 336559622.0, + "step": 13005 + }, + { + "epoch": 1.4282890401932793, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.443288803100586, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7170705795288086, + "num_tokens": 336586090.0, + "step": 13006 + }, + { + "epoch": 1.4283988578958928, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5859084129333496, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7114616632461548, + "num_tokens": 336610788.0, + "step": 13007 + }, + { + "epoch": 1.4285086755985064, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.42891526222229, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7124316692352295, + "num_tokens": 336637114.0, + "step": 13008 + }, + { + "epoch": 1.4286184933011201, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.620732545852661, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7039628028869629, + "num_tokens": 336662254.0, + "step": 13009 + }, + { + "epoch": 1.4287283110037339, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.320209503173828, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6903758645057678, + "num_tokens": 336692324.0, + "step": 13010 + }, + { + "epoch": 1.4288381287063474, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.653498649597168, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6936764121055603, + "num_tokens": 336719362.0, + "step": 13011 + }, + { + "epoch": 1.4289479464089612, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 3.080587148666382, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7073873281478882, + "num_tokens": 336739122.0, + "step": 13012 + }, + { + "epoch": 1.4290577641115747, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3497912883758545, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.6842151284217834, + "num_tokens": 336772121.0, + "step": 13013 + }, + { + "epoch": 1.4291675818141885, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.5158934593200684, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6950114965438843, + "num_tokens": 336796925.0, + "step": 13014 + }, + { + "epoch": 1.4292773995168022, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.705183506011963, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.69901442527771, + "num_tokens": 336821568.0, + "step": 13015 + }, + { + "epoch": 1.4293872172194158, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.607492685317993, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7032738327980042, + "num_tokens": 336847377.0, + "step": 13016 + }, + { + "epoch": 1.4294970349220295, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.599217653274536, + "learning_rate": 1e-06, + "loss": 1.1098, + "mean_token_accuracy": 0.6790554523468018, + "num_tokens": 336872968.0, + "step": 13017 + }, + { + "epoch": 1.429606852624643, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.7447972297668457, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6918275356292725, + "num_tokens": 336897484.0, + "step": 13018 + }, + { + "epoch": 1.4297166703272568, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.3895554542541504, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.7259776592254639, + "num_tokens": 336925818.0, + "step": 13019 + }, + { + "epoch": 1.4298264880298703, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.6546783447265625, + "learning_rate": 1e-06, + "loss": 0.9421, + "mean_token_accuracy": 0.7198657989501953, + "num_tokens": 336949216.0, + "step": 13020 + }, + { + "epoch": 1.429936305732484, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.897397518157959, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6938014626502991, + "num_tokens": 336970982.0, + "step": 13021 + }, + { + "epoch": 1.4300461234350976, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.362252950668335, + "learning_rate": 1e-06, + "loss": 1.1233, + "mean_token_accuracy": 0.6905168294906616, + "num_tokens": 337003995.0, + "step": 13022 + }, + { + "epoch": 1.4301559411377114, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.990339994430542, + "learning_rate": 1e-06, + "loss": 0.9271, + "mean_token_accuracy": 0.72980797290802, + "num_tokens": 337022070.0, + "step": 13023 + }, + { + "epoch": 1.4302657588403251, + "ewc_loss": 9.357929229736328e-06, + "grad_norm": 2.648585319519043, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7103195786476135, + "num_tokens": 337044754.0, + "step": 13024 + }, + { + "epoch": 1.4303755765429387, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7766263484954834, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6892895698547363, + "num_tokens": 337071397.0, + "step": 13025 + }, + { + "epoch": 1.4304853942455524, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6124372482299805, + "learning_rate": 1e-06, + "loss": 1.0994, + "mean_token_accuracy": 0.6877161860466003, + "num_tokens": 337097440.0, + "step": 13026 + }, + { + "epoch": 1.430595211948166, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.687070369720459, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.705374538898468, + "num_tokens": 337122037.0, + "step": 13027 + }, + { + "epoch": 1.4307050296507797, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.537369966506958, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7041854858398438, + "num_tokens": 337147695.0, + "step": 13028 + }, + { + "epoch": 1.4308148473533935, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.691896438598633, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7133373022079468, + "num_tokens": 337174018.0, + "step": 13029 + }, + { + "epoch": 1.430924665056007, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5280234813690186, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7008376121520996, + "num_tokens": 337201036.0, + "step": 13030 + }, + { + "epoch": 1.4310344827586206, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5274503231048584, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7105368375778198, + "num_tokens": 337227358.0, + "step": 13031 + }, + { + "epoch": 1.4311443004612343, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5966858863830566, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7037575244903564, + "num_tokens": 337251693.0, + "step": 13032 + }, + { + "epoch": 1.431254118163848, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5996809005737305, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7120853662490845, + "num_tokens": 337278719.0, + "step": 13033 + }, + { + "epoch": 1.4313639358664616, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.423945426940918, + "learning_rate": 1e-06, + "loss": 1.1115, + "mean_token_accuracy": 0.6800363063812256, + "num_tokens": 337310203.0, + "step": 13034 + }, + { + "epoch": 1.4314737535690754, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7223849296569824, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.705552339553833, + "num_tokens": 337332944.0, + "step": 13035 + }, + { + "epoch": 1.431583571271689, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.728799819946289, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7046636343002319, + "num_tokens": 337353080.0, + "step": 13036 + }, + { + "epoch": 1.4316933889743026, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.214343309402466, + "learning_rate": 1e-06, + "loss": 1.1326, + "mean_token_accuracy": 0.6731694936752319, + "num_tokens": 337388005.0, + "step": 13037 + }, + { + "epoch": 1.4318032066769164, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.706526041030884, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7056005597114563, + "num_tokens": 337412486.0, + "step": 13038 + }, + { + "epoch": 1.43191302437953, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.415505886077881, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6918438076972961, + "num_tokens": 337442562.0, + "step": 13039 + }, + { + "epoch": 1.4320228420821437, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9271490573883057, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.726871907711029, + "num_tokens": 337462022.0, + "step": 13040 + }, + { + "epoch": 1.4321326597847572, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5196309089660645, + "learning_rate": 1e-06, + "loss": 1.1317, + "mean_token_accuracy": 0.6726028919219971, + "num_tokens": 337489573.0, + "step": 13041 + }, + { + "epoch": 1.432242477487371, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5015923976898193, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6938748359680176, + "num_tokens": 337517170.0, + "step": 13042 + }, + { + "epoch": 1.4323522951899847, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7303459644317627, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.7005356550216675, + "num_tokens": 337550404.0, + "step": 13043 + }, + { + "epoch": 1.4324621128925983, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.681196928024292, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7099007368087769, + "num_tokens": 337573747.0, + "step": 13044 + }, + { + "epoch": 1.4325719305952118, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.881268262863159, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.716011643409729, + "num_tokens": 337596783.0, + "step": 13045 + }, + { + "epoch": 1.4326817482978256, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.749706983566284, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7096711993217468, + "num_tokens": 337619950.0, + "step": 13046 + }, + { + "epoch": 1.4327915660004393, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.299197196960449, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7236777544021606, + "num_tokens": 337650293.0, + "step": 13047 + }, + { + "epoch": 1.4329013837030529, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.484241008758545, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7069184184074402, + "num_tokens": 337680476.0, + "step": 13048 + }, + { + "epoch": 1.4330112014056666, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6477973461151123, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7055679559707642, + "num_tokens": 337702464.0, + "step": 13049 + }, + { + "epoch": 1.4331210191082802, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4140872955322266, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7055686712265015, + "num_tokens": 337729715.0, + "step": 13050 + }, + { + "epoch": 1.433230836810894, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4376111030578613, + "learning_rate": 1e-06, + "loss": 1.0679, + "mean_token_accuracy": 0.6856457591056824, + "num_tokens": 337757375.0, + "step": 13051 + }, + { + "epoch": 1.4333406545135077, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.586670398712158, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7149802446365356, + "num_tokens": 337784525.0, + "step": 13052 + }, + { + "epoch": 1.4334504722161212, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8829808235168457, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7138547301292419, + "num_tokens": 337807150.0, + "step": 13053 + }, + { + "epoch": 1.433560289918735, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7040367126464844, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.695398211479187, + "num_tokens": 337830400.0, + "step": 13054 + }, + { + "epoch": 1.4336701076213485, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.590144634246826, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7151954174041748, + "num_tokens": 337854411.0, + "step": 13055 + }, + { + "epoch": 1.4337799253239623, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.807138442993164, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7157081365585327, + "num_tokens": 337880853.0, + "step": 13056 + }, + { + "epoch": 1.433889743026576, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7888195514678955, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7089661359786987, + "num_tokens": 337906239.0, + "step": 13057 + }, + { + "epoch": 1.4339995607291895, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6771769523620605, + "learning_rate": 1e-06, + "loss": 0.8939, + "mean_token_accuracy": 0.7317652702331543, + "num_tokens": 337928118.0, + "step": 13058 + }, + { + "epoch": 1.434109378431803, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4814860820770264, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6954034566879272, + "num_tokens": 337957483.0, + "step": 13059 + }, + { + "epoch": 1.4342191961344168, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.442173480987549, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7112799882888794, + "num_tokens": 337986256.0, + "step": 13060 + }, + { + "epoch": 1.4343290138370306, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3972713947296143, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.719092845916748, + "num_tokens": 338016447.0, + "step": 13061 + }, + { + "epoch": 1.4344388315396441, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.796570062637329, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6810139417648315, + "num_tokens": 338041049.0, + "step": 13062 + }, + { + "epoch": 1.434548649242258, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4460275173187256, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7265762090682983, + "num_tokens": 338069754.0, + "step": 13063 + }, + { + "epoch": 1.4346584669448714, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7520675659179688, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7165100574493408, + "num_tokens": 338090336.0, + "step": 13064 + }, + { + "epoch": 1.4347682846474852, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5192408561706543, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7189990282058716, + "num_tokens": 338118398.0, + "step": 13065 + }, + { + "epoch": 1.434878102350099, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.470329761505127, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.718588650226593, + "num_tokens": 338144635.0, + "step": 13066 + }, + { + "epoch": 1.4349879200527125, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5319900512695312, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6931061744689941, + "num_tokens": 338172175.0, + "step": 13067 + }, + { + "epoch": 1.4350977377553262, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5093302726745605, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7019174098968506, + "num_tokens": 338200321.0, + "step": 13068 + }, + { + "epoch": 1.4352075554579398, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6047887802124023, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7115054130554199, + "num_tokens": 338224147.0, + "step": 13069 + }, + { + "epoch": 1.4353173731605535, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.327562093734741, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7138567566871643, + "num_tokens": 338253884.0, + "step": 13070 + }, + { + "epoch": 1.4354271908631673, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3562963008880615, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.6997852921485901, + "num_tokens": 338281900.0, + "step": 13071 + }, + { + "epoch": 1.4355370085657808, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7476725578308105, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7073412537574768, + "num_tokens": 338305899.0, + "step": 13072 + }, + { + "epoch": 1.4356468262683943, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.216048002243042, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6955879330635071, + "num_tokens": 338344227.0, + "step": 13073 + }, + { + "epoch": 1.435756643971008, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.810044527053833, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.6992481350898743, + "num_tokens": 338369871.0, + "step": 13074 + }, + { + "epoch": 1.4358664616736219, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6306450366973877, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7013938426971436, + "num_tokens": 338396378.0, + "step": 13075 + }, + { + "epoch": 1.4359762793762354, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6803863048553467, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7050879001617432, + "num_tokens": 338418917.0, + "step": 13076 + }, + { + "epoch": 1.4360860970788492, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.582797050476074, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7100784778594971, + "num_tokens": 338443366.0, + "step": 13077 + }, + { + "epoch": 1.4361959147814627, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7319681644439697, + "learning_rate": 1e-06, + "loss": 0.8984, + "mean_token_accuracy": 0.7318453788757324, + "num_tokens": 338464714.0, + "step": 13078 + }, + { + "epoch": 1.4363057324840764, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5159504413604736, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6926400661468506, + "num_tokens": 338496171.0, + "step": 13079 + }, + { + "epoch": 1.4364155501866902, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.670051097869873, + "learning_rate": 1e-06, + "loss": 0.922, + "mean_token_accuracy": 0.7259738445281982, + "num_tokens": 338521962.0, + "step": 13080 + }, + { + "epoch": 1.4365253678893037, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5417582988739014, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7050678730010986, + "num_tokens": 338547278.0, + "step": 13081 + }, + { + "epoch": 1.4366351855919175, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6310362815856934, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7081019282341003, + "num_tokens": 338574124.0, + "step": 13082 + }, + { + "epoch": 1.436745003294531, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8037943840026855, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7150284647941589, + "num_tokens": 338598037.0, + "step": 13083 + }, + { + "epoch": 1.4368548209971448, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7192494869232178, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6913058161735535, + "num_tokens": 338629611.0, + "step": 13084 + }, + { + "epoch": 1.4369646386997583, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.667232036590576, + "learning_rate": 1e-06, + "loss": 0.8741, + "mean_token_accuracy": 0.7458380460739136, + "num_tokens": 338652735.0, + "step": 13085 + }, + { + "epoch": 1.437074456402372, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8972346782684326, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7128945589065552, + "num_tokens": 338673571.0, + "step": 13086 + }, + { + "epoch": 1.4371842741049856, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5093207359313965, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6931341290473938, + "num_tokens": 338700594.0, + "step": 13087 + }, + { + "epoch": 1.4372940918075994, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8322644233703613, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7138204574584961, + "num_tokens": 338721880.0, + "step": 13088 + }, + { + "epoch": 1.4374039095102131, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.772862434387207, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7198995351791382, + "num_tokens": 338745508.0, + "step": 13089 + }, + { + "epoch": 1.4375137272128267, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7290031909942627, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7064754962921143, + "num_tokens": 338769394.0, + "step": 13090 + }, + { + "epoch": 1.4376235449154404, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.477854013442993, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.7020643949508667, + "num_tokens": 338799700.0, + "step": 13091 + }, + { + "epoch": 1.437733362618054, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7555363178253174, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7053373456001282, + "num_tokens": 338823255.0, + "step": 13092 + }, + { + "epoch": 1.4378431803206677, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.442065715789795, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6988651752471924, + "num_tokens": 338855495.0, + "step": 13093 + }, + { + "epoch": 1.4379529980232815, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.758296012878418, + "learning_rate": 1e-06, + "loss": 0.8206, + "mean_token_accuracy": 0.7566865086555481, + "num_tokens": 338873672.0, + "step": 13094 + }, + { + "epoch": 1.438062815725895, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.546783208847046, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6850270628929138, + "num_tokens": 338900914.0, + "step": 13095 + }, + { + "epoch": 1.4381726334285085, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.823380708694458, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7021621465682983, + "num_tokens": 338926093.0, + "step": 13096 + }, + { + "epoch": 1.4382824511311223, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.248061180114746, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.694223165512085, + "num_tokens": 338959557.0, + "step": 13097 + }, + { + "epoch": 1.438392268833736, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5411646366119385, + "learning_rate": 1e-06, + "loss": 1.1033, + "mean_token_accuracy": 0.6836292743682861, + "num_tokens": 338995283.0, + "step": 13098 + }, + { + "epoch": 1.4385020865363496, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8975038528442383, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6886325478553772, + "num_tokens": 339017549.0, + "step": 13099 + }, + { + "epoch": 1.4386119042389633, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6841495037078857, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7032092213630676, + "num_tokens": 339045339.0, + "step": 13100 + }, + { + "epoch": 1.4387217219415769, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4529850482940674, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6994307041168213, + "num_tokens": 339076105.0, + "step": 13101 + }, + { + "epoch": 1.4388315396441906, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5868077278137207, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6889245510101318, + "num_tokens": 339101917.0, + "step": 13102 + }, + { + "epoch": 1.4389413573468044, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6295480728149414, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7127758264541626, + "num_tokens": 339125277.0, + "step": 13103 + }, + { + "epoch": 1.439051175049418, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.08206844329834, + "learning_rate": 1e-06, + "loss": 0.9195, + "mean_token_accuracy": 0.7257518768310547, + "num_tokens": 339142988.0, + "step": 13104 + }, + { + "epoch": 1.4391609927520317, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8339784145355225, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.6984449625015259, + "num_tokens": 339164220.0, + "step": 13105 + }, + { + "epoch": 1.4392708104546452, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4247124195098877, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6945693492889404, + "num_tokens": 339194424.0, + "step": 13106 + }, + { + "epoch": 1.439380628157259, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.676748275756836, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6895454525947571, + "num_tokens": 339219591.0, + "step": 13107 + }, + { + "epoch": 1.4394904458598727, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4328792095184326, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6939152479171753, + "num_tokens": 339250645.0, + "step": 13108 + }, + { + "epoch": 1.4396002635624863, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.597990036010742, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6965996623039246, + "num_tokens": 339276743.0, + "step": 13109 + }, + { + "epoch": 1.4397100812650998, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7903356552124023, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7188436985015869, + "num_tokens": 339297691.0, + "step": 13110 + }, + { + "epoch": 1.4398198989677136, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.635580539703369, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7092139720916748, + "num_tokens": 339320684.0, + "step": 13111 + }, + { + "epoch": 1.4399297166703273, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6826279163360596, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7070299386978149, + "num_tokens": 339345478.0, + "step": 13112 + }, + { + "epoch": 1.4400395343729409, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.749058723449707, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7096775770187378, + "num_tokens": 339368782.0, + "step": 13113 + }, + { + "epoch": 1.4401493520755546, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6139047145843506, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.689662516117096, + "num_tokens": 339396266.0, + "step": 13114 + }, + { + "epoch": 1.4402591697781681, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6283113956451416, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7144391536712646, + "num_tokens": 339425471.0, + "step": 13115 + }, + { + "epoch": 1.440368987480782, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5181853771209717, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7265679240226746, + "num_tokens": 339448977.0, + "step": 13116 + }, + { + "epoch": 1.4404788051833957, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.642962694168091, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7142841815948486, + "num_tokens": 339471164.0, + "step": 13117 + }, + { + "epoch": 1.4405886228860092, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.552597999572754, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7095786929130554, + "num_tokens": 339497571.0, + "step": 13118 + }, + { + "epoch": 1.440698440588623, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.672649621963501, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7124931812286377, + "num_tokens": 339522326.0, + "step": 13119 + }, + { + "epoch": 1.4408082582912365, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5687508583068848, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.6988544464111328, + "num_tokens": 339547243.0, + "step": 13120 + }, + { + "epoch": 1.4409180759938502, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6878910064697266, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.706304669380188, + "num_tokens": 339570558.0, + "step": 13121 + }, + { + "epoch": 1.441027893696464, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6134896278381348, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7237958908081055, + "num_tokens": 339592445.0, + "step": 13122 + }, + { + "epoch": 1.4411377113990775, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.579352378845215, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6965857744216919, + "num_tokens": 339617239.0, + "step": 13123 + }, + { + "epoch": 1.441247529101691, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.759613037109375, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7034905552864075, + "num_tokens": 339639633.0, + "step": 13124 + }, + { + "epoch": 1.4413573468043048, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.63429856300354, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7117288112640381, + "num_tokens": 339663745.0, + "step": 13125 + }, + { + "epoch": 1.4414671645069186, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.818498373031616, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7129837274551392, + "num_tokens": 339683723.0, + "step": 13126 + }, + { + "epoch": 1.4415769822095321, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.942396640777588, + "learning_rate": 1e-06, + "loss": 0.9227, + "mean_token_accuracy": 0.7281313538551331, + "num_tokens": 339702235.0, + "step": 13127 + }, + { + "epoch": 1.4416867999121459, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5006771087646484, + "learning_rate": 1e-06, + "loss": 0.9089, + "mean_token_accuracy": 0.729537844657898, + "num_tokens": 339727218.0, + "step": 13128 + }, + { + "epoch": 1.4417966176147594, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3482563495635986, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7117151021957397, + "num_tokens": 339755696.0, + "step": 13129 + }, + { + "epoch": 1.4419064353173732, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.467298984527588, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7023611068725586, + "num_tokens": 339781899.0, + "step": 13130 + }, + { + "epoch": 1.442016253019987, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5801916122436523, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7274566292762756, + "num_tokens": 339805352.0, + "step": 13131 + }, + { + "epoch": 1.4421260707226005, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6059517860412598, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7246540188789368, + "num_tokens": 339829741.0, + "step": 13132 + }, + { + "epoch": 1.4422358884252142, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.645054578781128, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7149527072906494, + "num_tokens": 339852460.0, + "step": 13133 + }, + { + "epoch": 1.4423457061278278, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6132631301879883, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6942052841186523, + "num_tokens": 339877706.0, + "step": 13134 + }, + { + "epoch": 1.4424555238304415, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.594589948654175, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7183294892311096, + "num_tokens": 339901952.0, + "step": 13135 + }, + { + "epoch": 1.442565341533055, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6540729999542236, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7115681171417236, + "num_tokens": 339924641.0, + "step": 13136 + }, + { + "epoch": 1.4426751592356688, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.323395252227783, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7077276706695557, + "num_tokens": 339955650.0, + "step": 13137 + }, + { + "epoch": 1.4427849769382823, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5369043350219727, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.7007142305374146, + "num_tokens": 339982083.0, + "step": 13138 + }, + { + "epoch": 1.442894794640896, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7709028720855713, + "learning_rate": 1e-06, + "loss": 1.1221, + "mean_token_accuracy": 0.6814432144165039, + "num_tokens": 340003890.0, + "step": 13139 + }, + { + "epoch": 1.4430046123435099, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.627908945083618, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7025755047798157, + "num_tokens": 340028979.0, + "step": 13140 + }, + { + "epoch": 1.4431144300461234, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.755079746246338, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7083420753479004, + "num_tokens": 340051821.0, + "step": 13141 + }, + { + "epoch": 1.4432242477487371, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.503197431564331, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7164043188095093, + "num_tokens": 340077118.0, + "step": 13142 + }, + { + "epoch": 1.4433340654513507, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.533501148223877, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6948065757751465, + "num_tokens": 340103599.0, + "step": 13143 + }, + { + "epoch": 1.4434438831539644, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.653932571411133, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7238563299179077, + "num_tokens": 340125652.0, + "step": 13144 + }, + { + "epoch": 1.4435537008565782, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6296627521514893, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7151442766189575, + "num_tokens": 340150073.0, + "step": 13145 + }, + { + "epoch": 1.4436635185591917, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7619919776916504, + "learning_rate": 1e-06, + "loss": 0.9326, + "mean_token_accuracy": 0.7220388650894165, + "num_tokens": 340172018.0, + "step": 13146 + }, + { + "epoch": 1.4437733362618055, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.572974920272827, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6969073414802551, + "num_tokens": 340197922.0, + "step": 13147 + }, + { + "epoch": 1.443883153964419, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.288586139678955, + "learning_rate": 1e-06, + "loss": 1.1137, + "mean_token_accuracy": 0.6765995025634766, + "num_tokens": 340232043.0, + "step": 13148 + }, + { + "epoch": 1.4439929716670328, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.284705400466919, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6946946978569031, + "num_tokens": 340263950.0, + "step": 13149 + }, + { + "epoch": 1.4441027893696463, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5859525203704834, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7046098709106445, + "num_tokens": 340290039.0, + "step": 13150 + }, + { + "epoch": 1.44421260707226, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7204530239105225, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6947574615478516, + "num_tokens": 340312673.0, + "step": 13151 + }, + { + "epoch": 1.4443224247748736, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.54868221282959, + "learning_rate": 1e-06, + "loss": 0.9726, + "mean_token_accuracy": 0.7199632525444031, + "num_tokens": 340339078.0, + "step": 13152 + }, + { + "epoch": 1.4444322424774874, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.579710006713867, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7033904790878296, + "num_tokens": 340364148.0, + "step": 13153 + }, + { + "epoch": 1.4445420601801011, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6111388206481934, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7130855917930603, + "num_tokens": 340390357.0, + "step": 13154 + }, + { + "epoch": 1.4446518778827147, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6890006065368652, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7162774205207825, + "num_tokens": 340414598.0, + "step": 13155 + }, + { + "epoch": 1.4447616955853284, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.664098024368286, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7026652097702026, + "num_tokens": 340436416.0, + "step": 13156 + }, + { + "epoch": 1.444871513287942, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.452025890350342, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.7008280754089355, + "num_tokens": 340468002.0, + "step": 13157 + }, + { + "epoch": 1.4449813309905557, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.873971462249756, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.6985270977020264, + "num_tokens": 340490276.0, + "step": 13158 + }, + { + "epoch": 1.4450911486931695, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.841190814971924, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7172555923461914, + "num_tokens": 340511926.0, + "step": 13159 + }, + { + "epoch": 1.445200966395783, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5340256690979004, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7133159637451172, + "num_tokens": 340537993.0, + "step": 13160 + }, + { + "epoch": 1.4453107840983965, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5741114616394043, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6992490887641907, + "num_tokens": 340562051.0, + "step": 13161 + }, + { + "epoch": 1.4454206018010103, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.70963454246521, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6957874894142151, + "num_tokens": 340584541.0, + "step": 13162 + }, + { + "epoch": 1.445530419503624, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.67922043800354, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7173499464988708, + "num_tokens": 340607199.0, + "step": 13163 + }, + { + "epoch": 1.4456402372062376, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5468592643737793, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7025128602981567, + "num_tokens": 340631608.0, + "step": 13164 + }, + { + "epoch": 1.4457500549088513, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.389460802078247, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6889452934265137, + "num_tokens": 340659819.0, + "step": 13165 + }, + { + "epoch": 1.4458598726114649, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0899925231933594, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7233009338378906, + "num_tokens": 340678889.0, + "step": 13166 + }, + { + "epoch": 1.4459696903140786, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.60852313041687, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7057143449783325, + "num_tokens": 340704348.0, + "step": 13167 + }, + { + "epoch": 1.4460795080166924, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.50380802154541, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7157871723175049, + "num_tokens": 340728379.0, + "step": 13168 + }, + { + "epoch": 1.446189325719306, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.654242992401123, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.6947400569915771, + "num_tokens": 340753423.0, + "step": 13169 + }, + { + "epoch": 1.4462991434219197, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4461288452148438, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6954346895217896, + "num_tokens": 340779801.0, + "step": 13170 + }, + { + "epoch": 1.4464089611245332, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7122321128845215, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6931490302085876, + "num_tokens": 340805614.0, + "step": 13171 + }, + { + "epoch": 1.446518778827147, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2410542964935303, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.698922872543335, + "num_tokens": 340840393.0, + "step": 13172 + }, + { + "epoch": 1.4466285965297607, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3346290588378906, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6936142444610596, + "num_tokens": 340871054.0, + "step": 13173 + }, + { + "epoch": 1.4467384142323743, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.607300043106079, + "learning_rate": 1e-06, + "loss": 1.1077, + "mean_token_accuracy": 0.6780865788459778, + "num_tokens": 340898566.0, + "step": 13174 + }, + { + "epoch": 1.4468482319349878, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 80.54161071777344, + "learning_rate": 1e-06, + "loss": 0.9159, + "mean_token_accuracy": 0.7326589822769165, + "num_tokens": 340921348.0, + "step": 13175 + }, + { + "epoch": 1.4469580496376016, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5398447513580322, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7040162086486816, + "num_tokens": 340948757.0, + "step": 13176 + }, + { + "epoch": 1.4470678673402153, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7379608154296875, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6983537077903748, + "num_tokens": 340975774.0, + "step": 13177 + }, + { + "epoch": 1.4471776850428288, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0225040912628174, + "learning_rate": 1e-06, + "loss": 0.9092, + "mean_token_accuracy": 0.7275906205177307, + "num_tokens": 340996031.0, + "step": 13178 + }, + { + "epoch": 1.4472875027454426, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7726540565490723, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.7014299631118774, + "num_tokens": 341020197.0, + "step": 13179 + }, + { + "epoch": 1.4473973204480561, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6103153228759766, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7079150080680847, + "num_tokens": 341047284.0, + "step": 13180 + }, + { + "epoch": 1.44750713815067, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.609299659729004, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6933605670928955, + "num_tokens": 341074120.0, + "step": 13181 + }, + { + "epoch": 1.4476169558532836, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.421225070953369, + "learning_rate": 1e-06, + "loss": 1.134, + "mean_token_accuracy": 0.6844501495361328, + "num_tokens": 341106844.0, + "step": 13182 + }, + { + "epoch": 1.4477267735558972, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5093436241149902, + "learning_rate": 1e-06, + "loss": 1.0986, + "mean_token_accuracy": 0.6809648275375366, + "num_tokens": 341133552.0, + "step": 13183 + }, + { + "epoch": 1.447836591258511, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5611963272094727, + "learning_rate": 1e-06, + "loss": 0.9156, + "mean_token_accuracy": 0.7269823551177979, + "num_tokens": 341159404.0, + "step": 13184 + }, + { + "epoch": 1.4479464089611245, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.4793779850006104, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.7027783393859863, + "num_tokens": 341188828.0, + "step": 13185 + }, + { + "epoch": 1.4480562266637382, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6137566566467285, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7164863348007202, + "num_tokens": 341212905.0, + "step": 13186 + }, + { + "epoch": 1.448166044366352, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6353402137756348, + "learning_rate": 1e-06, + "loss": 0.9087, + "mean_token_accuracy": 0.7360595464706421, + "num_tokens": 341237002.0, + "step": 13187 + }, + { + "epoch": 1.4482758620689655, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.365466594696045, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6982763409614563, + "num_tokens": 341268766.0, + "step": 13188 + }, + { + "epoch": 1.448385679771579, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.840866804122925, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7117730379104614, + "num_tokens": 341290040.0, + "step": 13189 + }, + { + "epoch": 1.4484954974741928, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3037972450256348, + "learning_rate": 1e-06, + "loss": 1.1046, + "mean_token_accuracy": 0.6825467348098755, + "num_tokens": 341322524.0, + "step": 13190 + }, + { + "epoch": 1.4486053151768066, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8820412158966064, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6915218830108643, + "num_tokens": 341345369.0, + "step": 13191 + }, + { + "epoch": 1.44871513287942, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.703927516937256, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7047240138053894, + "num_tokens": 341370538.0, + "step": 13192 + }, + { + "epoch": 1.4488249505820339, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 16.688547134399414, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6932587027549744, + "num_tokens": 341402489.0, + "step": 13193 + }, + { + "epoch": 1.4489347682846474, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7340242862701416, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7066284418106079, + "num_tokens": 341428296.0, + "step": 13194 + }, + { + "epoch": 1.4490445859872612, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.603947401046753, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6907583475112915, + "num_tokens": 341455598.0, + "step": 13195 + }, + { + "epoch": 1.449154403689875, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.505936622619629, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7027063965797424, + "num_tokens": 341485051.0, + "step": 13196 + }, + { + "epoch": 1.4492642213924884, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.630192518234253, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7033736109733582, + "num_tokens": 341508687.0, + "step": 13197 + }, + { + "epoch": 1.4493740390951022, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4955434799194336, + "learning_rate": 1e-06, + "loss": 0.9141, + "mean_token_accuracy": 0.735153079032898, + "num_tokens": 341536130.0, + "step": 13198 + }, + { + "epoch": 1.4494838567977157, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6227307319641113, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7073343992233276, + "num_tokens": 341560952.0, + "step": 13199 + }, + { + "epoch": 1.4495936745003295, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.41337251663208, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6971625089645386, + "num_tokens": 341591777.0, + "step": 13200 + }, + { + "epoch": 1.449703492202943, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4932796955108643, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7007982730865479, + "num_tokens": 341621448.0, + "step": 13201 + }, + { + "epoch": 1.4498133099055568, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6567509174346924, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6858123540878296, + "num_tokens": 341647764.0, + "step": 13202 + }, + { + "epoch": 1.4499231276081703, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4575231075286865, + "learning_rate": 1e-06, + "loss": 1.1419, + "mean_token_accuracy": 0.6719162464141846, + "num_tokens": 341677722.0, + "step": 13203 + }, + { + "epoch": 1.450032945310784, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.753031015396118, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6948122382164001, + "num_tokens": 341699976.0, + "step": 13204 + }, + { + "epoch": 1.4501427630133978, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.5697238445281982, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.718056321144104, + "num_tokens": 341727137.0, + "step": 13205 + }, + { + "epoch": 1.4502525807160114, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6133460998535156, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7047560214996338, + "num_tokens": 341756721.0, + "step": 13206 + }, + { + "epoch": 1.4503623984186251, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.705806255340576, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6866289377212524, + "num_tokens": 341781790.0, + "step": 13207 + }, + { + "epoch": 1.4504722161212387, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.669769763946533, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7142489552497864, + "num_tokens": 341804509.0, + "step": 13208 + }, + { + "epoch": 1.4505820338238524, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7972841262817383, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6915172934532166, + "num_tokens": 341828413.0, + "step": 13209 + }, + { + "epoch": 1.4506918515264662, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9099156856536865, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7194414734840393, + "num_tokens": 341851612.0, + "step": 13210 + }, + { + "epoch": 1.4508016692290797, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.612128734588623, + "learning_rate": 1e-06, + "loss": 0.9068, + "mean_token_accuracy": 0.7402823567390442, + "num_tokens": 341873403.0, + "step": 13211 + }, + { + "epoch": 1.4509114869316933, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.399186372756958, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7095032334327698, + "num_tokens": 341903244.0, + "step": 13212 + }, + { + "epoch": 1.451021304634307, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.210721015930176, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7084678411483765, + "num_tokens": 341921413.0, + "step": 13213 + }, + { + "epoch": 1.4511311223369208, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.68992280960083, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7143781185150146, + "num_tokens": 341946752.0, + "step": 13214 + }, + { + "epoch": 1.4512409400395343, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.740368366241455, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.708817720413208, + "num_tokens": 341974994.0, + "step": 13215 + }, + { + "epoch": 1.451350757742148, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6268880367279053, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6975746750831604, + "num_tokens": 342001400.0, + "step": 13216 + }, + { + "epoch": 1.4514605754447616, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7135982513427734, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6974756717681885, + "num_tokens": 342025161.0, + "step": 13217 + }, + { + "epoch": 1.4515703931473753, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5404410362243652, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.7135604619979858, + "num_tokens": 342052036.0, + "step": 13218 + }, + { + "epoch": 1.451680210849989, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.2836151123046875, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6982190012931824, + "num_tokens": 342081345.0, + "step": 13219 + }, + { + "epoch": 1.4517900285526026, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6958253383636475, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7065122723579407, + "num_tokens": 342103168.0, + "step": 13220 + }, + { + "epoch": 1.4518998462552164, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.712127447128296, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7024165391921997, + "num_tokens": 342128722.0, + "step": 13221 + }, + { + "epoch": 1.45200966395783, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5569241046905518, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6981799602508545, + "num_tokens": 342153962.0, + "step": 13222 + }, + { + "epoch": 1.4521194816604437, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.769106149673462, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6853145360946655, + "num_tokens": 342177007.0, + "step": 13223 + }, + { + "epoch": 1.4522292993630574, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6930313110351562, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7093989849090576, + "num_tokens": 342204681.0, + "step": 13224 + }, + { + "epoch": 1.452339117065671, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8542871475219727, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7334874868392944, + "num_tokens": 342223924.0, + "step": 13225 + }, + { + "epoch": 1.4524489347682845, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4737918376922607, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7206722497940063, + "num_tokens": 342250807.0, + "step": 13226 + }, + { + "epoch": 1.4525587524708983, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.514444589614868, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7141562700271606, + "num_tokens": 342275965.0, + "step": 13227 + }, + { + "epoch": 1.452668570173512, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4153189659118652, + "learning_rate": 1e-06, + "loss": 1.14, + "mean_token_accuracy": 0.6722090840339661, + "num_tokens": 342307194.0, + "step": 13228 + }, + { + "epoch": 1.4527783878761256, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.425725221633911, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7023789286613464, + "num_tokens": 342335970.0, + "step": 13229 + }, + { + "epoch": 1.4528882055787393, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6521217823028564, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7055511474609375, + "num_tokens": 342359907.0, + "step": 13230 + }, + { + "epoch": 1.4529980232813529, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6020450592041016, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7130732536315918, + "num_tokens": 342384495.0, + "step": 13231 + }, + { + "epoch": 1.4531078409839666, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3479716777801514, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6870405673980713, + "num_tokens": 342419582.0, + "step": 13232 + }, + { + "epoch": 1.4532176586865804, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5491294860839844, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.682945728302002, + "num_tokens": 342446851.0, + "step": 13233 + }, + { + "epoch": 1.453327476389194, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.840491533279419, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7217755317687988, + "num_tokens": 342470904.0, + "step": 13234 + }, + { + "epoch": 1.4534372940918077, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9261627197265625, + "learning_rate": 1e-06, + "loss": 0.9443, + "mean_token_accuracy": 0.7170885801315308, + "num_tokens": 342491848.0, + "step": 13235 + }, + { + "epoch": 1.4535471117944212, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0094642639160156, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.702631413936615, + "num_tokens": 342521773.0, + "step": 13236 + }, + { + "epoch": 1.453656929497035, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6014013290405273, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6833909749984741, + "num_tokens": 342548782.0, + "step": 13237 + }, + { + "epoch": 1.4537667471996487, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3647146224975586, + "learning_rate": 1e-06, + "loss": 1.1044, + "mean_token_accuracy": 0.6740970611572266, + "num_tokens": 342581713.0, + "step": 13238 + }, + { + "epoch": 1.4538765649022622, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4458487033843994, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7171305418014526, + "num_tokens": 342608817.0, + "step": 13239 + }, + { + "epoch": 1.4539863826048758, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4868314266204834, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7109116911888123, + "num_tokens": 342635871.0, + "step": 13240 + }, + { + "epoch": 1.4540962003074895, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.468635082244873, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7067300081253052, + "num_tokens": 342660502.0, + "step": 13241 + }, + { + "epoch": 1.4542060180101033, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7841854095458984, + "learning_rate": 1e-06, + "loss": 0.977, + "mean_token_accuracy": 0.7140470743179321, + "num_tokens": 342681603.0, + "step": 13242 + }, + { + "epoch": 1.4543158357127168, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5368263721466064, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7169976234436035, + "num_tokens": 342707872.0, + "step": 13243 + }, + { + "epoch": 1.4544256534153306, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5258073806762695, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.682124674320221, + "num_tokens": 342738552.0, + "step": 13244 + }, + { + "epoch": 1.4545354711179441, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5591952800750732, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7115007638931274, + "num_tokens": 342762795.0, + "step": 13245 + }, + { + "epoch": 1.4546452888205579, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.598191022872925, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7172662615776062, + "num_tokens": 342787005.0, + "step": 13246 + }, + { + "epoch": 1.4547551065231716, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8448829650878906, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.697874128818512, + "num_tokens": 342807908.0, + "step": 13247 + }, + { + "epoch": 1.4548649242257852, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3862369060516357, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6893993616104126, + "num_tokens": 342836125.0, + "step": 13248 + }, + { + "epoch": 1.454974741928399, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.393781900405884, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.718377947807312, + "num_tokens": 342868122.0, + "step": 13249 + }, + { + "epoch": 1.4550845596310125, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4939355850219727, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7151403427124023, + "num_tokens": 342895977.0, + "step": 13250 + }, + { + "epoch": 1.4551943773336262, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.667578935623169, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6935182213783264, + "num_tokens": 342921265.0, + "step": 13251 + }, + { + "epoch": 1.45530419503624, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.489504337310791, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7078375816345215, + "num_tokens": 342948013.0, + "step": 13252 + }, + { + "epoch": 1.4554140127388535, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.04537034034729, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7061567902565002, + "num_tokens": 342979030.0, + "step": 13253 + }, + { + "epoch": 1.455523830441467, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6758742332458496, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7147424221038818, + "num_tokens": 343001334.0, + "step": 13254 + }, + { + "epoch": 1.4556336481440808, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.553046703338623, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6956214308738708, + "num_tokens": 343030971.0, + "step": 13255 + }, + { + "epoch": 1.4557434658466946, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.65417742729187, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6880902051925659, + "num_tokens": 343057286.0, + "step": 13256 + }, + { + "epoch": 1.455853283549308, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.411298990249634, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6936739087104797, + "num_tokens": 343087943.0, + "step": 13257 + }, + { + "epoch": 1.4559631012519219, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.667161226272583, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7163866758346558, + "num_tokens": 343110953.0, + "step": 13258 + }, + { + "epoch": 1.4560729189545354, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.701321840286255, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.7050985097885132, + "num_tokens": 343133106.0, + "step": 13259 + }, + { + "epoch": 1.4561827366571491, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.608025550842285, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7126812934875488, + "num_tokens": 343157036.0, + "step": 13260 + }, + { + "epoch": 1.456292554359763, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4451026916503906, + "learning_rate": 1e-06, + "loss": 1.1071, + "mean_token_accuracy": 0.6790667772293091, + "num_tokens": 343186413.0, + "step": 13261 + }, + { + "epoch": 1.4564023720623764, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4513065814971924, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6863290667533875, + "num_tokens": 343218981.0, + "step": 13262 + }, + { + "epoch": 1.4565121897649902, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.248914957046509, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6912650465965271, + "num_tokens": 343245759.0, + "step": 13263 + }, + { + "epoch": 1.4566220074676037, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.851764440536499, + "learning_rate": 1e-06, + "loss": 0.835, + "mean_token_accuracy": 0.748750627040863, + "num_tokens": 343265382.0, + "step": 13264 + }, + { + "epoch": 1.4567318251702175, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.850356101989746, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6933476328849792, + "num_tokens": 343287786.0, + "step": 13265 + }, + { + "epoch": 1.456841642872831, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8392794132232666, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7140346169471741, + "num_tokens": 343312571.0, + "step": 13266 + }, + { + "epoch": 1.4569514605754448, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.439725399017334, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6904472708702087, + "num_tokens": 343338845.0, + "step": 13267 + }, + { + "epoch": 1.4570612782780583, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5488736629486084, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6960289478302002, + "num_tokens": 343366259.0, + "step": 13268 + }, + { + "epoch": 1.457171095980672, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5893614292144775, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6954841613769531, + "num_tokens": 343389709.0, + "step": 13269 + }, + { + "epoch": 1.4572809136832858, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.0934436321258545, + "learning_rate": 1e-06, + "loss": 0.8722, + "mean_token_accuracy": 0.7407362461090088, + "num_tokens": 343409779.0, + "step": 13270 + }, + { + "epoch": 1.4573907313858994, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6997501850128174, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7095082402229309, + "num_tokens": 343433056.0, + "step": 13271 + }, + { + "epoch": 1.4575005490885131, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.67517352104187, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7038406133651733, + "num_tokens": 343455858.0, + "step": 13272 + }, + { + "epoch": 1.4576103667911267, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.5703072547912598, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7120019197463989, + "num_tokens": 343474974.0, + "step": 13273 + }, + { + "epoch": 1.4577201844937404, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9166719913482666, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7066528797149658, + "num_tokens": 343494711.0, + "step": 13274 + }, + { + "epoch": 1.4578300021963542, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5219321250915527, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.699541449546814, + "num_tokens": 343523387.0, + "step": 13275 + }, + { + "epoch": 1.4579398198989677, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8204991817474365, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7216482162475586, + "num_tokens": 343544079.0, + "step": 13276 + }, + { + "epoch": 1.4580496376015812, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5111470222473145, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7084444761276245, + "num_tokens": 343571845.0, + "step": 13277 + }, + { + "epoch": 1.458159455304195, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5883188247680664, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7141725420951843, + "num_tokens": 343596591.0, + "step": 13278 + }, + { + "epoch": 1.4582692730068088, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.470781087875366, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7139480113983154, + "num_tokens": 343622319.0, + "step": 13279 + }, + { + "epoch": 1.4583790907094223, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6337757110595703, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.70706707239151, + "num_tokens": 343645477.0, + "step": 13280 + }, + { + "epoch": 1.458488908412036, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.050083875656128, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7021821737289429, + "num_tokens": 343664693.0, + "step": 13281 + }, + { + "epoch": 1.4585987261146496, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8726794719696045, + "learning_rate": 1e-06, + "loss": 0.9108, + "mean_token_accuracy": 0.7312122583389282, + "num_tokens": 343683758.0, + "step": 13282 + }, + { + "epoch": 1.4587085438172633, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.843803882598877, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6937467455863953, + "num_tokens": 343708334.0, + "step": 13283 + }, + { + "epoch": 1.458818361519877, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.289066791534424, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.7013925909996033, + "num_tokens": 343740386.0, + "step": 13284 + }, + { + "epoch": 1.4589281792224906, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6798834800720215, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7177286744117737, + "num_tokens": 343765478.0, + "step": 13285 + }, + { + "epoch": 1.4590379969251044, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.819765567779541, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7177155017852783, + "num_tokens": 343790678.0, + "step": 13286 + }, + { + "epoch": 1.459147814627718, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7335400581359863, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.699173092842102, + "num_tokens": 343818608.0, + "step": 13287 + }, + { + "epoch": 1.4592576323303317, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4799211025238037, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.710834264755249, + "num_tokens": 343844894.0, + "step": 13288 + }, + { + "epoch": 1.4593674500329454, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7156875133514404, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7081148624420166, + "num_tokens": 343866536.0, + "step": 13289 + }, + { + "epoch": 1.459477267735559, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.275392532348633, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7066727876663208, + "num_tokens": 343894798.0, + "step": 13290 + }, + { + "epoch": 1.4595870854381725, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4572482109069824, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6900606155395508, + "num_tokens": 343921430.0, + "step": 13291 + }, + { + "epoch": 1.4596969031407863, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.499119997024536, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.7199798226356506, + "num_tokens": 343949715.0, + "step": 13292 + }, + { + "epoch": 1.4598067208434, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.530452013015747, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7219442129135132, + "num_tokens": 343974416.0, + "step": 13293 + }, + { + "epoch": 1.4599165385460136, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.423675537109375, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6869630813598633, + "num_tokens": 344004519.0, + "step": 13294 + }, + { + "epoch": 1.4600263562486273, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.716801166534424, + "learning_rate": 1e-06, + "loss": 0.8488, + "mean_token_accuracy": 0.7510744333267212, + "num_tokens": 344028304.0, + "step": 13295 + }, + { + "epoch": 1.4601361739512408, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.534642219543457, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7206634283065796, + "num_tokens": 344053256.0, + "step": 13296 + }, + { + "epoch": 1.4602459916538546, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9102632999420166, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7119354009628296, + "num_tokens": 344073480.0, + "step": 13297 + }, + { + "epoch": 1.4603558093564684, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.40529465675354, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6864277720451355, + "num_tokens": 344103404.0, + "step": 13298 + }, + { + "epoch": 1.460465627059082, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.509058713912964, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.711158812046051, + "num_tokens": 344129978.0, + "step": 13299 + }, + { + "epoch": 1.4605754447616957, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.62546968460083, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7290893793106079, + "num_tokens": 344153913.0, + "step": 13300 + }, + { + "epoch": 1.4606852624643092, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5614962577819824, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7100456953048706, + "num_tokens": 344179762.0, + "step": 13301 + }, + { + "epoch": 1.460795080166923, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5859317779541016, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7115200161933899, + "num_tokens": 344203479.0, + "step": 13302 + }, + { + "epoch": 1.4609048978695367, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.682255506515503, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6994878053665161, + "num_tokens": 344227702.0, + "step": 13303 + }, + { + "epoch": 1.4610147155721502, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6446759700775146, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7093576192855835, + "num_tokens": 344250615.0, + "step": 13304 + }, + { + "epoch": 1.4611245332747638, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.607355833053589, + "learning_rate": 1e-06, + "loss": 1.0916, + "mean_token_accuracy": 0.688177227973938, + "num_tokens": 344279535.0, + "step": 13305 + }, + { + "epoch": 1.4612343509773775, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.553377628326416, + "learning_rate": 1e-06, + "loss": 1.2103, + "mean_token_accuracy": 0.6564205288887024, + "num_tokens": 344315144.0, + "step": 13306 + }, + { + "epoch": 1.4613441686799913, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6882212162017822, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7020872235298157, + "num_tokens": 344338533.0, + "step": 13307 + }, + { + "epoch": 1.4614539863826048, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.978891611099243, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.696847677230835, + "num_tokens": 344364248.0, + "step": 13308 + }, + { + "epoch": 1.4615638040852186, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.1455516815185547, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7160010933876038, + "num_tokens": 344386105.0, + "step": 13309 + }, + { + "epoch": 1.4616736217878321, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8115110397338867, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.708443284034729, + "num_tokens": 344407088.0, + "step": 13310 + }, + { + "epoch": 1.4617834394904459, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0393669605255127, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.700693666934967, + "num_tokens": 344428868.0, + "step": 13311 + }, + { + "epoch": 1.4618932571930596, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5822622776031494, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.692084550857544, + "num_tokens": 344454809.0, + "step": 13312 + }, + { + "epoch": 1.4620030748956732, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3040122985839844, + "learning_rate": 1e-06, + "loss": 1.1038, + "mean_token_accuracy": 0.6798506379127502, + "num_tokens": 344487596.0, + "step": 13313 + }, + { + "epoch": 1.462112892598287, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5680289268493652, + "learning_rate": 1e-06, + "loss": 1.0829, + "mean_token_accuracy": 0.6849123239517212, + "num_tokens": 344515855.0, + "step": 13314 + }, + { + "epoch": 1.4622227103009005, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.632854461669922, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7144860029220581, + "num_tokens": 344540736.0, + "step": 13315 + }, + { + "epoch": 1.4623325280035142, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.55730938911438, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6851294040679932, + "num_tokens": 344569990.0, + "step": 13316 + }, + { + "epoch": 1.462442345706128, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3743956089019775, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7134872674942017, + "num_tokens": 344598249.0, + "step": 13317 + }, + { + "epoch": 1.4625521634087415, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6333606243133545, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6853868365287781, + "num_tokens": 344623371.0, + "step": 13318 + }, + { + "epoch": 1.462661981111355, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.947479248046875, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7177268862724304, + "num_tokens": 344645429.0, + "step": 13319 + }, + { + "epoch": 1.4627717988139688, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5439841747283936, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7156863212585449, + "num_tokens": 344668941.0, + "step": 13320 + }, + { + "epoch": 1.4628816165165826, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8900017738342285, + "learning_rate": 1e-06, + "loss": 0.9002, + "mean_token_accuracy": 0.7351287603378296, + "num_tokens": 344687947.0, + "step": 13321 + }, + { + "epoch": 1.462991434219196, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5241353511810303, + "learning_rate": 1e-06, + "loss": 0.9155, + "mean_token_accuracy": 0.7304935455322266, + "num_tokens": 344714531.0, + "step": 13322 + }, + { + "epoch": 1.4631012519218098, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.296544313430786, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7086282968521118, + "num_tokens": 344745952.0, + "step": 13323 + }, + { + "epoch": 1.4632110696244234, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.621731758117676, + "learning_rate": 1e-06, + "loss": 0.977, + "mean_token_accuracy": 0.709851086139679, + "num_tokens": 344769489.0, + "step": 13324 + }, + { + "epoch": 1.4633208873270371, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5574493408203125, + "learning_rate": 1e-06, + "loss": 1.12, + "mean_token_accuracy": 0.6759929060935974, + "num_tokens": 344798325.0, + "step": 13325 + }, + { + "epoch": 1.463430705029651, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.442089080810547, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7054228782653809, + "num_tokens": 344826816.0, + "step": 13326 + }, + { + "epoch": 1.4635405227322644, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.340817451477051, + "learning_rate": 1e-06, + "loss": 1.1244, + "mean_token_accuracy": 0.6868348717689514, + "num_tokens": 344857874.0, + "step": 13327 + }, + { + "epoch": 1.4636503404348782, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.842442035675049, + "learning_rate": 1e-06, + "loss": 1.0753, + "mean_token_accuracy": 0.6950424909591675, + "num_tokens": 344883472.0, + "step": 13328 + }, + { + "epoch": 1.4637601581374917, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6063663959503174, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7032661437988281, + "num_tokens": 344911717.0, + "step": 13329 + }, + { + "epoch": 1.4638699758401055, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.720031976699829, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6962864398956299, + "num_tokens": 344935917.0, + "step": 13330 + }, + { + "epoch": 1.463979793542719, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9824531078338623, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7119764089584351, + "num_tokens": 344960203.0, + "step": 13331 + }, + { + "epoch": 1.4640896112453328, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5552687644958496, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7182093858718872, + "num_tokens": 344983619.0, + "step": 13332 + }, + { + "epoch": 1.4641994289479463, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3048017024993896, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.7018225789070129, + "num_tokens": 345015122.0, + "step": 13333 + }, + { + "epoch": 1.46430924665056, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4678125381469727, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.701826810836792, + "num_tokens": 345045631.0, + "step": 13334 + }, + { + "epoch": 1.4644190643531738, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3987138271331787, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7040004134178162, + "num_tokens": 345076630.0, + "step": 13335 + }, + { + "epoch": 1.4645288820557874, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5164222717285156, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7260732650756836, + "num_tokens": 345101823.0, + "step": 13336 + }, + { + "epoch": 1.464638699758401, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5453150272369385, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7045376300811768, + "num_tokens": 345128795.0, + "step": 13337 + }, + { + "epoch": 1.4647485174610146, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.569370746612549, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.703076183795929, + "num_tokens": 345156852.0, + "step": 13338 + }, + { + "epoch": 1.4648583351636284, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6578023433685303, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7146216630935669, + "num_tokens": 345182349.0, + "step": 13339 + }, + { + "epoch": 1.4649681528662422, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.419666051864624, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.720204770565033, + "num_tokens": 345210182.0, + "step": 13340 + }, + { + "epoch": 1.4650779705688557, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7737667560577393, + "learning_rate": 1e-06, + "loss": 0.8725, + "mean_token_accuracy": 0.7407863140106201, + "num_tokens": 345230677.0, + "step": 13341 + }, + { + "epoch": 1.4651877882714692, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.843475103378296, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7093673348426819, + "num_tokens": 345256864.0, + "step": 13342 + }, + { + "epoch": 1.465297605974083, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.598294734954834, + "learning_rate": 1e-06, + "loss": 0.9136, + "mean_token_accuracy": 0.7285686731338501, + "num_tokens": 345283478.0, + "step": 13343 + }, + { + "epoch": 1.4654074236766967, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6626670360565186, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.6956973671913147, + "num_tokens": 345307144.0, + "step": 13344 + }, + { + "epoch": 1.4655172413793103, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3588366508483887, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6899968385696411, + "num_tokens": 345336359.0, + "step": 13345 + }, + { + "epoch": 1.465627059081924, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.1538286209106445, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7199190855026245, + "num_tokens": 345359927.0, + "step": 13346 + }, + { + "epoch": 1.4657368767845376, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7536044120788574, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7176409363746643, + "num_tokens": 345380985.0, + "step": 13347 + }, + { + "epoch": 1.4658466944871513, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5602478981018066, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6789538860321045, + "num_tokens": 345409928.0, + "step": 13348 + }, + { + "epoch": 1.465956512189765, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.782028913497925, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7078498601913452, + "num_tokens": 345435622.0, + "step": 13349 + }, + { + "epoch": 1.4660663298923786, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9110970497131348, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7033877372741699, + "num_tokens": 345457381.0, + "step": 13350 + }, + { + "epoch": 1.4661761475949924, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.639540910720825, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7183166146278381, + "num_tokens": 345480374.0, + "step": 13351 + }, + { + "epoch": 1.466285965297606, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5636491775512695, + "learning_rate": 1e-06, + "loss": 0.9182, + "mean_token_accuracy": 0.7389795780181885, + "num_tokens": 345505238.0, + "step": 13352 + }, + { + "epoch": 1.4663957830002197, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.543952465057373, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6841139793395996, + "num_tokens": 345531950.0, + "step": 13353 + }, + { + "epoch": 1.4665056007028334, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.1946072578430176, + "learning_rate": 1e-06, + "loss": 1.1067, + "mean_token_accuracy": 0.6824002265930176, + "num_tokens": 345556257.0, + "step": 13354 + }, + { + "epoch": 1.466615418405447, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6921660900115967, + "learning_rate": 1e-06, + "loss": 1.1543, + "mean_token_accuracy": 0.6802242994308472, + "num_tokens": 345586451.0, + "step": 13355 + }, + { + "epoch": 1.4667252361080605, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7993581295013428, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.7044394612312317, + "num_tokens": 345610752.0, + "step": 13356 + }, + { + "epoch": 1.4668350538106742, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.801884174346924, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7144649028778076, + "num_tokens": 345631872.0, + "step": 13357 + }, + { + "epoch": 1.466944871513288, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.450860023498535, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6886488199234009, + "num_tokens": 345661342.0, + "step": 13358 + }, + { + "epoch": 1.4670546892159015, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6911983489990234, + "learning_rate": 1e-06, + "loss": 0.8978, + "mean_token_accuracy": 0.740408182144165, + "num_tokens": 345683908.0, + "step": 13359 + }, + { + "epoch": 1.4671645069185153, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7310147285461426, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6954171657562256, + "num_tokens": 345709061.0, + "step": 13360 + }, + { + "epoch": 1.4672743246211288, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.645883798599243, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.705469012260437, + "num_tokens": 345734093.0, + "step": 13361 + }, + { + "epoch": 1.4673841423237426, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.577554225921631, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6920424103736877, + "num_tokens": 345762128.0, + "step": 13362 + }, + { + "epoch": 1.4674939600263563, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 7.985687255859375, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7085217237472534, + "num_tokens": 345785034.0, + "step": 13363 + }, + { + "epoch": 1.4676037777289699, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7408204078674316, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7065122723579407, + "num_tokens": 345808752.0, + "step": 13364 + }, + { + "epoch": 1.4677135954315836, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.522995948791504, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6971082091331482, + "num_tokens": 345838267.0, + "step": 13365 + }, + { + "epoch": 1.4678234131341972, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5554282665252686, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.724317193031311, + "num_tokens": 345861573.0, + "step": 13366 + }, + { + "epoch": 1.467933230836811, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4310739040374756, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7027503848075867, + "num_tokens": 345888270.0, + "step": 13367 + }, + { + "epoch": 1.4680430485394247, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0931284427642822, + "learning_rate": 1e-06, + "loss": 0.9362, + "mean_token_accuracy": 0.7186914682388306, + "num_tokens": 345909122.0, + "step": 13368 + }, + { + "epoch": 1.4681528662420382, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.279289722442627, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6940615177154541, + "num_tokens": 345941407.0, + "step": 13369 + }, + { + "epoch": 1.4682626839446518, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5769236087799072, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.6991575956344604, + "num_tokens": 345967749.0, + "step": 13370 + }, + { + "epoch": 1.4683725016472655, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.995614767074585, + "learning_rate": 1e-06, + "loss": 0.924, + "mean_token_accuracy": 0.72929847240448, + "num_tokens": 345986427.0, + "step": 13371 + }, + { + "epoch": 1.4684823193498793, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6172378063201904, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7064493298530579, + "num_tokens": 346014159.0, + "step": 13372 + }, + { + "epoch": 1.4685921370524928, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3836708068847656, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7108979225158691, + "num_tokens": 346043607.0, + "step": 13373 + }, + { + "epoch": 1.4687019547551066, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.649158239364624, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.710863471031189, + "num_tokens": 346068148.0, + "step": 13374 + }, + { + "epoch": 1.46881177245772, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7375988960266113, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7067649364471436, + "num_tokens": 346089553.0, + "step": 13375 + }, + { + "epoch": 1.4689215901603339, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5202009677886963, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6954560279846191, + "num_tokens": 346119179.0, + "step": 13376 + }, + { + "epoch": 1.4690314078629476, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8222110271453857, + "learning_rate": 1e-06, + "loss": 0.8905, + "mean_token_accuracy": 0.7325968742370605, + "num_tokens": 346137783.0, + "step": 13377 + }, + { + "epoch": 1.4691412255655611, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5389585494995117, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.7004299163818359, + "num_tokens": 346163726.0, + "step": 13378 + }, + { + "epoch": 1.469251043268175, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3758254051208496, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6976934671401978, + "num_tokens": 346192312.0, + "step": 13379 + }, + { + "epoch": 1.4693608609707884, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5691158771514893, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6918550729751587, + "num_tokens": 346221468.0, + "step": 13380 + }, + { + "epoch": 1.4694706786734022, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5213050842285156, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7043067216873169, + "num_tokens": 346247660.0, + "step": 13381 + }, + { + "epoch": 1.4695804963760157, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.479832410812378, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7011763453483582, + "num_tokens": 346275419.0, + "step": 13382 + }, + { + "epoch": 1.4696903140786295, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.685364246368408, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.71677565574646, + "num_tokens": 346299583.0, + "step": 13383 + }, + { + "epoch": 1.469800131781243, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.017604112625122, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.7233731746673584, + "num_tokens": 346319314.0, + "step": 13384 + }, + { + "epoch": 1.4699099494838568, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9121346473693848, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6881356239318848, + "num_tokens": 346344487.0, + "step": 13385 + }, + { + "epoch": 1.4700197671864705, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6576271057128906, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7147713303565979, + "num_tokens": 346371288.0, + "step": 13386 + }, + { + "epoch": 1.470129584889084, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5327060222625732, + "learning_rate": 1e-06, + "loss": 0.9368, + "mean_token_accuracy": 0.7194181680679321, + "num_tokens": 346397949.0, + "step": 13387 + }, + { + "epoch": 1.4702394025916978, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.890824317932129, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7143760919570923, + "num_tokens": 346418085.0, + "step": 13388 + }, + { + "epoch": 1.4703492202943114, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4054970741271973, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7012902498245239, + "num_tokens": 346445774.0, + "step": 13389 + }, + { + "epoch": 1.4704590379969251, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 7.8234357833862305, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6985281705856323, + "num_tokens": 346479092.0, + "step": 13390 + }, + { + "epoch": 1.4705688556995389, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0641562938690186, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6944493055343628, + "num_tokens": 346500975.0, + "step": 13391 + }, + { + "epoch": 1.4706786734021524, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7272613048553467, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7064900398254395, + "num_tokens": 346524780.0, + "step": 13392 + }, + { + "epoch": 1.4707884911047662, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.408918619155884, + "learning_rate": 1e-06, + "loss": 1.0844, + "mean_token_accuracy": 0.6885978579521179, + "num_tokens": 346554229.0, + "step": 13393 + }, + { + "epoch": 1.4708983088073797, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.717541456222534, + "learning_rate": 1e-06, + "loss": 0.9418, + "mean_token_accuracy": 0.7181495428085327, + "num_tokens": 346581471.0, + "step": 13394 + }, + { + "epoch": 1.4710081265099935, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.557236909866333, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7137134671211243, + "num_tokens": 346606506.0, + "step": 13395 + }, + { + "epoch": 1.471117944212607, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.589888572692871, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6933663487434387, + "num_tokens": 346632543.0, + "step": 13396 + }, + { + "epoch": 1.4712277619152208, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7370152473449707, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7134543061256409, + "num_tokens": 346654268.0, + "step": 13397 + }, + { + "epoch": 1.4713375796178343, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4214906692504883, + "learning_rate": 1e-06, + "loss": 0.8948, + "mean_token_accuracy": 0.7407400608062744, + "num_tokens": 346681670.0, + "step": 13398 + }, + { + "epoch": 1.471447397320448, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.576704502105713, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7269318699836731, + "num_tokens": 346706343.0, + "step": 13399 + }, + { + "epoch": 1.4715572150230618, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.627073049545288, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7108099460601807, + "num_tokens": 346731063.0, + "step": 13400 + }, + { + "epoch": 1.4716670327256753, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5618503093719482, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7104507088661194, + "num_tokens": 346755843.0, + "step": 13401 + }, + { + "epoch": 1.471776850428289, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5747971534729004, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.6983112096786499, + "num_tokens": 346782314.0, + "step": 13402 + }, + { + "epoch": 1.4718866681309026, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7163712978363037, + "learning_rate": 1e-06, + "loss": 0.9442, + "mean_token_accuracy": 0.7244818210601807, + "num_tokens": 346804881.0, + "step": 13403 + }, + { + "epoch": 1.4719964858335164, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6736950874328613, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6806766986846924, + "num_tokens": 346833038.0, + "step": 13404 + }, + { + "epoch": 1.4721063035361301, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.706139087677002, + "learning_rate": 1e-06, + "loss": 0.9307, + "mean_token_accuracy": 0.7256813645362854, + "num_tokens": 346854806.0, + "step": 13405 + }, + { + "epoch": 1.4722161212387437, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3627967834472656, + "learning_rate": 1e-06, + "loss": 1.0845, + "mean_token_accuracy": 0.6872636675834656, + "num_tokens": 346886827.0, + "step": 13406 + }, + { + "epoch": 1.4723259389413572, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7828431129455566, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7067131996154785, + "num_tokens": 346909719.0, + "step": 13407 + }, + { + "epoch": 1.472435756643971, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4805920124053955, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7107790112495422, + "num_tokens": 346936087.0, + "step": 13408 + }, + { + "epoch": 1.4725455743465847, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.674691915512085, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6874509453773499, + "num_tokens": 346962253.0, + "step": 13409 + }, + { + "epoch": 1.4726553920491983, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.571673631668091, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.6988994479179382, + "num_tokens": 346987083.0, + "step": 13410 + }, + { + "epoch": 1.472765209751812, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6516528129577637, + "learning_rate": 1e-06, + "loss": 0.9269, + "mean_token_accuracy": 0.7250945568084717, + "num_tokens": 347008954.0, + "step": 13411 + }, + { + "epoch": 1.4728750274544256, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5079760551452637, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6949758529663086, + "num_tokens": 347037526.0, + "step": 13412 + }, + { + "epoch": 1.4729848451570393, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.660874605178833, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7174273133277893, + "num_tokens": 347060196.0, + "step": 13413 + }, + { + "epoch": 1.473094662859653, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.059253692626953, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.7218282222747803, + "num_tokens": 347079279.0, + "step": 13414 + }, + { + "epoch": 1.4732044805622666, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6307036876678467, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7067493796348572, + "num_tokens": 347101296.0, + "step": 13415 + }, + { + "epoch": 1.4733142982648804, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6133840084075928, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.7137885689735413, + "num_tokens": 347124779.0, + "step": 13416 + }, + { + "epoch": 1.473424115967494, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8431968688964844, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7159860134124756, + "num_tokens": 347148769.0, + "step": 13417 + }, + { + "epoch": 1.4735339336701077, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.611771821975708, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7196913957595825, + "num_tokens": 347176103.0, + "step": 13418 + }, + { + "epoch": 1.4736437513727214, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7815072536468506, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7075412273406982, + "num_tokens": 347201415.0, + "step": 13419 + }, + { + "epoch": 1.473753569075335, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.471909761428833, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7230019569396973, + "num_tokens": 347228122.0, + "step": 13420 + }, + { + "epoch": 1.4738633867779485, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4843852519989014, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7037518620491028, + "num_tokens": 347254149.0, + "step": 13421 + }, + { + "epoch": 1.4739732044805622, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8208179473876953, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7168935537338257, + "num_tokens": 347278195.0, + "step": 13422 + }, + { + "epoch": 1.474083022183176, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6601178646087646, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7350034117698669, + "num_tokens": 347301112.0, + "step": 13423 + }, + { + "epoch": 1.4741928398857895, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.533710241317749, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.704952597618103, + "num_tokens": 347326955.0, + "step": 13424 + }, + { + "epoch": 1.4743026575884033, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8165557384490967, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7056801319122314, + "num_tokens": 347352680.0, + "step": 13425 + }, + { + "epoch": 1.4744124752910168, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.689257860183716, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.7039786577224731, + "num_tokens": 347379559.0, + "step": 13426 + }, + { + "epoch": 1.4745222929936306, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4405839443206787, + "learning_rate": 1e-06, + "loss": 0.9421, + "mean_token_accuracy": 0.7251667976379395, + "num_tokens": 347404330.0, + "step": 13427 + }, + { + "epoch": 1.4746321106962443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.37158203125, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7041822075843811, + "num_tokens": 347432882.0, + "step": 13428 + }, + { + "epoch": 1.4747419283988579, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.415313720703125, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7098610997200012, + "num_tokens": 347460572.0, + "step": 13429 + }, + { + "epoch": 1.4748517461014716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7348592281341553, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.6969519257545471, + "num_tokens": 347486650.0, + "step": 13430 + }, + { + "epoch": 1.4749615638040852, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4770052433013916, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7168287038803101, + "num_tokens": 347514758.0, + "step": 13431 + }, + { + "epoch": 1.475071381506699, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6070549488067627, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7043211460113525, + "num_tokens": 347539986.0, + "step": 13432 + }, + { + "epoch": 1.4751811992093127, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.619447708129883, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7024365067481995, + "num_tokens": 347563513.0, + "step": 13433 + }, + { + "epoch": 1.4752910169119262, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5063867568969727, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.716140627861023, + "num_tokens": 347591159.0, + "step": 13434 + }, + { + "epoch": 1.4754008346145397, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4502053260803223, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7097831964492798, + "num_tokens": 347618018.0, + "step": 13435 + }, + { + "epoch": 1.4755106523171535, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4038538932800293, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6895793080329895, + "num_tokens": 347645281.0, + "step": 13436 + }, + { + "epoch": 1.4756204700197673, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.619030237197876, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7085176110267639, + "num_tokens": 347669321.0, + "step": 13437 + }, + { + "epoch": 1.4757302877223808, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6810669898986816, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6936001777648926, + "num_tokens": 347696628.0, + "step": 13438 + }, + { + "epoch": 1.4758401054249946, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.654527187347412, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6962674260139465, + "num_tokens": 347720666.0, + "step": 13439 + }, + { + "epoch": 1.475949923127608, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6521565914154053, + "learning_rate": 1e-06, + "loss": 1.1361, + "mean_token_accuracy": 0.6777139902114868, + "num_tokens": 347748398.0, + "step": 13440 + }, + { + "epoch": 1.4760597408302218, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.424943208694458, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6846050024032593, + "num_tokens": 347779219.0, + "step": 13441 + }, + { + "epoch": 1.4761695585328356, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.732907772064209, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7079037427902222, + "num_tokens": 347800401.0, + "step": 13442 + }, + { + "epoch": 1.4762793762354491, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6089534759521484, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7221890091896057, + "num_tokens": 347824024.0, + "step": 13443 + }, + { + "epoch": 1.476389193938063, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.467106342315674, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7230509519577026, + "num_tokens": 347850139.0, + "step": 13444 + }, + { + "epoch": 1.4764990116406764, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4695122241973877, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7055755257606506, + "num_tokens": 347875516.0, + "step": 13445 + }, + { + "epoch": 1.4766088293432902, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3328304290771484, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6865488290786743, + "num_tokens": 347906133.0, + "step": 13446 + }, + { + "epoch": 1.4767186470459037, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5339126586914062, + "learning_rate": 1e-06, + "loss": 1.1066, + "mean_token_accuracy": 0.6825957298278809, + "num_tokens": 347938177.0, + "step": 13447 + }, + { + "epoch": 1.4768284647485175, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.63486647605896, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6902732849121094, + "num_tokens": 347964338.0, + "step": 13448 + }, + { + "epoch": 1.476938282451131, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.969520092010498, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7097281217575073, + "num_tokens": 347983881.0, + "step": 13449 + }, + { + "epoch": 1.4770481001537448, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4183762073516846, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7019621729850769, + "num_tokens": 348011842.0, + "step": 13450 + }, + { + "epoch": 1.4771579178563585, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.645024538040161, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7086920738220215, + "num_tokens": 348037002.0, + "step": 13451 + }, + { + "epoch": 1.477267735558972, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.585275888442993, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7085086107254028, + "num_tokens": 348063768.0, + "step": 13452 + }, + { + "epoch": 1.4773775532615858, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.368351697921753, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7295694351196289, + "num_tokens": 348092857.0, + "step": 13453 + }, + { + "epoch": 1.4774873709641994, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.581778049468994, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7043030858039856, + "num_tokens": 348116796.0, + "step": 13454 + }, + { + "epoch": 1.477597188666813, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.527810573577881, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7128164172172546, + "num_tokens": 348142158.0, + "step": 13455 + }, + { + "epoch": 1.4777070063694269, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.488513231277466, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.6977237462997437, + "num_tokens": 348170191.0, + "step": 13456 + }, + { + "epoch": 1.4778168240720404, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5658817291259766, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7248510718345642, + "num_tokens": 348194238.0, + "step": 13457 + }, + { + "epoch": 1.477926641774654, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.2846806049346924, + "learning_rate": 1e-06, + "loss": 0.9215, + "mean_token_accuracy": 0.7315819263458252, + "num_tokens": 348226443.0, + "step": 13458 + }, + { + "epoch": 1.4780364594772677, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4798922538757324, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7046302556991577, + "num_tokens": 348253733.0, + "step": 13459 + }, + { + "epoch": 1.4781462771798815, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.60390043258667, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7043161392211914, + "num_tokens": 348278012.0, + "step": 13460 + }, + { + "epoch": 1.478256094882495, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8721296787261963, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7126889228820801, + "num_tokens": 348298231.0, + "step": 13461 + }, + { + "epoch": 1.4783659125851087, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.252870798110962, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7121388912200928, + "num_tokens": 348330234.0, + "step": 13462 + }, + { + "epoch": 1.4784757302877223, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.413935661315918, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6911415457725525, + "num_tokens": 348359195.0, + "step": 13463 + }, + { + "epoch": 1.478585547990336, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6863760948181152, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7065351009368896, + "num_tokens": 348385778.0, + "step": 13464 + }, + { + "epoch": 1.4786953656929498, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7479288578033447, + "learning_rate": 1e-06, + "loss": 1.1051, + "mean_token_accuracy": 0.6777682304382324, + "num_tokens": 348409555.0, + "step": 13465 + }, + { + "epoch": 1.4788051833955633, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.612133264541626, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7118057012557983, + "num_tokens": 348436878.0, + "step": 13466 + }, + { + "epoch": 1.478915001098177, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.477935552597046, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7098007202148438, + "num_tokens": 348460782.0, + "step": 13467 + }, + { + "epoch": 1.4790248188007906, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.2832562923431396, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.6744794845581055, + "num_tokens": 348493364.0, + "step": 13468 + }, + { + "epoch": 1.4791346365034044, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4481263160705566, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7147705554962158, + "num_tokens": 348522052.0, + "step": 13469 + }, + { + "epoch": 1.4792444542060181, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.428544044494629, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.698365330696106, + "num_tokens": 348550787.0, + "step": 13470 + }, + { + "epoch": 1.4793542719086317, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7520363330841064, + "learning_rate": 1e-06, + "loss": 0.9202, + "mean_token_accuracy": 0.7346075773239136, + "num_tokens": 348569893.0, + "step": 13471 + }, + { + "epoch": 1.4794640896112452, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.921196222305298, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7255345582962036, + "num_tokens": 348589231.0, + "step": 13472 + }, + { + "epoch": 1.479573907313859, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6742022037506104, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7301772832870483, + "num_tokens": 348612204.0, + "step": 13473 + }, + { + "epoch": 1.4796837250164727, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7102396488189697, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.6930503249168396, + "num_tokens": 348635785.0, + "step": 13474 + }, + { + "epoch": 1.4797935427190863, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3503949642181396, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6917986869812012, + "num_tokens": 348667896.0, + "step": 13475 + }, + { + "epoch": 1.4799033604217, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.585470676422119, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.71222984790802, + "num_tokens": 348693407.0, + "step": 13476 + }, + { + "epoch": 1.4800131781243135, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.240750789642334, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7252097129821777, + "num_tokens": 348716369.0, + "step": 13477 + }, + { + "epoch": 1.4801229958269273, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3887133598327637, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6939952969551086, + "num_tokens": 348746705.0, + "step": 13478 + }, + { + "epoch": 1.480232813529541, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.1043436527252197, + "learning_rate": 1e-06, + "loss": 0.8894, + "mean_token_accuracy": 0.735317051410675, + "num_tokens": 348766027.0, + "step": 13479 + }, + { + "epoch": 1.4803426312321546, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4457759857177734, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7085977792739868, + "num_tokens": 348793646.0, + "step": 13480 + }, + { + "epoch": 1.4804524489347684, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5799152851104736, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7061596512794495, + "num_tokens": 348820525.0, + "step": 13481 + }, + { + "epoch": 1.4805622666373819, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.537076950073242, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6985804438591003, + "num_tokens": 348847130.0, + "step": 13482 + }, + { + "epoch": 1.4806720843399956, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.700347661972046, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7128741145133972, + "num_tokens": 348870238.0, + "step": 13483 + }, + { + "epoch": 1.4807819020426094, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6235759258270264, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.684017539024353, + "num_tokens": 348898170.0, + "step": 13484 + }, + { + "epoch": 1.480891719745223, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.736659288406372, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.6857783198356628, + "num_tokens": 348928003.0, + "step": 13485 + }, + { + "epoch": 1.4810015374478365, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8680076599121094, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7102131247520447, + "num_tokens": 348953028.0, + "step": 13486 + }, + { + "epoch": 1.4811113551504502, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4657931327819824, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7257746458053589, + "num_tokens": 348981766.0, + "step": 13487 + }, + { + "epoch": 1.481221172853064, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.41117000579834, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.6969876885414124, + "num_tokens": 349010130.0, + "step": 13488 + }, + { + "epoch": 1.4813309905556775, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.751821279525757, + "learning_rate": 1e-06, + "loss": 1.1149, + "mean_token_accuracy": 0.6817015409469604, + "num_tokens": 349037423.0, + "step": 13489 + }, + { + "epoch": 1.4814408082582913, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7593986988067627, + "learning_rate": 1e-06, + "loss": 1.0965, + "mean_token_accuracy": 0.6906590461730957, + "num_tokens": 349063582.0, + "step": 13490 + }, + { + "epoch": 1.4815506259609048, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5249433517456055, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6948456764221191, + "num_tokens": 349097443.0, + "step": 13491 + }, + { + "epoch": 1.4816604436635186, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6671862602233887, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6924678087234497, + "num_tokens": 349124836.0, + "step": 13492 + }, + { + "epoch": 1.4817702613661323, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6853208541870117, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.6872180700302124, + "num_tokens": 349150018.0, + "step": 13493 + }, + { + "epoch": 1.4818800790687459, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5333380699157715, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7050044536590576, + "num_tokens": 349178029.0, + "step": 13494 + }, + { + "epoch": 1.4819898967713596, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.74228572845459, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6938584446907043, + "num_tokens": 349201132.0, + "step": 13495 + }, + { + "epoch": 1.4820997144739732, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.722292184829712, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7187974452972412, + "num_tokens": 349222121.0, + "step": 13496 + }, + { + "epoch": 1.482209532176587, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3801212310791016, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7279731035232544, + "num_tokens": 349250114.0, + "step": 13497 + }, + { + "epoch": 1.4823193498792007, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.306584596633911, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.69441819190979, + "num_tokens": 349278670.0, + "step": 13498 + }, + { + "epoch": 1.4824291675818142, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5738134384155273, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6958553194999695, + "num_tokens": 349304750.0, + "step": 13499 + }, + { + "epoch": 1.4825389852844277, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.418175220489502, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7209600806236267, + "num_tokens": 349332424.0, + "step": 13500 + }, + { + "epoch": 1.4826488029870415, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6458847522735596, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6928644180297852, + "num_tokens": 349359251.0, + "step": 13501 + }, + { + "epoch": 1.4827586206896552, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3624746799468994, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.689731240272522, + "num_tokens": 349390802.0, + "step": 13502 + }, + { + "epoch": 1.4828684383922688, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6897313594818115, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7233204245567322, + "num_tokens": 349411462.0, + "step": 13503 + }, + { + "epoch": 1.4829782560948825, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3642282485961914, + "learning_rate": 1e-06, + "loss": 0.886, + "mean_token_accuracy": 0.7399503588676453, + "num_tokens": 349437804.0, + "step": 13504 + }, + { + "epoch": 1.483088073797496, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.698944568634033, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7157393097877502, + "num_tokens": 349461433.0, + "step": 13505 + }, + { + "epoch": 1.4831978915001098, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3669092655181885, + "learning_rate": 1e-06, + "loss": 1.1245, + "mean_token_accuracy": 0.6798238158226013, + "num_tokens": 349492785.0, + "step": 13506 + }, + { + "epoch": 1.4833077092027236, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.078683614730835, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7120530605316162, + "num_tokens": 349514255.0, + "step": 13507 + }, + { + "epoch": 1.4834175269053371, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.579552173614502, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7076167464256287, + "num_tokens": 349540250.0, + "step": 13508 + }, + { + "epoch": 1.4835273446079509, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.2452402114868164, + "learning_rate": 1e-06, + "loss": 0.9174, + "mean_token_accuracy": 0.7198910713195801, + "num_tokens": 349555718.0, + "step": 13509 + }, + { + "epoch": 1.4836371623105644, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5478124618530273, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.711296796798706, + "num_tokens": 349581064.0, + "step": 13510 + }, + { + "epoch": 1.4837469800131782, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.733002185821533, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6897889971733093, + "num_tokens": 349610750.0, + "step": 13511 + }, + { + "epoch": 1.4838567977157917, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.971247911453247, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7084962129592896, + "num_tokens": 349631750.0, + "step": 13512 + }, + { + "epoch": 1.4839666154184055, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5305099487304688, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6922297477722168, + "num_tokens": 349660878.0, + "step": 13513 + }, + { + "epoch": 1.484076433121019, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6199517250061035, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7192106246948242, + "num_tokens": 349684386.0, + "step": 13514 + }, + { + "epoch": 1.4841862508236328, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.558039903640747, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6799649000167847, + "num_tokens": 349713907.0, + "step": 13515 + }, + { + "epoch": 1.4842960685262465, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.361384630203247, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7023369073867798, + "num_tokens": 349744778.0, + "step": 13516 + }, + { + "epoch": 1.48440588622886, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.77616548538208, + "learning_rate": 1e-06, + "loss": 0.9024, + "mean_token_accuracy": 0.7404725551605225, + "num_tokens": 349767661.0, + "step": 13517 + }, + { + "epoch": 1.4845157039314738, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6068904399871826, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6935789585113525, + "num_tokens": 349794478.0, + "step": 13518 + }, + { + "epoch": 1.4846255216340873, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7656197547912598, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.6961824893951416, + "num_tokens": 349819844.0, + "step": 13519 + }, + { + "epoch": 1.484735339336701, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6848597526550293, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.710331916809082, + "num_tokens": 349847165.0, + "step": 13520 + }, + { + "epoch": 1.4848451570393149, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.501868963241577, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6877762079238892, + "num_tokens": 349874515.0, + "step": 13521 + }, + { + "epoch": 1.4849549747419284, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5958211421966553, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7109431028366089, + "num_tokens": 349899695.0, + "step": 13522 + }, + { + "epoch": 1.485064792444542, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3921167850494385, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6935117244720459, + "num_tokens": 349930611.0, + "step": 13523 + }, + { + "epoch": 1.4851746101471557, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5951712131500244, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6914975643157959, + "num_tokens": 349958000.0, + "step": 13524 + }, + { + "epoch": 1.4852844278497694, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.565561294555664, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.700963020324707, + "num_tokens": 349984827.0, + "step": 13525 + }, + { + "epoch": 1.485394245552383, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.1010079383850098, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7058143019676208, + "num_tokens": 350008778.0, + "step": 13526 + }, + { + "epoch": 1.4855040632549967, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.521174430847168, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7048328518867493, + "num_tokens": 350034827.0, + "step": 13527 + }, + { + "epoch": 1.4856138809576103, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4512994289398193, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7181464433670044, + "num_tokens": 350062547.0, + "step": 13528 + }, + { + "epoch": 1.485723698660224, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7518482208251953, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.706847071647644, + "num_tokens": 350089143.0, + "step": 13529 + }, + { + "epoch": 1.4858335163628378, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6744112968444824, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6942263841629028, + "num_tokens": 350113252.0, + "step": 13530 + }, + { + "epoch": 1.4859433340654513, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3676235675811768, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7028489112854004, + "num_tokens": 350145085.0, + "step": 13531 + }, + { + "epoch": 1.486053151768065, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6667187213897705, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7149374485015869, + "num_tokens": 350167183.0, + "step": 13532 + }, + { + "epoch": 1.4861629694706786, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6328649520874023, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.7143456935882568, + "num_tokens": 350189984.0, + "step": 13533 + }, + { + "epoch": 1.4862727871732924, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7127492427825928, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7020634412765503, + "num_tokens": 350216411.0, + "step": 13534 + }, + { + "epoch": 1.4863826048759061, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.3760855197906494, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7112404704093933, + "num_tokens": 350247603.0, + "step": 13535 + }, + { + "epoch": 1.4864924225785197, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.061446189880371, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.719275712966919, + "num_tokens": 350273919.0, + "step": 13536 + }, + { + "epoch": 1.4866022402811332, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.550347328186035, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7046067118644714, + "num_tokens": 350300025.0, + "step": 13537 + }, + { + "epoch": 1.486712057983747, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.792233943939209, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7235013246536255, + "num_tokens": 350323919.0, + "step": 13538 + }, + { + "epoch": 1.4868218756863607, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5667035579681396, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7069367170333862, + "num_tokens": 350349168.0, + "step": 13539 + }, + { + "epoch": 1.4869316933889742, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.873647928237915, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7127990126609802, + "num_tokens": 350370466.0, + "step": 13540 + }, + { + "epoch": 1.487041511091588, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.14546799659729, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7174188494682312, + "num_tokens": 350392292.0, + "step": 13541 + }, + { + "epoch": 1.4871513287942015, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7289066314697266, + "learning_rate": 1e-06, + "loss": 1.1062, + "mean_token_accuracy": 0.6809865832328796, + "num_tokens": 350418165.0, + "step": 13542 + }, + { + "epoch": 1.4872611464968153, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9144768714904785, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.7048552632331848, + "num_tokens": 350438677.0, + "step": 13543 + }, + { + "epoch": 1.487370964199429, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.740903377532959, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7041491866111755, + "num_tokens": 350463147.0, + "step": 13544 + }, + { + "epoch": 1.4874807819020426, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3555612564086914, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6909038424491882, + "num_tokens": 350493784.0, + "step": 13545 + }, + { + "epoch": 1.4875905996046563, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8180673122406006, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7021876573562622, + "num_tokens": 350515693.0, + "step": 13546 + }, + { + "epoch": 1.4877004173072699, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2195305824279785, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7083656787872314, + "num_tokens": 350549384.0, + "step": 13547 + }, + { + "epoch": 1.4878102350098836, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.545515775680542, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6870883703231812, + "num_tokens": 350580496.0, + "step": 13548 + }, + { + "epoch": 1.4879200527124974, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.719494581222534, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7042129635810852, + "num_tokens": 350602962.0, + "step": 13549 + }, + { + "epoch": 1.488029870415111, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5473382472991943, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7033650279045105, + "num_tokens": 350628895.0, + "step": 13550 + }, + { + "epoch": 1.4881396881177245, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.89569091796875, + "learning_rate": 1e-06, + "loss": 0.8812, + "mean_token_accuracy": 0.74061518907547, + "num_tokens": 350646645.0, + "step": 13551 + }, + { + "epoch": 1.4882495058203382, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.6739025115966797, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.695504903793335, + "num_tokens": 350671701.0, + "step": 13552 + }, + { + "epoch": 1.488359323522952, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2650153636932373, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6994283199310303, + "num_tokens": 350703873.0, + "step": 13553 + }, + { + "epoch": 1.4884691412255655, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4327392578125, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6974767446517944, + "num_tokens": 350733475.0, + "step": 13554 + }, + { + "epoch": 1.4885789589281793, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.471073865890503, + "learning_rate": 1e-06, + "loss": 0.9095, + "mean_token_accuracy": 0.733002781867981, + "num_tokens": 350758833.0, + "step": 13555 + }, + { + "epoch": 1.4886887766307928, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.717956781387329, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7065486907958984, + "num_tokens": 350780400.0, + "step": 13556 + }, + { + "epoch": 1.4887985943334066, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.619365692138672, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7230983972549438, + "num_tokens": 350808141.0, + "step": 13557 + }, + { + "epoch": 1.4889084120360203, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6188013553619385, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7073834538459778, + "num_tokens": 350835002.0, + "step": 13558 + }, + { + "epoch": 1.4890182297386338, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.71036696434021, + "learning_rate": 1e-06, + "loss": 0.8791, + "mean_token_accuracy": 0.7452366948127747, + "num_tokens": 350856990.0, + "step": 13559 + }, + { + "epoch": 1.4891280474412476, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4467244148254395, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6933500170707703, + "num_tokens": 350886291.0, + "step": 13560 + }, + { + "epoch": 1.4892378651438611, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.359729051589966, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7066343426704407, + "num_tokens": 350914005.0, + "step": 13561 + }, + { + "epoch": 1.489347682846475, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.621774911880493, + "learning_rate": 1e-06, + "loss": 1.0831, + "mean_token_accuracy": 0.677149772644043, + "num_tokens": 350941507.0, + "step": 13562 + }, + { + "epoch": 1.4894575005490884, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.061584711074829, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7214231491088867, + "num_tokens": 350960208.0, + "step": 13563 + }, + { + "epoch": 1.4895673182517022, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5480775833129883, + "learning_rate": 1e-06, + "loss": 1.1006, + "mean_token_accuracy": 0.678130567073822, + "num_tokens": 350988155.0, + "step": 13564 + }, + { + "epoch": 1.4896771359543157, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.682255744934082, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7132171392440796, + "num_tokens": 351013036.0, + "step": 13565 + }, + { + "epoch": 1.4897869536569295, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5892598628997803, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7142746448516846, + "num_tokens": 351036133.0, + "step": 13566 + }, + { + "epoch": 1.4898967713595432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.560046434402466, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7110062837600708, + "num_tokens": 351069401.0, + "step": 13567 + }, + { + "epoch": 1.4900065890621568, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.38696026802063, + "learning_rate": 1e-06, + "loss": 1.1259, + "mean_token_accuracy": 0.6838739514350891, + "num_tokens": 351102336.0, + "step": 13568 + }, + { + "epoch": 1.4901164067647705, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.957841396331787, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.719219446182251, + "num_tokens": 351120748.0, + "step": 13569 + }, + { + "epoch": 1.490226224467384, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.858358860015869, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.7008029222488403, + "num_tokens": 351140821.0, + "step": 13570 + }, + { + "epoch": 1.4903360421699978, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4149489402770996, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6837821006774902, + "num_tokens": 351170688.0, + "step": 13571 + }, + { + "epoch": 1.4904458598726116, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.507596492767334, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.6994447112083435, + "num_tokens": 351195904.0, + "step": 13572 + }, + { + "epoch": 1.4905556775752251, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7379581928253174, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.707940399646759, + "num_tokens": 351218093.0, + "step": 13573 + }, + { + "epoch": 1.4906654952778389, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5587849617004395, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6945367455482483, + "num_tokens": 351243442.0, + "step": 13574 + }, + { + "epoch": 1.4907753129804524, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.104128360748291, + "learning_rate": 1e-06, + "loss": 1.0915, + "mean_token_accuracy": 0.6848998665809631, + "num_tokens": 351264376.0, + "step": 13575 + }, + { + "epoch": 1.4908851306830662, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.9757328033447266, + "learning_rate": 1e-06, + "loss": 0.8761, + "mean_token_accuracy": 0.7392650842666626, + "num_tokens": 351282166.0, + "step": 13576 + }, + { + "epoch": 1.4909949483856797, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4608664512634277, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6907527446746826, + "num_tokens": 351310966.0, + "step": 13577 + }, + { + "epoch": 1.4911047660882935, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4728970527648926, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7161725759506226, + "num_tokens": 351338849.0, + "step": 13578 + }, + { + "epoch": 1.491214583790907, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5647480487823486, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7061742544174194, + "num_tokens": 351364478.0, + "step": 13579 + }, + { + "epoch": 1.4913244014935207, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7154903411865234, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7056407332420349, + "num_tokens": 351389319.0, + "step": 13580 + }, + { + "epoch": 1.4914342191961345, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.690939426422119, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7020678520202637, + "num_tokens": 351413848.0, + "step": 13581 + }, + { + "epoch": 1.491544036898748, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.362886428833008, + "learning_rate": 1e-06, + "loss": 1.1124, + "mean_token_accuracy": 0.6876729130744934, + "num_tokens": 351443166.0, + "step": 13582 + }, + { + "epoch": 1.4916538546013618, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.943835735321045, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7080508470535278, + "num_tokens": 351462274.0, + "step": 13583 + }, + { + "epoch": 1.4917636723039753, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7247326374053955, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.712550938129425, + "num_tokens": 351487010.0, + "step": 13584 + }, + { + "epoch": 1.491873490006589, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.439067840576172, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.730850100517273, + "num_tokens": 351514148.0, + "step": 13585 + }, + { + "epoch": 1.4919833077092028, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.329338788986206, + "learning_rate": 1e-06, + "loss": 1.118, + "mean_token_accuracy": 0.6807066202163696, + "num_tokens": 351548859.0, + "step": 13586 + }, + { + "epoch": 1.4920931254118164, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.785869598388672, + "learning_rate": 1e-06, + "loss": 0.897, + "mean_token_accuracy": 0.7346673011779785, + "num_tokens": 351568450.0, + "step": 13587 + }, + { + "epoch": 1.49220294311443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.687175989151001, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7079744338989258, + "num_tokens": 351590179.0, + "step": 13588 + }, + { + "epoch": 1.4923127608170437, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4440767765045166, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7219055891036987, + "num_tokens": 351617740.0, + "step": 13589 + }, + { + "epoch": 1.4924225785196574, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5164999961853027, + "learning_rate": 1e-06, + "loss": 1.1488, + "mean_token_accuracy": 0.6779487133026123, + "num_tokens": 351649879.0, + "step": 13590 + }, + { + "epoch": 1.492532396222271, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.882390022277832, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7029045820236206, + "num_tokens": 351671227.0, + "step": 13591 + }, + { + "epoch": 1.4926422139248847, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7365777492523193, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7043197154998779, + "num_tokens": 351693296.0, + "step": 13592 + }, + { + "epoch": 1.4927520316274983, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8082430362701416, + "learning_rate": 1e-06, + "loss": 0.7791, + "mean_token_accuracy": 0.7634198665618896, + "num_tokens": 351711924.0, + "step": 13593 + }, + { + "epoch": 1.492861849330112, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4960548877716064, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7206516265869141, + "num_tokens": 351737262.0, + "step": 13594 + }, + { + "epoch": 1.4929716670327258, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6748244762420654, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7186078429222107, + "num_tokens": 351760464.0, + "step": 13595 + }, + { + "epoch": 1.4930814847353393, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6922500133514404, + "learning_rate": 1e-06, + "loss": 1.0794, + "mean_token_accuracy": 0.6846271753311157, + "num_tokens": 351784671.0, + "step": 13596 + }, + { + "epoch": 1.493191302437953, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6296253204345703, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6950583457946777, + "num_tokens": 351813130.0, + "step": 13597 + }, + { + "epoch": 1.4933011201405666, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6993014812469482, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.697607159614563, + "num_tokens": 351836354.0, + "step": 13598 + }, + { + "epoch": 1.4934109378431804, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3987748622894287, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7181839942932129, + "num_tokens": 351864756.0, + "step": 13599 + }, + { + "epoch": 1.493520755545794, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2695462703704834, + "learning_rate": 1e-06, + "loss": 1.0796, + "mean_token_accuracy": 0.6887611746788025, + "num_tokens": 351900574.0, + "step": 13600 + }, + { + "epoch": 1.4936305732484076, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.736678123474121, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7191880941390991, + "num_tokens": 351922710.0, + "step": 13601 + }, + { + "epoch": 1.4937403909510212, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5933549404144287, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7092553377151489, + "num_tokens": 351949225.0, + "step": 13602 + }, + { + "epoch": 1.493850208653635, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.575810432434082, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.6991559267044067, + "num_tokens": 351974234.0, + "step": 13603 + }, + { + "epoch": 1.4939600263562487, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7016539573669434, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.6981266140937805, + "num_tokens": 351996602.0, + "step": 13604 + }, + { + "epoch": 1.4940698440588622, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7122323513031006, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6980716586112976, + "num_tokens": 352020711.0, + "step": 13605 + }, + { + "epoch": 1.494179661761476, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4381728172302246, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6808332204818726, + "num_tokens": 352052927.0, + "step": 13606 + }, + { + "epoch": 1.4942894794640895, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7149341106414795, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.704505443572998, + "num_tokens": 352077195.0, + "step": 13607 + }, + { + "epoch": 1.4943992971667033, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4482553005218506, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.690751850605011, + "num_tokens": 352106228.0, + "step": 13608 + }, + { + "epoch": 1.494509114869317, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4351131916046143, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7024380564689636, + "num_tokens": 352133965.0, + "step": 13609 + }, + { + "epoch": 1.4946189325719306, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.647138833999634, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6937105059623718, + "num_tokens": 352158760.0, + "step": 13610 + }, + { + "epoch": 1.4947287502745443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0654568672180176, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7172733545303345, + "num_tokens": 352177506.0, + "step": 13611 + }, + { + "epoch": 1.4948385679771579, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.011967658996582, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7101386785507202, + "num_tokens": 352198833.0, + "step": 13612 + }, + { + "epoch": 1.4949483856797716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8051130771636963, + "learning_rate": 1e-06, + "loss": 0.8838, + "mean_token_accuracy": 0.7260787487030029, + "num_tokens": 352220545.0, + "step": 13613 + }, + { + "epoch": 1.4950582033823854, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.616467237472534, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6958369016647339, + "num_tokens": 352251267.0, + "step": 13614 + }, + { + "epoch": 1.495168021084999, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.738731861114502, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6894348859786987, + "num_tokens": 352277812.0, + "step": 13615 + }, + { + "epoch": 1.4952778387876124, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5809388160705566, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.6977251768112183, + "num_tokens": 352302730.0, + "step": 13616 + }, + { + "epoch": 1.4953876564902262, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7877330780029297, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.6945257782936096, + "num_tokens": 352323936.0, + "step": 13617 + }, + { + "epoch": 1.49549747419284, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.40254282951355, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6989989280700684, + "num_tokens": 352353122.0, + "step": 13618 + }, + { + "epoch": 1.4956072918954535, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8525774478912354, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7030183672904968, + "num_tokens": 352379397.0, + "step": 13619 + }, + { + "epoch": 1.4957171095980673, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7072479724884033, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7153042554855347, + "num_tokens": 352405394.0, + "step": 13620 + }, + { + "epoch": 1.4958269273006808, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.777249336242676, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7217336893081665, + "num_tokens": 352427832.0, + "step": 13621 + }, + { + "epoch": 1.4959367450032945, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6190884113311768, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.718512237071991, + "num_tokens": 352451288.0, + "step": 13622 + }, + { + "epoch": 1.4960465627059083, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4062373638153076, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7092883586883545, + "num_tokens": 352480422.0, + "step": 13623 + }, + { + "epoch": 1.4961563804085218, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.805687665939331, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7056782245635986, + "num_tokens": 352502417.0, + "step": 13624 + }, + { + "epoch": 1.4962661981111356, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.543369770050049, + "learning_rate": 1e-06, + "loss": 1.1118, + "mean_token_accuracy": 0.6769030094146729, + "num_tokens": 352534885.0, + "step": 13625 + }, + { + "epoch": 1.4963760158137491, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3636393547058105, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7149448394775391, + "num_tokens": 352564471.0, + "step": 13626 + }, + { + "epoch": 1.4964858335163629, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.833894968032837, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6827110052108765, + "num_tokens": 352586736.0, + "step": 13627 + }, + { + "epoch": 1.4965956512189764, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4750313758850098, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7089455127716064, + "num_tokens": 352615469.0, + "step": 13628 + }, + { + "epoch": 1.4967054689215902, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5162746906280518, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7093662619590759, + "num_tokens": 352643181.0, + "step": 13629 + }, + { + "epoch": 1.4968152866242037, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5325801372528076, + "learning_rate": 1e-06, + "loss": 1.1179, + "mean_token_accuracy": 0.675230085849762, + "num_tokens": 352673504.0, + "step": 13630 + }, + { + "epoch": 1.4969251043268175, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.543701171875, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7132209539413452, + "num_tokens": 352701778.0, + "step": 13631 + }, + { + "epoch": 1.4970349220294312, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.586496114730835, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.7046380043029785, + "num_tokens": 352728066.0, + "step": 13632 + }, + { + "epoch": 1.4971447397320448, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.700986623764038, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6980979442596436, + "num_tokens": 352752048.0, + "step": 13633 + }, + { + "epoch": 1.4972545574346585, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.410597562789917, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6865153312683105, + "num_tokens": 352786096.0, + "step": 13634 + }, + { + "epoch": 1.497364375137272, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6195950508117676, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7242209911346436, + "num_tokens": 352812176.0, + "step": 13635 + }, + { + "epoch": 1.4974741928398858, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 8.853311538696289, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7073808908462524, + "num_tokens": 352835227.0, + "step": 13636 + }, + { + "epoch": 1.4975840105424996, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.757906436920166, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7064459919929504, + "num_tokens": 352861196.0, + "step": 13637 + }, + { + "epoch": 1.497693828245113, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.895719528198242, + "learning_rate": 1e-06, + "loss": 0.8322, + "mean_token_accuracy": 0.7519779801368713, + "num_tokens": 352882435.0, + "step": 13638 + }, + { + "epoch": 1.4978036459477266, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.711074113845825, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6834225654602051, + "num_tokens": 352907753.0, + "step": 13639 + }, + { + "epoch": 1.4979134636503404, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7895455360412598, + "learning_rate": 1e-06, + "loss": 0.8758, + "mean_token_accuracy": 0.7332429885864258, + "num_tokens": 352929689.0, + "step": 13640 + }, + { + "epoch": 1.4980232813529542, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2072906494140625, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7131916880607605, + "num_tokens": 352950689.0, + "step": 13641 + }, + { + "epoch": 1.4981330990555677, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5373706817626953, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7097947001457214, + "num_tokens": 352978013.0, + "step": 13642 + }, + { + "epoch": 1.4982429167581814, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.603381872177124, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.7049839496612549, + "num_tokens": 353007336.0, + "step": 13643 + }, + { + "epoch": 1.498352734460795, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.553006172180176, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7131547927856445, + "num_tokens": 353032226.0, + "step": 13644 + }, + { + "epoch": 1.4984625521634087, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.71903920173645, + "learning_rate": 1e-06, + "loss": 0.934, + "mean_token_accuracy": 0.7210147976875305, + "num_tokens": 353054202.0, + "step": 13645 + }, + { + "epoch": 1.4985723698660225, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.709233045578003, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7106130719184875, + "num_tokens": 353077624.0, + "step": 13646 + }, + { + "epoch": 1.498682187568636, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.531031608581543, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7096829414367676, + "num_tokens": 353105436.0, + "step": 13647 + }, + { + "epoch": 1.4987920052712498, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.54472279548645, + "learning_rate": 1e-06, + "loss": 1.0929, + "mean_token_accuracy": 0.6911382675170898, + "num_tokens": 353133795.0, + "step": 13648 + }, + { + "epoch": 1.4989018229738633, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.79775333404541, + "learning_rate": 1e-06, + "loss": 0.9111, + "mean_token_accuracy": 0.7263478636741638, + "num_tokens": 353153980.0, + "step": 13649 + }, + { + "epoch": 1.499011640676477, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.576876163482666, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7123334407806396, + "num_tokens": 353180072.0, + "step": 13650 + }, + { + "epoch": 1.4991214583790908, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5637447834014893, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6912534832954407, + "num_tokens": 353209854.0, + "step": 13651 + }, + { + "epoch": 1.4992312760817044, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.0927488803863525, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7144057154655457, + "num_tokens": 353227772.0, + "step": 13652 + }, + { + "epoch": 1.499341093784318, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9042065143585205, + "learning_rate": 1e-06, + "loss": 0.9033, + "mean_token_accuracy": 0.735580325126648, + "num_tokens": 353248130.0, + "step": 13653 + }, + { + "epoch": 1.4994509114869317, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 3.0620715618133545, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7008223533630371, + "num_tokens": 353268005.0, + "step": 13654 + }, + { + "epoch": 1.4995607291895454, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4786176681518555, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.7240776419639587, + "num_tokens": 353293531.0, + "step": 13655 + }, + { + "epoch": 1.499670546892159, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.298692226409912, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6852408647537231, + "num_tokens": 353323734.0, + "step": 13656 + }, + { + "epoch": 1.4997803645947727, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.859797477722168, + "learning_rate": 1e-06, + "loss": 1.1102, + "mean_token_accuracy": 0.6960249543190002, + "num_tokens": 353345463.0, + "step": 13657 + }, + { + "epoch": 1.4998901822973862, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.750283718109131, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.7025542259216309, + "num_tokens": 353369808.0, + "step": 13658 + }, + { + "epoch": 1.5, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.628089666366577, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.731255292892456, + "num_tokens": 353393328.0, + "step": 13659 + }, + { + "epoch": 1.5001098177026138, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.754127264022827, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7062304019927979, + "num_tokens": 353417908.0, + "step": 13660 + }, + { + "epoch": 1.5002196354052273, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5769331455230713, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7040984630584717, + "num_tokens": 353443856.0, + "step": 13661 + }, + { + "epoch": 1.5003294531078408, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4875407218933105, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7105597257614136, + "num_tokens": 353469890.0, + "step": 13662 + }, + { + "epoch": 1.5004392708104546, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7618114948272705, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6890164613723755, + "num_tokens": 353496955.0, + "step": 13663 + }, + { + "epoch": 1.5005490885130683, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.314976692199707, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6980084180831909, + "num_tokens": 353527810.0, + "step": 13664 + }, + { + "epoch": 1.500658906215682, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.514265298843384, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.6946032643318176, + "num_tokens": 353554731.0, + "step": 13665 + }, + { + "epoch": 1.5007687239182956, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.243079423904419, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6959624886512756, + "num_tokens": 353586332.0, + "step": 13666 + }, + { + "epoch": 1.5008785416209092, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4662551879882812, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6910249590873718, + "num_tokens": 353612790.0, + "step": 13667 + }, + { + "epoch": 1.500988359323523, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3685450553894043, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.70183265209198, + "num_tokens": 353640385.0, + "step": 13668 + }, + { + "epoch": 1.5010981770261367, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5964884757995605, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6881725788116455, + "num_tokens": 353665468.0, + "step": 13669 + }, + { + "epoch": 1.5012079947287504, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.544443130493164, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7215903401374817, + "num_tokens": 353690606.0, + "step": 13670 + }, + { + "epoch": 1.501317812431364, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.667081356048584, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7134090065956116, + "num_tokens": 353713806.0, + "step": 13671 + }, + { + "epoch": 1.5014276301339775, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5179786682128906, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7129627466201782, + "num_tokens": 353738924.0, + "step": 13672 + }, + { + "epoch": 1.5015374478365913, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.546354055404663, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6983193755149841, + "num_tokens": 353763454.0, + "step": 13673 + }, + { + "epoch": 1.501647265539205, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.598451852798462, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.7261425256729126, + "num_tokens": 353787635.0, + "step": 13674 + }, + { + "epoch": 1.5017570832418186, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4991958141326904, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7021990418434143, + "num_tokens": 353815796.0, + "step": 13675 + }, + { + "epoch": 1.501866900944432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5417892932891846, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7305738925933838, + "num_tokens": 353840383.0, + "step": 13676 + }, + { + "epoch": 1.5019767186470458, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.536186456680298, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7063943147659302, + "num_tokens": 353866233.0, + "step": 13677 + }, + { + "epoch": 1.5020865363496596, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3565146923065186, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7019944190979004, + "num_tokens": 353894026.0, + "step": 13678 + }, + { + "epoch": 1.5021963540522734, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1099822521209717, + "learning_rate": 1e-06, + "loss": 1.1223, + "mean_token_accuracy": 0.6782701015472412, + "num_tokens": 353923991.0, + "step": 13679 + }, + { + "epoch": 1.502306171754887, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3811962604522705, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7178619503974915, + "num_tokens": 353954798.0, + "step": 13680 + }, + { + "epoch": 1.5024159894575004, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.057263135910034, + "learning_rate": 1e-06, + "loss": 0.9726, + "mean_token_accuracy": 0.7117239832878113, + "num_tokens": 353975069.0, + "step": 13681 + }, + { + "epoch": 1.5025258071601142, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.964005708694458, + "learning_rate": 1e-06, + "loss": 0.9112, + "mean_token_accuracy": 0.7289941310882568, + "num_tokens": 353994106.0, + "step": 13682 + }, + { + "epoch": 1.502635624862728, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6310079097747803, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6860331892967224, + "num_tokens": 354018331.0, + "step": 13683 + }, + { + "epoch": 1.5027454425653415, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6684460639953613, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7245234251022339, + "num_tokens": 354040243.0, + "step": 13684 + }, + { + "epoch": 1.5028552602679552, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6589138507843018, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.713510274887085, + "num_tokens": 354064205.0, + "step": 13685 + }, + { + "epoch": 1.5029650779705688, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.77640700340271, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6991447806358337, + "num_tokens": 354089638.0, + "step": 13686 + }, + { + "epoch": 1.5030748956731825, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.713428258895874, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7205044031143188, + "num_tokens": 354113131.0, + "step": 13687 + }, + { + "epoch": 1.5031847133757963, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6771864891052246, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7173968553543091, + "num_tokens": 354137117.0, + "step": 13688 + }, + { + "epoch": 1.5032945310784098, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7303075790405273, + "learning_rate": 1e-06, + "loss": 0.947, + "mean_token_accuracy": 0.7225511074066162, + "num_tokens": 354161436.0, + "step": 13689 + }, + { + "epoch": 1.5034043487810234, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7293689250946045, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6980727910995483, + "num_tokens": 354186678.0, + "step": 13690 + }, + { + "epoch": 1.5035141664836371, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.421544313430786, + "learning_rate": 1e-06, + "loss": 0.9145, + "mean_token_accuracy": 0.7231432199478149, + "num_tokens": 354214855.0, + "step": 13691 + }, + { + "epoch": 1.5036239841862509, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3373525142669678, + "learning_rate": 1e-06, + "loss": 1.1249, + "mean_token_accuracy": 0.668752908706665, + "num_tokens": 354247092.0, + "step": 13692 + }, + { + "epoch": 1.5037338018888646, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8004705905914307, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.7228268384933472, + "num_tokens": 354268746.0, + "step": 13693 + }, + { + "epoch": 1.5038436195914782, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4484798908233643, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7093671560287476, + "num_tokens": 354296795.0, + "step": 13694 + }, + { + "epoch": 1.5039534372940917, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4302117824554443, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.715631365776062, + "num_tokens": 354325884.0, + "step": 13695 + }, + { + "epoch": 1.5040632549967055, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.721365213394165, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7175635099411011, + "num_tokens": 354348878.0, + "step": 13696 + }, + { + "epoch": 1.5041730726993192, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6191117763519287, + "learning_rate": 1e-06, + "loss": 0.9023, + "mean_token_accuracy": 0.7331997752189636, + "num_tokens": 354373386.0, + "step": 13697 + }, + { + "epoch": 1.5042828904019327, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.470280647277832, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7068626284599304, + "num_tokens": 354399582.0, + "step": 13698 + }, + { + "epoch": 1.5043927081045465, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4792580604553223, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6716158986091614, + "num_tokens": 354427759.0, + "step": 13699 + }, + { + "epoch": 1.50450252580716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8551573753356934, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6889806389808655, + "num_tokens": 354450286.0, + "step": 13700 + }, + { + "epoch": 1.5046123435097738, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7010130882263184, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6923837661743164, + "num_tokens": 354476107.0, + "step": 13701 + }, + { + "epoch": 1.5047221612123876, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.531538963317871, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6824979186058044, + "num_tokens": 354505040.0, + "step": 13702 + }, + { + "epoch": 1.504831978915001, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5259344577789307, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6999291181564331, + "num_tokens": 354534336.0, + "step": 13703 + }, + { + "epoch": 1.5049417966176146, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7430520057678223, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7051697373390198, + "num_tokens": 354559588.0, + "step": 13704 + }, + { + "epoch": 1.5050516143202284, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.748059034347534, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7006809711456299, + "num_tokens": 354583966.0, + "step": 13705 + }, + { + "epoch": 1.5051614320228421, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5829060077667236, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7097688913345337, + "num_tokens": 354608592.0, + "step": 13706 + }, + { + "epoch": 1.505271249725456, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8842198848724365, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7184004783630371, + "num_tokens": 354631381.0, + "step": 13707 + }, + { + "epoch": 1.5053810674280694, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7969651222229004, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7072039246559143, + "num_tokens": 354655687.0, + "step": 13708 + }, + { + "epoch": 1.505490885130683, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8678929805755615, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6957088112831116, + "num_tokens": 354680390.0, + "step": 13709 + }, + { + "epoch": 1.5056007028332967, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.679647445678711, + "learning_rate": 1e-06, + "loss": 1.0854, + "mean_token_accuracy": 0.6849144697189331, + "num_tokens": 354706633.0, + "step": 13710 + }, + { + "epoch": 1.5057105205359105, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6413302421569824, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7101767659187317, + "num_tokens": 354730625.0, + "step": 13711 + }, + { + "epoch": 1.505820338238524, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 8.116055488586426, + "learning_rate": 1e-06, + "loss": 0.9241, + "mean_token_accuracy": 0.7285792827606201, + "num_tokens": 354747197.0, + "step": 13712 + }, + { + "epoch": 1.5059301559411375, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6564207077026367, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7118346691131592, + "num_tokens": 354771713.0, + "step": 13713 + }, + { + "epoch": 1.5060399736437513, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5984978675842285, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7113158106803894, + "num_tokens": 354799653.0, + "step": 13714 + }, + { + "epoch": 1.506149791346365, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7264175415039062, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6935515403747559, + "num_tokens": 354827490.0, + "step": 13715 + }, + { + "epoch": 1.5062596090489788, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6693618297576904, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6956996917724609, + "num_tokens": 354854128.0, + "step": 13716 + }, + { + "epoch": 1.5063694267515924, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6166718006134033, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.6987345218658447, + "num_tokens": 354879134.0, + "step": 13717 + }, + { + "epoch": 1.506479244454206, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7446229457855225, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.7008464336395264, + "num_tokens": 354902971.0, + "step": 13718 + }, + { + "epoch": 1.5065890621568196, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7128217220306396, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7149548530578613, + "num_tokens": 354926622.0, + "step": 13719 + }, + { + "epoch": 1.5066988798594334, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.649181604385376, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7073959708213806, + "num_tokens": 354950837.0, + "step": 13720 + }, + { + "epoch": 1.5068086975620472, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6438987255096436, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6859661340713501, + "num_tokens": 354978412.0, + "step": 13721 + }, + { + "epoch": 1.5069185152646607, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7201778888702393, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7119989395141602, + "num_tokens": 355003622.0, + "step": 13722 + }, + { + "epoch": 1.5070283329672742, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9903242588043213, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7062975168228149, + "num_tokens": 355023549.0, + "step": 13723 + }, + { + "epoch": 1.507138150669888, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.557079792022705, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.6951379776000977, + "num_tokens": 355048551.0, + "step": 13724 + }, + { + "epoch": 1.5072479683725017, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6214540004730225, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7055239081382751, + "num_tokens": 355075336.0, + "step": 13725 + }, + { + "epoch": 1.5073577860751153, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.386326551437378, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7006590962409973, + "num_tokens": 355105548.0, + "step": 13726 + }, + { + "epoch": 1.5074676037777288, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.331693410873413, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.717435896396637, + "num_tokens": 355133482.0, + "step": 13727 + }, + { + "epoch": 1.5075774214803426, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5879361629486084, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7007878422737122, + "num_tokens": 355158234.0, + "step": 13728 + }, + { + "epoch": 1.5076872391829563, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5222980976104736, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.7036416530609131, + "num_tokens": 355184906.0, + "step": 13729 + }, + { + "epoch": 1.50779705688557, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7828361988067627, + "learning_rate": 1e-06, + "loss": 1.1066, + "mean_token_accuracy": 0.6834853291511536, + "num_tokens": 355206627.0, + "step": 13730 + }, + { + "epoch": 1.5079068745881836, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.393890619277954, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.6998347043991089, + "num_tokens": 355236100.0, + "step": 13731 + }, + { + "epoch": 1.5080166922907972, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3193490505218506, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6806987524032593, + "num_tokens": 355270428.0, + "step": 13732 + }, + { + "epoch": 1.508126509993411, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.385833263397217, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.6996663212776184, + "num_tokens": 355303500.0, + "step": 13733 + }, + { + "epoch": 1.5082363276960247, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6481056213378906, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6957502365112305, + "num_tokens": 355333276.0, + "step": 13734 + }, + { + "epoch": 1.5083461453986382, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6181211471557617, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7225051522254944, + "num_tokens": 355356832.0, + "step": 13735 + }, + { + "epoch": 1.508455963101252, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8547275066375732, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7108882665634155, + "num_tokens": 355379912.0, + "step": 13736 + }, + { + "epoch": 1.5085657808038655, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.53406023979187, + "learning_rate": 1e-06, + "loss": 1.0797, + "mean_token_accuracy": 0.6885802149772644, + "num_tokens": 355408113.0, + "step": 13737 + }, + { + "epoch": 1.5086755985064793, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.741834878921509, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.717158317565918, + "num_tokens": 355432286.0, + "step": 13738 + }, + { + "epoch": 1.508785416209093, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8156023025512695, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6881053447723389, + "num_tokens": 355457847.0, + "step": 13739 + }, + { + "epoch": 1.5088952339117065, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6545097827911377, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7139862775802612, + "num_tokens": 355483148.0, + "step": 13740 + }, + { + "epoch": 1.50900505161432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.460728406906128, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7158659100532532, + "num_tokens": 355510583.0, + "step": 13741 + }, + { + "epoch": 1.5091148693169338, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.607170343399048, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7304498553276062, + "num_tokens": 355535520.0, + "step": 13742 + }, + { + "epoch": 1.5092246870195476, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4486613273620605, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.710263192653656, + "num_tokens": 355562645.0, + "step": 13743 + }, + { + "epoch": 1.5093345047221614, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4846556186676025, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6959798336029053, + "num_tokens": 355590352.0, + "step": 13744 + }, + { + "epoch": 1.5094443224247749, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6147704124450684, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6897916793823242, + "num_tokens": 355616447.0, + "step": 13745 + }, + { + "epoch": 1.5095541401273884, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.584237575531006, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6948868036270142, + "num_tokens": 355643921.0, + "step": 13746 + }, + { + "epoch": 1.5096639578300022, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.843938112258911, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.7203848361968994, + "num_tokens": 355663688.0, + "step": 13747 + }, + { + "epoch": 1.509773775532616, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0856306552886963, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7023249864578247, + "num_tokens": 355684287.0, + "step": 13748 + }, + { + "epoch": 1.5098835932352295, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.355391025543213, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6964612007141113, + "num_tokens": 355714256.0, + "step": 13749 + }, + { + "epoch": 1.5099934109378432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.301820993423462, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.701380729675293, + "num_tokens": 355741462.0, + "step": 13750 + }, + { + "epoch": 1.5101032286404568, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.703704833984375, + "learning_rate": 1e-06, + "loss": 0.8972, + "mean_token_accuracy": 0.7324097156524658, + "num_tokens": 355762296.0, + "step": 13751 + }, + { + "epoch": 1.5102130463430705, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3997671604156494, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6845303773880005, + "num_tokens": 355792887.0, + "step": 13752 + }, + { + "epoch": 1.5103228640456843, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6016836166381836, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6992136240005493, + "num_tokens": 355821379.0, + "step": 13753 + }, + { + "epoch": 1.5104326817482978, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8720526695251465, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7127506136894226, + "num_tokens": 355843447.0, + "step": 13754 + }, + { + "epoch": 1.5105424994509113, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.593012809753418, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.7011141777038574, + "num_tokens": 355867998.0, + "step": 13755 + }, + { + "epoch": 1.510652317153525, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8838443756103516, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7221126556396484, + "num_tokens": 355888455.0, + "step": 13756 + }, + { + "epoch": 1.5107621348561389, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.499887228012085, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7243602871894836, + "num_tokens": 355913597.0, + "step": 13757 + }, + { + "epoch": 1.5108719525587526, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.651977777481079, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.7002714276313782, + "num_tokens": 355939107.0, + "step": 13758 + }, + { + "epoch": 1.5109817702613662, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4018616676330566, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7060515880584717, + "num_tokens": 355969652.0, + "step": 13759 + }, + { + "epoch": 1.5110915879639797, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.560918092727661, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7257399559020996, + "num_tokens": 355995923.0, + "step": 13760 + }, + { + "epoch": 1.5112014056665934, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4760522842407227, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7044476866722107, + "num_tokens": 356020220.0, + "step": 13761 + }, + { + "epoch": 1.5113112233692072, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4466662406921387, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6965277791023254, + "num_tokens": 356047911.0, + "step": 13762 + }, + { + "epoch": 1.5114210410718207, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6472413539886475, + "learning_rate": 1e-06, + "loss": 1.1472, + "mean_token_accuracy": 0.6758793592453003, + "num_tokens": 356074462.0, + "step": 13763 + }, + { + "epoch": 1.5115308587744345, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5723459720611572, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.717296302318573, + "num_tokens": 356101105.0, + "step": 13764 + }, + { + "epoch": 1.511640676477048, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5479016304016113, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.6875312924385071, + "num_tokens": 356128051.0, + "step": 13765 + }, + { + "epoch": 1.5117504941796618, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.783174991607666, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7131966948509216, + "num_tokens": 356151505.0, + "step": 13766 + }, + { + "epoch": 1.5118603118822755, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6072943210601807, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7086772918701172, + "num_tokens": 356177332.0, + "step": 13767 + }, + { + "epoch": 1.511970129584889, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.334721803665161, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7355798482894897, + "num_tokens": 356205591.0, + "step": 13768 + }, + { + "epoch": 1.5120799472875026, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7217979431152344, + "learning_rate": 1e-06, + "loss": 0.9098, + "mean_token_accuracy": 0.7316296100616455, + "num_tokens": 356226094.0, + "step": 13769 + }, + { + "epoch": 1.5121897649901164, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7619740962982178, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6804393529891968, + "num_tokens": 356252556.0, + "step": 13770 + }, + { + "epoch": 1.5122995826927301, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7315008640289307, + "learning_rate": 1e-06, + "loss": 0.9126, + "mean_token_accuracy": 0.726543664932251, + "num_tokens": 356276592.0, + "step": 13771 + }, + { + "epoch": 1.5124094003953439, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8119397163391113, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7101340293884277, + "num_tokens": 356298457.0, + "step": 13772 + }, + { + "epoch": 1.5125192180979574, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.612210750579834, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7009122371673584, + "num_tokens": 356324337.0, + "step": 13773 + }, + { + "epoch": 1.512629035800571, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.869551181793213, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7121996879577637, + "num_tokens": 356345853.0, + "step": 13774 + }, + { + "epoch": 1.5127388535031847, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4718542098999023, + "learning_rate": 1e-06, + "loss": 1.126, + "mean_token_accuracy": 0.6720533967018127, + "num_tokens": 356379117.0, + "step": 13775 + }, + { + "epoch": 1.5128486712057985, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7059240341186523, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7021096348762512, + "num_tokens": 356405071.0, + "step": 13776 + }, + { + "epoch": 1.512958488908412, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7648916244506836, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7101624011993408, + "num_tokens": 356425984.0, + "step": 13777 + }, + { + "epoch": 1.5130683066110255, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7293334007263184, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7105385065078735, + "num_tokens": 356448982.0, + "step": 13778 + }, + { + "epoch": 1.5131781243136393, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8131558895111084, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7174583673477173, + "num_tokens": 356468840.0, + "step": 13779 + }, + { + "epoch": 1.513287942016253, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.522777557373047, + "learning_rate": 1e-06, + "loss": 1.1199, + "mean_token_accuracy": 0.6826980113983154, + "num_tokens": 356496916.0, + "step": 13780 + }, + { + "epoch": 1.5133977597188668, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.909834861755371, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7285547256469727, + "num_tokens": 356519639.0, + "step": 13781 + }, + { + "epoch": 1.5135075774214803, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.690239667892456, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6998717188835144, + "num_tokens": 356543945.0, + "step": 13782 + }, + { + "epoch": 1.5136173951240939, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.717024326324463, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6982302069664001, + "num_tokens": 356571028.0, + "step": 13783 + }, + { + "epoch": 1.5137272128267076, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.427800178527832, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7100942730903625, + "num_tokens": 356600256.0, + "step": 13784 + }, + { + "epoch": 1.5138370305293214, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7331812381744385, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.690372884273529, + "num_tokens": 356625344.0, + "step": 13785 + }, + { + "epoch": 1.5139468482319351, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4739081859588623, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7133354544639587, + "num_tokens": 356650057.0, + "step": 13786 + }, + { + "epoch": 1.5140566659345487, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6378581523895264, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7065787315368652, + "num_tokens": 356675194.0, + "step": 13787 + }, + { + "epoch": 1.5141664836371622, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.627056121826172, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7208893299102783, + "num_tokens": 356698600.0, + "step": 13788 + }, + { + "epoch": 1.514276301339776, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.349144458770752, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6919962167739868, + "num_tokens": 356729849.0, + "step": 13789 + }, + { + "epoch": 1.5143861190423897, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7923178672790527, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7113466262817383, + "num_tokens": 356755212.0, + "step": 13790 + }, + { + "epoch": 1.5144959367450033, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3608310222625732, + "learning_rate": 1e-06, + "loss": 1.12, + "mean_token_accuracy": 0.6887215375900269, + "num_tokens": 356786954.0, + "step": 13791 + }, + { + "epoch": 1.5146057544476168, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5782177448272705, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6930820941925049, + "num_tokens": 356811653.0, + "step": 13792 + }, + { + "epoch": 1.5147155721502306, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.481844663619995, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.7008919715881348, + "num_tokens": 356841369.0, + "step": 13793 + }, + { + "epoch": 1.5148253898528443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.593545436859131, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6928614377975464, + "num_tokens": 356870485.0, + "step": 13794 + }, + { + "epoch": 1.514935207555458, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.30279278755188, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.692732572555542, + "num_tokens": 356904064.0, + "step": 13795 + }, + { + "epoch": 1.5150450252580716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.828331470489502, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7228913307189941, + "num_tokens": 356924167.0, + "step": 13796 + }, + { + "epoch": 1.5151548429606851, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.928253412246704, + "learning_rate": 1e-06, + "loss": 0.9249, + "mean_token_accuracy": 0.7279316186904907, + "num_tokens": 356944442.0, + "step": 13797 + }, + { + "epoch": 1.515264660663299, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.958665370941162, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.7045103311538696, + "num_tokens": 356963483.0, + "step": 13798 + }, + { + "epoch": 1.5153744783659127, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.727811098098755, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.6896010041236877, + "num_tokens": 356986322.0, + "step": 13799 + }, + { + "epoch": 1.5154842960685262, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.617244243621826, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.6763307452201843, + "num_tokens": 357019143.0, + "step": 13800 + }, + { + "epoch": 1.51559411377114, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.779261827468872, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7100226879119873, + "num_tokens": 357041884.0, + "step": 13801 + }, + { + "epoch": 1.5157039314737535, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3268704414367676, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6916051506996155, + "num_tokens": 357072800.0, + "step": 13802 + }, + { + "epoch": 1.5158137491763672, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6120188236236572, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7096372246742249, + "num_tokens": 357096559.0, + "step": 13803 + }, + { + "epoch": 1.515923566878981, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.820087194442749, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6844266057014465, + "num_tokens": 357118220.0, + "step": 13804 + }, + { + "epoch": 1.5160333845815945, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.635181188583374, + "learning_rate": 1e-06, + "loss": 1.1535, + "mean_token_accuracy": 0.671454668045044, + "num_tokens": 357150262.0, + "step": 13805 + }, + { + "epoch": 1.516143202284208, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.667173147201538, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7170612812042236, + "num_tokens": 357176008.0, + "step": 13806 + }, + { + "epoch": 1.5162530199868218, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6001951694488525, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6943022012710571, + "num_tokens": 357202813.0, + "step": 13807 + }, + { + "epoch": 1.5163628376894356, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5480754375457764, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7092217206954956, + "num_tokens": 357229265.0, + "step": 13808 + }, + { + "epoch": 1.5164726553920493, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4388620853424072, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7225095629692078, + "num_tokens": 357256720.0, + "step": 13809 + }, + { + "epoch": 1.5165824730946629, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5418825149536133, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7066447138786316, + "num_tokens": 357281438.0, + "step": 13810 + }, + { + "epoch": 1.5166922907972764, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4333107471466064, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7106075882911682, + "num_tokens": 357309993.0, + "step": 13811 + }, + { + "epoch": 1.5168021084998902, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5087947845458984, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6916230320930481, + "num_tokens": 357338261.0, + "step": 13812 + }, + { + "epoch": 1.516911926202504, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5443661212921143, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7090060114860535, + "num_tokens": 357364164.0, + "step": 13813 + }, + { + "epoch": 1.5170217439051175, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5642173290252686, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7185384631156921, + "num_tokens": 357390889.0, + "step": 13814 + }, + { + "epoch": 1.5171315616077312, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5214879512786865, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7007076740264893, + "num_tokens": 357417326.0, + "step": 13815 + }, + { + "epoch": 1.5172413793103448, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9028120040893555, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7188828587532043, + "num_tokens": 357437130.0, + "step": 13816 + }, + { + "epoch": 1.5173511970129585, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6393110752105713, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7090166807174683, + "num_tokens": 357462575.0, + "step": 13817 + }, + { + "epoch": 1.5174610147155723, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6132216453552246, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7212299108505249, + "num_tokens": 357485898.0, + "step": 13818 + }, + { + "epoch": 1.5175708324181858, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8241119384765625, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7072163224220276, + "num_tokens": 357508517.0, + "step": 13819 + }, + { + "epoch": 1.5176806501207993, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5726232528686523, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7169027328491211, + "num_tokens": 357535443.0, + "step": 13820 + }, + { + "epoch": 1.517790467823413, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3667449951171875, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7033854126930237, + "num_tokens": 357565210.0, + "step": 13821 + }, + { + "epoch": 1.5179002855260268, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.566011667251587, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6976465582847595, + "num_tokens": 357589676.0, + "step": 13822 + }, + { + "epoch": 1.5180101032286406, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4293439388275146, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7264112234115601, + "num_tokens": 357616481.0, + "step": 13823 + }, + { + "epoch": 1.5181199209312541, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6962766647338867, + "learning_rate": 1e-06, + "loss": 0.8596, + "mean_token_accuracy": 0.7463781833648682, + "num_tokens": 357637844.0, + "step": 13824 + }, + { + "epoch": 1.5182297386338677, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4235522747039795, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7044761776924133, + "num_tokens": 357665617.0, + "step": 13825 + }, + { + "epoch": 1.5183395563364814, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.317899703979492, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7050561904907227, + "num_tokens": 357693346.0, + "step": 13826 + }, + { + "epoch": 1.5184493740390952, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6198651790618896, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7032638788223267, + "num_tokens": 357716822.0, + "step": 13827 + }, + { + "epoch": 1.5185591917417087, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6593077182769775, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7104412317276001, + "num_tokens": 357743016.0, + "step": 13828 + }, + { + "epoch": 1.5186690094443223, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6062257289886475, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6984469890594482, + "num_tokens": 357767385.0, + "step": 13829 + }, + { + "epoch": 1.518778827146936, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 7.934982776641846, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.6949611902236938, + "num_tokens": 357794899.0, + "step": 13830 + }, + { + "epoch": 1.5188886448495498, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7944793701171875, + "learning_rate": 1e-06, + "loss": 0.9254, + "mean_token_accuracy": 0.7216916084289551, + "num_tokens": 357817416.0, + "step": 13831 + }, + { + "epoch": 1.5189984625521635, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6564908027648926, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6929648518562317, + "num_tokens": 357843749.0, + "step": 13832 + }, + { + "epoch": 1.519108280254777, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7570676803588867, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7169395685195923, + "num_tokens": 357867032.0, + "step": 13833 + }, + { + "epoch": 1.5192180979573906, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 4.317569732666016, + "learning_rate": 1e-06, + "loss": 0.8895, + "mean_token_accuracy": 0.7338054776191711, + "num_tokens": 357888379.0, + "step": 13834 + }, + { + "epoch": 1.5193279156600044, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6468276977539062, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.69774329662323, + "num_tokens": 357911911.0, + "step": 13835 + }, + { + "epoch": 1.5194377333626181, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9770596027374268, + "learning_rate": 1e-06, + "loss": 0.855, + "mean_token_accuracy": 0.7426566481590271, + "num_tokens": 357929136.0, + "step": 13836 + }, + { + "epoch": 1.5195475510652319, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4955809116363525, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7083655595779419, + "num_tokens": 357955890.0, + "step": 13837 + }, + { + "epoch": 1.5196573687678454, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 5.190907955169678, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.71541428565979, + "num_tokens": 357978799.0, + "step": 13838 + }, + { + "epoch": 1.519767186470459, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0571770668029785, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7053053379058838, + "num_tokens": 357997097.0, + "step": 13839 + }, + { + "epoch": 1.5198770041730727, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4805798530578613, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7029685378074646, + "num_tokens": 358023895.0, + "step": 13840 + }, + { + "epoch": 1.5199868218756865, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4408042430877686, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6944496035575867, + "num_tokens": 358055221.0, + "step": 13841 + }, + { + "epoch": 1.5200966395783, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.277310848236084, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7140978574752808, + "num_tokens": 358086249.0, + "step": 13842 + }, + { + "epoch": 1.5202064572809135, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5818660259246826, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7210440635681152, + "num_tokens": 358108163.0, + "step": 13843 + }, + { + "epoch": 1.5203162749835273, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5243418216705322, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6980831623077393, + "num_tokens": 358136746.0, + "step": 13844 + }, + { + "epoch": 1.520426092686141, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6967127323150635, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7017614245414734, + "num_tokens": 358161247.0, + "step": 13845 + }, + { + "epoch": 1.5205359103887548, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5921061038970947, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.695564329624176, + "num_tokens": 358185895.0, + "step": 13846 + }, + { + "epoch": 1.5206457280913683, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.629725933074951, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7241642475128174, + "num_tokens": 358207700.0, + "step": 13847 + }, + { + "epoch": 1.5207555457939819, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7396814823150635, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7143738269805908, + "num_tokens": 358229003.0, + "step": 13848 + }, + { + "epoch": 1.5208653634965956, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.4916398525238037, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7115286588668823, + "num_tokens": 358254281.0, + "step": 13849 + }, + { + "epoch": 1.5209751811992094, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.5989716053009033, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6882582902908325, + "num_tokens": 358280670.0, + "step": 13850 + }, + { + "epoch": 1.5210849989018231, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.549818992614746, + "learning_rate": 1e-06, + "loss": 1.1101, + "mean_token_accuracy": 0.686095654964447, + "num_tokens": 358310030.0, + "step": 13851 + }, + { + "epoch": 1.5211948166044367, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7483267784118652, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6933391094207764, + "num_tokens": 358335096.0, + "step": 13852 + }, + { + "epoch": 1.5213046343070502, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.7100436687469482, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7160006761550903, + "num_tokens": 358356406.0, + "step": 13853 + }, + { + "epoch": 1.521414452009664, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.818002700805664, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7219111919403076, + "num_tokens": 358377855.0, + "step": 13854 + }, + { + "epoch": 1.5215242697122777, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.54675030708313, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7274606227874756, + "num_tokens": 358405613.0, + "step": 13855 + }, + { + "epoch": 1.5216340874148913, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4359548091888428, + "learning_rate": 1e-06, + "loss": 1.1479, + "mean_token_accuracy": 0.6729768514633179, + "num_tokens": 358434728.0, + "step": 13856 + }, + { + "epoch": 1.5217439051175048, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7459003925323486, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7166489362716675, + "num_tokens": 358458558.0, + "step": 13857 + }, + { + "epoch": 1.5218537228201185, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9537651538848877, + "learning_rate": 1e-06, + "loss": 0.8649, + "mean_token_accuracy": 0.7460785508155823, + "num_tokens": 358475837.0, + "step": 13858 + }, + { + "epoch": 1.5219635405227323, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.69413685798645, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7182546257972717, + "num_tokens": 358500531.0, + "step": 13859 + }, + { + "epoch": 1.522073358225346, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0385377407073975, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7203225493431091, + "num_tokens": 358519057.0, + "step": 13860 + }, + { + "epoch": 1.5221831759279596, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6917777061462402, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7088363170623779, + "num_tokens": 358543699.0, + "step": 13861 + }, + { + "epoch": 1.5222929936305731, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8171305656433105, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7112926244735718, + "num_tokens": 358564328.0, + "step": 13862 + }, + { + "epoch": 1.522402811333187, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6590116024017334, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7218267917633057, + "num_tokens": 358587141.0, + "step": 13863 + }, + { + "epoch": 1.5225126290358006, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3983805179595947, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7139042615890503, + "num_tokens": 358616127.0, + "step": 13864 + }, + { + "epoch": 1.5226224467384142, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7357847690582275, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7008514404296875, + "num_tokens": 358639200.0, + "step": 13865 + }, + { + "epoch": 1.522732264441028, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6343672275543213, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6988292932510376, + "num_tokens": 358666300.0, + "step": 13866 + }, + { + "epoch": 1.5228420821436415, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.494722604751587, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.692156195640564, + "num_tokens": 358696624.0, + "step": 13867 + }, + { + "epoch": 1.5229518998462552, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4603376388549805, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6981909871101379, + "num_tokens": 358725844.0, + "step": 13868 + }, + { + "epoch": 1.523061717548869, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7782747745513916, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7007120251655579, + "num_tokens": 358746996.0, + "step": 13869 + }, + { + "epoch": 1.5231715352514825, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5188169479370117, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.712437629699707, + "num_tokens": 358772315.0, + "step": 13870 + }, + { + "epoch": 1.523281352954096, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.485236883163452, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7055349349975586, + "num_tokens": 358798304.0, + "step": 13871 + }, + { + "epoch": 1.5233911706567098, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.502281665802002, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.7082430720329285, + "num_tokens": 358823430.0, + "step": 13872 + }, + { + "epoch": 1.5235009883593236, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.802949905395508, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.6923315525054932, + "num_tokens": 358845496.0, + "step": 13873 + }, + { + "epoch": 1.5236108060619373, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3811216354370117, + "learning_rate": 1e-06, + "loss": 0.977, + "mean_token_accuracy": 0.7129698395729065, + "num_tokens": 358874770.0, + "step": 13874 + }, + { + "epoch": 1.5237206237645509, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6263325214385986, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7128587365150452, + "num_tokens": 358898948.0, + "step": 13875 + }, + { + "epoch": 1.5238304414671644, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.402261972427368, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7279043197631836, + "num_tokens": 358925509.0, + "step": 13876 + }, + { + "epoch": 1.5239402591697782, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3427538871765137, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7049270868301392, + "num_tokens": 358956547.0, + "step": 13877 + }, + { + "epoch": 1.524050076872392, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3809406757354736, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7110486030578613, + "num_tokens": 358985992.0, + "step": 13878 + }, + { + "epoch": 1.5241598945750054, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6733200550079346, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.7039110064506531, + "num_tokens": 359011171.0, + "step": 13879 + }, + { + "epoch": 1.5242697122776192, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0419702529907227, + "learning_rate": 1e-06, + "loss": 0.8743, + "mean_token_accuracy": 0.7356944680213928, + "num_tokens": 359029728.0, + "step": 13880 + }, + { + "epoch": 1.5243795299802327, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.832332134246826, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.6963406801223755, + "num_tokens": 359051188.0, + "step": 13881 + }, + { + "epoch": 1.5244893476828465, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7646055221557617, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7072599530220032, + "num_tokens": 359075009.0, + "step": 13882 + }, + { + "epoch": 1.5245991653854603, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 8.204280853271484, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.731865406036377, + "num_tokens": 359096328.0, + "step": 13883 + }, + { + "epoch": 1.5247089830880738, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2568085193634033, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7192592024803162, + "num_tokens": 359120262.0, + "step": 13884 + }, + { + "epoch": 1.5248188007906873, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7821080684661865, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7113081216812134, + "num_tokens": 359140696.0, + "step": 13885 + }, + { + "epoch": 1.524928618493301, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7746779918670654, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6969724893569946, + "num_tokens": 359164468.0, + "step": 13886 + }, + { + "epoch": 1.5250384361959148, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.42869234085083, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6900745630264282, + "num_tokens": 359195179.0, + "step": 13887 + }, + { + "epoch": 1.5251482538985286, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5734024047851562, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6811423301696777, + "num_tokens": 359222376.0, + "step": 13888 + }, + { + "epoch": 1.5252580716011421, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.588468313217163, + "learning_rate": 1e-06, + "loss": 1.0984, + "mean_token_accuracy": 0.6848158836364746, + "num_tokens": 359251589.0, + "step": 13889 + }, + { + "epoch": 1.5253678893037557, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.319636344909668, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6904009580612183, + "num_tokens": 359283194.0, + "step": 13890 + }, + { + "epoch": 1.5254777070063694, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.643925428390503, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7196903228759766, + "num_tokens": 359306658.0, + "step": 13891 + }, + { + "epoch": 1.5255875247089832, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.530827283859253, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.684402346611023, + "num_tokens": 359338558.0, + "step": 13892 + }, + { + "epoch": 1.5256973424115967, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6448826789855957, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7089186906814575, + "num_tokens": 359365423.0, + "step": 13893 + }, + { + "epoch": 1.5258071601142102, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8953092098236084, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6900612115859985, + "num_tokens": 359387069.0, + "step": 13894 + }, + { + "epoch": 1.525916977816824, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.657367467880249, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.702567458152771, + "num_tokens": 359412170.0, + "step": 13895 + }, + { + "epoch": 1.5260267955194378, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.430140495300293, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.683353066444397, + "num_tokens": 359441555.0, + "step": 13896 + }, + { + "epoch": 1.5261366132220515, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7522683143615723, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.7314895391464233, + "num_tokens": 359463418.0, + "step": 13897 + }, + { + "epoch": 1.526246430924665, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.319023847579956, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7136982679367065, + "num_tokens": 359493384.0, + "step": 13898 + }, + { + "epoch": 1.5263562486272786, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4119820594787598, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7110520601272583, + "num_tokens": 359525207.0, + "step": 13899 + }, + { + "epoch": 1.5264660663298923, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6655521392822266, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7070406079292297, + "num_tokens": 359554399.0, + "step": 13900 + }, + { + "epoch": 1.526575884032506, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7810909748077393, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7143033742904663, + "num_tokens": 359574869.0, + "step": 13901 + }, + { + "epoch": 1.5266857017351199, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7211618423461914, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7063570022583008, + "num_tokens": 359598923.0, + "step": 13902 + }, + { + "epoch": 1.5267955194377334, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.381545305252075, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.693178117275238, + "num_tokens": 359633756.0, + "step": 13903 + }, + { + "epoch": 1.526905337140347, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6895761489868164, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7074451446533203, + "num_tokens": 359655317.0, + "step": 13904 + }, + { + "epoch": 1.5270151548429607, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.1203925609588623, + "learning_rate": 1e-06, + "loss": 1.1046, + "mean_token_accuracy": 0.6761485934257507, + "num_tokens": 359679187.0, + "step": 13905 + }, + { + "epoch": 1.5271249725455744, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.82362961769104, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7097440958023071, + "num_tokens": 359701225.0, + "step": 13906 + }, + { + "epoch": 1.527234790248188, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 16.670928955078125, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7103391885757446, + "num_tokens": 359731034.0, + "step": 13907 + }, + { + "epoch": 1.5273446079508015, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2851526737213135, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6917581558227539, + "num_tokens": 359759267.0, + "step": 13908 + }, + { + "epoch": 1.5274544256534153, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 4.3814005851745605, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7115014791488647, + "num_tokens": 359783366.0, + "step": 13909 + }, + { + "epoch": 1.527564243356029, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4160242080688477, + "learning_rate": 1e-06, + "loss": 1.0797, + "mean_token_accuracy": 0.6865681409835815, + "num_tokens": 359813623.0, + "step": 13910 + }, + { + "epoch": 1.5276740610586428, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5707666873931885, + "learning_rate": 1e-06, + "loss": 0.931, + "mean_token_accuracy": 0.7269467711448669, + "num_tokens": 359839622.0, + "step": 13911 + }, + { + "epoch": 1.5277838787612563, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8202638626098633, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.7062616348266602, + "num_tokens": 359861511.0, + "step": 13912 + }, + { + "epoch": 1.5278936964638699, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.828911304473877, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.6966434717178345, + "num_tokens": 359882319.0, + "step": 13913 + }, + { + "epoch": 1.5280035141664836, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5929033756256104, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.6992534399032593, + "num_tokens": 359907346.0, + "step": 13914 + }, + { + "epoch": 1.5281133318690974, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6301801204681396, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7138731479644775, + "num_tokens": 359937308.0, + "step": 13915 + }, + { + "epoch": 1.5282231495717111, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3093528747558594, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7087941765785217, + "num_tokens": 359969915.0, + "step": 13916 + }, + { + "epoch": 1.5283329672743247, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8167240619659424, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7080415487289429, + "num_tokens": 359992266.0, + "step": 13917 + }, + { + "epoch": 1.5284427849769382, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7585556507110596, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7185822129249573, + "num_tokens": 360014096.0, + "step": 13918 + }, + { + "epoch": 1.528552602679552, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3834550380706787, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6864746809005737, + "num_tokens": 360044076.0, + "step": 13919 + }, + { + "epoch": 1.5286624203821657, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3126842975616455, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7064308524131775, + "num_tokens": 360073970.0, + "step": 13920 + }, + { + "epoch": 1.5287722380847792, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.541618824005127, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7122796773910522, + "num_tokens": 360099803.0, + "step": 13921 + }, + { + "epoch": 1.5288820557873928, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5762171745300293, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7024081349372864, + "num_tokens": 360124665.0, + "step": 13922 + }, + { + "epoch": 1.5289918734900065, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.373593330383301, + "learning_rate": 1e-06, + "loss": 1.1261, + "mean_token_accuracy": 0.6838020086288452, + "num_tokens": 360154624.0, + "step": 13923 + }, + { + "epoch": 1.5291016911926203, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.420607805252075, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7140865921974182, + "num_tokens": 360181063.0, + "step": 13924 + }, + { + "epoch": 1.529211508895234, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.337531805038452, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.7004107236862183, + "num_tokens": 360209406.0, + "step": 13925 + }, + { + "epoch": 1.5293213265978476, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.733177423477173, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7243136167526245, + "num_tokens": 360231664.0, + "step": 13926 + }, + { + "epoch": 1.5294311443004611, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.445708990097046, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7101322412490845, + "num_tokens": 360259971.0, + "step": 13927 + }, + { + "epoch": 1.5295409620030749, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6315064430236816, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7190876007080078, + "num_tokens": 360283538.0, + "step": 13928 + }, + { + "epoch": 1.5296507797056886, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.501532793045044, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.690265417098999, + "num_tokens": 360311061.0, + "step": 13929 + }, + { + "epoch": 1.5297605974083022, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.1817948818206787, + "learning_rate": 1e-06, + "loss": 0.9094, + "mean_token_accuracy": 0.7257070541381836, + "num_tokens": 360327539.0, + "step": 13930 + }, + { + "epoch": 1.529870415110916, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.649937868118286, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7015313506126404, + "num_tokens": 360352858.0, + "step": 13931 + }, + { + "epoch": 1.5299802328135295, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.418545961380005, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.700782060623169, + "num_tokens": 360383213.0, + "step": 13932 + }, + { + "epoch": 1.5300900505161432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.582854986190796, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6959948539733887, + "num_tokens": 360411710.0, + "step": 13933 + }, + { + "epoch": 1.530199868218757, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.63814115524292, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.6956403255462646, + "num_tokens": 360434777.0, + "step": 13934 + }, + { + "epoch": 1.5303096859213705, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 80.537353515625, + "learning_rate": 1e-06, + "loss": 1.1148, + "mean_token_accuracy": 0.6781058311462402, + "num_tokens": 360467550.0, + "step": 13935 + }, + { + "epoch": 1.530419503623984, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.661285638809204, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6900796890258789, + "num_tokens": 360494011.0, + "step": 13936 + }, + { + "epoch": 1.5305293213265978, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.3037126064300537, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.707656741142273, + "num_tokens": 360521569.0, + "step": 13937 + }, + { + "epoch": 1.5306391390292116, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8503224849700928, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7090998888015747, + "num_tokens": 360548911.0, + "step": 13938 + }, + { + "epoch": 1.5307489567318253, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6286370754241943, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.697269082069397, + "num_tokens": 360574672.0, + "step": 13939 + }, + { + "epoch": 1.5308587744344389, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6187660694122314, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7106362581253052, + "num_tokens": 360599959.0, + "step": 13940 + }, + { + "epoch": 1.5309685921370524, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4463272094726562, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7058196067810059, + "num_tokens": 360626176.0, + "step": 13941 + }, + { + "epoch": 1.5310784098396661, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.580484628677368, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7139204740524292, + "num_tokens": 360650766.0, + "step": 13942 + }, + { + "epoch": 1.53118822754228, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.598917245864868, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6962770223617554, + "num_tokens": 360674516.0, + "step": 13943 + }, + { + "epoch": 1.5312980452448934, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4096381664276123, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6956068873405457, + "num_tokens": 360702725.0, + "step": 13944 + }, + { + "epoch": 1.5314078629475072, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.548335075378418, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6945106983184814, + "num_tokens": 360727098.0, + "step": 13945 + }, + { + "epoch": 1.5315176806501207, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.526784658432007, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7054064273834229, + "num_tokens": 360755180.0, + "step": 13946 + }, + { + "epoch": 1.5316274983527345, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6522116661071777, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7209979295730591, + "num_tokens": 360779517.0, + "step": 13947 + }, + { + "epoch": 1.5317373160553482, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9079740047454834, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7013359069824219, + "num_tokens": 360799708.0, + "step": 13948 + }, + { + "epoch": 1.5318471337579618, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.8807897567749023, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.689107358455658, + "num_tokens": 360825254.0, + "step": 13949 + }, + { + "epoch": 1.5319569514605753, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.015885353088379, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7022349834442139, + "num_tokens": 360844500.0, + "step": 13950 + }, + { + "epoch": 1.532066769163189, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9013097286224365, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6967920660972595, + "num_tokens": 360864445.0, + "step": 13951 + }, + { + "epoch": 1.5321765868658028, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7890398502349854, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7287490963935852, + "num_tokens": 360887670.0, + "step": 13952 + }, + { + "epoch": 1.5322864045684166, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 16.609031677246094, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7265636920928955, + "num_tokens": 360911609.0, + "step": 13953 + }, + { + "epoch": 1.5323962222710301, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.128392219543457, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.7063060998916626, + "num_tokens": 360935516.0, + "step": 13954 + }, + { + "epoch": 1.5325060399736437, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0446488857269287, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.699962317943573, + "num_tokens": 360957172.0, + "step": 13955 + }, + { + "epoch": 1.5326158576762574, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.509777069091797, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7297117710113525, + "num_tokens": 360984712.0, + "step": 13956 + }, + { + "epoch": 1.5327256753788712, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.468813419342041, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6850191354751587, + "num_tokens": 361013040.0, + "step": 13957 + }, + { + "epoch": 1.5328354930814847, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4157943725585938, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6869407892227173, + "num_tokens": 361045049.0, + "step": 13958 + }, + { + "epoch": 1.5329453107840982, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.56916880607605, + "learning_rate": 1e-06, + "loss": 1.1083, + "mean_token_accuracy": 0.6792760491371155, + "num_tokens": 361074053.0, + "step": 13959 + }, + { + "epoch": 1.533055128486712, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5571415424346924, + "learning_rate": 1e-06, + "loss": 1.1287, + "mean_token_accuracy": 0.673032283782959, + "num_tokens": 361104096.0, + "step": 13960 + }, + { + "epoch": 1.5331649461893258, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.630254030227661, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7098830342292786, + "num_tokens": 361127413.0, + "step": 13961 + }, + { + "epoch": 1.5332747638919395, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3647115230560303, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6983612179756165, + "num_tokens": 361156264.0, + "step": 13962 + }, + { + "epoch": 1.533384581594553, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.753492593765259, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.701170802116394, + "num_tokens": 361179506.0, + "step": 13963 + }, + { + "epoch": 1.5334943992971666, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.510960340499878, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6980376243591309, + "num_tokens": 361208503.0, + "step": 13964 + }, + { + "epoch": 1.5336042169997803, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4727060794830322, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6953264474868774, + "num_tokens": 361236807.0, + "step": 13965 + }, + { + "epoch": 1.533714034702394, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4948670864105225, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7285268306732178, + "num_tokens": 361260891.0, + "step": 13966 + }, + { + "epoch": 1.5338238524050078, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8064348697662354, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7388156652450562, + "num_tokens": 361280470.0, + "step": 13967 + }, + { + "epoch": 1.5339336701076214, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5025289058685303, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7022937536239624, + "num_tokens": 361308667.0, + "step": 13968 + }, + { + "epoch": 1.534043487810235, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7102880477905273, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7144287824630737, + "num_tokens": 361333878.0, + "step": 13969 + }, + { + "epoch": 1.5341533055128487, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.707077741622925, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6888374090194702, + "num_tokens": 361360833.0, + "step": 13970 + }, + { + "epoch": 1.5342631232154624, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4354352951049805, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.698980450630188, + "num_tokens": 361388075.0, + "step": 13971 + }, + { + "epoch": 1.534372940918076, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8175580501556396, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7138839960098267, + "num_tokens": 361413148.0, + "step": 13972 + }, + { + "epoch": 1.5344827586206895, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6147525310516357, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.6984171867370605, + "num_tokens": 361441051.0, + "step": 13973 + }, + { + "epoch": 1.5345925763233033, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.773984670639038, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7132670879364014, + "num_tokens": 361464033.0, + "step": 13974 + }, + { + "epoch": 1.534702394025917, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5396127700805664, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7137042880058289, + "num_tokens": 361490130.0, + "step": 13975 + }, + { + "epoch": 1.5348122117285308, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.388737201690674, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7213926315307617, + "num_tokens": 361516786.0, + "step": 13976 + }, + { + "epoch": 1.5349220294311443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7149407863616943, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7182668447494507, + "num_tokens": 361541102.0, + "step": 13977 + }, + { + "epoch": 1.5350318471337578, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8154447078704834, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6917916536331177, + "num_tokens": 361564405.0, + "step": 13978 + }, + { + "epoch": 1.5351416648363716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.747769594192505, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7218730449676514, + "num_tokens": 361587221.0, + "step": 13979 + }, + { + "epoch": 1.5352514825389854, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5423552989959717, + "learning_rate": 1e-06, + "loss": 1.1158, + "mean_token_accuracy": 0.6761592030525208, + "num_tokens": 361617470.0, + "step": 13980 + }, + { + "epoch": 1.535361300241599, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.701491355895996, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6963249444961548, + "num_tokens": 361641579.0, + "step": 13981 + }, + { + "epoch": 1.5354711179442126, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 16.773958206176758, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7058576345443726, + "num_tokens": 361660463.0, + "step": 13982 + }, + { + "epoch": 1.5355809356468262, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6237049102783203, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7037777900695801, + "num_tokens": 361687467.0, + "step": 13983 + }, + { + "epoch": 1.53569075334944, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.731243848800659, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7179086208343506, + "num_tokens": 361711970.0, + "step": 13984 + }, + { + "epoch": 1.5358005710520537, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8497862815856934, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7116417288780212, + "num_tokens": 361734584.0, + "step": 13985 + }, + { + "epoch": 1.5359103887546672, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.385378837585449, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7026780247688293, + "num_tokens": 361764865.0, + "step": 13986 + }, + { + "epoch": 1.5360202064572808, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3887414932250977, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.6987030506134033, + "num_tokens": 361794556.0, + "step": 13987 + }, + { + "epoch": 1.5361300241598945, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.465005874633789, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.711006760597229, + "num_tokens": 361820100.0, + "step": 13988 + }, + { + "epoch": 1.5362398418625083, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4967095851898193, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6988718509674072, + "num_tokens": 361847555.0, + "step": 13989 + }, + { + "epoch": 1.536349659565122, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5399937629699707, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7109510898590088, + "num_tokens": 361873202.0, + "step": 13990 + }, + { + "epoch": 1.5364594772677356, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4997239112854004, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7139263153076172, + "num_tokens": 361900641.0, + "step": 13991 + }, + { + "epoch": 1.536569294970349, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9613521099090576, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7083815336227417, + "num_tokens": 361920081.0, + "step": 13992 + }, + { + "epoch": 1.5366791126729629, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3300161361694336, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.7001127600669861, + "num_tokens": 361951476.0, + "step": 13993 + }, + { + "epoch": 1.5367889303755766, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.58457612991333, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.7023225426673889, + "num_tokens": 361978693.0, + "step": 13994 + }, + { + "epoch": 1.5368987480781902, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.836707592010498, + "learning_rate": 1e-06, + "loss": 1.1743, + "mean_token_accuracy": 0.6692830324172974, + "num_tokens": 362001740.0, + "step": 13995 + }, + { + "epoch": 1.537008565780804, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6744062900543213, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7062512040138245, + "num_tokens": 362024253.0, + "step": 13996 + }, + { + "epoch": 1.5371183834834174, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7059051990509033, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.7000831365585327, + "num_tokens": 362052889.0, + "step": 13997 + }, + { + "epoch": 1.5372282011860312, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.403123140335083, + "learning_rate": 1e-06, + "loss": 0.9206, + "mean_token_accuracy": 0.7290257215499878, + "num_tokens": 362081419.0, + "step": 13998 + }, + { + "epoch": 1.537338018888645, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9139721393585205, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.7198996543884277, + "num_tokens": 362101127.0, + "step": 13999 + }, + { + "epoch": 1.5374478365912585, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4977145195007324, + "learning_rate": 1e-06, + "loss": 0.9106, + "mean_token_accuracy": 0.7357876896858215, + "num_tokens": 362129145.0, + "step": 14000 + }, + { + "epoch": 1.537557654293872, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.935737371444702, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7206789255142212, + "num_tokens": 362148434.0, + "step": 14001 + }, + { + "epoch": 1.5376674719964858, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.432996988296509, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6924697756767273, + "num_tokens": 362177949.0, + "step": 14002 + }, + { + "epoch": 1.5377772896990995, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.658531904220581, + "learning_rate": 1e-06, + "loss": 1.0609, + "mean_token_accuracy": 0.6883129477500916, + "num_tokens": 362202940.0, + "step": 14003 + }, + { + "epoch": 1.5378871074017133, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.699110507965088, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7031124830245972, + "num_tokens": 362227009.0, + "step": 14004 + }, + { + "epoch": 1.5379969251043268, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.752288579940796, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7189182043075562, + "num_tokens": 362251623.0, + "step": 14005 + }, + { + "epoch": 1.5381067428069404, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8451552391052246, + "learning_rate": 1e-06, + "loss": 1.115, + "mean_token_accuracy": 0.6739925146102905, + "num_tokens": 362282130.0, + "step": 14006 + }, + { + "epoch": 1.5382165605095541, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.502103090286255, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6853930950164795, + "num_tokens": 362309947.0, + "step": 14007 + }, + { + "epoch": 1.5383263782121679, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.817192316055298, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.708294153213501, + "num_tokens": 362331758.0, + "step": 14008 + }, + { + "epoch": 1.5384361959147814, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.594426393508911, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.6990941762924194, + "num_tokens": 362355589.0, + "step": 14009 + }, + { + "epoch": 1.538546013617395, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7240989208221436, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7208923697471619, + "num_tokens": 362378441.0, + "step": 14010 + }, + { + "epoch": 1.5386558313200087, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4440460205078125, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7062386870384216, + "num_tokens": 362406264.0, + "step": 14011 + }, + { + "epoch": 1.5387656490226225, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.638230323791504, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7149348258972168, + "num_tokens": 362429698.0, + "step": 14012 + }, + { + "epoch": 1.5388754667252362, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.737633466720581, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7222540378570557, + "num_tokens": 362451815.0, + "step": 14013 + }, + { + "epoch": 1.5389852844278498, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.718160629272461, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.696053683757782, + "num_tokens": 362475655.0, + "step": 14014 + }, + { + "epoch": 1.5390951021304633, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2932136058807373, + "learning_rate": 1e-06, + "loss": 1.1153, + "mean_token_accuracy": 0.6844609975814819, + "num_tokens": 362509816.0, + "step": 14015 + }, + { + "epoch": 1.539204919833077, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.571495294570923, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7151979207992554, + "num_tokens": 362536931.0, + "step": 14016 + }, + { + "epoch": 1.5393147375356908, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.859191417694092, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7138824462890625, + "num_tokens": 362558784.0, + "step": 14017 + }, + { + "epoch": 1.5394245552383046, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.758296489715576, + "learning_rate": 1e-06, + "loss": 0.887, + "mean_token_accuracy": 0.739453911781311, + "num_tokens": 362579862.0, + "step": 14018 + }, + { + "epoch": 1.539534372940918, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7606730461120605, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7286913394927979, + "num_tokens": 362604486.0, + "step": 14019 + }, + { + "epoch": 1.5396441906435316, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3677401542663574, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7163100838661194, + "num_tokens": 362632912.0, + "step": 14020 + }, + { + "epoch": 1.5397540083461454, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5111255645751953, + "learning_rate": 1e-06, + "loss": 0.9429, + "mean_token_accuracy": 0.7219816446304321, + "num_tokens": 362658011.0, + "step": 14021 + }, + { + "epoch": 1.5398638260487592, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.853916883468628, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.6949348449707031, + "num_tokens": 362678574.0, + "step": 14022 + }, + { + "epoch": 1.5399736437513727, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.599851131439209, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6916863322257996, + "num_tokens": 362707936.0, + "step": 14023 + }, + { + "epoch": 1.5400834614539862, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.542618989944458, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6887755393981934, + "num_tokens": 362736487.0, + "step": 14024 + }, + { + "epoch": 1.5401932791566, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7892377376556396, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6996159553527832, + "num_tokens": 362759878.0, + "step": 14025 + }, + { + "epoch": 1.5403030968592137, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.468963146209717, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6883291006088257, + "num_tokens": 362788417.0, + "step": 14026 + }, + { + "epoch": 1.5404129145618275, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.469452381134033, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7086336016654968, + "num_tokens": 362817928.0, + "step": 14027 + }, + { + "epoch": 1.540522732264441, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3886287212371826, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.6762294769287109, + "num_tokens": 362846569.0, + "step": 14028 + }, + { + "epoch": 1.5406325499670546, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5939390659332275, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7211720943450928, + "num_tokens": 362872228.0, + "step": 14029 + }, + { + "epoch": 1.5407423676696683, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6627819538116455, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7200479507446289, + "num_tokens": 362894173.0, + "step": 14030 + }, + { + "epoch": 1.540852185372282, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.155750274658203, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7038075923919678, + "num_tokens": 362920259.0, + "step": 14031 + }, + { + "epoch": 1.5409620030748958, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8605263233184814, + "learning_rate": 1e-06, + "loss": 0.9023, + "mean_token_accuracy": 0.7309096455574036, + "num_tokens": 362938237.0, + "step": 14032 + }, + { + "epoch": 1.5410718207775094, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6944832801818848, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.706865668296814, + "num_tokens": 362961984.0, + "step": 14033 + }, + { + "epoch": 1.541181638480123, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.776047468185425, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.680500864982605, + "num_tokens": 362985064.0, + "step": 14034 + }, + { + "epoch": 1.5412914561827367, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.528904438018799, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7026292085647583, + "num_tokens": 363010611.0, + "step": 14035 + }, + { + "epoch": 1.5414012738853504, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8915328979492188, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.7160143852233887, + "num_tokens": 363032997.0, + "step": 14036 + }, + { + "epoch": 1.541511091587964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5015904903411865, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6918035745620728, + "num_tokens": 363060709.0, + "step": 14037 + }, + { + "epoch": 1.5416209092905775, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5009284019470215, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7019150853157043, + "num_tokens": 363087751.0, + "step": 14038 + }, + { + "epoch": 1.5417307269931912, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7655603885650635, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7029582262039185, + "num_tokens": 363109821.0, + "step": 14039 + }, + { + "epoch": 1.541840544695805, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.425123929977417, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6900897026062012, + "num_tokens": 363140215.0, + "step": 14040 + }, + { + "epoch": 1.5419503623984188, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7596871852874756, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6999674439430237, + "num_tokens": 363164361.0, + "step": 14041 + }, + { + "epoch": 1.5420601801010323, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.756558418273926, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6913143396377563, + "num_tokens": 363199268.0, + "step": 14042 + }, + { + "epoch": 1.5421699978036458, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.447629690170288, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6857788562774658, + "num_tokens": 363229026.0, + "step": 14043 + }, + { + "epoch": 1.5422798155062596, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4459383487701416, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7147446870803833, + "num_tokens": 363257039.0, + "step": 14044 + }, + { + "epoch": 1.5423896332088733, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.539069890975952, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.6933330297470093, + "num_tokens": 363282409.0, + "step": 14045 + }, + { + "epoch": 1.5424994509114869, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.671076536178589, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.698966383934021, + "num_tokens": 363311899.0, + "step": 14046 + }, + { + "epoch": 1.5426092686141006, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.3647031784057617, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7115511298179626, + "num_tokens": 363337782.0, + "step": 14047 + }, + { + "epoch": 1.5427190863167142, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5766890048980713, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7175476551055908, + "num_tokens": 363361218.0, + "step": 14048 + }, + { + "epoch": 1.542828904019328, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4481313228607178, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.718465268611908, + "num_tokens": 363390987.0, + "step": 14049 + }, + { + "epoch": 1.5429387217219417, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.1762664318084717, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7230603694915771, + "num_tokens": 363423598.0, + "step": 14050 + }, + { + "epoch": 1.5430485394245552, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.614823579788208, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7148364186286926, + "num_tokens": 363445154.0, + "step": 14051 + }, + { + "epoch": 1.5431583571271688, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.333364486694336, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6906609535217285, + "num_tokens": 363475840.0, + "step": 14052 + }, + { + "epoch": 1.5432681748297825, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.690687656402588, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7254266738891602, + "num_tokens": 363497933.0, + "step": 14053 + }, + { + "epoch": 1.5433779925323963, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4586338996887207, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7030014991760254, + "num_tokens": 363524024.0, + "step": 14054 + }, + { + "epoch": 1.54348781023501, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.657618284225464, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7039226293563843, + "num_tokens": 363549265.0, + "step": 14055 + }, + { + "epoch": 1.5435976279376236, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7812819480895996, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7115497589111328, + "num_tokens": 363571016.0, + "step": 14056 + }, + { + "epoch": 1.543707445640237, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.390688419342041, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6853093504905701, + "num_tokens": 363600326.0, + "step": 14057 + }, + { + "epoch": 1.5438172633428509, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.471937417984009, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7013317346572876, + "num_tokens": 363626848.0, + "step": 14058 + }, + { + "epoch": 1.5439270810454646, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.310976505279541, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.70909583568573, + "num_tokens": 363656209.0, + "step": 14059 + }, + { + "epoch": 1.5440368987480781, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.563253402709961, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7012604475021362, + "num_tokens": 363683720.0, + "step": 14060 + }, + { + "epoch": 1.544146716450692, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.733879327774048, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7013822793960571, + "num_tokens": 363705783.0, + "step": 14061 + }, + { + "epoch": 1.5442565341533054, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6828644275665283, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.7038935422897339, + "num_tokens": 363729198.0, + "step": 14062 + }, + { + "epoch": 1.5443663518559192, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.452791213989258, + "learning_rate": 1e-06, + "loss": 0.938, + "mean_token_accuracy": 0.7251279950141907, + "num_tokens": 363757236.0, + "step": 14063 + }, + { + "epoch": 1.544476169558533, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4666860103607178, + "learning_rate": 1e-06, + "loss": 1.1133, + "mean_token_accuracy": 0.6776612401008606, + "num_tokens": 363784799.0, + "step": 14064 + }, + { + "epoch": 1.5445859872611465, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6544981002807617, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7131260633468628, + "num_tokens": 363806497.0, + "step": 14065 + }, + { + "epoch": 1.54469580496376, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5086452960968018, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6884137392044067, + "num_tokens": 363833430.0, + "step": 14066 + }, + { + "epoch": 1.5448056226663738, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.688103437423706, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6960644721984863, + "num_tokens": 363856361.0, + "step": 14067 + }, + { + "epoch": 1.5449154403689875, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6735575199127197, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7140687704086304, + "num_tokens": 363881557.0, + "step": 14068 + }, + { + "epoch": 1.5450252580716013, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.902925968170166, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7017520666122437, + "num_tokens": 363901078.0, + "step": 14069 + }, + { + "epoch": 1.5451350757742148, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7343573570251465, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7271939516067505, + "num_tokens": 363923541.0, + "step": 14070 + }, + { + "epoch": 1.5452448934768284, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.673328399658203, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7045320272445679, + "num_tokens": 363947757.0, + "step": 14071 + }, + { + "epoch": 1.5453547111794421, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.850581169128418, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6963486671447754, + "num_tokens": 363969680.0, + "step": 14072 + }, + { + "epoch": 1.5454645288820559, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6354026794433594, + "learning_rate": 1e-06, + "loss": 1.0978, + "mean_token_accuracy": 0.6854774951934814, + "num_tokens": 363995485.0, + "step": 14073 + }, + { + "epoch": 1.5455743465846694, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8754594326019287, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6929978132247925, + "num_tokens": 364017312.0, + "step": 14074 + }, + { + "epoch": 1.545684164287283, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.404717206954956, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7015054225921631, + "num_tokens": 364043938.0, + "step": 14075 + }, + { + "epoch": 1.5457939819898967, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6911208629608154, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7217229604721069, + "num_tokens": 364066453.0, + "step": 14076 + }, + { + "epoch": 1.5459037996925105, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.645108461380005, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.689520537853241, + "num_tokens": 364089828.0, + "step": 14077 + }, + { + "epoch": 1.5460136173951242, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6670081615448, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.7089729309082031, + "num_tokens": 364117049.0, + "step": 14078 + }, + { + "epoch": 1.5461234350977378, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6267499923706055, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7083857655525208, + "num_tokens": 364141653.0, + "step": 14079 + }, + { + "epoch": 1.5462332528003513, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5556881427764893, + "learning_rate": 1e-06, + "loss": 0.9113, + "mean_token_accuracy": 0.7344480156898499, + "num_tokens": 364167485.0, + "step": 14080 + }, + { + "epoch": 1.546343070502965, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.112189531326294, + "learning_rate": 1e-06, + "loss": 0.8771, + "mean_token_accuracy": 0.7379459142684937, + "num_tokens": 364183903.0, + "step": 14081 + }, + { + "epoch": 1.5464528882055788, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6431243419647217, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6949769258499146, + "num_tokens": 364210196.0, + "step": 14082 + }, + { + "epoch": 1.5465627059081926, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.609139919281006, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7186789512634277, + "num_tokens": 364235182.0, + "step": 14083 + }, + { + "epoch": 1.546672523610806, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.633782148361206, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7028210759162903, + "num_tokens": 364260900.0, + "step": 14084 + }, + { + "epoch": 1.5467823413134196, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.713472366333008, + "learning_rate": 1e-06, + "loss": 1.0874, + "mean_token_accuracy": 0.6852527856826782, + "num_tokens": 364288023.0, + "step": 14085 + }, + { + "epoch": 1.5468921590160334, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.327486515045166, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7028762102127075, + "num_tokens": 364318840.0, + "step": 14086 + }, + { + "epoch": 1.5470019767186471, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4652841091156006, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.705162763595581, + "num_tokens": 364346546.0, + "step": 14087 + }, + { + "epoch": 1.5471117944212607, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.948502779006958, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.717130184173584, + "num_tokens": 364367853.0, + "step": 14088 + }, + { + "epoch": 1.5472216121238742, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4599287509918213, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7017074823379517, + "num_tokens": 364395855.0, + "step": 14089 + }, + { + "epoch": 1.547331429826488, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6269094944000244, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6823113560676575, + "num_tokens": 364421719.0, + "step": 14090 + }, + { + "epoch": 1.5474412475291017, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.799041509628296, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7219256162643433, + "num_tokens": 364444366.0, + "step": 14091 + }, + { + "epoch": 1.5475510652317155, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6794495582580566, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6993119716644287, + "num_tokens": 364467061.0, + "step": 14092 + }, + { + "epoch": 1.547660882934329, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7594711780548096, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7159782648086548, + "num_tokens": 364488964.0, + "step": 14093 + }, + { + "epoch": 1.5477707006369426, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.832881450653076, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7014200687408447, + "num_tokens": 364511490.0, + "step": 14094 + }, + { + "epoch": 1.5478805183395563, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5918452739715576, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7121454477310181, + "num_tokens": 364537702.0, + "step": 14095 + }, + { + "epoch": 1.54799033604217, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4006388187408447, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7094249725341797, + "num_tokens": 364566065.0, + "step": 14096 + }, + { + "epoch": 1.5481001537447838, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6204328536987305, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7118914127349854, + "num_tokens": 364591670.0, + "step": 14097 + }, + { + "epoch": 1.5482099714473974, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4097485542297363, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7189883589744568, + "num_tokens": 364622290.0, + "step": 14098 + }, + { + "epoch": 1.548319789150011, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9155280590057373, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7085676789283752, + "num_tokens": 364643694.0, + "step": 14099 + }, + { + "epoch": 1.5484296068526247, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.466925859451294, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.7017968893051147, + "num_tokens": 364671334.0, + "step": 14100 + }, + { + "epoch": 1.5485394245552384, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.820371150970459, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6988210678100586, + "num_tokens": 364694191.0, + "step": 14101 + }, + { + "epoch": 1.548649242257852, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7466518878936768, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6993552446365356, + "num_tokens": 364718287.0, + "step": 14102 + }, + { + "epoch": 1.5487590599604655, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6872715950012207, + "learning_rate": 1e-06, + "loss": 0.9205, + "mean_token_accuracy": 0.7357164621353149, + "num_tokens": 364742007.0, + "step": 14103 + }, + { + "epoch": 1.5488688776630792, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.396792411804199, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.7018023133277893, + "num_tokens": 364772320.0, + "step": 14104 + }, + { + "epoch": 1.548978695365693, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7216742038726807, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7096061110496521, + "num_tokens": 364794693.0, + "step": 14105 + }, + { + "epoch": 1.5490885130683067, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.518441915512085, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.6992954015731812, + "num_tokens": 364821681.0, + "step": 14106 + }, + { + "epoch": 1.5491983307709203, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.317918300628662, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.6790245771408081, + "num_tokens": 364854392.0, + "step": 14107 + }, + { + "epoch": 1.5493081484735338, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6655566692352295, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7081170082092285, + "num_tokens": 364882341.0, + "step": 14108 + }, + { + "epoch": 1.5494179661761476, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6661441326141357, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.7006446123123169, + "num_tokens": 364905926.0, + "step": 14109 + }, + { + "epoch": 1.5495277838787613, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.636554002761841, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6883931159973145, + "num_tokens": 364934990.0, + "step": 14110 + }, + { + "epoch": 1.5496376015813749, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.591653347015381, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6967158317565918, + "num_tokens": 364960789.0, + "step": 14111 + }, + { + "epoch": 1.5497474192839886, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.665461301803589, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.6973192691802979, + "num_tokens": 364986802.0, + "step": 14112 + }, + { + "epoch": 1.5498572369866022, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.677887439727783, + "learning_rate": 1e-06, + "loss": 1.0609, + "mean_token_accuracy": 0.7022722363471985, + "num_tokens": 365012297.0, + "step": 14113 + }, + { + "epoch": 1.549967054689216, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6405551433563232, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.7206142544746399, + "num_tokens": 365037997.0, + "step": 14114 + }, + { + "epoch": 1.5500768723918297, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9649109840393066, + "learning_rate": 1e-06, + "loss": 0.9207, + "mean_token_accuracy": 0.7366840839385986, + "num_tokens": 365058239.0, + "step": 14115 + }, + { + "epoch": 1.5501866900944432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.644681453704834, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6946715712547302, + "num_tokens": 365086407.0, + "step": 14116 + }, + { + "epoch": 1.5502965077970567, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6653285026550293, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7248018383979797, + "num_tokens": 365109176.0, + "step": 14117 + }, + { + "epoch": 1.5504063254996705, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8402347564697266, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.7078166007995605, + "num_tokens": 365131615.0, + "step": 14118 + }, + { + "epoch": 1.5505161432022843, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6526541709899902, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6917741298675537, + "num_tokens": 365158257.0, + "step": 14119 + }, + { + "epoch": 1.550625960904898, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0183815956115723, + "learning_rate": 1e-06, + "loss": 0.8623, + "mean_token_accuracy": 0.7433122396469116, + "num_tokens": 365175178.0, + "step": 14120 + }, + { + "epoch": 1.5507357786075116, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.675126552581787, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7121479511260986, + "num_tokens": 365197933.0, + "step": 14121 + }, + { + "epoch": 1.550845596310125, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.540706157684326, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6916440725326538, + "num_tokens": 365233179.0, + "step": 14122 + }, + { + "epoch": 1.5509554140127388, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6173038482666016, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6956836581230164, + "num_tokens": 365259249.0, + "step": 14123 + }, + { + "epoch": 1.5510652317153526, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.61999773979187, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7090109586715698, + "num_tokens": 365282260.0, + "step": 14124 + }, + { + "epoch": 1.5511750494179661, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6262171268463135, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.699863851070404, + "num_tokens": 365308282.0, + "step": 14125 + }, + { + "epoch": 1.55128486712058, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6613855361938477, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.7081723213195801, + "num_tokens": 365335380.0, + "step": 14126 + }, + { + "epoch": 1.5513946848231934, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.950709819793701, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.7040657997131348, + "num_tokens": 365360887.0, + "step": 14127 + }, + { + "epoch": 1.5515045025258072, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8098537921905518, + "learning_rate": 1e-06, + "loss": 0.9127, + "mean_token_accuracy": 0.7286250591278076, + "num_tokens": 365380668.0, + "step": 14128 + }, + { + "epoch": 1.551614320228421, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.650933027267456, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7136257886886597, + "num_tokens": 365407323.0, + "step": 14129 + }, + { + "epoch": 1.5517241379310345, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.979015827178955, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7270199060440063, + "num_tokens": 365426765.0, + "step": 14130 + }, + { + "epoch": 1.551833955633648, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5754709243774414, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.708776593208313, + "num_tokens": 365452730.0, + "step": 14131 + }, + { + "epoch": 1.5519437733362618, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7896921634674072, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7141286134719849, + "num_tokens": 365478368.0, + "step": 14132 + }, + { + "epoch": 1.5520535910388755, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.658106803894043, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.701693058013916, + "num_tokens": 365502634.0, + "step": 14133 + }, + { + "epoch": 1.5521634087414893, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.75079607963562, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7183347940444946, + "num_tokens": 365523409.0, + "step": 14134 + }, + { + "epoch": 1.5522732264441028, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5772204399108887, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7109935879707336, + "num_tokens": 365550306.0, + "step": 14135 + }, + { + "epoch": 1.5523830441467164, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.47007417678833, + "learning_rate": 1e-06, + "loss": 0.9063, + "mean_token_accuracy": 0.7324573993682861, + "num_tokens": 365575175.0, + "step": 14136 + }, + { + "epoch": 1.55249286184933, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.985840320587158, + "learning_rate": 1e-06, + "loss": 0.8918, + "mean_token_accuracy": 0.7352436780929565, + "num_tokens": 365592589.0, + "step": 14137 + }, + { + "epoch": 1.5526026795519439, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7895729541778564, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7161458730697632, + "num_tokens": 365615450.0, + "step": 14138 + }, + { + "epoch": 1.5527124972545574, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4853501319885254, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7098464369773865, + "num_tokens": 365643137.0, + "step": 14139 + }, + { + "epoch": 1.552822314957171, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4882678985595703, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.711715042591095, + "num_tokens": 365670653.0, + "step": 14140 + }, + { + "epoch": 1.5529321326597847, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.51727294921875, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6864933371543884, + "num_tokens": 365699304.0, + "step": 14141 + }, + { + "epoch": 1.5530419503623984, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6469757556915283, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6898350715637207, + "num_tokens": 365727050.0, + "step": 14142 + }, + { + "epoch": 1.5531517680650122, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5140204429626465, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7087918519973755, + "num_tokens": 365752438.0, + "step": 14143 + }, + { + "epoch": 1.5532615857676257, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6087992191314697, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7038382291793823, + "num_tokens": 365775100.0, + "step": 14144 + }, + { + "epoch": 1.5533714034702393, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9289724826812744, + "learning_rate": 1e-06, + "loss": 1.1196, + "mean_token_accuracy": 0.6763413548469543, + "num_tokens": 365804574.0, + "step": 14145 + }, + { + "epoch": 1.553481221172853, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6723532676696777, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6931217908859253, + "num_tokens": 365827418.0, + "step": 14146 + }, + { + "epoch": 1.5535910388754668, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7896218299865723, + "learning_rate": 1e-06, + "loss": 0.8946, + "mean_token_accuracy": 0.7394090294837952, + "num_tokens": 365847419.0, + "step": 14147 + }, + { + "epoch": 1.5537008565780805, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.502052068710327, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7094271183013916, + "num_tokens": 365874118.0, + "step": 14148 + }, + { + "epoch": 1.553810674280694, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.508430004119873, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6999983191490173, + "num_tokens": 365902006.0, + "step": 14149 + }, + { + "epoch": 1.5539204919833076, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.896652936935425, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7061911821365356, + "num_tokens": 365922732.0, + "step": 14150 + }, + { + "epoch": 1.5540303096859214, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5166618824005127, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6816335320472717, + "num_tokens": 365951237.0, + "step": 14151 + }, + { + "epoch": 1.5541401273885351, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6374385356903076, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7098931670188904, + "num_tokens": 365976298.0, + "step": 14152 + }, + { + "epoch": 1.5542499450911487, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8661699295043945, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6894097328186035, + "num_tokens": 366003615.0, + "step": 14153 + }, + { + "epoch": 1.5543597627937622, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.682976007461548, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7040033340454102, + "num_tokens": 366028463.0, + "step": 14154 + }, + { + "epoch": 1.554469580496376, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.518049716949463, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7021410465240479, + "num_tokens": 366054708.0, + "step": 14155 + }, + { + "epoch": 1.5545793981989897, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3644654750823975, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7073976993560791, + "num_tokens": 366084130.0, + "step": 14156 + }, + { + "epoch": 1.5546892159016035, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8742010593414307, + "learning_rate": 1e-06, + "loss": 0.9237, + "mean_token_accuracy": 0.7288745641708374, + "num_tokens": 366105798.0, + "step": 14157 + }, + { + "epoch": 1.554799033604217, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.390371084213257, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6922601461410522, + "num_tokens": 366138830.0, + "step": 14158 + }, + { + "epoch": 1.5549088513068305, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6956520080566406, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7201909422874451, + "num_tokens": 366164338.0, + "step": 14159 + }, + { + "epoch": 1.5550186690094443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.715550422668457, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.7033383846282959, + "num_tokens": 366188310.0, + "step": 14160 + }, + { + "epoch": 1.555128486712058, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3688881397247314, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7192400097846985, + "num_tokens": 366218560.0, + "step": 14161 + }, + { + "epoch": 1.5552383044146716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.790178060531616, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.7120321989059448, + "num_tokens": 366239808.0, + "step": 14162 + }, + { + "epoch": 1.5553481221172853, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6920788288116455, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7173267602920532, + "num_tokens": 366265356.0, + "step": 14163 + }, + { + "epoch": 1.5554579398198989, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4986042976379395, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7061816453933716, + "num_tokens": 366294312.0, + "step": 14164 + }, + { + "epoch": 1.5555677575225126, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.928162097930908, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.697591245174408, + "num_tokens": 366315345.0, + "step": 14165 + }, + { + "epoch": 1.5556775752251264, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9087257385253906, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.695142924785614, + "num_tokens": 366338596.0, + "step": 14166 + }, + { + "epoch": 1.55578739292774, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8588528633117676, + "learning_rate": 1e-06, + "loss": 0.902, + "mean_token_accuracy": 0.7311276197433472, + "num_tokens": 366360721.0, + "step": 14167 + }, + { + "epoch": 1.5558972106303535, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7530410289764404, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.719942569732666, + "num_tokens": 366381748.0, + "step": 14168 + }, + { + "epoch": 1.5560070283329672, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6375057697296143, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7087020874023438, + "num_tokens": 366408008.0, + "step": 14169 + }, + { + "epoch": 1.556116846035581, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.792792558670044, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7077901363372803, + "num_tokens": 366431701.0, + "step": 14170 + }, + { + "epoch": 1.5562266637381947, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6483254432678223, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7151470184326172, + "num_tokens": 366457522.0, + "step": 14171 + }, + { + "epoch": 1.5563364814408083, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.915006399154663, + "learning_rate": 1e-06, + "loss": 0.9286, + "mean_token_accuracy": 0.7233086824417114, + "num_tokens": 366481980.0, + "step": 14172 + }, + { + "epoch": 1.5564462991434218, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8142974376678467, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6850991249084473, + "num_tokens": 366504694.0, + "step": 14173 + }, + { + "epoch": 1.5565561168460356, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7534821033477783, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.6999263167381287, + "num_tokens": 366526182.0, + "step": 14174 + }, + { + "epoch": 1.5566659345486493, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.316032886505127, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7158910632133484, + "num_tokens": 366555008.0, + "step": 14175 + }, + { + "epoch": 1.5567757522512629, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.453225612640381, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7106467485427856, + "num_tokens": 366580796.0, + "step": 14176 + }, + { + "epoch": 1.5568855699538766, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5237252712249756, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6993463039398193, + "num_tokens": 366606322.0, + "step": 14177 + }, + { + "epoch": 1.5569953876564901, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5968315601348877, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7088507413864136, + "num_tokens": 366629526.0, + "step": 14178 + }, + { + "epoch": 1.557105205359104, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.791527509689331, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7202504277229309, + "num_tokens": 366650121.0, + "step": 14179 + }, + { + "epoch": 1.5572150230617177, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7827093601226807, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7057934999465942, + "num_tokens": 366673256.0, + "step": 14180 + }, + { + "epoch": 1.5573248407643312, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4187352657318115, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7023328542709351, + "num_tokens": 366703456.0, + "step": 14181 + }, + { + "epoch": 1.5574346584669447, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4618539810180664, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6926940679550171, + "num_tokens": 366734352.0, + "step": 14182 + }, + { + "epoch": 1.5575444761695585, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.548163414001465, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7091583609580994, + "num_tokens": 366761205.0, + "step": 14183 + }, + { + "epoch": 1.5576542938721722, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.541914224624634, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7126554846763611, + "num_tokens": 366785165.0, + "step": 14184 + }, + { + "epoch": 1.557764111574786, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.47109055519104, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.71290123462677, + "num_tokens": 366809717.0, + "step": 14185 + }, + { + "epoch": 1.5578739292773995, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7065670490264893, + "learning_rate": 1e-06, + "loss": 0.8839, + "mean_token_accuracy": 0.7341043949127197, + "num_tokens": 366832449.0, + "step": 14186 + }, + { + "epoch": 1.557983746980013, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6976234912872314, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7117615938186646, + "num_tokens": 366857295.0, + "step": 14187 + }, + { + "epoch": 1.5580935646826268, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6679272651672363, + "learning_rate": 1e-06, + "loss": 1.1254, + "mean_token_accuracy": 0.6787022352218628, + "num_tokens": 366885833.0, + "step": 14188 + }, + { + "epoch": 1.5582033823852406, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.646289587020874, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.7006867527961731, + "num_tokens": 366912767.0, + "step": 14189 + }, + { + "epoch": 1.5583132000878541, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.728116035461426, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6981186270713806, + "num_tokens": 366935518.0, + "step": 14190 + }, + { + "epoch": 1.5584230177904677, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.629330635070801, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7097562551498413, + "num_tokens": 366960821.0, + "step": 14191 + }, + { + "epoch": 1.5585328354930814, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2725353240966797, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.6847390532493591, + "num_tokens": 366996506.0, + "step": 14192 + }, + { + "epoch": 1.5586426531956952, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6066946983337402, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6974673271179199, + "num_tokens": 367021721.0, + "step": 14193 + }, + { + "epoch": 1.558752470898309, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5773909091949463, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7160450220108032, + "num_tokens": 367047808.0, + "step": 14194 + }, + { + "epoch": 1.5588622886009225, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.566699743270874, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6895318627357483, + "num_tokens": 367073388.0, + "step": 14195 + }, + { + "epoch": 1.558972106303536, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4630606174468994, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6929147243499756, + "num_tokens": 367101125.0, + "step": 14196 + }, + { + "epoch": 1.5590819240061498, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.024930715560913, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6905854940414429, + "num_tokens": 367126849.0, + "step": 14197 + }, + { + "epoch": 1.5591917417087635, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.039217948913574, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.6990577578544617, + "num_tokens": 367150035.0, + "step": 14198 + }, + { + "epoch": 1.5593015594113773, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.774467945098877, + "learning_rate": 1e-06, + "loss": 0.9311, + "mean_token_accuracy": 0.7257601618766785, + "num_tokens": 367171590.0, + "step": 14199 + }, + { + "epoch": 1.5594113771139908, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3816492557525635, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7167893648147583, + "num_tokens": 367199356.0, + "step": 14200 + }, + { + "epoch": 1.5595211948166043, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5415759086608887, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6869152188301086, + "num_tokens": 367228959.0, + "step": 14201 + }, + { + "epoch": 1.559631012519218, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.488056182861328, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7049994468688965, + "num_tokens": 367256230.0, + "step": 14202 + }, + { + "epoch": 1.5597408302218319, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3750882148742676, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.695202112197876, + "num_tokens": 367287741.0, + "step": 14203 + }, + { + "epoch": 1.5598506479244454, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8138985633850098, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7224467992782593, + "num_tokens": 367310656.0, + "step": 14204 + }, + { + "epoch": 1.559960465627059, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.462820529937744, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.710364580154419, + "num_tokens": 367337462.0, + "step": 14205 + }, + { + "epoch": 1.5600702833296727, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4341609477996826, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7133791446685791, + "num_tokens": 367363685.0, + "step": 14206 + }, + { + "epoch": 1.5601801010322864, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.745765447616577, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7039997577667236, + "num_tokens": 367386120.0, + "step": 14207 + }, + { + "epoch": 1.5602899187349002, + "ewc_loss": 9.417533874511719e-06, + "grad_norm": 2.712935209274292, + "learning_rate": 1e-06, + "loss": 1.1144, + "mean_token_accuracy": 0.6747102737426758, + "num_tokens": 367417056.0, + "step": 14208 + }, + { + "epoch": 1.5603997364375137, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5124034881591797, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7102495431900024, + "num_tokens": 367442342.0, + "step": 14209 + }, + { + "epoch": 1.5605095541401273, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2668755054473877, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7099540829658508, + "num_tokens": 367472690.0, + "step": 14210 + }, + { + "epoch": 1.560619371842741, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.576219320297241, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6968083381652832, + "num_tokens": 367498412.0, + "step": 14211 + }, + { + "epoch": 1.5607291895453548, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.464345932006836, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7212851047515869, + "num_tokens": 367525955.0, + "step": 14212 + }, + { + "epoch": 1.5608390072479685, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3555092811584473, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7230793237686157, + "num_tokens": 367554794.0, + "step": 14213 + }, + { + "epoch": 1.560948824950582, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3771045207977295, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6975647211074829, + "num_tokens": 367586005.0, + "step": 14214 + }, + { + "epoch": 1.5610586426531956, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.498365640640259, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7191894054412842, + "num_tokens": 367612133.0, + "step": 14215 + }, + { + "epoch": 1.5611684603558094, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7382519245147705, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.710396409034729, + "num_tokens": 367633180.0, + "step": 14216 + }, + { + "epoch": 1.5612782780584231, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4499027729034424, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6952642798423767, + "num_tokens": 367663191.0, + "step": 14217 + }, + { + "epoch": 1.5613880957610367, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.576526403427124, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6958411335945129, + "num_tokens": 367689478.0, + "step": 14218 + }, + { + "epoch": 1.5614979134636502, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.535694122314453, + "learning_rate": 1e-06, + "loss": 0.9245, + "mean_token_accuracy": 0.7282481789588928, + "num_tokens": 367713709.0, + "step": 14219 + }, + { + "epoch": 1.561607731166264, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4775547981262207, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6885796189308167, + "num_tokens": 367744479.0, + "step": 14220 + }, + { + "epoch": 1.5617175488688777, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.789494276046753, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.7231955528259277, + "num_tokens": 367765403.0, + "step": 14221 + }, + { + "epoch": 1.5618273665714915, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.528686046600342, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7132726311683655, + "num_tokens": 367792333.0, + "step": 14222 + }, + { + "epoch": 1.561937184274105, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.586864471435547, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7136332988739014, + "num_tokens": 367817925.0, + "step": 14223 + }, + { + "epoch": 1.5620470019767185, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3931479454040527, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7071912288665771, + "num_tokens": 367846064.0, + "step": 14224 + }, + { + "epoch": 1.5621568196793323, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7596912384033203, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.719086766242981, + "num_tokens": 367871089.0, + "step": 14225 + }, + { + "epoch": 1.562266637381946, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.924147129058838, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7145102620124817, + "num_tokens": 367891372.0, + "step": 14226 + }, + { + "epoch": 1.5623764550845596, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7523090839385986, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.7164218425750732, + "num_tokens": 367914436.0, + "step": 14227 + }, + { + "epoch": 1.5624862727871733, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9695329666137695, + "learning_rate": 1e-06, + "loss": 0.8763, + "mean_token_accuracy": 0.7382799386978149, + "num_tokens": 367931973.0, + "step": 14228 + }, + { + "epoch": 1.5625960904897869, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.570409059524536, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7097780704498291, + "num_tokens": 367957829.0, + "step": 14229 + }, + { + "epoch": 1.5627059081924006, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.696643114089966, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6864689588546753, + "num_tokens": 367981594.0, + "step": 14230 + }, + { + "epoch": 1.5628157258950144, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7648403644561768, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.706392765045166, + "num_tokens": 368012875.0, + "step": 14231 + }, + { + "epoch": 1.562925543597628, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7969064712524414, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7113060355186462, + "num_tokens": 368038115.0, + "step": 14232 + }, + { + "epoch": 1.5630353613002415, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.756892681121826, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6851240396499634, + "num_tokens": 368064082.0, + "step": 14233 + }, + { + "epoch": 1.5631451790028552, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.820122480392456, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7106481790542603, + "num_tokens": 368085281.0, + "step": 14234 + }, + { + "epoch": 1.563254996705469, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.735797643661499, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7237356901168823, + "num_tokens": 368108808.0, + "step": 14235 + }, + { + "epoch": 1.5633648144080827, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8474042415618896, + "learning_rate": 1e-06, + "loss": 0.8656, + "mean_token_accuracy": 0.7428126335144043, + "num_tokens": 368127261.0, + "step": 14236 + }, + { + "epoch": 1.5634746321106963, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.410341739654541, + "learning_rate": 1e-06, + "loss": 0.9151, + "mean_token_accuracy": 0.7335065007209778, + "num_tokens": 368152865.0, + "step": 14237 + }, + { + "epoch": 1.5635844498133098, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3090386390686035, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6924622058868408, + "num_tokens": 368183795.0, + "step": 14238 + }, + { + "epoch": 1.5636942675159236, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.496964454650879, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.6950299739837646, + "num_tokens": 368209576.0, + "step": 14239 + }, + { + "epoch": 1.5638040852185373, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4243288040161133, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7129764556884766, + "num_tokens": 368240376.0, + "step": 14240 + }, + { + "epoch": 1.5639139029211508, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3267006874084473, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7115763425827026, + "num_tokens": 368269971.0, + "step": 14241 + }, + { + "epoch": 1.5640237206237646, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.006865978240967, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7186989784240723, + "num_tokens": 368288264.0, + "step": 14242 + }, + { + "epoch": 1.5641335383263781, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.532824754714966, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7180508375167847, + "num_tokens": 368313174.0, + "step": 14243 + }, + { + "epoch": 1.564243356028992, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.156790018081665, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7098606824874878, + "num_tokens": 368338117.0, + "step": 14244 + }, + { + "epoch": 1.5643531737316057, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5800838470458984, + "learning_rate": 1e-06, + "loss": 0.95, + "mean_token_accuracy": 0.7176535129547119, + "num_tokens": 368363618.0, + "step": 14245 + }, + { + "epoch": 1.5644629914342192, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.000239849090576, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6858389973640442, + "num_tokens": 368384049.0, + "step": 14246 + }, + { + "epoch": 1.5645728091368327, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9747002124786377, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.728023886680603, + "num_tokens": 368404667.0, + "step": 14247 + }, + { + "epoch": 1.5646826268394465, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.640444755554199, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7095261216163635, + "num_tokens": 368429066.0, + "step": 14248 + }, + { + "epoch": 1.5647924445420602, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0983338356018066, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6961718797683716, + "num_tokens": 368455380.0, + "step": 14249 + }, + { + "epoch": 1.564902262244674, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8909270763397217, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7166200280189514, + "num_tokens": 368474136.0, + "step": 14250 + }, + { + "epoch": 1.5650120799472875, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3021669387817383, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.6941756010055542, + "num_tokens": 368505615.0, + "step": 14251 + }, + { + "epoch": 1.565121897649901, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6539957523345947, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6911860108375549, + "num_tokens": 368530575.0, + "step": 14252 + }, + { + "epoch": 1.5652317153525148, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7730603218078613, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7150002121925354, + "num_tokens": 368550127.0, + "step": 14253 + }, + { + "epoch": 1.5653415330551286, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8995609283447266, + "learning_rate": 1e-06, + "loss": 0.9442, + "mean_token_accuracy": 0.7196523547172546, + "num_tokens": 368570110.0, + "step": 14254 + }, + { + "epoch": 1.565451350757742, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.690916061401367, + "learning_rate": 1e-06, + "loss": 0.8894, + "mean_token_accuracy": 0.7437542080879211, + "num_tokens": 368593773.0, + "step": 14255 + }, + { + "epoch": 1.5655611684603556, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6107428073883057, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7303299903869629, + "num_tokens": 368620602.0, + "step": 14256 + }, + { + "epoch": 1.5656709861629694, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7869043350219727, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7160114645957947, + "num_tokens": 368645461.0, + "step": 14257 + }, + { + "epoch": 1.5657808038655832, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5536727905273438, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6943435668945312, + "num_tokens": 368671850.0, + "step": 14258 + }, + { + "epoch": 1.565890621568197, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.450911521911621, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7127657532691956, + "num_tokens": 368697958.0, + "step": 14259 + }, + { + "epoch": 1.5660004392708105, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8736536502838135, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6993999481201172, + "num_tokens": 368725058.0, + "step": 14260 + }, + { + "epoch": 1.566110256973424, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.3161938190460205, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6901092529296875, + "num_tokens": 368750045.0, + "step": 14261 + }, + { + "epoch": 1.5662200746760377, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5238304138183594, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6948161125183105, + "num_tokens": 368773970.0, + "step": 14262 + }, + { + "epoch": 1.5663298923786515, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6179912090301514, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7174644470214844, + "num_tokens": 368797353.0, + "step": 14263 + }, + { + "epoch": 1.5664397100812653, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.767080783843994, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7058591246604919, + "num_tokens": 368822956.0, + "step": 14264 + }, + { + "epoch": 1.5665495277838788, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.9580979347229004, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6904053688049316, + "num_tokens": 368844391.0, + "step": 14265 + }, + { + "epoch": 1.5666593454864923, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3990890979766846, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7139824032783508, + "num_tokens": 368871579.0, + "step": 14266 + }, + { + "epoch": 1.566769163189106, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6399221420288086, + "learning_rate": 1e-06, + "loss": 0.8209, + "mean_token_accuracy": 0.7498154640197754, + "num_tokens": 368892953.0, + "step": 14267 + }, + { + "epoch": 1.5668789808917198, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.638484001159668, + "learning_rate": 1e-06, + "loss": 0.9407, + "mean_token_accuracy": 0.7213456034660339, + "num_tokens": 368915633.0, + "step": 14268 + }, + { + "epoch": 1.5669887985943334, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7620151042938232, + "learning_rate": 1e-06, + "loss": 0.9378, + "mean_token_accuracy": 0.73832106590271, + "num_tokens": 368937851.0, + "step": 14269 + }, + { + "epoch": 1.567098616296947, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7636256217956543, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.713805079460144, + "num_tokens": 368959935.0, + "step": 14270 + }, + { + "epoch": 1.5672084339995607, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7054405212402344, + "learning_rate": 1e-06, + "loss": 1.1183, + "mean_token_accuracy": 0.6841284036636353, + "num_tokens": 368983887.0, + "step": 14271 + }, + { + "epoch": 1.5673182517021744, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.601679563522339, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6959733963012695, + "num_tokens": 369007119.0, + "step": 14272 + }, + { + "epoch": 1.5674280694047882, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4853224754333496, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.7047007083892822, + "num_tokens": 369038244.0, + "step": 14273 + }, + { + "epoch": 1.5675378871074017, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6007704734802246, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7068837285041809, + "num_tokens": 369064040.0, + "step": 14274 + }, + { + "epoch": 1.5676477048100153, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.719017267227173, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7151666879653931, + "num_tokens": 369099983.0, + "step": 14275 + }, + { + "epoch": 1.567757522512629, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.688870668411255, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.705985963344574, + "num_tokens": 369124905.0, + "step": 14276 + }, + { + "epoch": 1.5678673402152428, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.50472092628479, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6930828094482422, + "num_tokens": 369152553.0, + "step": 14277 + }, + { + "epoch": 1.5679771579178565, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7434496879577637, + "learning_rate": 1e-06, + "loss": 0.91, + "mean_token_accuracy": 0.7231727838516235, + "num_tokens": 369173755.0, + "step": 14278 + }, + { + "epoch": 1.56808697562047, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.260106086730957, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7168475985527039, + "num_tokens": 369204835.0, + "step": 14279 + }, + { + "epoch": 1.5681967933230836, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6076977252960205, + "learning_rate": 1e-06, + "loss": 1.0846, + "mean_token_accuracy": 0.6883232593536377, + "num_tokens": 369229517.0, + "step": 14280 + }, + { + "epoch": 1.5683066110256974, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6618735790252686, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6948850154876709, + "num_tokens": 369251994.0, + "step": 14281 + }, + { + "epoch": 1.568416428728311, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.382293224334717, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.6948786377906799, + "num_tokens": 369278656.0, + "step": 14282 + }, + { + "epoch": 1.5685262464309246, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7449798583984375, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6987544298171997, + "num_tokens": 369307060.0, + "step": 14283 + }, + { + "epoch": 1.5686360641335382, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5120718479156494, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6950033903121948, + "num_tokens": 369334937.0, + "step": 14284 + }, + { + "epoch": 1.568745881836152, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.727639675140381, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7087044715881348, + "num_tokens": 369359214.0, + "step": 14285 + }, + { + "epoch": 1.5688556995387657, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8112452030181885, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7187183499336243, + "num_tokens": 369387553.0, + "step": 14286 + }, + { + "epoch": 1.5689655172413794, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.702630043029785, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7145059704780579, + "num_tokens": 369411021.0, + "step": 14287 + }, + { + "epoch": 1.569075334943993, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.423884153366089, + "learning_rate": 1e-06, + "loss": 0.9066, + "mean_token_accuracy": 0.7381842732429504, + "num_tokens": 369433815.0, + "step": 14288 + }, + { + "epoch": 1.5691851526466065, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7415952682495117, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7101037502288818, + "num_tokens": 369456333.0, + "step": 14289 + }, + { + "epoch": 1.5692949703492203, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.61958646774292, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7069011926651001, + "num_tokens": 369479927.0, + "step": 14290 + }, + { + "epoch": 1.569404788051834, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.423367977142334, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6979577541351318, + "num_tokens": 369509725.0, + "step": 14291 + }, + { + "epoch": 1.5695146057544476, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5384697914123535, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7114273905754089, + "num_tokens": 369533129.0, + "step": 14292 + }, + { + "epoch": 1.5696244234570613, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2563889026641846, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.6938058137893677, + "num_tokens": 369566005.0, + "step": 14293 + }, + { + "epoch": 1.5697342411596749, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8982701301574707, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6951099038124084, + "num_tokens": 369594485.0, + "step": 14294 + }, + { + "epoch": 1.5698440588622886, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6983397006988525, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7126460671424866, + "num_tokens": 369616813.0, + "step": 14295 + }, + { + "epoch": 1.5699538765649024, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.460576057434082, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6988319754600525, + "num_tokens": 369644035.0, + "step": 14296 + }, + { + "epoch": 1.570063694267516, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.588813304901123, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7009690403938293, + "num_tokens": 369671675.0, + "step": 14297 + }, + { + "epoch": 1.5701735119701294, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7116780281066895, + "learning_rate": 1e-06, + "loss": 0.9378, + "mean_token_accuracy": 0.7228271961212158, + "num_tokens": 369697141.0, + "step": 14298 + }, + { + "epoch": 1.5702833296727432, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.553603410720825, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7091537117958069, + "num_tokens": 369723377.0, + "step": 14299 + }, + { + "epoch": 1.570393147375357, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.687736988067627, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.7061278820037842, + "num_tokens": 369746629.0, + "step": 14300 + }, + { + "epoch": 1.5705029650779707, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.578129768371582, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7136127352714539, + "num_tokens": 369772301.0, + "step": 14301 + }, + { + "epoch": 1.5706127827805842, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7008540630340576, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7057262659072876, + "num_tokens": 369794383.0, + "step": 14302 + }, + { + "epoch": 1.5707226004831978, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.457056760787964, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7103064060211182, + "num_tokens": 369820624.0, + "step": 14303 + }, + { + "epoch": 1.5708324181858115, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.576547384262085, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6973952054977417, + "num_tokens": 369847660.0, + "step": 14304 + }, + { + "epoch": 1.5709422358884253, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.876376152038574, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7119214534759521, + "num_tokens": 369869638.0, + "step": 14305 + }, + { + "epoch": 1.5710520535910388, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.400137424468994, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7312030792236328, + "num_tokens": 369894309.0, + "step": 14306 + }, + { + "epoch": 1.5711618712936526, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.493345260620117, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7100279331207275, + "num_tokens": 369921955.0, + "step": 14307 + }, + { + "epoch": 1.5712716889962661, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.531963348388672, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.7040002346038818, + "num_tokens": 369947219.0, + "step": 14308 + }, + { + "epoch": 1.5713815066988799, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5767109394073486, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7323927879333496, + "num_tokens": 369972667.0, + "step": 14309 + }, + { + "epoch": 1.5714913244014936, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5861754417419434, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7146110534667969, + "num_tokens": 369998788.0, + "step": 14310 + }, + { + "epoch": 1.5716011421041072, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.343451738357544, + "learning_rate": 1e-06, + "loss": 0.9282, + "mean_token_accuracy": 0.7446743249893188, + "num_tokens": 370029169.0, + "step": 14311 + }, + { + "epoch": 1.5717109598067207, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5713462829589844, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.70367830991745, + "num_tokens": 370053808.0, + "step": 14312 + }, + { + "epoch": 1.5718207775093345, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.585833787918091, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7082374691963196, + "num_tokens": 370078890.0, + "step": 14313 + }, + { + "epoch": 1.5719305952119482, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7705283164978027, + "learning_rate": 1e-06, + "loss": 1.153, + "mean_token_accuracy": 0.6702152490615845, + "num_tokens": 370104070.0, + "step": 14314 + }, + { + "epoch": 1.572040412914562, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.625908136367798, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7170637845993042, + "num_tokens": 370128170.0, + "step": 14315 + }, + { + "epoch": 1.5721502306171755, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.2613422870635986, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.708050549030304, + "num_tokens": 370159853.0, + "step": 14316 + }, + { + "epoch": 1.572260048319789, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4811933040618896, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7124120593070984, + "num_tokens": 370188294.0, + "step": 14317 + }, + { + "epoch": 1.5723698660224028, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.61696720123291, + "learning_rate": 1e-06, + "loss": 1.1065, + "mean_token_accuracy": 0.6816832423210144, + "num_tokens": 370213186.0, + "step": 14318 + }, + { + "epoch": 1.5724796837250166, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 9.674795150756836, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7142035961151123, + "num_tokens": 370242167.0, + "step": 14319 + }, + { + "epoch": 1.57258950142763, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1569247245788574, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7086571455001831, + "num_tokens": 370275462.0, + "step": 14320 + }, + { + "epoch": 1.5726993191302436, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5150699615478516, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7083050608634949, + "num_tokens": 370301238.0, + "step": 14321 + }, + { + "epoch": 1.5728091368328574, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7787179946899414, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7136101722717285, + "num_tokens": 370322420.0, + "step": 14322 + }, + { + "epoch": 1.5729189545354711, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.484771490097046, + "learning_rate": 1e-06, + "loss": 1.1114, + "mean_token_accuracy": 0.6786829233169556, + "num_tokens": 370350201.0, + "step": 14323 + }, + { + "epoch": 1.573028772238085, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5913658142089844, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6854215860366821, + "num_tokens": 370379132.0, + "step": 14324 + }, + { + "epoch": 1.5731385899406984, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6264028549194336, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6871054172515869, + "num_tokens": 370403156.0, + "step": 14325 + }, + { + "epoch": 1.573248407643312, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8336617946624756, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7119407653808594, + "num_tokens": 370425779.0, + "step": 14326 + }, + { + "epoch": 1.5733582253459257, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.461179494857788, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.7092623710632324, + "num_tokens": 370456410.0, + "step": 14327 + }, + { + "epoch": 1.5734680430485395, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.512977361679077, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7250496745109558, + "num_tokens": 370482807.0, + "step": 14328 + }, + { + "epoch": 1.5735778607511532, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.733583927154541, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7126755118370056, + "num_tokens": 370504454.0, + "step": 14329 + }, + { + "epoch": 1.5736876784537668, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.546013355255127, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6934216022491455, + "num_tokens": 370531949.0, + "step": 14330 + }, + { + "epoch": 1.5737974961563803, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6007962226867676, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7179266214370728, + "num_tokens": 370557677.0, + "step": 14331 + }, + { + "epoch": 1.573907313858994, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.749528408050537, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6878060102462769, + "num_tokens": 370584060.0, + "step": 14332 + }, + { + "epoch": 1.5740171315616078, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.653193473815918, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7149347066879272, + "num_tokens": 370608927.0, + "step": 14333 + }, + { + "epoch": 1.5741269492642214, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7086455821990967, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7092294692993164, + "num_tokens": 370634430.0, + "step": 14334 + }, + { + "epoch": 1.574236766966835, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7246315479278564, + "learning_rate": 1e-06, + "loss": 0.9565, + "mean_token_accuracy": 0.715703010559082, + "num_tokens": 370655568.0, + "step": 14335 + }, + { + "epoch": 1.5743465846694487, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.1610875129699707, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7007138729095459, + "num_tokens": 370681602.0, + "step": 14336 + }, + { + "epoch": 1.5744564023720624, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.233875036239624, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7067931890487671, + "num_tokens": 370713455.0, + "step": 14337 + }, + { + "epoch": 1.5745662200746762, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5336802005767822, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.705926775932312, + "num_tokens": 370739938.0, + "step": 14338 + }, + { + "epoch": 1.5746760377772897, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6272919178009033, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.7021250128746033, + "num_tokens": 370763405.0, + "step": 14339 + }, + { + "epoch": 1.5747858554799032, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.523538827896118, + "learning_rate": 1e-06, + "loss": 1.1201, + "mean_token_accuracy": 0.6726226806640625, + "num_tokens": 370795013.0, + "step": 14340 + }, + { + "epoch": 1.574895673182517, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.486417531967163, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.703945517539978, + "num_tokens": 370821450.0, + "step": 14341 + }, + { + "epoch": 1.5750054908851308, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6251466274261475, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7067949175834656, + "num_tokens": 370846368.0, + "step": 14342 + }, + { + "epoch": 1.5751153085877443, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.516458034515381, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7155650854110718, + "num_tokens": 370872383.0, + "step": 14343 + }, + { + "epoch": 1.575225126290358, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5349133014678955, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6930709481239319, + "num_tokens": 370901133.0, + "step": 14344 + }, + { + "epoch": 1.5753349439929716, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.604480028152466, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6885610818862915, + "num_tokens": 370926893.0, + "step": 14345 + }, + { + "epoch": 1.5754447616955853, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4835805892944336, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7135354280471802, + "num_tokens": 370953501.0, + "step": 14346 + }, + { + "epoch": 1.575554579398199, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.248692274093628, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7176100015640259, + "num_tokens": 370983747.0, + "step": 14347 + }, + { + "epoch": 1.5756643971008126, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3821351528167725, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.7016146183013916, + "num_tokens": 371012900.0, + "step": 14348 + }, + { + "epoch": 1.5757742148034262, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6715898513793945, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7036943435668945, + "num_tokens": 371037220.0, + "step": 14349 + }, + { + "epoch": 1.57588403250604, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8969502449035645, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.6996821165084839, + "num_tokens": 371059267.0, + "step": 14350 + }, + { + "epoch": 1.5759938502086537, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.557811737060547, + "learning_rate": 1e-06, + "loss": 1.1566, + "mean_token_accuracy": 0.6726676821708679, + "num_tokens": 371090398.0, + "step": 14351 + }, + { + "epoch": 1.5761036679112674, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.516814708709717, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6888724565505981, + "num_tokens": 371119403.0, + "step": 14352 + }, + { + "epoch": 1.576213485613881, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.286017656326294, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6922374367713928, + "num_tokens": 371154075.0, + "step": 14353 + }, + { + "epoch": 1.5763233033164945, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8247337341308594, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.706855058670044, + "num_tokens": 371176912.0, + "step": 14354 + }, + { + "epoch": 1.5764331210191083, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.4708900451660156, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6971182227134705, + "num_tokens": 371202272.0, + "step": 14355 + }, + { + "epoch": 1.576542938721722, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6429827213287354, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.7101944088935852, + "num_tokens": 371227268.0, + "step": 14356 + }, + { + "epoch": 1.5766527564243356, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6727352142333984, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7194753885269165, + "num_tokens": 371252192.0, + "step": 14357 + }, + { + "epoch": 1.5767625741269493, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.3942205905914307, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7074406147003174, + "num_tokens": 371274291.0, + "step": 14358 + }, + { + "epoch": 1.5768723918295628, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3999733924865723, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6862196922302246, + "num_tokens": 371304575.0, + "step": 14359 + }, + { + "epoch": 1.5769822095321766, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4044883251190186, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7175170183181763, + "num_tokens": 371333681.0, + "step": 14360 + }, + { + "epoch": 1.5770920272347904, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3231360912323, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.7001730799674988, + "num_tokens": 371367450.0, + "step": 14361 + }, + { + "epoch": 1.577201844937404, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.445504665374756, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.6996828317642212, + "num_tokens": 371393930.0, + "step": 14362 + }, + { + "epoch": 1.5773116626400174, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6515250205993652, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7044729590415955, + "num_tokens": 371415346.0, + "step": 14363 + }, + { + "epoch": 1.5774214803426312, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6920325756073, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6939058899879456, + "num_tokens": 371436946.0, + "step": 14364 + }, + { + "epoch": 1.577531298045245, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.531712532043457, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6995713114738464, + "num_tokens": 371466547.0, + "step": 14365 + }, + { + "epoch": 1.5776411157478587, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5051443576812744, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7134521007537842, + "num_tokens": 371498664.0, + "step": 14366 + }, + { + "epoch": 1.5777509334504722, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.579235792160034, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7124263048171997, + "num_tokens": 371526832.0, + "step": 14367 + }, + { + "epoch": 1.5778607511530858, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3146071434020996, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7086522579193115, + "num_tokens": 371556697.0, + "step": 14368 + }, + { + "epoch": 1.5779705688556995, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3297157287597656, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6909658908843994, + "num_tokens": 371587747.0, + "step": 14369 + }, + { + "epoch": 1.5780803865583133, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.649928331375122, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7000477910041809, + "num_tokens": 371611011.0, + "step": 14370 + }, + { + "epoch": 1.5781902042609268, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4961695671081543, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.702878475189209, + "num_tokens": 371636310.0, + "step": 14371 + }, + { + "epoch": 1.5783000219635406, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8009679317474365, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7001640796661377, + "num_tokens": 371664235.0, + "step": 14372 + }, + { + "epoch": 1.5784098396661541, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5322530269622803, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7008358240127563, + "num_tokens": 371689834.0, + "step": 14373 + }, + { + "epoch": 1.5785196573687679, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.231260061264038, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6923079490661621, + "num_tokens": 371725226.0, + "step": 14374 + }, + { + "epoch": 1.5786294750713816, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6935956478118896, + "learning_rate": 1e-06, + "loss": 0.92, + "mean_token_accuracy": 0.7256121635437012, + "num_tokens": 371748342.0, + "step": 14375 + }, + { + "epoch": 1.5787392927739952, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5493626594543457, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7068246006965637, + "num_tokens": 371771183.0, + "step": 14376 + }, + { + "epoch": 1.5788491104766087, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5286717414855957, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7024778127670288, + "num_tokens": 371797106.0, + "step": 14377 + }, + { + "epoch": 1.5789589281792225, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5479209423065186, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6938716173171997, + "num_tokens": 371824608.0, + "step": 14378 + }, + { + "epoch": 1.5790687458818362, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.634467363357544, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.698776364326477, + "num_tokens": 371849493.0, + "step": 14379 + }, + { + "epoch": 1.57917856358445, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4412081241607666, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7120763063430786, + "num_tokens": 371877672.0, + "step": 14380 + }, + { + "epoch": 1.5792883812870635, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.364170789718628, + "learning_rate": 1e-06, + "loss": 1.1348, + "mean_token_accuracy": 0.6693316698074341, + "num_tokens": 371909281.0, + "step": 14381 + }, + { + "epoch": 1.579398198989677, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6476550102233887, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.6964763402938843, + "num_tokens": 371932126.0, + "step": 14382 + }, + { + "epoch": 1.5795080166922908, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7188198566436768, + "learning_rate": 1e-06, + "loss": 0.916, + "mean_token_accuracy": 0.7284818291664124, + "num_tokens": 371952465.0, + "step": 14383 + }, + { + "epoch": 1.5796178343949046, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 3.0280826091766357, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7255554795265198, + "num_tokens": 371971853.0, + "step": 14384 + }, + { + "epoch": 1.579727652097518, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8859071731567383, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6853611469268799, + "num_tokens": 371999889.0, + "step": 14385 + }, + { + "epoch": 1.5798374698001316, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.289242744445801, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6923117637634277, + "num_tokens": 372031413.0, + "step": 14386 + }, + { + "epoch": 1.5799472875027454, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5306038856506348, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7142422795295715, + "num_tokens": 372055163.0, + "step": 14387 + }, + { + "epoch": 1.5800571052053591, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6413724422454834, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6927415728569031, + "num_tokens": 372081104.0, + "step": 14388 + }, + { + "epoch": 1.580166922907973, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7708044052124023, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7149202823638916, + "num_tokens": 372105152.0, + "step": 14389 + }, + { + "epoch": 1.5802767406105864, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.414841651916504, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7023804783821106, + "num_tokens": 372134673.0, + "step": 14390 + }, + { + "epoch": 1.5803865583132, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.838545560836792, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7062820196151733, + "num_tokens": 372157655.0, + "step": 14391 + }, + { + "epoch": 1.5804963760158137, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.509819507598877, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7211195826530457, + "num_tokens": 372181746.0, + "step": 14392 + }, + { + "epoch": 1.5806061937184275, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6919498443603516, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6920285224914551, + "num_tokens": 372209356.0, + "step": 14393 + }, + { + "epoch": 1.5807160114210412, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.5744333267211914, + "learning_rate": 1e-06, + "loss": 1.1246, + "mean_token_accuracy": 0.6818268299102783, + "num_tokens": 372233574.0, + "step": 14394 + }, + { + "epoch": 1.5808258291236548, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6125621795654297, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6929675340652466, + "num_tokens": 372261343.0, + "step": 14395 + }, + { + "epoch": 1.5809356468262683, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7788281440734863, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7024500370025635, + "num_tokens": 372284609.0, + "step": 14396 + }, + { + "epoch": 1.581045464528882, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7151784896850586, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7107411623001099, + "num_tokens": 372309407.0, + "step": 14397 + }, + { + "epoch": 1.5811552822314958, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.871281147003174, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7179266810417175, + "num_tokens": 372330438.0, + "step": 14398 + }, + { + "epoch": 1.5812650999341094, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.608950138092041, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7095255851745605, + "num_tokens": 372358343.0, + "step": 14399 + }, + { + "epoch": 1.5813749176367229, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.910817861557007, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7302041053771973, + "num_tokens": 372377730.0, + "step": 14400 + }, + { + "epoch": 1.5814847353393366, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.539689779281616, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.6950796246528625, + "num_tokens": 372410573.0, + "step": 14401 + }, + { + "epoch": 1.5815945530419504, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4558639526367188, + "learning_rate": 1e-06, + "loss": 1.1057, + "mean_token_accuracy": 0.687411904335022, + "num_tokens": 372438940.0, + "step": 14402 + }, + { + "epoch": 1.5817043707445642, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7374110221862793, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6966062784194946, + "num_tokens": 372462985.0, + "step": 14403 + }, + { + "epoch": 1.5818141884471777, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.7434089183807373, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7010586261749268, + "num_tokens": 372488502.0, + "step": 14404 + }, + { + "epoch": 1.5819240061497912, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4643428325653076, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6994524598121643, + "num_tokens": 372514893.0, + "step": 14405 + }, + { + "epoch": 1.582033823852405, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.788637399673462, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7013373970985413, + "num_tokens": 372538957.0, + "step": 14406 + }, + { + "epoch": 1.5821436415550187, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4552838802337646, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6907503604888916, + "num_tokens": 372568819.0, + "step": 14407 + }, + { + "epoch": 1.5822534592576323, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4530763626098633, + "learning_rate": 1e-06, + "loss": 1.1043, + "mean_token_accuracy": 0.6806011199951172, + "num_tokens": 372599091.0, + "step": 14408 + }, + { + "epoch": 1.582363276960246, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.3582348823547363, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7258738279342651, + "num_tokens": 372627719.0, + "step": 14409 + }, + { + "epoch": 1.5824730946628596, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.772184371948242, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6915575861930847, + "num_tokens": 372650903.0, + "step": 14410 + }, + { + "epoch": 1.5825829123654733, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.581826686859131, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7022217512130737, + "num_tokens": 372678165.0, + "step": 14411 + }, + { + "epoch": 1.582692730068087, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.304288148880005, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.690009355545044, + "num_tokens": 372713150.0, + "step": 14412 + }, + { + "epoch": 1.5828025477707006, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4807348251342773, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6995022296905518, + "num_tokens": 372739846.0, + "step": 14413 + }, + { + "epoch": 1.5829123654733142, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6061184406280518, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6919771432876587, + "num_tokens": 372765815.0, + "step": 14414 + }, + { + "epoch": 1.583022183175928, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.291064977645874, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6890637874603271, + "num_tokens": 372797191.0, + "step": 14415 + }, + { + "epoch": 1.5831320008785417, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.4100801944732666, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7154072523117065, + "num_tokens": 372824195.0, + "step": 14416 + }, + { + "epoch": 1.5832418185811554, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.748600482940674, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.7003926634788513, + "num_tokens": 372847841.0, + "step": 14417 + }, + { + "epoch": 1.583351636283769, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8754336833953857, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.713764488697052, + "num_tokens": 372867002.0, + "step": 14418 + }, + { + "epoch": 1.5834614539863825, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.997356414794922, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.7012473344802856, + "num_tokens": 372892138.0, + "step": 14419 + }, + { + "epoch": 1.5835712716889963, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.8996939659118652, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.701915442943573, + "num_tokens": 372917333.0, + "step": 14420 + }, + { + "epoch": 1.58368108939161, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6548657417297363, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7131901979446411, + "num_tokens": 372941779.0, + "step": 14421 + }, + { + "epoch": 1.5837909070942235, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.731229066848755, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7051423788070679, + "num_tokens": 372962844.0, + "step": 14422 + }, + { + "epoch": 1.5839007247968373, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6067087650299072, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6927492022514343, + "num_tokens": 372989413.0, + "step": 14423 + }, + { + "epoch": 1.5840105424994508, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.668565273284912, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7102113366127014, + "num_tokens": 373014865.0, + "step": 14424 + }, + { + "epoch": 1.5841203602020646, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 6.579122066497803, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.6887809634208679, + "num_tokens": 373043360.0, + "step": 14425 + }, + { + "epoch": 1.5842301779046783, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4553375244140625, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6976314783096313, + "num_tokens": 373071402.0, + "step": 14426 + }, + { + "epoch": 1.5843399956072919, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.996825933456421, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7014920115470886, + "num_tokens": 373092339.0, + "step": 14427 + }, + { + "epoch": 1.5844498133099054, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.795236349105835, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6980456113815308, + "num_tokens": 373113931.0, + "step": 14428 + }, + { + "epoch": 1.5845596310125192, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.739527940750122, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7213080525398254, + "num_tokens": 373137454.0, + "step": 14429 + }, + { + "epoch": 1.584669448715133, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.465195894241333, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6879065036773682, + "num_tokens": 373163741.0, + "step": 14430 + }, + { + "epoch": 1.5847792664177467, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.522775650024414, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6996999382972717, + "num_tokens": 373191778.0, + "step": 14431 + }, + { + "epoch": 1.5848890841203602, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.642531156539917, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6969324350357056, + "num_tokens": 373215503.0, + "step": 14432 + }, + { + "epoch": 1.5849989018229738, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6390490531921387, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7046855688095093, + "num_tokens": 373239727.0, + "step": 14433 + }, + { + "epoch": 1.5851087195255875, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.649604320526123, + "learning_rate": 1e-06, + "loss": 1.1181, + "mean_token_accuracy": 0.6769442558288574, + "num_tokens": 373268032.0, + "step": 14434 + }, + { + "epoch": 1.5852185372282013, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5153558254241943, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7009599208831787, + "num_tokens": 373295587.0, + "step": 14435 + }, + { + "epoch": 1.5853283549308148, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.797668933868408, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.6991753578186035, + "num_tokens": 373316872.0, + "step": 14436 + }, + { + "epoch": 1.5854381726334283, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0316736698150635, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7256516814231873, + "num_tokens": 373337786.0, + "step": 14437 + }, + { + "epoch": 1.585547990336042, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7540640830993652, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.7061139345169067, + "num_tokens": 373359915.0, + "step": 14438 + }, + { + "epoch": 1.5856578080386559, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6882283687591553, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7084773778915405, + "num_tokens": 373384787.0, + "step": 14439 + }, + { + "epoch": 1.5857676257412696, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.755018711090088, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7101724147796631, + "num_tokens": 373414612.0, + "step": 14440 + }, + { + "epoch": 1.5858774434438832, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.581691026687622, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.7072797417640686, + "num_tokens": 373439049.0, + "step": 14441 + }, + { + "epoch": 1.5859872611464967, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5879483222961426, + "learning_rate": 1e-06, + "loss": 1.1213, + "mean_token_accuracy": 0.6781589984893799, + "num_tokens": 373474120.0, + "step": 14442 + }, + { + "epoch": 1.5860970788491104, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.597888708114624, + "learning_rate": 1e-06, + "loss": 0.9105, + "mean_token_accuracy": 0.7269309759140015, + "num_tokens": 373497659.0, + "step": 14443 + }, + { + "epoch": 1.5862068965517242, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5792667865753174, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7260957956314087, + "num_tokens": 373522404.0, + "step": 14444 + }, + { + "epoch": 1.586316714254338, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4109795093536377, + "learning_rate": 1e-06, + "loss": 0.9353, + "mean_token_accuracy": 0.724981427192688, + "num_tokens": 373549575.0, + "step": 14445 + }, + { + "epoch": 1.5864265319569515, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.67936372756958, + "learning_rate": 1e-06, + "loss": 0.9358, + "mean_token_accuracy": 0.7282849550247192, + "num_tokens": 373572238.0, + "step": 14446 + }, + { + "epoch": 1.586536349659565, + "ewc_loss": 9.47713851928711e-06, + "grad_norm": 2.6452057361602783, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7064988613128662, + "num_tokens": 373597048.0, + "step": 14447 + }, + { + "epoch": 1.5866461673621788, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6348540782928467, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7057756185531616, + "num_tokens": 373620945.0, + "step": 14448 + }, + { + "epoch": 1.5867559850647925, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7821621894836426, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7004114985466003, + "num_tokens": 373644170.0, + "step": 14449 + }, + { + "epoch": 1.586865802767406, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.82867431640625, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7007299661636353, + "num_tokens": 373665553.0, + "step": 14450 + }, + { + "epoch": 1.5869756204700196, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.702233076095581, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.6964156031608582, + "num_tokens": 373693133.0, + "step": 14451 + }, + { + "epoch": 1.5870854381726334, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.746692419052124, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.70670485496521, + "num_tokens": 373716781.0, + "step": 14452 + }, + { + "epoch": 1.5871952558752471, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.602442502975464, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7078883051872253, + "num_tokens": 373742232.0, + "step": 14453 + }, + { + "epoch": 1.5873050735778609, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5723960399627686, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.7035933136940002, + "num_tokens": 373768072.0, + "step": 14454 + }, + { + "epoch": 1.5874148912804744, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.725428581237793, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7111011147499084, + "num_tokens": 373794979.0, + "step": 14455 + }, + { + "epoch": 1.587524708983088, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8846776485443115, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7119982242584229, + "num_tokens": 373817546.0, + "step": 14456 + }, + { + "epoch": 1.5876345266857017, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.645676612854004, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7084414958953857, + "num_tokens": 373845958.0, + "step": 14457 + }, + { + "epoch": 1.5877443443883155, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.751610517501831, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7156326174736023, + "num_tokens": 373869453.0, + "step": 14458 + }, + { + "epoch": 1.5878541620909292, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6453471183776855, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7021214962005615, + "num_tokens": 373894136.0, + "step": 14459 + }, + { + "epoch": 1.5879639797935428, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9062812328338623, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6998481750488281, + "num_tokens": 373914939.0, + "step": 14460 + }, + { + "epoch": 1.5880737974961563, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.568577527999878, + "learning_rate": 1e-06, + "loss": 1.1002, + "mean_token_accuracy": 0.6799967885017395, + "num_tokens": 373942048.0, + "step": 14461 + }, + { + "epoch": 1.58818361519877, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.527176856994629, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7023217678070068, + "num_tokens": 373971014.0, + "step": 14462 + }, + { + "epoch": 1.5882934329013838, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5793392658233643, + "learning_rate": 1e-06, + "loss": 0.8661, + "mean_token_accuracy": 0.7465107440948486, + "num_tokens": 373994789.0, + "step": 14463 + }, + { + "epoch": 1.5884032506039973, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7380292415618896, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7161871194839478, + "num_tokens": 374016747.0, + "step": 14464 + }, + { + "epoch": 1.5885130683066109, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8154492378234863, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7199051380157471, + "num_tokens": 374039477.0, + "step": 14465 + }, + { + "epoch": 1.5886228860092246, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.615849733352661, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6957660913467407, + "num_tokens": 374068549.0, + "step": 14466 + }, + { + "epoch": 1.5887327037118384, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.74733567237854, + "learning_rate": 1e-06, + "loss": 0.939, + "mean_token_accuracy": 0.7198804020881653, + "num_tokens": 374090115.0, + "step": 14467 + }, + { + "epoch": 1.5888425214144521, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6206917762756348, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6953535676002502, + "num_tokens": 374114433.0, + "step": 14468 + }, + { + "epoch": 1.5889523391170657, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3921918869018555, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.6995342969894409, + "num_tokens": 374144386.0, + "step": 14469 + }, + { + "epoch": 1.5890621568196792, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.680729389190674, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7204673886299133, + "num_tokens": 374166726.0, + "step": 14470 + }, + { + "epoch": 1.589171974522293, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.393810749053955, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7005879878997803, + "num_tokens": 374197369.0, + "step": 14471 + }, + { + "epoch": 1.5892817922249067, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.720153331756592, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6989188194274902, + "num_tokens": 374220021.0, + "step": 14472 + }, + { + "epoch": 1.5893916099275203, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4323623180389404, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6910462379455566, + "num_tokens": 374252238.0, + "step": 14473 + }, + { + "epoch": 1.589501427630134, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.556840181350708, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7022286653518677, + "num_tokens": 374279155.0, + "step": 14474 + }, + { + "epoch": 1.5896112453327476, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.562697649002075, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7053719758987427, + "num_tokens": 374308634.0, + "step": 14475 + }, + { + "epoch": 1.5897210630353613, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.699979305267334, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7048400640487671, + "num_tokens": 374331359.0, + "step": 14476 + }, + { + "epoch": 1.589830880737975, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7319865226745605, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6822776198387146, + "num_tokens": 374355995.0, + "step": 14477 + }, + { + "epoch": 1.5899406984405886, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.907132148742676, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7087909579277039, + "num_tokens": 374377534.0, + "step": 14478 + }, + { + "epoch": 1.5900505161432021, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5271031856536865, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7108601331710815, + "num_tokens": 374403849.0, + "step": 14479 + }, + { + "epoch": 1.590160333845816, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5962045192718506, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.722687840461731, + "num_tokens": 374429684.0, + "step": 14480 + }, + { + "epoch": 1.5902701515484297, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.508836507797241, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7151398062705994, + "num_tokens": 374455969.0, + "step": 14481 + }, + { + "epoch": 1.5903799692510434, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5990912914276123, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7074326872825623, + "num_tokens": 374481316.0, + "step": 14482 + }, + { + "epoch": 1.590489786953657, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5031745433807373, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6925144195556641, + "num_tokens": 374505819.0, + "step": 14483 + }, + { + "epoch": 1.5905996046562705, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5268540382385254, + "learning_rate": 1e-06, + "loss": 0.88, + "mean_token_accuracy": 0.7394189238548279, + "num_tokens": 374529943.0, + "step": 14484 + }, + { + "epoch": 1.5907094223588842, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.74355411529541, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.711635947227478, + "num_tokens": 374551720.0, + "step": 14485 + }, + { + "epoch": 1.590819240061498, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4361071586608887, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6970791816711426, + "num_tokens": 374580820.0, + "step": 14486 + }, + { + "epoch": 1.5909290577641115, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6163742542266846, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.715698778629303, + "num_tokens": 374603548.0, + "step": 14487 + }, + { + "epoch": 1.5910388754667253, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.70106840133667, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.7005593180656433, + "num_tokens": 374629400.0, + "step": 14488 + }, + { + "epoch": 1.5911486931693388, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7078022956848145, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7045110464096069, + "num_tokens": 374651432.0, + "step": 14489 + }, + { + "epoch": 1.5912585108719526, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3584632873535156, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7010249495506287, + "num_tokens": 374684310.0, + "step": 14490 + }, + { + "epoch": 1.5913683285745663, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.483635187149048, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.696255624294281, + "num_tokens": 374709123.0, + "step": 14491 + }, + { + "epoch": 1.5914781462771799, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3516032695770264, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7041822075843811, + "num_tokens": 374737090.0, + "step": 14492 + }, + { + "epoch": 1.5915879639797934, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.529104232788086, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7051227688789368, + "num_tokens": 374762696.0, + "step": 14493 + }, + { + "epoch": 1.5916977816824072, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8258087635040283, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7136760950088501, + "num_tokens": 374787539.0, + "step": 14494 + }, + { + "epoch": 1.591807599385021, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4637370109558105, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6953349709510803, + "num_tokens": 374815548.0, + "step": 14495 + }, + { + "epoch": 1.5919174170876347, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4294745922088623, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6957919001579285, + "num_tokens": 374842855.0, + "step": 14496 + }, + { + "epoch": 1.5920272347902482, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2496514320373535, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6979349851608276, + "num_tokens": 374874927.0, + "step": 14497 + }, + { + "epoch": 1.5921370524928617, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.430650234222412, + "learning_rate": 1e-06, + "loss": 0.9126, + "mean_token_accuracy": 0.7284603118896484, + "num_tokens": 374895034.0, + "step": 14498 + }, + { + "epoch": 1.5922468701954755, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6993396282196045, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6960960626602173, + "num_tokens": 374919496.0, + "step": 14499 + }, + { + "epoch": 1.5923566878980893, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6452219486236572, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6979013681411743, + "num_tokens": 374943244.0, + "step": 14500 + }, + { + "epoch": 1.5924665056007028, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.444741725921631, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7002451419830322, + "num_tokens": 374971146.0, + "step": 14501 + }, + { + "epoch": 1.5925763233033163, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.414921998977661, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7120705842971802, + "num_tokens": 374997752.0, + "step": 14502 + }, + { + "epoch": 1.59268614100593, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4636316299438477, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6905295252799988, + "num_tokens": 375025536.0, + "step": 14503 + }, + { + "epoch": 1.5927959587085438, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.08697772026062, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7060203552246094, + "num_tokens": 375049295.0, + "step": 14504 + }, + { + "epoch": 1.5929057764111576, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5092155933380127, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7179670929908752, + "num_tokens": 375073356.0, + "step": 14505 + }, + { + "epoch": 1.5930155941137711, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.547218084335327, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7200644612312317, + "num_tokens": 375097520.0, + "step": 14506 + }, + { + "epoch": 1.5931254118163847, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5630688667297363, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7029761672019958, + "num_tokens": 375124237.0, + "step": 14507 + }, + { + "epoch": 1.5932352295189984, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.581634283065796, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7334643602371216, + "num_tokens": 375148049.0, + "step": 14508 + }, + { + "epoch": 1.5933450472216122, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.427241325378418, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7055619955062866, + "num_tokens": 375177807.0, + "step": 14509 + }, + { + "epoch": 1.593454864924226, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7230348587036133, + "learning_rate": 1e-06, + "loss": 0.9395, + "mean_token_accuracy": 0.724378764629364, + "num_tokens": 375201645.0, + "step": 14510 + }, + { + "epoch": 1.5935646826268395, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3920300006866455, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7001014351844788, + "num_tokens": 375229559.0, + "step": 14511 + }, + { + "epoch": 1.593674500329453, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7318813800811768, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.699612021446228, + "num_tokens": 375259333.0, + "step": 14512 + }, + { + "epoch": 1.5937843180320668, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.452418088912964, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.6829620003700256, + "num_tokens": 375288550.0, + "step": 14513 + }, + { + "epoch": 1.5938941357346805, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2788681983947754, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7223158478736877, + "num_tokens": 375320715.0, + "step": 14514 + }, + { + "epoch": 1.594003953437294, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4242401123046875, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6831098794937134, + "num_tokens": 375350974.0, + "step": 14515 + }, + { + "epoch": 1.5941137711399076, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8719136714935303, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6934768557548523, + "num_tokens": 375380124.0, + "step": 14516 + }, + { + "epoch": 1.5942235888425214, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6741135120391846, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7250651717185974, + "num_tokens": 375403688.0, + "step": 14517 + }, + { + "epoch": 1.5943334065451351, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8251307010650635, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7138819694519043, + "num_tokens": 375424204.0, + "step": 14518 + }, + { + "epoch": 1.5944432242477489, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3474884033203125, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6858857274055481, + "num_tokens": 375454822.0, + "step": 14519 + }, + { + "epoch": 1.5945530419503624, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.654207229614258, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6969172358512878, + "num_tokens": 375481295.0, + "step": 14520 + }, + { + "epoch": 1.594662859652976, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.452409267425537, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7147592902183533, + "num_tokens": 375505553.0, + "step": 14521 + }, + { + "epoch": 1.5947726773555897, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.602452039718628, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7083672285079956, + "num_tokens": 375529493.0, + "step": 14522 + }, + { + "epoch": 1.5948824950582035, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.343075752258301, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.7059313654899597, + "num_tokens": 375560012.0, + "step": 14523 + }, + { + "epoch": 1.5949923127608172, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6054694652557373, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7076525092124939, + "num_tokens": 375584353.0, + "step": 14524 + }, + { + "epoch": 1.5951021304634307, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4212987422943115, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7135431170463562, + "num_tokens": 375610895.0, + "step": 14525 + }, + { + "epoch": 1.5952119481660443, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 12.893235206604004, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7165285348892212, + "num_tokens": 375632870.0, + "step": 14526 + }, + { + "epoch": 1.595321765868658, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8878233432769775, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.6967651844024658, + "num_tokens": 375658316.0, + "step": 14527 + }, + { + "epoch": 1.5954315835712718, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.577298402786255, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6973065137863159, + "num_tokens": 375684916.0, + "step": 14528 + }, + { + "epoch": 1.5955414012738853, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.725994348526001, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.705306887626648, + "num_tokens": 375711341.0, + "step": 14529 + }, + { + "epoch": 1.5956512189764989, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9085423946380615, + "learning_rate": 1e-06, + "loss": 1.0854, + "mean_token_accuracy": 0.6863614320755005, + "num_tokens": 375738416.0, + "step": 14530 + }, + { + "epoch": 1.5957610366791126, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.688180685043335, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7282346487045288, + "num_tokens": 375760713.0, + "step": 14531 + }, + { + "epoch": 1.5958708543817264, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.561318874359131, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.728049635887146, + "num_tokens": 375786840.0, + "step": 14532 + }, + { + "epoch": 1.5959806720843401, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.626291275024414, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7053382992744446, + "num_tokens": 375811614.0, + "step": 14533 + }, + { + "epoch": 1.5960904897869537, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.536567211151123, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7052434682846069, + "num_tokens": 375838020.0, + "step": 14534 + }, + { + "epoch": 1.5962003074895672, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6800591945648193, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7064149379730225, + "num_tokens": 375861796.0, + "step": 14535 + }, + { + "epoch": 1.596310125192181, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5634071826934814, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7109729051589966, + "num_tokens": 375887156.0, + "step": 14536 + }, + { + "epoch": 1.5964199428947947, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5287437438964844, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7267357707023621, + "num_tokens": 375911343.0, + "step": 14537 + }, + { + "epoch": 1.5965297605974083, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.584852933883667, + "learning_rate": 1e-06, + "loss": 1.0847, + "mean_token_accuracy": 0.6857678294181824, + "num_tokens": 375936167.0, + "step": 14538 + }, + { + "epoch": 1.596639578300022, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 5.262316703796387, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6998456716537476, + "num_tokens": 375967462.0, + "step": 14539 + }, + { + "epoch": 1.5967493960026355, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6159844398498535, + "learning_rate": 1e-06, + "loss": 0.9431, + "mean_token_accuracy": 0.7196138501167297, + "num_tokens": 375990868.0, + "step": 14540 + }, + { + "epoch": 1.5968592137052493, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6579363346099854, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.6925605535507202, + "num_tokens": 376016493.0, + "step": 14541 + }, + { + "epoch": 1.596969031407863, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7990550994873047, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7081010341644287, + "num_tokens": 376039072.0, + "step": 14542 + }, + { + "epoch": 1.5970788491104766, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4717578887939453, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7103532552719116, + "num_tokens": 376067183.0, + "step": 14543 + }, + { + "epoch": 1.5971886668130901, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.770467519760132, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7154344916343689, + "num_tokens": 376091007.0, + "step": 14544 + }, + { + "epoch": 1.5972984845157039, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7987313270568848, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7265165448188782, + "num_tokens": 376110858.0, + "step": 14545 + }, + { + "epoch": 1.5974083022183176, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.753002882003784, + "learning_rate": 1e-06, + "loss": 0.9142, + "mean_token_accuracy": 0.7284177541732788, + "num_tokens": 376131065.0, + "step": 14546 + }, + { + "epoch": 1.5975181199209314, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 5.089259147644043, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7033768892288208, + "num_tokens": 376156691.0, + "step": 14547 + }, + { + "epoch": 1.597627937623545, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6211018562316895, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6951040029525757, + "num_tokens": 376180701.0, + "step": 14548 + }, + { + "epoch": 1.5977377553261585, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.466566324234009, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.6958022713661194, + "num_tokens": 376209886.0, + "step": 14549 + }, + { + "epoch": 1.5978475730287722, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6081807613372803, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7230810523033142, + "num_tokens": 376234719.0, + "step": 14550 + }, + { + "epoch": 1.597957390731386, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.821246862411499, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7172741293907166, + "num_tokens": 376255215.0, + "step": 14551 + }, + { + "epoch": 1.5980672084339995, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.511284828186035, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7066093683242798, + "num_tokens": 376280985.0, + "step": 14552 + }, + { + "epoch": 1.5981770261366133, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6369001865386963, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7073070406913757, + "num_tokens": 376305600.0, + "step": 14553 + }, + { + "epoch": 1.5982868438392268, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.756234884262085, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.7055654525756836, + "num_tokens": 376329147.0, + "step": 14554 + }, + { + "epoch": 1.5983966615418406, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6927683353424072, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7022745013237, + "num_tokens": 376353320.0, + "step": 14555 + }, + { + "epoch": 1.5985064792444543, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4895095825195312, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6923274993896484, + "num_tokens": 376380025.0, + "step": 14556 + }, + { + "epoch": 1.5986162969470679, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2554407119750977, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7002019286155701, + "num_tokens": 376411452.0, + "step": 14557 + }, + { + "epoch": 1.5987261146496814, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.24432635307312, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.693638801574707, + "num_tokens": 376444923.0, + "step": 14558 + }, + { + "epoch": 1.5988359323522952, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.777033567428589, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7014838457107544, + "num_tokens": 376467832.0, + "step": 14559 + }, + { + "epoch": 1.598945750054909, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5146067142486572, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7021224498748779, + "num_tokens": 376492742.0, + "step": 14560 + }, + { + "epoch": 1.5990555677575227, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.576288938522339, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7097896337509155, + "num_tokens": 376525201.0, + "step": 14561 + }, + { + "epoch": 1.5991653854601362, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5824568271636963, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7192179560661316, + "num_tokens": 376549642.0, + "step": 14562 + }, + { + "epoch": 1.5992752031627497, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.081563711166382, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.6976718902587891, + "num_tokens": 376568606.0, + "step": 14563 + }, + { + "epoch": 1.5993850208653635, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7164528369903564, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.6886842846870422, + "num_tokens": 376593160.0, + "step": 14564 + }, + { + "epoch": 1.5994948385679773, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6964149475097656, + "learning_rate": 1e-06, + "loss": 0.9298, + "mean_token_accuracy": 0.7272455096244812, + "num_tokens": 376615368.0, + "step": 14565 + }, + { + "epoch": 1.5996046562705908, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.743971347808838, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6821535229682922, + "num_tokens": 376639264.0, + "step": 14566 + }, + { + "epoch": 1.5997144739732043, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.4405019283294678, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6955485343933105, + "num_tokens": 376669428.0, + "step": 14567 + }, + { + "epoch": 1.599824291675818, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8104255199432373, + "learning_rate": 1e-06, + "loss": 0.9078, + "mean_token_accuracy": 0.7273105382919312, + "num_tokens": 376694719.0, + "step": 14568 + }, + { + "epoch": 1.5999341093784318, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5946295261383057, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7030407190322876, + "num_tokens": 376718948.0, + "step": 14569 + }, + { + "epoch": 1.6000439270810456, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.464365005493164, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7042210102081299, + "num_tokens": 376747031.0, + "step": 14570 + }, + { + "epoch": 1.6001537447836591, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.676996946334839, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7218195199966431, + "num_tokens": 376770942.0, + "step": 14571 + }, + { + "epoch": 1.6002635624862727, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4016225337982178, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7082883715629578, + "num_tokens": 376800528.0, + "step": 14572 + }, + { + "epoch": 1.6003733801888864, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2998619079589844, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7016962766647339, + "num_tokens": 376833529.0, + "step": 14573 + }, + { + "epoch": 1.6004831978915002, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.756121873855591, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7088637351989746, + "num_tokens": 376856645.0, + "step": 14574 + }, + { + "epoch": 1.600593015594114, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.359957695007324, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.7078696489334106, + "num_tokens": 376888611.0, + "step": 14575 + }, + { + "epoch": 1.6007028332967275, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5539891719818115, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7034369111061096, + "num_tokens": 376912936.0, + "step": 14576 + }, + { + "epoch": 1.600812650999341, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.675614595413208, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.6957905888557434, + "num_tokens": 376937269.0, + "step": 14577 + }, + { + "epoch": 1.6009224687019548, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9503731727600098, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7113929390907288, + "num_tokens": 376960702.0, + "step": 14578 + }, + { + "epoch": 1.6010322864045685, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7817177772521973, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7133864164352417, + "num_tokens": 376988619.0, + "step": 14579 + }, + { + "epoch": 1.601142104107182, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.344465732574463, + "learning_rate": 1e-06, + "loss": 1.1304, + "mean_token_accuracy": 0.6771966814994812, + "num_tokens": 377020953.0, + "step": 14580 + }, + { + "epoch": 1.6012519218097956, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6294846534729004, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7139616012573242, + "num_tokens": 377044028.0, + "step": 14581 + }, + { + "epoch": 1.6013617395124093, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.47833514213562, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.7010465860366821, + "num_tokens": 377076356.0, + "step": 14582 + }, + { + "epoch": 1.601471557215023, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.896688938140869, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7130255699157715, + "num_tokens": 377100286.0, + "step": 14583 + }, + { + "epoch": 1.6015813749176369, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3392035961151123, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7152578830718994, + "num_tokens": 377129006.0, + "step": 14584 + }, + { + "epoch": 1.6016911926202504, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0679917335510254, + "learning_rate": 1e-06, + "loss": 0.9324, + "mean_token_accuracy": 0.7250633239746094, + "num_tokens": 377147317.0, + "step": 14585 + }, + { + "epoch": 1.601801010322864, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.455986976623535, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6977555155754089, + "num_tokens": 377174880.0, + "step": 14586 + }, + { + "epoch": 1.6019108280254777, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6547415256500244, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7042778730392456, + "num_tokens": 377197324.0, + "step": 14587 + }, + { + "epoch": 1.6020206457280914, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8010804653167725, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6990769505500793, + "num_tokens": 377218568.0, + "step": 14588 + }, + { + "epoch": 1.602130463430705, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6557252407073975, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.708186149597168, + "num_tokens": 377242710.0, + "step": 14589 + }, + { + "epoch": 1.6022402811333187, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.489586114883423, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.7055138349533081, + "num_tokens": 377282627.0, + "step": 14590 + }, + { + "epoch": 1.6023500988359323, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.787212610244751, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.718353271484375, + "num_tokens": 377304607.0, + "step": 14591 + }, + { + "epoch": 1.602459916538546, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.827690839767456, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7246540784835815, + "num_tokens": 377324519.0, + "step": 14592 + }, + { + "epoch": 1.6025697342411598, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.4123497009277344, + "learning_rate": 1e-06, + "loss": 1.1096, + "mean_token_accuracy": 0.6763828992843628, + "num_tokens": 377360265.0, + "step": 14593 + }, + { + "epoch": 1.6026795519437733, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.232297897338867, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7054146528244019, + "num_tokens": 377377975.0, + "step": 14594 + }, + { + "epoch": 1.6027893696463869, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5620028972625732, + "learning_rate": 1e-06, + "loss": 1.0837, + "mean_token_accuracy": 0.7101945281028748, + "num_tokens": 377407738.0, + "step": 14595 + }, + { + "epoch": 1.6028991873490006, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6731231212615967, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7086038589477539, + "num_tokens": 377429190.0, + "step": 14596 + }, + { + "epoch": 1.6030090050516144, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.697728157043457, + "learning_rate": 1e-06, + "loss": 0.939, + "mean_token_accuracy": 0.7292057275772095, + "num_tokens": 377451639.0, + "step": 14597 + }, + { + "epoch": 1.6031188227542281, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.623344898223877, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.697227954864502, + "num_tokens": 377477193.0, + "step": 14598 + }, + { + "epoch": 1.6032286404568417, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6548399925231934, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7104184031486511, + "num_tokens": 377503375.0, + "step": 14599 + }, + { + "epoch": 1.6033384581594552, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8884832859039307, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7171435356140137, + "num_tokens": 377523637.0, + "step": 14600 + }, + { + "epoch": 1.603448275862069, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.1750519275665283, + "learning_rate": 1e-06, + "loss": 1.09, + "mean_token_accuracy": 0.6864790320396423, + "num_tokens": 377553443.0, + "step": 14601 + }, + { + "epoch": 1.6035580935646827, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4814226627349854, + "learning_rate": 1e-06, + "loss": 0.8942, + "mean_token_accuracy": 0.742617130279541, + "num_tokens": 377582274.0, + "step": 14602 + }, + { + "epoch": 1.6036679112672962, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9430203437805176, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.69715815782547, + "num_tokens": 377604082.0, + "step": 14603 + }, + { + "epoch": 1.60377772896991, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.538123607635498, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7266002893447876, + "num_tokens": 377629865.0, + "step": 14604 + }, + { + "epoch": 1.6038875466725235, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.756261110305786, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7254980802536011, + "num_tokens": 377651745.0, + "step": 14605 + }, + { + "epoch": 1.6039973643751373, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3281049728393555, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6905813813209534, + "num_tokens": 377683487.0, + "step": 14606 + }, + { + "epoch": 1.604107182077751, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6304919719696045, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6860216856002808, + "num_tokens": 377708915.0, + "step": 14607 + }, + { + "epoch": 1.6042169997803646, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.640502452850342, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7055301666259766, + "num_tokens": 377733857.0, + "step": 14608 + }, + { + "epoch": 1.6043268174829781, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.091963768005371, + "learning_rate": 1e-06, + "loss": 0.907, + "mean_token_accuracy": 0.7322115302085876, + "num_tokens": 377752263.0, + "step": 14609 + }, + { + "epoch": 1.6044366351855919, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6475250720977783, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7050734162330627, + "num_tokens": 377775949.0, + "step": 14610 + }, + { + "epoch": 1.6045464528882056, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8992557525634766, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7216074466705322, + "num_tokens": 377797937.0, + "step": 14611 + }, + { + "epoch": 1.6046562705908194, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.534855604171753, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7104636430740356, + "num_tokens": 377828982.0, + "step": 14612 + }, + { + "epoch": 1.604766088293433, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2787935733795166, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.6898378729820251, + "num_tokens": 377863360.0, + "step": 14613 + }, + { + "epoch": 1.6048759059960465, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.398186445236206, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6884545087814331, + "num_tokens": 377891909.0, + "step": 14614 + }, + { + "epoch": 1.6049857236986602, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5482177734375, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7076516151428223, + "num_tokens": 377916869.0, + "step": 14615 + }, + { + "epoch": 1.605095541401274, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.76944899559021, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6777287721633911, + "num_tokens": 377941425.0, + "step": 14616 + }, + { + "epoch": 1.6052053591038875, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6095967292785645, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7215840220451355, + "num_tokens": 377966567.0, + "step": 14617 + }, + { + "epoch": 1.605315176806501, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.496213436126709, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6825185418128967, + "num_tokens": 377993994.0, + "step": 14618 + }, + { + "epoch": 1.6054249945091148, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6430342197418213, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7088050842285156, + "num_tokens": 378018392.0, + "step": 14619 + }, + { + "epoch": 1.6055348122117286, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8551998138427734, + "learning_rate": 1e-06, + "loss": 0.8519, + "mean_token_accuracy": 0.7453114986419678, + "num_tokens": 378038143.0, + "step": 14620 + }, + { + "epoch": 1.6056446299143423, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.597071647644043, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6959052085876465, + "num_tokens": 378063431.0, + "step": 14621 + }, + { + "epoch": 1.6057544476169558, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4724349975585938, + "learning_rate": 1e-06, + "loss": 0.9044, + "mean_token_accuracy": 0.7332692742347717, + "num_tokens": 378086320.0, + "step": 14622 + }, + { + "epoch": 1.6058642653195694, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.897857666015625, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7051486968994141, + "num_tokens": 378107634.0, + "step": 14623 + }, + { + "epoch": 1.6059740830221831, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.612776041030884, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7104906439781189, + "num_tokens": 378134425.0, + "step": 14624 + }, + { + "epoch": 1.606083900724797, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8810973167419434, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6847718358039856, + "num_tokens": 378163654.0, + "step": 14625 + }, + { + "epoch": 1.6061937184274107, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5339550971984863, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7078676223754883, + "num_tokens": 378190317.0, + "step": 14626 + }, + { + "epoch": 1.6063035361300242, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3931074142456055, + "learning_rate": 1e-06, + "loss": 1.1005, + "mean_token_accuracy": 0.6774709224700928, + "num_tokens": 378219657.0, + "step": 14627 + }, + { + "epoch": 1.6064133538326377, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9408962726593018, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.712578535079956, + "num_tokens": 378239882.0, + "step": 14628 + }, + { + "epoch": 1.6065231715352515, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4587020874023438, + "learning_rate": 1e-06, + "loss": 1.1315, + "mean_token_accuracy": 0.6791560649871826, + "num_tokens": 378267995.0, + "step": 14629 + }, + { + "epoch": 1.6066329892378652, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6338164806365967, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6972005367279053, + "num_tokens": 378291931.0, + "step": 14630 + }, + { + "epoch": 1.6067428069404788, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.881150245666504, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6999366879463196, + "num_tokens": 378315418.0, + "step": 14631 + }, + { + "epoch": 1.6068526246430923, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.836970090866089, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7132583856582642, + "num_tokens": 378336854.0, + "step": 14632 + }, + { + "epoch": 1.606962442345706, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.403165340423584, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7031509876251221, + "num_tokens": 378365320.0, + "step": 14633 + }, + { + "epoch": 1.6070722600483198, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.003314971923828, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7143585085868835, + "num_tokens": 378387196.0, + "step": 14634 + }, + { + "epoch": 1.6071820777509336, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9401469230651855, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7229297161102295, + "num_tokens": 378406099.0, + "step": 14635 + }, + { + "epoch": 1.6072918954535471, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.447730779647827, + "learning_rate": 1e-06, + "loss": 1.1029, + "mean_token_accuracy": 0.6849550604820251, + "num_tokens": 378434516.0, + "step": 14636 + }, + { + "epoch": 1.6074017131561606, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4038476943969727, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7187612056732178, + "num_tokens": 378463211.0, + "step": 14637 + }, + { + "epoch": 1.6075115308587744, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6799724102020264, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.697035551071167, + "num_tokens": 378487206.0, + "step": 14638 + }, + { + "epoch": 1.6076213485613882, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5987017154693604, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.703967809677124, + "num_tokens": 378509810.0, + "step": 14639 + }, + { + "epoch": 1.607731166264002, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6642324924468994, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.697284460067749, + "num_tokens": 378535236.0, + "step": 14640 + }, + { + "epoch": 1.6078409839666155, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6234803199768066, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.703567385673523, + "num_tokens": 378564805.0, + "step": 14641 + }, + { + "epoch": 1.607950801669229, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6253092288970947, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6995989084243774, + "num_tokens": 378590294.0, + "step": 14642 + }, + { + "epoch": 1.6080606193718427, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5630970001220703, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6811408996582031, + "num_tokens": 378616209.0, + "step": 14643 + }, + { + "epoch": 1.6081704370744565, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7338156700134277, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.693687915802002, + "num_tokens": 378642574.0, + "step": 14644 + }, + { + "epoch": 1.60828025477707, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5668540000915527, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.702181339263916, + "num_tokens": 378667975.0, + "step": 14645 + }, + { + "epoch": 1.6083900724796836, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.415501832962036, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7181650400161743, + "num_tokens": 378697179.0, + "step": 14646 + }, + { + "epoch": 1.6084998901822973, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.847527027130127, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7125658392906189, + "num_tokens": 378718268.0, + "step": 14647 + }, + { + "epoch": 1.608609707884911, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5978684425354004, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.701661229133606, + "num_tokens": 378745182.0, + "step": 14648 + }, + { + "epoch": 1.6087195255875248, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6101796627044678, + "learning_rate": 1e-06, + "loss": 0.9419, + "mean_token_accuracy": 0.720733642578125, + "num_tokens": 378770063.0, + "step": 14649 + }, + { + "epoch": 1.6088293432901384, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5857608318328857, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7221958637237549, + "num_tokens": 378794104.0, + "step": 14650 + }, + { + "epoch": 1.608939160992752, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.423699378967285, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.703211784362793, + "num_tokens": 378820354.0, + "step": 14651 + }, + { + "epoch": 1.6090489786953657, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6211042404174805, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.702474057674408, + "num_tokens": 378847049.0, + "step": 14652 + }, + { + "epoch": 1.6091587963979794, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5537216663360596, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7259126901626587, + "num_tokens": 378871601.0, + "step": 14653 + }, + { + "epoch": 1.609268614100593, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.721834659576416, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7199040651321411, + "num_tokens": 378895536.0, + "step": 14654 + }, + { + "epoch": 1.6093784318032067, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4312822818756104, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.702262282371521, + "num_tokens": 378923499.0, + "step": 14655 + }, + { + "epoch": 1.6094882495058203, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.386868715286255, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6967002153396606, + "num_tokens": 378952595.0, + "step": 14656 + }, + { + "epoch": 1.609598067208434, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.446990728378296, + "learning_rate": 1e-06, + "loss": 1.0905, + "mean_token_accuracy": 0.6917377710342407, + "num_tokens": 378981185.0, + "step": 14657 + }, + { + "epoch": 1.6097078849110478, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5929155349731445, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6954765915870667, + "num_tokens": 379009117.0, + "step": 14658 + }, + { + "epoch": 1.6098177026136613, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8310928344726562, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7233920097351074, + "num_tokens": 379028374.0, + "step": 14659 + }, + { + "epoch": 1.6099275203162748, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.476442575454712, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6846708655357361, + "num_tokens": 379058265.0, + "step": 14660 + }, + { + "epoch": 1.6100373380188886, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.802891969680786, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7041647434234619, + "num_tokens": 379078961.0, + "step": 14661 + }, + { + "epoch": 1.6101471557215024, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.633612632751465, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7090060710906982, + "num_tokens": 379104371.0, + "step": 14662 + }, + { + "epoch": 1.6102569734241161, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.4226906299591064, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.716261625289917, + "num_tokens": 379125368.0, + "step": 14663 + }, + { + "epoch": 1.6103667911267296, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.595156669616699, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.702435314655304, + "num_tokens": 379148661.0, + "step": 14664 + }, + { + "epoch": 1.6104766088293432, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4199059009552, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6950275897979736, + "num_tokens": 379180087.0, + "step": 14665 + }, + { + "epoch": 1.610586426531957, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.626554489135742, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6824929714202881, + "num_tokens": 379208009.0, + "step": 14666 + }, + { + "epoch": 1.6106962442345707, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.741657257080078, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7038611769676208, + "num_tokens": 379229757.0, + "step": 14667 + }, + { + "epoch": 1.6108060619371842, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7511682510375977, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7219557762145996, + "num_tokens": 379252466.0, + "step": 14668 + }, + { + "epoch": 1.610915879639798, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.436889171600342, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7162562608718872, + "num_tokens": 379281949.0, + "step": 14669 + }, + { + "epoch": 1.6110256973424115, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4088363647460938, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.6831852197647095, + "num_tokens": 379314100.0, + "step": 14670 + }, + { + "epoch": 1.6111355150450253, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4684641361236572, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7066076993942261, + "num_tokens": 379341840.0, + "step": 14671 + }, + { + "epoch": 1.611245332747639, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.561798334121704, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7185800075531006, + "num_tokens": 379368847.0, + "step": 14672 + }, + { + "epoch": 1.6113551504502526, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9165263175964355, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7213560938835144, + "num_tokens": 379391704.0, + "step": 14673 + }, + { + "epoch": 1.611464968152866, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.700376272201538, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7159263491630554, + "num_tokens": 379415996.0, + "step": 14674 + }, + { + "epoch": 1.6115747858554799, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.49983549118042, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7150585055351257, + "num_tokens": 379445398.0, + "step": 14675 + }, + { + "epoch": 1.6116846035580936, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.464782476425171, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7091680765151978, + "num_tokens": 379473108.0, + "step": 14676 + }, + { + "epoch": 1.6117944212607074, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.192589282989502, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.713995099067688, + "num_tokens": 379497755.0, + "step": 14677 + }, + { + "epoch": 1.611904238963321, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8515896797180176, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.714015543460846, + "num_tokens": 379520189.0, + "step": 14678 + }, + { + "epoch": 1.6120140566659344, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7505414485931396, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.6992689371109009, + "num_tokens": 379543320.0, + "step": 14679 + }, + { + "epoch": 1.6121238743685482, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.638293981552124, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6947584748268127, + "num_tokens": 379566938.0, + "step": 14680 + }, + { + "epoch": 1.612233692071162, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5211737155914307, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.704519510269165, + "num_tokens": 379591567.0, + "step": 14681 + }, + { + "epoch": 1.6123435097737755, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5358405113220215, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7174211740493774, + "num_tokens": 379617964.0, + "step": 14682 + }, + { + "epoch": 1.612453327476389, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4416375160217285, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.704745888710022, + "num_tokens": 379645750.0, + "step": 14683 + }, + { + "epoch": 1.6125631451790028, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6421687602996826, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7259029746055603, + "num_tokens": 379669620.0, + "step": 14684 + }, + { + "epoch": 1.6126729628816165, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6200807094573975, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.7029033899307251, + "num_tokens": 379695559.0, + "step": 14685 + }, + { + "epoch": 1.6127827805842303, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.698552370071411, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7024132609367371, + "num_tokens": 379719060.0, + "step": 14686 + }, + { + "epoch": 1.6128925982868438, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4874415397644043, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6886125802993774, + "num_tokens": 379746301.0, + "step": 14687 + }, + { + "epoch": 1.6130024159894574, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3801400661468506, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6954942941665649, + "num_tokens": 379774412.0, + "step": 14688 + }, + { + "epoch": 1.6131122336920711, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.465071439743042, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.6852632164955139, + "num_tokens": 379802433.0, + "step": 14689 + }, + { + "epoch": 1.6132220513946849, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.247150421142578, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6990846395492554, + "num_tokens": 379834583.0, + "step": 14690 + }, + { + "epoch": 1.6133318690972986, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9911391735076904, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7079550623893738, + "num_tokens": 379854369.0, + "step": 14691 + }, + { + "epoch": 1.6134416867999122, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.588385820388794, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7055657505989075, + "num_tokens": 379880685.0, + "step": 14692 + }, + { + "epoch": 1.6135515045025257, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.512223482131958, + "learning_rate": 1e-06, + "loss": 1.1288, + "mean_token_accuracy": 0.6748424768447876, + "num_tokens": 379910731.0, + "step": 14693 + }, + { + "epoch": 1.6136613222051395, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.749812126159668, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7140528559684753, + "num_tokens": 379934116.0, + "step": 14694 + }, + { + "epoch": 1.6137711399077532, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6308133602142334, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7160524129867554, + "num_tokens": 379960684.0, + "step": 14695 + }, + { + "epoch": 1.6138809576103668, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.602728843688965, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6933552026748657, + "num_tokens": 379987816.0, + "step": 14696 + }, + { + "epoch": 1.6139907753129803, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.40873646736145, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.706325113773346, + "num_tokens": 380017144.0, + "step": 14697 + }, + { + "epoch": 1.614100593015594, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6660420894622803, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7131860256195068, + "num_tokens": 380039497.0, + "step": 14698 + }, + { + "epoch": 1.6142104107182078, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.447185754776001, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7049200534820557, + "num_tokens": 380068009.0, + "step": 14699 + }, + { + "epoch": 1.6143202284208216, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3196685314178467, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6954067945480347, + "num_tokens": 380101244.0, + "step": 14700 + }, + { + "epoch": 1.614430046123435, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4420082569122314, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7216331362724304, + "num_tokens": 380129242.0, + "step": 14701 + }, + { + "epoch": 1.6145398638260486, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7049496173858643, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6919587850570679, + "num_tokens": 380151774.0, + "step": 14702 + }, + { + "epoch": 1.6146496815286624, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4151201248168945, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7217625975608826, + "num_tokens": 380178855.0, + "step": 14703 + }, + { + "epoch": 1.6147594992312762, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7261204719543457, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7081878185272217, + "num_tokens": 380200391.0, + "step": 14704 + }, + { + "epoch": 1.61486931693389, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3849785327911377, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.6912847757339478, + "num_tokens": 380231561.0, + "step": 14705 + }, + { + "epoch": 1.6149791346365034, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.58842396736145, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7216877341270447, + "num_tokens": 380255188.0, + "step": 14706 + }, + { + "epoch": 1.615088952339117, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.728734016418457, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7095960974693298, + "num_tokens": 380278009.0, + "step": 14707 + }, + { + "epoch": 1.6151987700417307, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.296349287033081, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6830997467041016, + "num_tokens": 380311382.0, + "step": 14708 + }, + { + "epoch": 1.6153085877443445, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2242066860198975, + "learning_rate": 1e-06, + "loss": 1.1277, + "mean_token_accuracy": 0.6759673953056335, + "num_tokens": 380348900.0, + "step": 14709 + }, + { + "epoch": 1.615418405446958, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.619471788406372, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7246385812759399, + "num_tokens": 380373273.0, + "step": 14710 + }, + { + "epoch": 1.6155282231495716, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.25044322013855, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7138828039169312, + "num_tokens": 380404588.0, + "step": 14711 + }, + { + "epoch": 1.6156380408521853, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.338441848754883, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.689353883266449, + "num_tokens": 380435705.0, + "step": 14712 + }, + { + "epoch": 1.615747858554799, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.584441900253296, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6972000002861023, + "num_tokens": 380459583.0, + "step": 14713 + }, + { + "epoch": 1.6158576762574128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.573808431625366, + "learning_rate": 1e-06, + "loss": 0.8987, + "mean_token_accuracy": 0.73490971326828, + "num_tokens": 380483648.0, + "step": 14714 + }, + { + "epoch": 1.6159674939600264, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9028549194335938, + "learning_rate": 1e-06, + "loss": 1.0978, + "mean_token_accuracy": 0.6872988343238831, + "num_tokens": 380510835.0, + "step": 14715 + }, + { + "epoch": 1.61607731166264, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.822589159011841, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6958909034729004, + "num_tokens": 380534479.0, + "step": 14716 + }, + { + "epoch": 1.6161871293652537, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6345460414886475, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.70200514793396, + "num_tokens": 380558985.0, + "step": 14717 + }, + { + "epoch": 1.6162969470678674, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.694409132003784, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.6976301074028015, + "num_tokens": 380582954.0, + "step": 14718 + }, + { + "epoch": 1.616406764770481, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4196972846984863, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6951959133148193, + "num_tokens": 380609904.0, + "step": 14719 + }, + { + "epoch": 1.6165165824730947, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.765742301940918, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7018130421638489, + "num_tokens": 380633853.0, + "step": 14720 + }, + { + "epoch": 1.6166264001757082, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.63615083694458, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.7045803070068359, + "num_tokens": 380658581.0, + "step": 14721 + }, + { + "epoch": 1.616736217878322, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.521897792816162, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7200230956077576, + "num_tokens": 380684309.0, + "step": 14722 + }, + { + "epoch": 1.6168460355809358, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6330928802490234, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6968306303024292, + "num_tokens": 380714010.0, + "step": 14723 + }, + { + "epoch": 1.6169558532835493, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6407105922698975, + "learning_rate": 1e-06, + "loss": 0.8359, + "mean_token_accuracy": 0.7545092105865479, + "num_tokens": 380736775.0, + "step": 14724 + }, + { + "epoch": 1.6170656709861628, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.812549591064453, + "learning_rate": 1e-06, + "loss": 0.9112, + "mean_token_accuracy": 0.7303348779678345, + "num_tokens": 380758733.0, + "step": 14725 + }, + { + "epoch": 1.6171754886887766, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8531973361968994, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.7182618975639343, + "num_tokens": 380781061.0, + "step": 14726 + }, + { + "epoch": 1.6172853063913903, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8552298545837402, + "learning_rate": 1e-06, + "loss": 0.8835, + "mean_token_accuracy": 0.7360097765922546, + "num_tokens": 380800537.0, + "step": 14727 + }, + { + "epoch": 1.617395124094004, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.884169816970825, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6973969340324402, + "num_tokens": 380822551.0, + "step": 14728 + }, + { + "epoch": 1.6175049417966176, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5174877643585205, + "learning_rate": 1e-06, + "loss": 1.1104, + "mean_token_accuracy": 0.6799513101577759, + "num_tokens": 380850508.0, + "step": 14729 + }, + { + "epoch": 1.6176147594992312, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.634979248046875, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7027881145477295, + "num_tokens": 380873596.0, + "step": 14730 + }, + { + "epoch": 1.617724577201845, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7200980186462402, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7020269632339478, + "num_tokens": 380900552.0, + "step": 14731 + }, + { + "epoch": 1.6178343949044587, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.451211452484131, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7106350064277649, + "num_tokens": 380927255.0, + "step": 14732 + }, + { + "epoch": 1.6179442126070722, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6918270587921143, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6896625757217407, + "num_tokens": 380952323.0, + "step": 14733 + }, + { + "epoch": 1.618054030309686, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6612348556518555, + "learning_rate": 1e-06, + "loss": 1.1071, + "mean_token_accuracy": 0.6818453669548035, + "num_tokens": 380978041.0, + "step": 14734 + }, + { + "epoch": 1.6181638480122995, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7151598930358887, + "learning_rate": 1e-06, + "loss": 0.8986, + "mean_token_accuracy": 0.7296812534332275, + "num_tokens": 380998558.0, + "step": 14735 + }, + { + "epoch": 1.6182736657149133, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.305893898010254, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6914069652557373, + "num_tokens": 381029442.0, + "step": 14736 + }, + { + "epoch": 1.618383483417527, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6059136390686035, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7214841246604919, + "num_tokens": 381053022.0, + "step": 14737 + }, + { + "epoch": 1.6184933011201406, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.959663152694702, + "learning_rate": 1e-06, + "loss": 1.1172, + "mean_token_accuracy": 0.6907576322555542, + "num_tokens": 381078358.0, + "step": 14738 + }, + { + "epoch": 1.618603118822754, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.424116611480713, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.7070261836051941, + "num_tokens": 381105573.0, + "step": 14739 + }, + { + "epoch": 1.6187129365253679, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4341771602630615, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6998419165611267, + "num_tokens": 381136566.0, + "step": 14740 + }, + { + "epoch": 1.6188227542279816, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.637259006500244, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.700865626335144, + "num_tokens": 381159490.0, + "step": 14741 + }, + { + "epoch": 1.6189325719305954, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.479525089263916, + "learning_rate": 1e-06, + "loss": 0.9217, + "mean_token_accuracy": 0.7241981029510498, + "num_tokens": 381184986.0, + "step": 14742 + }, + { + "epoch": 1.619042389633209, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.525477409362793, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7071634531021118, + "num_tokens": 381210746.0, + "step": 14743 + }, + { + "epoch": 1.6191522073358224, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.273815870285034, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6943851113319397, + "num_tokens": 381241557.0, + "step": 14744 + }, + { + "epoch": 1.6192620250384362, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4860904216766357, + "learning_rate": 1e-06, + "loss": 0.8977, + "mean_token_accuracy": 0.7319161295890808, + "num_tokens": 381266486.0, + "step": 14745 + }, + { + "epoch": 1.61937184274105, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.501164436340332, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7204322218894958, + "num_tokens": 381292874.0, + "step": 14746 + }, + { + "epoch": 1.6194816604436635, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4666130542755127, + "learning_rate": 1e-06, + "loss": 1.1471, + "mean_token_accuracy": 0.6833165884017944, + "num_tokens": 381323286.0, + "step": 14747 + }, + { + "epoch": 1.619591478146277, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6985838413238525, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.716401219367981, + "num_tokens": 381344594.0, + "step": 14748 + }, + { + "epoch": 1.6197012958488908, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6114306449890137, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.707579493522644, + "num_tokens": 381371462.0, + "step": 14749 + }, + { + "epoch": 1.6198111135515045, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.650526523590088, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.689055323600769, + "num_tokens": 381397383.0, + "step": 14750 + }, + { + "epoch": 1.6199209312541183, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3567311763763428, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.708475649356842, + "num_tokens": 381426112.0, + "step": 14751 + }, + { + "epoch": 1.6200307489567318, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6389565467834473, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.693658173084259, + "num_tokens": 381452303.0, + "step": 14752 + }, + { + "epoch": 1.6201405666593454, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7855384349823, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6941390633583069, + "num_tokens": 381476583.0, + "step": 14753 + }, + { + "epoch": 1.6202503843619591, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3696811199188232, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6939612627029419, + "num_tokens": 381509030.0, + "step": 14754 + }, + { + "epoch": 1.6203602020645729, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.791821241378784, + "learning_rate": 1e-06, + "loss": 0.9006, + "mean_token_accuracy": 0.7363924980163574, + "num_tokens": 381529557.0, + "step": 14755 + }, + { + "epoch": 1.6204700197671866, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.533405065536499, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.6877350807189941, + "num_tokens": 381558586.0, + "step": 14756 + }, + { + "epoch": 1.6205798374698002, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5679972171783447, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.6814321279525757, + "num_tokens": 381587177.0, + "step": 14757 + }, + { + "epoch": 1.6206896551724137, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.396247148513794, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6950290203094482, + "num_tokens": 381617923.0, + "step": 14758 + }, + { + "epoch": 1.6207994728750275, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.421761989593506, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6943879127502441, + "num_tokens": 381647988.0, + "step": 14759 + }, + { + "epoch": 1.6209092905776412, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.554741382598877, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7005045413970947, + "num_tokens": 381671221.0, + "step": 14760 + }, + { + "epoch": 1.6210191082802548, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4883604049682617, + "learning_rate": 1e-06, + "loss": 1.1261, + "mean_token_accuracy": 0.6852580308914185, + "num_tokens": 381698408.0, + "step": 14761 + }, + { + "epoch": 1.6211289259828683, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.915311336517334, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7175780534744263, + "num_tokens": 381717742.0, + "step": 14762 + }, + { + "epoch": 1.621238743685482, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4433963298797607, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7103525400161743, + "num_tokens": 381748107.0, + "step": 14763 + }, + { + "epoch": 1.6213485613880958, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8023481369018555, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7170431613922119, + "num_tokens": 381770856.0, + "step": 14764 + }, + { + "epoch": 1.6214583790907096, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5094261169433594, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7010140419006348, + "num_tokens": 381799450.0, + "step": 14765 + }, + { + "epoch": 1.621568196793323, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.646933078765869, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7015885710716248, + "num_tokens": 381825078.0, + "step": 14766 + }, + { + "epoch": 1.6216780144959366, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.650228261947632, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6949014663696289, + "num_tokens": 381853370.0, + "step": 14767 + }, + { + "epoch": 1.6217878321985504, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.1817572116851807, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7098603248596191, + "num_tokens": 381880615.0, + "step": 14768 + }, + { + "epoch": 1.6218976499011641, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7287473678588867, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.7033946514129639, + "num_tokens": 381904268.0, + "step": 14769 + }, + { + "epoch": 1.6220074676037777, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.656665086746216, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.6991559267044067, + "num_tokens": 381928517.0, + "step": 14770 + }, + { + "epoch": 1.6221172853063914, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7837727069854736, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7059251070022583, + "num_tokens": 381951430.0, + "step": 14771 + }, + { + "epoch": 1.622227103009005, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.702235698699951, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7069665193557739, + "num_tokens": 381976894.0, + "step": 14772 + }, + { + "epoch": 1.6223369207116187, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.873408555984497, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7220365405082703, + "num_tokens": 382002349.0, + "step": 14773 + }, + { + "epoch": 1.6224467384142325, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.61440110206604, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.6999406814575195, + "num_tokens": 382025309.0, + "step": 14774 + }, + { + "epoch": 1.622556556116846, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6037423610687256, + "learning_rate": 1e-06, + "loss": 1.0475, + "mean_token_accuracy": 0.7001892924308777, + "num_tokens": 382047531.0, + "step": 14775 + }, + { + "epoch": 1.6226663738194596, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.666133165359497, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6834965944290161, + "num_tokens": 382070872.0, + "step": 14776 + }, + { + "epoch": 1.6227761915220733, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4397389888763428, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.7071220874786377, + "num_tokens": 382100230.0, + "step": 14777 + }, + { + "epoch": 1.622886009224687, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3411271572113037, + "learning_rate": 1e-06, + "loss": 0.912, + "mean_token_accuracy": 0.7348689436912537, + "num_tokens": 382129822.0, + "step": 14778 + }, + { + "epoch": 1.6229958269273008, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5410001277923584, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7235108613967896, + "num_tokens": 382155243.0, + "step": 14779 + }, + { + "epoch": 1.6231056446299144, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.622018814086914, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7241307497024536, + "num_tokens": 382179428.0, + "step": 14780 + }, + { + "epoch": 1.623215462332528, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.560413360595703, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.7001706957817078, + "num_tokens": 382205016.0, + "step": 14781 + }, + { + "epoch": 1.6233252800351416, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4522786140441895, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.6878815293312073, + "num_tokens": 382236986.0, + "step": 14782 + }, + { + "epoch": 1.6234350977377554, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.470127820968628, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6895125508308411, + "num_tokens": 382268483.0, + "step": 14783 + }, + { + "epoch": 1.623544915440369, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.664546489715576, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7002294063568115, + "num_tokens": 382294735.0, + "step": 14784 + }, + { + "epoch": 1.6236547331429827, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4887852668762207, + "learning_rate": 1e-06, + "loss": 0.8914, + "mean_token_accuracy": 0.734857976436615, + "num_tokens": 382319364.0, + "step": 14785 + }, + { + "epoch": 1.6237645508455962, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.505560874938965, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6970908641815186, + "num_tokens": 382343621.0, + "step": 14786 + }, + { + "epoch": 1.62387436854821, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.633295774459839, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6936559677124023, + "num_tokens": 382371121.0, + "step": 14787 + }, + { + "epoch": 1.6239841862508237, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5614521503448486, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.698463499546051, + "num_tokens": 382397503.0, + "step": 14788 + }, + { + "epoch": 1.6240940039534373, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.484745740890503, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.6861603260040283, + "num_tokens": 382424574.0, + "step": 14789 + }, + { + "epoch": 1.6242038216560508, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.471143960952759, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6981300115585327, + "num_tokens": 382450711.0, + "step": 14790 + }, + { + "epoch": 1.6243136393586646, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6014437675476074, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.7079100608825684, + "num_tokens": 382476821.0, + "step": 14791 + }, + { + "epoch": 1.6244234570612783, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.446531057357788, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7026642560958862, + "num_tokens": 382503988.0, + "step": 14792 + }, + { + "epoch": 1.624533274763892, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.355933904647827, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6964384913444519, + "num_tokens": 382533558.0, + "step": 14793 + }, + { + "epoch": 1.6246430924665056, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7194933891296387, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.69917893409729, + "num_tokens": 382557276.0, + "step": 14794 + }, + { + "epoch": 1.6247529101691192, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.681971788406372, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6903414726257324, + "num_tokens": 382582157.0, + "step": 14795 + }, + { + "epoch": 1.624862727871733, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4598987102508545, + "learning_rate": 1e-06, + "loss": 1.1519, + "mean_token_accuracy": 0.6859495639801025, + "num_tokens": 382612789.0, + "step": 14796 + }, + { + "epoch": 1.6249725455743467, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3298065662384033, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.690030574798584, + "num_tokens": 382646381.0, + "step": 14797 + }, + { + "epoch": 1.6250823632769602, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.651193380355835, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.70526123046875, + "num_tokens": 382670294.0, + "step": 14798 + }, + { + "epoch": 1.6251921809795737, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.680666923522949, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.716979444026947, + "num_tokens": 382694925.0, + "step": 14799 + }, + { + "epoch": 1.6253019986821875, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5207691192626953, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6899527907371521, + "num_tokens": 382724352.0, + "step": 14800 + }, + { + "epoch": 1.6254118163848013, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.924893379211426, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7102816700935364, + "num_tokens": 382751825.0, + "step": 14801 + }, + { + "epoch": 1.625521634087415, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5193262100219727, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7084475755691528, + "num_tokens": 382777700.0, + "step": 14802 + }, + { + "epoch": 1.6256314517900285, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 16.726852416992188, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7013176083564758, + "num_tokens": 382801147.0, + "step": 14803 + }, + { + "epoch": 1.625741269492642, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.872443199157715, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.6960169076919556, + "num_tokens": 382823777.0, + "step": 14804 + }, + { + "epoch": 1.6258510871952558, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5473990440368652, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.702532172203064, + "num_tokens": 382851424.0, + "step": 14805 + }, + { + "epoch": 1.6259609048978696, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5130906105041504, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7203141450881958, + "num_tokens": 382879282.0, + "step": 14806 + }, + { + "epoch": 1.6260707226004834, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.679392099380493, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.6955792903900146, + "num_tokens": 382905436.0, + "step": 14807 + }, + { + "epoch": 1.6261805403030969, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4864442348480225, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7003108859062195, + "num_tokens": 382933757.0, + "step": 14808 + }, + { + "epoch": 1.6262903580057104, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.642458915710449, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6979566812515259, + "num_tokens": 382956239.0, + "step": 14809 + }, + { + "epoch": 1.6264001757083242, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3662302494049072, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7079035043716431, + "num_tokens": 382988165.0, + "step": 14810 + }, + { + "epoch": 1.626509993410938, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.654895782470703, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.700023889541626, + "num_tokens": 383014012.0, + "step": 14811 + }, + { + "epoch": 1.6266198111135515, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.724020481109619, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.6948584318161011, + "num_tokens": 383039558.0, + "step": 14812 + }, + { + "epoch": 1.626729628816165, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3997604846954346, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6938703060150146, + "num_tokens": 383068124.0, + "step": 14813 + }, + { + "epoch": 1.6268394465187788, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.587672472000122, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7157496213912964, + "num_tokens": 383091127.0, + "step": 14814 + }, + { + "epoch": 1.6269492642213925, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.479341983795166, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7068496942520142, + "num_tokens": 383117348.0, + "step": 14815 + }, + { + "epoch": 1.6270590819240063, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4793314933776855, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7154667377471924, + "num_tokens": 383144594.0, + "step": 14816 + }, + { + "epoch": 1.6271688996266198, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.565505027770996, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6973987221717834, + "num_tokens": 383173211.0, + "step": 14817 + }, + { + "epoch": 1.6272787173292333, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6218652725219727, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6978778839111328, + "num_tokens": 383196574.0, + "step": 14818 + }, + { + "epoch": 1.627388535031847, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6563167572021484, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6982247233390808, + "num_tokens": 383225233.0, + "step": 14819 + }, + { + "epoch": 1.6274983527344609, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7917373180389404, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7158046960830688, + "num_tokens": 383246301.0, + "step": 14820 + }, + { + "epoch": 1.6276081704370746, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4265878200531006, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6980462074279785, + "num_tokens": 383276684.0, + "step": 14821 + }, + { + "epoch": 1.6277179881396882, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7973291873931885, + "learning_rate": 1e-06, + "loss": 0.9382, + "mean_token_accuracy": 0.7213755249977112, + "num_tokens": 383298194.0, + "step": 14822 + }, + { + "epoch": 1.6278278058423017, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7348403930664062, + "learning_rate": 1e-06, + "loss": 0.8864, + "mean_token_accuracy": 0.7379169464111328, + "num_tokens": 383319361.0, + "step": 14823 + }, + { + "epoch": 1.6279376235449154, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3144121170043945, + "learning_rate": 1e-06, + "loss": 0.9775, + "mean_token_accuracy": 0.7125393152236938, + "num_tokens": 383349452.0, + "step": 14824 + }, + { + "epoch": 1.6280474412475292, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4502317905426025, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6944534778594971, + "num_tokens": 383380037.0, + "step": 14825 + }, + { + "epoch": 1.6281572589501427, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4145803451538086, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7270511984825134, + "num_tokens": 383405249.0, + "step": 14826 + }, + { + "epoch": 1.6282670766527563, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.828921318054199, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6902651786804199, + "num_tokens": 383430728.0, + "step": 14827 + }, + { + "epoch": 1.62837689435537, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.421694040298462, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.7039952874183655, + "num_tokens": 383460877.0, + "step": 14828 + }, + { + "epoch": 1.6284867120579838, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6931300163269043, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7104263305664062, + "num_tokens": 383484181.0, + "step": 14829 + }, + { + "epoch": 1.6285965297605975, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4350028038024902, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7025351524353027, + "num_tokens": 383513323.0, + "step": 14830 + }, + { + "epoch": 1.628706347463211, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.455915689468384, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7308602333068848, + "num_tokens": 383538472.0, + "step": 14831 + }, + { + "epoch": 1.6288161651658246, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.536590814590454, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7178665995597839, + "num_tokens": 383563054.0, + "step": 14832 + }, + { + "epoch": 1.6289259828684384, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 16.714345932006836, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7020947933197021, + "num_tokens": 383586167.0, + "step": 14833 + }, + { + "epoch": 1.6290358005710521, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.949533462524414, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7247907519340515, + "num_tokens": 383609904.0, + "step": 14834 + }, + { + "epoch": 1.6291456182736657, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.774425745010376, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7112082242965698, + "num_tokens": 383631601.0, + "step": 14835 + }, + { + "epoch": 1.6292554359762794, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.672710418701172, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7214958071708679, + "num_tokens": 383653670.0, + "step": 14836 + }, + { + "epoch": 1.629365253678893, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0130207538604736, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6909758448600769, + "num_tokens": 383683590.0, + "step": 14837 + }, + { + "epoch": 1.6294750713815067, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.371415138244629, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7202566862106323, + "num_tokens": 383716512.0, + "step": 14838 + }, + { + "epoch": 1.6295848890841205, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.47463059425354, + "learning_rate": 1e-06, + "loss": 1.0799, + "mean_token_accuracy": 0.6852664947509766, + "num_tokens": 383748673.0, + "step": 14839 + }, + { + "epoch": 1.629694706786734, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.526087760925293, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7209120392799377, + "num_tokens": 383775985.0, + "step": 14840 + }, + { + "epoch": 1.6298045244893475, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4005191326141357, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6945987939834595, + "num_tokens": 383803403.0, + "step": 14841 + }, + { + "epoch": 1.6299143421919613, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4737226963043213, + "learning_rate": 1e-06, + "loss": 0.9267, + "mean_token_accuracy": 0.7248412370681763, + "num_tokens": 383829954.0, + "step": 14842 + }, + { + "epoch": 1.630024159894575, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.628171682357788, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7029159665107727, + "num_tokens": 383852254.0, + "step": 14843 + }, + { + "epoch": 1.6301339775971888, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9812192916870117, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7001757621765137, + "num_tokens": 383873300.0, + "step": 14844 + }, + { + "epoch": 1.6302437952998023, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6442394256591797, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7155095338821411, + "num_tokens": 383898671.0, + "step": 14845 + }, + { + "epoch": 1.6303536130024159, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1114768981933594, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.706528902053833, + "num_tokens": 383924445.0, + "step": 14846 + }, + { + "epoch": 1.6304634307050296, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.412482976913452, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7086827754974365, + "num_tokens": 383952449.0, + "step": 14847 + }, + { + "epoch": 1.6305732484076434, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.650698184967041, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.711530327796936, + "num_tokens": 383976174.0, + "step": 14848 + }, + { + "epoch": 1.630683066110257, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.862079381942749, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7081066966056824, + "num_tokens": 384003191.0, + "step": 14849 + }, + { + "epoch": 1.6307928838128707, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.560680389404297, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7039023637771606, + "num_tokens": 384030783.0, + "step": 14850 + }, + { + "epoch": 1.6309027015154842, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4691312313079834, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6804162263870239, + "num_tokens": 384061920.0, + "step": 14851 + }, + { + "epoch": 1.631012519218098, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.666947841644287, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7206471562385559, + "num_tokens": 384086460.0, + "step": 14852 + }, + { + "epoch": 1.6311223369207117, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5889604091644287, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7178868055343628, + "num_tokens": 384110573.0, + "step": 14853 + }, + { + "epoch": 1.6312321546233253, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6682205200195312, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7120174169540405, + "num_tokens": 384135820.0, + "step": 14854 + }, + { + "epoch": 1.6313419723259388, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.592332363128662, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7040959596633911, + "num_tokens": 384161870.0, + "step": 14855 + }, + { + "epoch": 1.6314517900285526, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.87634015083313, + "learning_rate": 1e-06, + "loss": 1.0679, + "mean_token_accuracy": 0.7010431289672852, + "num_tokens": 384185167.0, + "step": 14856 + }, + { + "epoch": 1.6315616077311663, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5855154991149902, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7239609360694885, + "num_tokens": 384209478.0, + "step": 14857 + }, + { + "epoch": 1.63167142543378, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3730075359344482, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7018551826477051, + "num_tokens": 384238523.0, + "step": 14858 + }, + { + "epoch": 1.6317812431363936, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.362448215484619, + "learning_rate": 1e-06, + "loss": 0.9418, + "mean_token_accuracy": 0.7230192422866821, + "num_tokens": 384264608.0, + "step": 14859 + }, + { + "epoch": 1.6318910608390071, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.661430597305298, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7043473720550537, + "num_tokens": 384292304.0, + "step": 14860 + }, + { + "epoch": 1.632000878541621, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.592435598373413, + "learning_rate": 1e-06, + "loss": 1.1245, + "mean_token_accuracy": 0.6733742356300354, + "num_tokens": 384322313.0, + "step": 14861 + }, + { + "epoch": 1.6321106962442347, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5012264251708984, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6951463222503662, + "num_tokens": 384351098.0, + "step": 14862 + }, + { + "epoch": 1.6322205139468482, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6088058948516846, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7051133513450623, + "num_tokens": 384376359.0, + "step": 14863 + }, + { + "epoch": 1.6323303316494617, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4507274627685547, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7053133249282837, + "num_tokens": 384404628.0, + "step": 14864 + }, + { + "epoch": 1.6324401493520755, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5294551849365234, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6965280771255493, + "num_tokens": 384431541.0, + "step": 14865 + }, + { + "epoch": 1.6325499670546892, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4673218727111816, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6878519654273987, + "num_tokens": 384460588.0, + "step": 14866 + }, + { + "epoch": 1.632659784757303, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0195767879486084, + "learning_rate": 1e-06, + "loss": 0.9465, + "mean_token_accuracy": 0.7204755544662476, + "num_tokens": 384481212.0, + "step": 14867 + }, + { + "epoch": 1.6327696024599165, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.443920373916626, + "learning_rate": 1e-06, + "loss": 0.884, + "mean_token_accuracy": 0.7330566644668579, + "num_tokens": 384507741.0, + "step": 14868 + }, + { + "epoch": 1.63287942016253, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.471731424331665, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7099488973617554, + "num_tokens": 384532661.0, + "step": 14869 + }, + { + "epoch": 1.6329892378651438, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.61419677734375, + "learning_rate": 1e-06, + "loss": 1.1076, + "mean_token_accuracy": 0.677241325378418, + "num_tokens": 384560894.0, + "step": 14870 + }, + { + "epoch": 1.6330990555677576, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.618475914001465, + "learning_rate": 1e-06, + "loss": 1.0962, + "mean_token_accuracy": 0.6912299990653992, + "num_tokens": 384588914.0, + "step": 14871 + }, + { + "epoch": 1.6332088732703713, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6461915969848633, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6936481595039368, + "num_tokens": 384616629.0, + "step": 14872 + }, + { + "epoch": 1.6333186909729849, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4578771591186523, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6969091892242432, + "num_tokens": 384647781.0, + "step": 14873 + }, + { + "epoch": 1.6334285086755984, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8434760570526123, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6942697167396545, + "num_tokens": 384670700.0, + "step": 14874 + }, + { + "epoch": 1.6335383263782122, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.102591037750244, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7173134684562683, + "num_tokens": 384689528.0, + "step": 14875 + }, + { + "epoch": 1.633648144080826, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.52644944190979, + "learning_rate": 1e-06, + "loss": 1.083, + "mean_token_accuracy": 0.6883130073547363, + "num_tokens": 384723911.0, + "step": 14876 + }, + { + "epoch": 1.6337579617834395, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6626594066619873, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.7049705386161804, + "num_tokens": 384750167.0, + "step": 14877 + }, + { + "epoch": 1.633867779486053, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5580039024353027, + "learning_rate": 1e-06, + "loss": 0.8454, + "mean_token_accuracy": 0.7486953735351562, + "num_tokens": 384775381.0, + "step": 14878 + }, + { + "epoch": 1.6339775971886668, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7047107219696045, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7122164964675903, + "num_tokens": 384795819.0, + "step": 14879 + }, + { + "epoch": 1.6340874148912805, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 80.547607421875, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7028756737709045, + "num_tokens": 384818800.0, + "step": 14880 + }, + { + "epoch": 1.6341972325938943, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.132642984390259, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6958713531494141, + "num_tokens": 384844704.0, + "step": 14881 + }, + { + "epoch": 1.6343070502965078, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6122570037841797, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7069767713546753, + "num_tokens": 384871538.0, + "step": 14882 + }, + { + "epoch": 1.6344168679991213, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.76802134513855, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6909218430519104, + "num_tokens": 384899619.0, + "step": 14883 + }, + { + "epoch": 1.634526685701735, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6532788276672363, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7048541307449341, + "num_tokens": 384927967.0, + "step": 14884 + }, + { + "epoch": 1.6346365034043489, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7931227684020996, + "learning_rate": 1e-06, + "loss": 0.9329, + "mean_token_accuracy": 0.7281511425971985, + "num_tokens": 384950049.0, + "step": 14885 + }, + { + "epoch": 1.6347463211069626, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6549782752990723, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7087749242782593, + "num_tokens": 384974721.0, + "step": 14886 + }, + { + "epoch": 1.6348561388095761, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 16.99990463256836, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7130097150802612, + "num_tokens": 384997463.0, + "step": 14887 + }, + { + "epoch": 1.6349659565121897, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.367875337600708, + "learning_rate": 1e-06, + "loss": 0.9565, + "mean_token_accuracy": 0.7173753976821899, + "num_tokens": 385015660.0, + "step": 14888 + }, + { + "epoch": 1.6350757742148034, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0167927742004395, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7180927991867065, + "num_tokens": 385040549.0, + "step": 14889 + }, + { + "epoch": 1.6351855919174172, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9008779525756836, + "learning_rate": 1e-06, + "loss": 0.9308, + "mean_token_accuracy": 0.7406305074691772, + "num_tokens": 385061746.0, + "step": 14890 + }, + { + "epoch": 1.6352954096200307, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.002617597579956, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7009133100509644, + "num_tokens": 385082821.0, + "step": 14891 + }, + { + "epoch": 1.6354052273226443, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4443862438201904, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6899047493934631, + "num_tokens": 385115248.0, + "step": 14892 + }, + { + "epoch": 1.635515045025258, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.768247127532959, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7093101739883423, + "num_tokens": 385139636.0, + "step": 14893 + }, + { + "epoch": 1.6356248627278718, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.763139486312866, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.702616274356842, + "num_tokens": 385164473.0, + "step": 14894 + }, + { + "epoch": 1.6357346804304855, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5050368309020996, + "learning_rate": 1e-06, + "loss": 1.1036, + "mean_token_accuracy": 0.6829817295074463, + "num_tokens": 385192452.0, + "step": 14895 + }, + { + "epoch": 1.635844498133099, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6621313095092773, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7119522094726562, + "num_tokens": 385215938.0, + "step": 14896 + }, + { + "epoch": 1.6359543158357126, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 17.085437774658203, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7144507169723511, + "num_tokens": 385239699.0, + "step": 14897 + }, + { + "epoch": 1.6360641335383264, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5260767936706543, + "learning_rate": 1e-06, + "loss": 1.0969, + "mean_token_accuracy": 0.6850656270980835, + "num_tokens": 385271485.0, + "step": 14898 + }, + { + "epoch": 1.6361739512409401, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0391581058502197, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6966906189918518, + "num_tokens": 385292962.0, + "step": 14899 + }, + { + "epoch": 1.6362837689435537, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6531951427459717, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7092038989067078, + "num_tokens": 385318331.0, + "step": 14900 + }, + { + "epoch": 1.6363935866461674, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.660288095474243, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7033993601799011, + "num_tokens": 385342660.0, + "step": 14901 + }, + { + "epoch": 1.636503404348781, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.60545015335083, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7079334855079651, + "num_tokens": 385371735.0, + "step": 14902 + }, + { + "epoch": 1.6366132220513947, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.72299861907959, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7174603939056396, + "num_tokens": 385396048.0, + "step": 14903 + }, + { + "epoch": 1.6367230397540085, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6573781967163086, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.697894275188446, + "num_tokens": 385421690.0, + "step": 14904 + }, + { + "epoch": 1.636832857456622, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.597853660583496, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.704820990562439, + "num_tokens": 385444381.0, + "step": 14905 + }, + { + "epoch": 1.6369426751592355, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5291311740875244, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.712039589881897, + "num_tokens": 385470342.0, + "step": 14906 + }, + { + "epoch": 1.6370524928618493, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.675546884536743, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.7236162424087524, + "num_tokens": 385493926.0, + "step": 14907 + }, + { + "epoch": 1.637162310564463, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7074503898620605, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7098695635795593, + "num_tokens": 385520209.0, + "step": 14908 + }, + { + "epoch": 1.6372721282670768, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7465689182281494, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7056273221969604, + "num_tokens": 385546552.0, + "step": 14909 + }, + { + "epoch": 1.6373819459696903, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.545988082885742, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6942993402481079, + "num_tokens": 385573161.0, + "step": 14910 + }, + { + "epoch": 1.6374917636723039, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5468435287475586, + "learning_rate": 1e-06, + "loss": 0.8589, + "mean_token_accuracy": 0.7447957992553711, + "num_tokens": 385597104.0, + "step": 14911 + }, + { + "epoch": 1.6376015813749176, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6871843338012695, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7225548624992371, + "num_tokens": 385619874.0, + "step": 14912 + }, + { + "epoch": 1.6377113990775314, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7837493419647217, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6857523322105408, + "num_tokens": 385646407.0, + "step": 14913 + }, + { + "epoch": 1.637821216780145, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.731184482574463, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7063302397727966, + "num_tokens": 385671319.0, + "step": 14914 + }, + { + "epoch": 1.6379310344827587, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.604465961456299, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6839953660964966, + "num_tokens": 385701247.0, + "step": 14915 + }, + { + "epoch": 1.6380408521853722, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4937775135040283, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.685139536857605, + "num_tokens": 385732322.0, + "step": 14916 + }, + { + "epoch": 1.638150669887986, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8010129928588867, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.711780309677124, + "num_tokens": 385754702.0, + "step": 14917 + }, + { + "epoch": 1.6382604875905997, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5352351665496826, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6911014914512634, + "num_tokens": 385785464.0, + "step": 14918 + }, + { + "epoch": 1.6383703052932133, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7699382305145264, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7114152312278748, + "num_tokens": 385810503.0, + "step": 14919 + }, + { + "epoch": 1.6384801229958268, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2463138103485107, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7168158888816833, + "num_tokens": 385829758.0, + "step": 14920 + }, + { + "epoch": 1.6385899406984406, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.975092649459839, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7123972177505493, + "num_tokens": 385849368.0, + "step": 14921 + }, + { + "epoch": 1.6386997584010543, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4579336643218994, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.7019710540771484, + "num_tokens": 385876430.0, + "step": 14922 + }, + { + "epoch": 1.638809576103668, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.832329273223877, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7168475389480591, + "num_tokens": 385899573.0, + "step": 14923 + }, + { + "epoch": 1.6389193938062816, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.1354336738586426, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7108222246170044, + "num_tokens": 385922626.0, + "step": 14924 + }, + { + "epoch": 1.6390292115088951, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8086581230163574, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7091265916824341, + "num_tokens": 385947999.0, + "step": 14925 + }, + { + "epoch": 1.639139029211509, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9880435466766357, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7097360491752625, + "num_tokens": 385968647.0, + "step": 14926 + }, + { + "epoch": 1.6392488469141226, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7770893573760986, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7220540642738342, + "num_tokens": 385988918.0, + "step": 14927 + }, + { + "epoch": 1.6393586646167362, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5289268493652344, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7102546691894531, + "num_tokens": 386014473.0, + "step": 14928 + }, + { + "epoch": 1.6394684823193497, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3732967376708984, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7053695917129517, + "num_tokens": 386044083.0, + "step": 14929 + }, + { + "epoch": 1.6395783000219635, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7439169883728027, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7056061029434204, + "num_tokens": 386066503.0, + "step": 14930 + }, + { + "epoch": 1.6396881177245772, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3066084384918213, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.693957507610321, + "num_tokens": 386100803.0, + "step": 14931 + }, + { + "epoch": 1.639797935427191, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.344442129135132, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6954264640808105, + "num_tokens": 386133291.0, + "step": 14932 + }, + { + "epoch": 1.6399077531298045, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.37276291847229, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6874582171440125, + "num_tokens": 386165398.0, + "step": 14933 + }, + { + "epoch": 1.640017570832418, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6430349349975586, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6917193531990051, + "num_tokens": 386196459.0, + "step": 14934 + }, + { + "epoch": 1.6401273885350318, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6441335678100586, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.6930258274078369, + "num_tokens": 386221217.0, + "step": 14935 + }, + { + "epoch": 1.6402372062376456, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.652520179748535, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7101197242736816, + "num_tokens": 386244454.0, + "step": 14936 + }, + { + "epoch": 1.6403470239402593, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6142776012420654, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6985458731651306, + "num_tokens": 386271152.0, + "step": 14937 + }, + { + "epoch": 1.6404568416428729, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6249868869781494, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.6950000524520874, + "num_tokens": 386296627.0, + "step": 14938 + }, + { + "epoch": 1.6405666593454864, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6960248947143555, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7148815393447876, + "num_tokens": 386322958.0, + "step": 14939 + }, + { + "epoch": 1.6406764770481002, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7314963340759277, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7190004587173462, + "num_tokens": 386345027.0, + "step": 14940 + }, + { + "epoch": 1.640786294750714, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4586548805236816, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7099032402038574, + "num_tokens": 386375754.0, + "step": 14941 + }, + { + "epoch": 1.6408961124533274, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5693633556365967, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6903005242347717, + "num_tokens": 386405499.0, + "step": 14942 + }, + { + "epoch": 1.641005930155941, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.364276885986328, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7148735523223877, + "num_tokens": 386435079.0, + "step": 14943 + }, + { + "epoch": 1.6411157478585547, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4705309867858887, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6986722350120544, + "num_tokens": 386461966.0, + "step": 14944 + }, + { + "epoch": 1.6412255655611685, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7895073890686035, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7121471166610718, + "num_tokens": 386487258.0, + "step": 14945 + }, + { + "epoch": 1.6413353832637823, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8907928466796875, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.6957137584686279, + "num_tokens": 386511094.0, + "step": 14946 + }, + { + "epoch": 1.6414452009663958, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.710967779159546, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6961796283721924, + "num_tokens": 386537644.0, + "step": 14947 + }, + { + "epoch": 1.6415550186690093, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.656853675842285, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6978303790092468, + "num_tokens": 386563414.0, + "step": 14948 + }, + { + "epoch": 1.641664836371623, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.682668447494507, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.691523015499115, + "num_tokens": 386588261.0, + "step": 14949 + }, + { + "epoch": 1.6417746540742368, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4515457153320312, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7105497717857361, + "num_tokens": 386618288.0, + "step": 14950 + }, + { + "epoch": 1.6418844717768504, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7926828861236572, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6975632905960083, + "num_tokens": 386641683.0, + "step": 14951 + }, + { + "epoch": 1.6419942894794641, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3856446743011475, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.721992552280426, + "num_tokens": 386669054.0, + "step": 14952 + }, + { + "epoch": 1.6421041071820777, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2941231727600098, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7130054831504822, + "num_tokens": 386698654.0, + "step": 14953 + }, + { + "epoch": 1.6422139248846914, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.615840196609497, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6969267129898071, + "num_tokens": 386725386.0, + "step": 14954 + }, + { + "epoch": 1.6423237425873052, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.553677797317505, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7080175876617432, + "num_tokens": 386748385.0, + "step": 14955 + }, + { + "epoch": 1.6424335602899187, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5117719173431396, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7118921279907227, + "num_tokens": 386775453.0, + "step": 14956 + }, + { + "epoch": 1.6425433779925322, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4816176891326904, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7179691195487976, + "num_tokens": 386802118.0, + "step": 14957 + }, + { + "epoch": 1.642653195695146, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4362030029296875, + "learning_rate": 1e-06, + "loss": 1.1139, + "mean_token_accuracy": 0.6820293664932251, + "num_tokens": 386833173.0, + "step": 14958 + }, + { + "epoch": 1.6427630133977598, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.529966354370117, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.6937699317932129, + "num_tokens": 386858346.0, + "step": 14959 + }, + { + "epoch": 1.6428728311003735, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.473759174346924, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7074810862541199, + "num_tokens": 386886388.0, + "step": 14960 + }, + { + "epoch": 1.642982648802987, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8849146366119385, + "learning_rate": 1e-06, + "loss": 0.9328, + "mean_token_accuracy": 0.7258996367454529, + "num_tokens": 386906007.0, + "step": 14961 + }, + { + "epoch": 1.6430924665056006, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4640960693359375, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6962807178497314, + "num_tokens": 386933428.0, + "step": 14962 + }, + { + "epoch": 1.6432022842082143, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7417824268341064, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6926114559173584, + "num_tokens": 386964853.0, + "step": 14963 + }, + { + "epoch": 1.643312101910828, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.496103048324585, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6924698352813721, + "num_tokens": 386993593.0, + "step": 14964 + }, + { + "epoch": 1.6434219196134416, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7625412940979004, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6920093894004822, + "num_tokens": 387017013.0, + "step": 14965 + }, + { + "epoch": 1.6435317373160554, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7998619079589844, + "learning_rate": 1e-06, + "loss": 0.9349, + "mean_token_accuracy": 0.7237372994422913, + "num_tokens": 387037142.0, + "step": 14966 + }, + { + "epoch": 1.643641555018669, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8462135791778564, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7167046666145325, + "num_tokens": 387058297.0, + "step": 14967 + }, + { + "epoch": 1.6437513727212827, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7841787338256836, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7086472511291504, + "num_tokens": 387079291.0, + "step": 14968 + }, + { + "epoch": 1.6438611904238964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.883931875228882, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.6971735954284668, + "num_tokens": 387100506.0, + "step": 14969 + }, + { + "epoch": 1.64397100812651, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5366592407226562, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6949996948242188, + "num_tokens": 387128650.0, + "step": 14970 + }, + { + "epoch": 1.6440808258291235, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.72739577293396, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7102296352386475, + "num_tokens": 387151054.0, + "step": 14971 + }, + { + "epoch": 1.6441906435317373, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7143707275390625, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.713116466999054, + "num_tokens": 387173317.0, + "step": 14972 + }, + { + "epoch": 1.644300461234351, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1131532192230225, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6946885585784912, + "num_tokens": 387195398.0, + "step": 14973 + }, + { + "epoch": 1.6444102789369648, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9257707595825195, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7145023345947266, + "num_tokens": 387216806.0, + "step": 14974 + }, + { + "epoch": 1.6445200966395783, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4389424324035645, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6943141222000122, + "num_tokens": 387244637.0, + "step": 14975 + }, + { + "epoch": 1.6446299143421919, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9153051376342773, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.7036224007606506, + "num_tokens": 387271443.0, + "step": 14976 + }, + { + "epoch": 1.6447397320448056, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.397934675216675, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6916055679321289, + "num_tokens": 387300901.0, + "step": 14977 + }, + { + "epoch": 1.6448495497474194, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.1668899059295654, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7284442186355591, + "num_tokens": 387319865.0, + "step": 14978 + }, + { + "epoch": 1.644959367450033, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6314032077789307, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7004762291908264, + "num_tokens": 387343059.0, + "step": 14979 + }, + { + "epoch": 1.6450691851526464, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.593254804611206, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.711074948310852, + "num_tokens": 387372396.0, + "step": 14980 + }, + { + "epoch": 1.6451790028552602, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4708638191223145, + "learning_rate": 1e-06, + "loss": 1.1322, + "mean_token_accuracy": 0.6698555946350098, + "num_tokens": 387405565.0, + "step": 14981 + }, + { + "epoch": 1.645288820557874, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.770195245742798, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7251316905021667, + "num_tokens": 387425517.0, + "step": 14982 + }, + { + "epoch": 1.6453986382604877, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.752945899963379, + "learning_rate": 1e-06, + "loss": 0.9202, + "mean_token_accuracy": 0.7300628423690796, + "num_tokens": 387445500.0, + "step": 14983 + }, + { + "epoch": 1.6455084559631012, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.850250482559204, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7199548482894897, + "num_tokens": 387464702.0, + "step": 14984 + }, + { + "epoch": 1.6456182736657148, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6395320892333984, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.6976733207702637, + "num_tokens": 387489147.0, + "step": 14985 + }, + { + "epoch": 1.6457280913683285, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.588618755340576, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6962034702301025, + "num_tokens": 387516359.0, + "step": 14986 + }, + { + "epoch": 1.6458379090709423, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6703038215637207, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7039651870727539, + "num_tokens": 387539169.0, + "step": 14987 + }, + { + "epoch": 1.645947726773556, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.755470037460327, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7280619144439697, + "num_tokens": 387561416.0, + "step": 14988 + }, + { + "epoch": 1.6460575444761696, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.528388023376465, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.6807419061660767, + "num_tokens": 387588704.0, + "step": 14989 + }, + { + "epoch": 1.6461673621787831, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.775740146636963, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7040585875511169, + "num_tokens": 387609671.0, + "step": 14990 + }, + { + "epoch": 1.6462771798813969, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.402900218963623, + "learning_rate": 1e-06, + "loss": 1.091, + "mean_token_accuracy": 0.686445951461792, + "num_tokens": 387644406.0, + "step": 14991 + }, + { + "epoch": 1.6463869975840106, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3502039909362793, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7066009044647217, + "num_tokens": 387675385.0, + "step": 14992 + }, + { + "epoch": 1.6464968152866242, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5275719165802, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7158395051956177, + "num_tokens": 387702653.0, + "step": 14993 + }, + { + "epoch": 1.6466066329892377, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.591233968734741, + "learning_rate": 1e-06, + "loss": 0.9039, + "mean_token_accuracy": 0.7280672788619995, + "num_tokens": 387726744.0, + "step": 14994 + }, + { + "epoch": 1.6467164506918515, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6569089889526367, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7139102220535278, + "num_tokens": 387749051.0, + "step": 14995 + }, + { + "epoch": 1.6468262683944652, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.531416177749634, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.6985763311386108, + "num_tokens": 387773399.0, + "step": 14996 + }, + { + "epoch": 1.646936086097079, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.652581214904785, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7082468271255493, + "num_tokens": 387798562.0, + "step": 14997 + }, + { + "epoch": 1.6470459037996925, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6134088039398193, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.7016509771347046, + "num_tokens": 387825123.0, + "step": 14998 + }, + { + "epoch": 1.647155721502306, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4037814140319824, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6928507089614868, + "num_tokens": 387855785.0, + "step": 14999 + }, + { + "epoch": 1.6472655392049198, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3981146812438965, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7021573185920715, + "num_tokens": 387885037.0, + "step": 15000 + }, + { + "epoch": 1.6473753569075336, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1597988605499268, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.7218209505081177, + "num_tokens": 387903513.0, + "step": 15001 + }, + { + "epoch": 1.6474851746101473, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3343868255615234, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.6870792508125305, + "num_tokens": 387935721.0, + "step": 15002 + }, + { + "epoch": 1.6475949923127609, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.572354793548584, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7131341695785522, + "num_tokens": 387960680.0, + "step": 15003 + }, + { + "epoch": 1.6477048100153744, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.479109764099121, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6877771615982056, + "num_tokens": 387990376.0, + "step": 15004 + }, + { + "epoch": 1.6478146277179881, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6189098358154297, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7177800536155701, + "num_tokens": 388013280.0, + "step": 15005 + }, + { + "epoch": 1.647924445420602, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.412163257598877, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6894460320472717, + "num_tokens": 388044228.0, + "step": 15006 + }, + { + "epoch": 1.6480342631232154, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4037396907806396, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7036325335502625, + "num_tokens": 388069822.0, + "step": 15007 + }, + { + "epoch": 1.648144080825829, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.433706045150757, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6906577348709106, + "num_tokens": 388102749.0, + "step": 15008 + }, + { + "epoch": 1.6482538985284427, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.894190788269043, + "learning_rate": 1e-06, + "loss": 0.9555, + "mean_token_accuracy": 0.7214952707290649, + "num_tokens": 388122017.0, + "step": 15009 + }, + { + "epoch": 1.6483637162310565, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.770451784133911, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6954776048660278, + "num_tokens": 388146361.0, + "step": 15010 + }, + { + "epoch": 1.6484735339336702, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7643723487854004, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6835377216339111, + "num_tokens": 388170621.0, + "step": 15011 + }, + { + "epoch": 1.6485833516362838, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.668119192123413, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7105064392089844, + "num_tokens": 388198618.0, + "step": 15012 + }, + { + "epoch": 1.6486931693388973, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6182339191436768, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6941496729850769, + "num_tokens": 388224203.0, + "step": 15013 + }, + { + "epoch": 1.648802987041511, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4026360511779785, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6831773519515991, + "num_tokens": 388256049.0, + "step": 15014 + }, + { + "epoch": 1.6489128047441248, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6680469512939453, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7067000865936279, + "num_tokens": 388280606.0, + "step": 15015 + }, + { + "epoch": 1.6490226224467384, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.822608232498169, + "learning_rate": 1e-06, + "loss": 1.065, + "mean_token_accuracy": 0.692156195640564, + "num_tokens": 388305205.0, + "step": 15016 + }, + { + "epoch": 1.6491324401493521, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5384697914123535, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.702337384223938, + "num_tokens": 388335252.0, + "step": 15017 + }, + { + "epoch": 1.6492422578519657, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8351731300354004, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7141144275665283, + "num_tokens": 388358098.0, + "step": 15018 + }, + { + "epoch": 1.6493520755545794, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5601165294647217, + "learning_rate": 1e-06, + "loss": 0.9175, + "mean_token_accuracy": 0.7270049452781677, + "num_tokens": 388381800.0, + "step": 15019 + }, + { + "epoch": 1.6494618932571932, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6547465324401855, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6910771131515503, + "num_tokens": 388406651.0, + "step": 15020 + }, + { + "epoch": 1.6495717109598067, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7032015323638916, + "learning_rate": 1e-06, + "loss": 0.9154, + "mean_token_accuracy": 0.7355250120162964, + "num_tokens": 388431031.0, + "step": 15021 + }, + { + "epoch": 1.6496815286624202, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5016698837280273, + "learning_rate": 1e-06, + "loss": 1.1108, + "mean_token_accuracy": 0.6905158758163452, + "num_tokens": 388459428.0, + "step": 15022 + }, + { + "epoch": 1.649791346365034, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3855226039886475, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7149274349212646, + "num_tokens": 388488077.0, + "step": 15023 + }, + { + "epoch": 1.6499011640676478, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.464297294616699, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7087893486022949, + "num_tokens": 388516556.0, + "step": 15024 + }, + { + "epoch": 1.6500109817702615, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7004435062408447, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6946827173233032, + "num_tokens": 388542273.0, + "step": 15025 + }, + { + "epoch": 1.650120799472875, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5328333377838135, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7116869688034058, + "num_tokens": 388568517.0, + "step": 15026 + }, + { + "epoch": 1.6502306171754886, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.66583514213562, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7129614353179932, + "num_tokens": 388590410.0, + "step": 15027 + }, + { + "epoch": 1.6503404348781023, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.551788568496704, + "learning_rate": 1e-06, + "loss": 0.8946, + "mean_token_accuracy": 0.7318322658538818, + "num_tokens": 388614421.0, + "step": 15028 + }, + { + "epoch": 1.650450252580716, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.442045211791992, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7236368656158447, + "num_tokens": 388643445.0, + "step": 15029 + }, + { + "epoch": 1.6505600702833296, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.711695909500122, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6864719390869141, + "num_tokens": 388669637.0, + "step": 15030 + }, + { + "epoch": 1.6506698879859434, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.69118070602417, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7160243988037109, + "num_tokens": 388692316.0, + "step": 15031 + }, + { + "epoch": 1.650779705688557, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.812370777130127, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.7255983352661133, + "num_tokens": 388714361.0, + "step": 15032 + }, + { + "epoch": 1.6508895233911707, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3943705558776855, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6980870962142944, + "num_tokens": 388742386.0, + "step": 15033 + }, + { + "epoch": 1.6509993410937844, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6961021423339844, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.6984001398086548, + "num_tokens": 388765011.0, + "step": 15034 + }, + { + "epoch": 1.651109158796398, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8321750164031982, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6964131593704224, + "num_tokens": 388790939.0, + "step": 15035 + }, + { + "epoch": 1.6512189764990115, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.727633237838745, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7069658041000366, + "num_tokens": 388816139.0, + "step": 15036 + }, + { + "epoch": 1.6513287942016253, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.665647268295288, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7071303129196167, + "num_tokens": 388841709.0, + "step": 15037 + }, + { + "epoch": 1.651438611904239, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.697568416595459, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.6998955011367798, + "num_tokens": 388868760.0, + "step": 15038 + }, + { + "epoch": 1.6515484296068528, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8147101402282715, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7141465544700623, + "num_tokens": 388889984.0, + "step": 15039 + }, + { + "epoch": 1.6516582473094663, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6914639472961426, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7081389427185059, + "num_tokens": 388915538.0, + "step": 15040 + }, + { + "epoch": 1.6517680650120798, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0761826038360596, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7264838218688965, + "num_tokens": 388935827.0, + "step": 15041 + }, + { + "epoch": 1.6518778827146936, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7948737144470215, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7009574770927429, + "num_tokens": 388959234.0, + "step": 15042 + }, + { + "epoch": 1.6519877004173074, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.56704044342041, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6935385465621948, + "num_tokens": 388985059.0, + "step": 15043 + }, + { + "epoch": 1.652097518119921, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.69760799407959, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6834275126457214, + "num_tokens": 389008956.0, + "step": 15044 + }, + { + "epoch": 1.6522073358225344, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7082488536834717, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6922672986984253, + "num_tokens": 389035428.0, + "step": 15045 + }, + { + "epoch": 1.6523171535251482, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.883474826812744, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.702878475189209, + "num_tokens": 389060170.0, + "step": 15046 + }, + { + "epoch": 1.652426971227762, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.560800790786743, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6889652013778687, + "num_tokens": 389087000.0, + "step": 15047 + }, + { + "epoch": 1.6525367889303757, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.788182497024536, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6895632743835449, + "num_tokens": 389110650.0, + "step": 15048 + }, + { + "epoch": 1.6526466066329892, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8446004390716553, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.7125247716903687, + "num_tokens": 389131713.0, + "step": 15049 + }, + { + "epoch": 1.6527564243356028, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.607670307159424, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6860804557800293, + "num_tokens": 389155984.0, + "step": 15050 + }, + { + "epoch": 1.6528662420382165, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8911867141723633, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7055553197860718, + "num_tokens": 389175024.0, + "step": 15051 + }, + { + "epoch": 1.6529760597408303, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6593594551086426, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7110162973403931, + "num_tokens": 389198396.0, + "step": 15052 + }, + { + "epoch": 1.653085877443444, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6049134731292725, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.7052545547485352, + "num_tokens": 389224011.0, + "step": 15053 + }, + { + "epoch": 1.6531956951460576, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.462545156478882, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7113991379737854, + "num_tokens": 389250356.0, + "step": 15054 + }, + { + "epoch": 1.653305512848671, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6658623218536377, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7106029987335205, + "num_tokens": 389274008.0, + "step": 15055 + }, + { + "epoch": 1.6534153305512849, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.766340494155884, + "learning_rate": 1e-06, + "loss": 1.1157, + "mean_token_accuracy": 0.673593282699585, + "num_tokens": 389297790.0, + "step": 15056 + }, + { + "epoch": 1.6535251482538986, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.448965549468994, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6932172179222107, + "num_tokens": 389329161.0, + "step": 15057 + }, + { + "epoch": 1.6536349659565122, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4956891536712646, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.715095043182373, + "num_tokens": 389355768.0, + "step": 15058 + }, + { + "epoch": 1.6537447836591257, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.467869520187378, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.702381432056427, + "num_tokens": 389386013.0, + "step": 15059 + }, + { + "epoch": 1.6538546013617395, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.691159725189209, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6921108961105347, + "num_tokens": 389412209.0, + "step": 15060 + }, + { + "epoch": 1.6539644190643532, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2907612323760986, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6892357468605042, + "num_tokens": 389446399.0, + "step": 15061 + }, + { + "epoch": 1.654074236766967, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5809895992279053, + "learning_rate": 1e-06, + "loss": 0.8619, + "mean_token_accuracy": 0.7436820268630981, + "num_tokens": 389471148.0, + "step": 15062 + }, + { + "epoch": 1.6541840544695805, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.373448133468628, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6833193302154541, + "num_tokens": 389508397.0, + "step": 15063 + }, + { + "epoch": 1.654293872172194, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.55914568901062, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6976841688156128, + "num_tokens": 389535562.0, + "step": 15064 + }, + { + "epoch": 1.6544036898748078, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.481309413909912, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.7208792567253113, + "num_tokens": 389563004.0, + "step": 15065 + }, + { + "epoch": 1.6545135075774215, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3925247192382812, + "learning_rate": 1e-06, + "loss": 0.8599, + "mean_token_accuracy": 0.7429215312004089, + "num_tokens": 389586088.0, + "step": 15066 + }, + { + "epoch": 1.6546233252800353, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5240252017974854, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7229689359664917, + "num_tokens": 389612021.0, + "step": 15067 + }, + { + "epoch": 1.6547331429826488, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6391677856445312, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.711650013923645, + "num_tokens": 389642765.0, + "step": 15068 + }, + { + "epoch": 1.6548429606852624, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8133742809295654, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7227030992507935, + "num_tokens": 389665514.0, + "step": 15069 + }, + { + "epoch": 1.6549527783878761, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5970091819763184, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6934747695922852, + "num_tokens": 389692152.0, + "step": 15070 + }, + { + "epoch": 1.65506259609049, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8342247009277344, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7068821787834167, + "num_tokens": 389717493.0, + "step": 15071 + }, + { + "epoch": 1.6551724137931034, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8522582054138184, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6953936219215393, + "num_tokens": 389741451.0, + "step": 15072 + }, + { + "epoch": 1.655282231495717, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.50286602973938, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7042756676673889, + "num_tokens": 389767007.0, + "step": 15073 + }, + { + "epoch": 1.6553920491983307, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.857055187225342, + "learning_rate": 1e-06, + "loss": 1.0725, + "mean_token_accuracy": 0.683866024017334, + "num_tokens": 389788434.0, + "step": 15074 + }, + { + "epoch": 1.6555018669009445, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5913758277893066, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6901146173477173, + "num_tokens": 389815287.0, + "step": 15075 + }, + { + "epoch": 1.6556116846035582, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7591171264648438, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7209987640380859, + "num_tokens": 389836107.0, + "step": 15076 + }, + { + "epoch": 1.6557215023061718, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.468384265899658, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7149379253387451, + "num_tokens": 389860566.0, + "step": 15077 + }, + { + "epoch": 1.6558313200087853, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3541343212127686, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7171025276184082, + "num_tokens": 389891930.0, + "step": 15078 + }, + { + "epoch": 1.655941137711399, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6914865970611572, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6945605278015137, + "num_tokens": 389916670.0, + "step": 15079 + }, + { + "epoch": 1.6560509554140128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.378322124481201, + "learning_rate": 1e-06, + "loss": 1.0794, + "mean_token_accuracy": 0.6856517791748047, + "num_tokens": 389945835.0, + "step": 15080 + }, + { + "epoch": 1.6561607731166264, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5607566833496094, + "learning_rate": 1e-06, + "loss": 0.888, + "mean_token_accuracy": 0.7438554763793945, + "num_tokens": 389972301.0, + "step": 15081 + }, + { + "epoch": 1.65627059081924, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.891982078552246, + "learning_rate": 1e-06, + "loss": 0.9039, + "mean_token_accuracy": 0.7344847917556763, + "num_tokens": 389992002.0, + "step": 15082 + }, + { + "epoch": 1.6563804085218536, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6102259159088135, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6900486946105957, + "num_tokens": 390018505.0, + "step": 15083 + }, + { + "epoch": 1.6564902262244674, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.774280309677124, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7153499126434326, + "num_tokens": 390040048.0, + "step": 15084 + }, + { + "epoch": 1.6566000439270812, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6141138076782227, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6954014301300049, + "num_tokens": 390068920.0, + "step": 15085 + }, + { + "epoch": 1.6567098616296947, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7142159938812256, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6816197037696838, + "num_tokens": 390095080.0, + "step": 15086 + }, + { + "epoch": 1.6568196793323082, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.46386981010437, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7149887084960938, + "num_tokens": 390123902.0, + "step": 15087 + }, + { + "epoch": 1.656929497034922, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.52921986579895, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6935884952545166, + "num_tokens": 390149633.0, + "step": 15088 + }, + { + "epoch": 1.6570393147375357, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.634671688079834, + "learning_rate": 1e-06, + "loss": 1.1354, + "mean_token_accuracy": 0.689969539642334, + "num_tokens": 390173020.0, + "step": 15089 + }, + { + "epoch": 1.6571491324401495, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.638262987136841, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6979993581771851, + "num_tokens": 390198722.0, + "step": 15090 + }, + { + "epoch": 1.657258950142763, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.619943380355835, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7151986956596375, + "num_tokens": 390226112.0, + "step": 15091 + }, + { + "epoch": 1.6573687678453766, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4906435012817383, + "learning_rate": 1e-06, + "loss": 0.9365, + "mean_token_accuracy": 0.7183277010917664, + "num_tokens": 390250826.0, + "step": 15092 + }, + { + "epoch": 1.6574785855479903, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7545292377471924, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7118346095085144, + "num_tokens": 390273521.0, + "step": 15093 + }, + { + "epoch": 1.657588403250604, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5469424724578857, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.7162163257598877, + "num_tokens": 390300333.0, + "step": 15094 + }, + { + "epoch": 1.6576982209532176, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5457355976104736, + "learning_rate": 1e-06, + "loss": 0.9073, + "mean_token_accuracy": 0.7343671321868896, + "num_tokens": 390327090.0, + "step": 15095 + }, + { + "epoch": 1.6578080386558314, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6808574199676514, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.6917546391487122, + "num_tokens": 390352112.0, + "step": 15096 + }, + { + "epoch": 1.657917856358445, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.497606039047241, + "learning_rate": 1e-06, + "loss": 0.9319, + "mean_token_accuracy": 0.7237178087234497, + "num_tokens": 390379128.0, + "step": 15097 + }, + { + "epoch": 1.6580276740610587, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7537055015563965, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7043800354003906, + "num_tokens": 390402702.0, + "step": 15098 + }, + { + "epoch": 1.6581374917636724, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9183530807495117, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6975975632667542, + "num_tokens": 390425482.0, + "step": 15099 + }, + { + "epoch": 1.658247309466286, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5766565799713135, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7088789939880371, + "num_tokens": 390448680.0, + "step": 15100 + }, + { + "epoch": 1.6583571271688995, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5081207752227783, + "learning_rate": 1e-06, + "loss": 1.1016, + "mean_token_accuracy": 0.6853481531143188, + "num_tokens": 390475776.0, + "step": 15101 + }, + { + "epoch": 1.6584669448715132, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.950990676879883, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7028372287750244, + "num_tokens": 390499965.0, + "step": 15102 + }, + { + "epoch": 1.658576762574127, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.68608021736145, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.715634822845459, + "num_tokens": 390525663.0, + "step": 15103 + }, + { + "epoch": 1.6586865802767408, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6580145359039307, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6939196586608887, + "num_tokens": 390549509.0, + "step": 15104 + }, + { + "epoch": 1.6587963979793543, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.634594202041626, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6830711960792542, + "num_tokens": 390577154.0, + "step": 15105 + }, + { + "epoch": 1.6589062156819678, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5044126510620117, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6985399127006531, + "num_tokens": 390606950.0, + "step": 15106 + }, + { + "epoch": 1.6590160333845816, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8445258140563965, + "learning_rate": 1e-06, + "loss": 1.1055, + "mean_token_accuracy": 0.6928380727767944, + "num_tokens": 390630059.0, + "step": 15107 + }, + { + "epoch": 1.6591258510871953, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.593703508377075, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.707206130027771, + "num_tokens": 390656855.0, + "step": 15108 + }, + { + "epoch": 1.6592356687898089, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8262696266174316, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7326748371124268, + "num_tokens": 390681132.0, + "step": 15109 + }, + { + "epoch": 1.6593454864924224, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5263450145721436, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.7037309408187866, + "num_tokens": 390710201.0, + "step": 15110 + }, + { + "epoch": 1.6594553041950362, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.598877191543579, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7256497740745544, + "num_tokens": 390735305.0, + "step": 15111 + }, + { + "epoch": 1.65956512189765, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.735570192337036, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7046915292739868, + "num_tokens": 390760313.0, + "step": 15112 + }, + { + "epoch": 1.6596749396002637, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6965513229370117, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.710273265838623, + "num_tokens": 390782847.0, + "step": 15113 + }, + { + "epoch": 1.6597847573028772, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 20.543628692626953, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.7008845806121826, + "num_tokens": 390808668.0, + "step": 15114 + }, + { + "epoch": 1.6598945750054908, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.80120587348938, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6950172185897827, + "num_tokens": 390835133.0, + "step": 15115 + }, + { + "epoch": 1.6600043927081045, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6451833248138428, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7166531085968018, + "num_tokens": 390860844.0, + "step": 15116 + }, + { + "epoch": 1.6601142104107183, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.2090349197387695, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7038570046424866, + "num_tokens": 390881447.0, + "step": 15117 + }, + { + "epoch": 1.660224028113332, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3690035343170166, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.7039859294891357, + "num_tokens": 390914311.0, + "step": 15118 + }, + { + "epoch": 1.6603338458159456, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.2381606101989746, + "learning_rate": 1e-06, + "loss": 0.9014, + "mean_token_accuracy": 0.7283539772033691, + "num_tokens": 390936130.0, + "step": 15119 + }, + { + "epoch": 1.660443663518559, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5490000247955322, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7214690446853638, + "num_tokens": 390963900.0, + "step": 15120 + }, + { + "epoch": 1.6605534812211729, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5710670948028564, + "learning_rate": 1e-06, + "loss": 0.9361, + "mean_token_accuracy": 0.7287485599517822, + "num_tokens": 390988318.0, + "step": 15121 + }, + { + "epoch": 1.6606632989237866, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6257917881011963, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.7040980458259583, + "num_tokens": 391013088.0, + "step": 15122 + }, + { + "epoch": 1.6607731166264001, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.691495895385742, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.7029584050178528, + "num_tokens": 391038533.0, + "step": 15123 + }, + { + "epoch": 1.6608829343290137, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6659817695617676, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.7012405395507812, + "num_tokens": 391063810.0, + "step": 15124 + }, + { + "epoch": 1.6609927520316274, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.629751443862915, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6901473999023438, + "num_tokens": 391089542.0, + "step": 15125 + }, + { + "epoch": 1.6611025697342412, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5269176959991455, + "learning_rate": 1e-06, + "loss": 0.9323, + "mean_token_accuracy": 0.7294841408729553, + "num_tokens": 391113833.0, + "step": 15126 + }, + { + "epoch": 1.661212387436855, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5413312911987305, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7163910865783691, + "num_tokens": 391138425.0, + "step": 15127 + }, + { + "epoch": 1.6613222051394685, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6971561908721924, + "learning_rate": 1e-06, + "loss": 0.9309, + "mean_token_accuracy": 0.7303580045700073, + "num_tokens": 391161315.0, + "step": 15128 + }, + { + "epoch": 1.661432022842082, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6205687522888184, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7164015173912048, + "num_tokens": 391185436.0, + "step": 15129 + }, + { + "epoch": 1.6615418405446958, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7099876403808594, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7109940052032471, + "num_tokens": 391209066.0, + "step": 15130 + }, + { + "epoch": 1.6616516582473095, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9777684211730957, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7070295214653015, + "num_tokens": 391229470.0, + "step": 15131 + }, + { + "epoch": 1.661761475949923, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.483232259750366, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7046160697937012, + "num_tokens": 391259599.0, + "step": 15132 + }, + { + "epoch": 1.6618712936525368, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5583207607269287, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.7149801254272461, + "num_tokens": 391284759.0, + "step": 15133 + }, + { + "epoch": 1.6619811113551504, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5119528770446777, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7095794677734375, + "num_tokens": 391310964.0, + "step": 15134 + }, + { + "epoch": 1.6620909290577641, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.616428852081299, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.714774489402771, + "num_tokens": 391336718.0, + "step": 15135 + }, + { + "epoch": 1.6622007467603779, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6109237670898438, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6932680606842041, + "num_tokens": 391363522.0, + "step": 15136 + }, + { + "epoch": 1.6623105644629914, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4300529956817627, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.71189284324646, + "num_tokens": 391390172.0, + "step": 15137 + }, + { + "epoch": 1.662420382165605, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5560572147369385, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7021058797836304, + "num_tokens": 391416941.0, + "step": 15138 + }, + { + "epoch": 1.6625301998682187, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4679791927337646, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.691612184047699, + "num_tokens": 391444599.0, + "step": 15139 + }, + { + "epoch": 1.6626400175708325, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4193692207336426, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6935813426971436, + "num_tokens": 391471679.0, + "step": 15140 + }, + { + "epoch": 1.6627498352734462, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4944159984588623, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7013229727745056, + "num_tokens": 391500427.0, + "step": 15141 + }, + { + "epoch": 1.6628596529760598, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5940563678741455, + "learning_rate": 1e-06, + "loss": 0.896, + "mean_token_accuracy": 0.7374728918075562, + "num_tokens": 391526531.0, + "step": 15142 + }, + { + "epoch": 1.6629694706786733, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6808176040649414, + "learning_rate": 1e-06, + "loss": 0.9152, + "mean_token_accuracy": 0.7318513989448547, + "num_tokens": 391549471.0, + "step": 15143 + }, + { + "epoch": 1.663079288381287, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7504303455352783, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7151541113853455, + "num_tokens": 391572180.0, + "step": 15144 + }, + { + "epoch": 1.6631891060839008, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.530571222305298, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7093536853790283, + "num_tokens": 391598234.0, + "step": 15145 + }, + { + "epoch": 1.6632989237865143, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.512773275375366, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6915686726570129, + "num_tokens": 391627804.0, + "step": 15146 + }, + { + "epoch": 1.663408741489128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.853844165802002, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7169134616851807, + "num_tokens": 391650198.0, + "step": 15147 + }, + { + "epoch": 1.6635185591917416, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 20.522674560546875, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.707486629486084, + "num_tokens": 391679974.0, + "step": 15148 + }, + { + "epoch": 1.6636283768943554, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7590079307556152, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.7036298513412476, + "num_tokens": 391706421.0, + "step": 15149 + }, + { + "epoch": 1.6637381945969691, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6401641368865967, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7006388306617737, + "num_tokens": 391737789.0, + "step": 15150 + }, + { + "epoch": 1.6638480122995827, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.408395290374756, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7182024717330933, + "num_tokens": 391767656.0, + "step": 15151 + }, + { + "epoch": 1.6639578300021962, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7396433353424072, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7130991220474243, + "num_tokens": 391788356.0, + "step": 15152 + }, + { + "epoch": 1.66406764770481, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.398987293243408, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7020643949508667, + "num_tokens": 391817753.0, + "step": 15153 + }, + { + "epoch": 1.6641774654074237, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.530085563659668, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7087761759757996, + "num_tokens": 391842091.0, + "step": 15154 + }, + { + "epoch": 1.6642872831100375, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.542515516281128, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.6932944059371948, + "num_tokens": 391870270.0, + "step": 15155 + }, + { + "epoch": 1.664397100812651, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.285938262939453, + "learning_rate": 1e-06, + "loss": 1.1782, + "mean_token_accuracy": 0.6684891581535339, + "num_tokens": 391903803.0, + "step": 15156 + }, + { + "epoch": 1.6645069185152646, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6718289852142334, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7153047323226929, + "num_tokens": 391926968.0, + "step": 15157 + }, + { + "epoch": 1.6646167362178783, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.605766773223877, + "learning_rate": 1e-06, + "loss": 1.0837, + "mean_token_accuracy": 0.6921840310096741, + "num_tokens": 391954503.0, + "step": 15158 + }, + { + "epoch": 1.664726553920492, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.786654472351074, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7061561346054077, + "num_tokens": 391976220.0, + "step": 15159 + }, + { + "epoch": 1.6648363716231056, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3993451595306396, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.704497218132019, + "num_tokens": 392003237.0, + "step": 15160 + }, + { + "epoch": 1.6649461893257194, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.406682252883911, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7093943357467651, + "num_tokens": 392032988.0, + "step": 15161 + }, + { + "epoch": 1.665056007028333, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6062183380126953, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7255300283432007, + "num_tokens": 392057474.0, + "step": 15162 + }, + { + "epoch": 1.6651658247309467, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4786839485168457, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7061047554016113, + "num_tokens": 392084909.0, + "step": 15163 + }, + { + "epoch": 1.6652756424335604, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5924742221832275, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6923283338546753, + "num_tokens": 392111931.0, + "step": 15164 + }, + { + "epoch": 1.665385460136174, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4748802185058594, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7185968160629272, + "num_tokens": 392138413.0, + "step": 15165 + }, + { + "epoch": 1.6654952778387875, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4989266395568848, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7107656002044678, + "num_tokens": 392164030.0, + "step": 15166 + }, + { + "epoch": 1.6656050955414012, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.580850124359131, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7050336599349976, + "num_tokens": 392189971.0, + "step": 15167 + }, + { + "epoch": 1.665714913244015, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.608505964279175, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.709438145160675, + "num_tokens": 392217168.0, + "step": 15168 + }, + { + "epoch": 1.6658247309466288, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.439344644546509, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7055895328521729, + "num_tokens": 392245303.0, + "step": 15169 + }, + { + "epoch": 1.6659345486492423, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5099282264709473, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7197058200836182, + "num_tokens": 392270791.0, + "step": 15170 + }, + { + "epoch": 1.6660443663518558, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.608999252319336, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7007365822792053, + "num_tokens": 392295622.0, + "step": 15171 + }, + { + "epoch": 1.6661541840544696, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5923233032226562, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6981302499771118, + "num_tokens": 392322785.0, + "step": 15172 + }, + { + "epoch": 1.6662640017570833, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.013542413711548, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7117806077003479, + "num_tokens": 392343448.0, + "step": 15173 + }, + { + "epoch": 1.6663738194596969, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0430915355682373, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7204992771148682, + "num_tokens": 392362200.0, + "step": 15174 + }, + { + "epoch": 1.6664836371623104, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4093167781829834, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7029512524604797, + "num_tokens": 392390515.0, + "step": 15175 + }, + { + "epoch": 1.6665934548649242, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.577571153640747, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7035537958145142, + "num_tokens": 392416679.0, + "step": 15176 + }, + { + "epoch": 1.666703272567538, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7439346313476562, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7051700353622437, + "num_tokens": 392441347.0, + "step": 15177 + }, + { + "epoch": 1.6668130902701517, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5067615509033203, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.697647750377655, + "num_tokens": 392472060.0, + "step": 15178 + }, + { + "epoch": 1.6669229079727652, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.599407911300659, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6895022392272949, + "num_tokens": 392494595.0, + "step": 15179 + }, + { + "epoch": 1.6670327256753787, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5282156467437744, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.710242748260498, + "num_tokens": 392521890.0, + "step": 15180 + }, + { + "epoch": 1.6671425433779925, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5955522060394287, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7037005424499512, + "num_tokens": 392543064.0, + "step": 15181 + }, + { + "epoch": 1.6672523610806063, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8351378440856934, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7129141688346863, + "num_tokens": 392565624.0, + "step": 15182 + }, + { + "epoch": 1.66736217878322, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7264041900634766, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.723633885383606, + "num_tokens": 392588081.0, + "step": 15183 + }, + { + "epoch": 1.6674719964858336, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.61883282661438, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7114100456237793, + "num_tokens": 392615830.0, + "step": 15184 + }, + { + "epoch": 1.667581814188447, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5098679065704346, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7206186056137085, + "num_tokens": 392643642.0, + "step": 15185 + }, + { + "epoch": 1.6676916318910608, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6847949028015137, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7009738683700562, + "num_tokens": 392668842.0, + "step": 15186 + }, + { + "epoch": 1.6678014495936746, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8133585453033447, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7194629311561584, + "num_tokens": 392691727.0, + "step": 15187 + }, + { + "epoch": 1.6679112672962881, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4385721683502197, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7046542167663574, + "num_tokens": 392718206.0, + "step": 15188 + }, + { + "epoch": 1.6680210849989017, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2303249835968018, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.706513524055481, + "num_tokens": 392749720.0, + "step": 15189 + }, + { + "epoch": 1.6681309027015154, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.618577003479004, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6883171796798706, + "num_tokens": 392775450.0, + "step": 15190 + }, + { + "epoch": 1.6682407204041292, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4070613384246826, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7000287175178528, + "num_tokens": 392805725.0, + "step": 15191 + }, + { + "epoch": 1.668350538106743, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6606907844543457, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.6986469030380249, + "num_tokens": 392828935.0, + "step": 15192 + }, + { + "epoch": 1.6684603558093565, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6065890789031982, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7186135649681091, + "num_tokens": 392852371.0, + "step": 15193 + }, + { + "epoch": 1.66857017351197, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6789772510528564, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7098613381385803, + "num_tokens": 392876515.0, + "step": 15194 + }, + { + "epoch": 1.6686799912145838, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5135276317596436, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7007123231887817, + "num_tokens": 392909674.0, + "step": 15195 + }, + { + "epoch": 1.6687898089171975, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.228562831878662, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.690804123878479, + "num_tokens": 392942641.0, + "step": 15196 + }, + { + "epoch": 1.668899626619811, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.1509640216827393, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6928085088729858, + "num_tokens": 392976520.0, + "step": 15197 + }, + { + "epoch": 1.6690094443224248, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7959415912628174, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6965815424919128, + "num_tokens": 392999207.0, + "step": 15198 + }, + { + "epoch": 1.6691192620250384, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2906179428100586, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.6997342109680176, + "num_tokens": 393029558.0, + "step": 15199 + }, + { + "epoch": 1.669229079727652, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.603649139404297, + "learning_rate": 1e-06, + "loss": 0.8748, + "mean_token_accuracy": 0.7378346920013428, + "num_tokens": 393051392.0, + "step": 15200 + }, + { + "epoch": 1.6693388974302659, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5279173851013184, + "learning_rate": 1e-06, + "loss": 0.891, + "mean_token_accuracy": 0.7321646213531494, + "num_tokens": 393076234.0, + "step": 15201 + }, + { + "epoch": 1.6694487151328794, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5095109939575195, + "learning_rate": 1e-06, + "loss": 0.9196, + "mean_token_accuracy": 0.7279837727546692, + "num_tokens": 393099892.0, + "step": 15202 + }, + { + "epoch": 1.669558532835493, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4619140625, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7000333070755005, + "num_tokens": 393128596.0, + "step": 15203 + }, + { + "epoch": 1.6696683505381067, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.411740779876709, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6992177367210388, + "num_tokens": 393159301.0, + "step": 15204 + }, + { + "epoch": 1.6697781682407205, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.837246894836426, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7121349573135376, + "num_tokens": 393181250.0, + "step": 15205 + }, + { + "epoch": 1.6698879859433342, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4702067375183105, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6886348128318787, + "num_tokens": 393208502.0, + "step": 15206 + }, + { + "epoch": 1.6699978036459477, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4699745178222656, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7079479694366455, + "num_tokens": 393234328.0, + "step": 15207 + }, + { + "epoch": 1.6701076213485613, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6589133739471436, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7084972858428955, + "num_tokens": 393257177.0, + "step": 15208 + }, + { + "epoch": 1.670217439051175, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7059168815612793, + "learning_rate": 1e-06, + "loss": 1.114, + "mean_token_accuracy": 0.6781336069107056, + "num_tokens": 393283471.0, + "step": 15209 + }, + { + "epoch": 1.6703272567537888, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7597405910491943, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7185511589050293, + "num_tokens": 393305955.0, + "step": 15210 + }, + { + "epoch": 1.6704370744564023, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.442516803741455, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.7062015533447266, + "num_tokens": 393333716.0, + "step": 15211 + }, + { + "epoch": 1.670546892159016, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3976879119873047, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7156717777252197, + "num_tokens": 393362875.0, + "step": 15212 + }, + { + "epoch": 1.6706567098616296, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.850905179977417, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.710537314414978, + "num_tokens": 393383648.0, + "step": 15213 + }, + { + "epoch": 1.6707665275642434, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6886379718780518, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.7008354663848877, + "num_tokens": 393410602.0, + "step": 15214 + }, + { + "epoch": 1.6708763452668571, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9630866050720215, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7131559252738953, + "num_tokens": 393433729.0, + "step": 15215 + }, + { + "epoch": 1.6709861629694707, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6162683963775635, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7024900913238525, + "num_tokens": 393457823.0, + "step": 15216 + }, + { + "epoch": 1.6710959806720842, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3497304916381836, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6915861368179321, + "num_tokens": 393487977.0, + "step": 15217 + }, + { + "epoch": 1.671205798374698, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.399895191192627, + "learning_rate": 1e-06, + "loss": 1.1175, + "mean_token_accuracy": 0.6762792468070984, + "num_tokens": 393520805.0, + "step": 15218 + }, + { + "epoch": 1.6713156160773117, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3620383739471436, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6952180862426758, + "num_tokens": 393550646.0, + "step": 15219 + }, + { + "epoch": 1.6714254337799255, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6717870235443115, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7220023274421692, + "num_tokens": 393572080.0, + "step": 15220 + }, + { + "epoch": 1.671535251482539, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.663480520248413, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.693682074546814, + "num_tokens": 393599908.0, + "step": 15221 + }, + { + "epoch": 1.6716450691851525, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7399215698242188, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7185898423194885, + "num_tokens": 393622182.0, + "step": 15222 + }, + { + "epoch": 1.6717548868877663, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5778746604919434, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7063785195350647, + "num_tokens": 393647899.0, + "step": 15223 + }, + { + "epoch": 1.67186470459038, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6592397689819336, + "learning_rate": 1e-06, + "loss": 0.9428, + "mean_token_accuracy": 0.7213999032974243, + "num_tokens": 393671432.0, + "step": 15224 + }, + { + "epoch": 1.6719745222929936, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5157315731048584, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.719441294670105, + "num_tokens": 393695873.0, + "step": 15225 + }, + { + "epoch": 1.6720843399956071, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6110424995422363, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7097118496894836, + "num_tokens": 393721029.0, + "step": 15226 + }, + { + "epoch": 1.6721941576982209, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3275015354156494, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7028819918632507, + "num_tokens": 393751345.0, + "step": 15227 + }, + { + "epoch": 1.6723039754008346, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.661013126373291, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.6995189189910889, + "num_tokens": 393777302.0, + "step": 15228 + }, + { + "epoch": 1.6724137931034484, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5818426609039307, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6878820657730103, + "num_tokens": 393805391.0, + "step": 15229 + }, + { + "epoch": 1.672523610806062, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9181928634643555, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6928583979606628, + "num_tokens": 393825653.0, + "step": 15230 + }, + { + "epoch": 1.6726334285086755, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5365118980407715, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6903796195983887, + "num_tokens": 393852131.0, + "step": 15231 + }, + { + "epoch": 1.6727432462112892, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.387402057647705, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.711237370967865, + "num_tokens": 393879563.0, + "step": 15232 + }, + { + "epoch": 1.672853063913903, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.450906991958618, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6959254145622253, + "num_tokens": 393908384.0, + "step": 15233 + }, + { + "epoch": 1.6729628816165167, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.866712808609009, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7062156200408936, + "num_tokens": 393929500.0, + "step": 15234 + }, + { + "epoch": 1.6730726993191303, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.638939380645752, + "learning_rate": 1e-06, + "loss": 0.9286, + "mean_token_accuracy": 0.7342671155929565, + "num_tokens": 393952337.0, + "step": 15235 + }, + { + "epoch": 1.6731825170217438, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.491943359375, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.6893860697746277, + "num_tokens": 393978801.0, + "step": 15236 + }, + { + "epoch": 1.6732923347243576, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5694820880889893, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.700687050819397, + "num_tokens": 394006803.0, + "step": 15237 + }, + { + "epoch": 1.6734021524269713, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.513404369354248, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6976996660232544, + "num_tokens": 394031596.0, + "step": 15238 + }, + { + "epoch": 1.6735119701295849, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.779158353805542, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7227936387062073, + "num_tokens": 394053285.0, + "step": 15239 + }, + { + "epoch": 1.6736217878321984, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.480759620666504, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7215051054954529, + "num_tokens": 394081476.0, + "step": 15240 + }, + { + "epoch": 1.6737316055348122, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8474488258361816, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.709195613861084, + "num_tokens": 394109698.0, + "step": 15241 + }, + { + "epoch": 1.673841423237426, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6502685546875, + "learning_rate": 1e-06, + "loss": 0.8857, + "mean_token_accuracy": 0.7400580048561096, + "num_tokens": 394133480.0, + "step": 15242 + }, + { + "epoch": 1.6739512409400397, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6273353099823, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7143371105194092, + "num_tokens": 394159412.0, + "step": 15243 + }, + { + "epoch": 1.6740610586426532, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6846354007720947, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6976839303970337, + "num_tokens": 394185616.0, + "step": 15244 + }, + { + "epoch": 1.6741708763452667, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3278565406799316, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7168627977371216, + "num_tokens": 394214586.0, + "step": 15245 + }, + { + "epoch": 1.6742806940478805, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7529351711273193, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6967134475708008, + "num_tokens": 394241550.0, + "step": 15246 + }, + { + "epoch": 1.6743905117504942, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.686903238296509, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7199029326438904, + "num_tokens": 394266073.0, + "step": 15247 + }, + { + "epoch": 1.674500329453108, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.57770037651062, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.7040194272994995, + "num_tokens": 394292595.0, + "step": 15248 + }, + { + "epoch": 1.6746101471557215, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5268547534942627, + "learning_rate": 1e-06, + "loss": 1.1261, + "mean_token_accuracy": 0.6813085675239563, + "num_tokens": 394323940.0, + "step": 15249 + }, + { + "epoch": 1.674719964858335, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5880982875823975, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6945058107376099, + "num_tokens": 394349290.0, + "step": 15250 + }, + { + "epoch": 1.6748297825609488, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0093445777893066, + "learning_rate": 1e-06, + "loss": 0.8842, + "mean_token_accuracy": 0.7448652982711792, + "num_tokens": 394368900.0, + "step": 15251 + }, + { + "epoch": 1.6749396002635626, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6795225143432617, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7115523815155029, + "num_tokens": 394390688.0, + "step": 15252 + }, + { + "epoch": 1.6750494179661761, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4991133213043213, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6910940408706665, + "num_tokens": 394419230.0, + "step": 15253 + }, + { + "epoch": 1.6751592356687897, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3205723762512207, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6826939582824707, + "num_tokens": 394458588.0, + "step": 15254 + }, + { + "epoch": 1.6752690533714034, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6291089057922363, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6916053891181946, + "num_tokens": 394486642.0, + "step": 15255 + }, + { + "epoch": 1.6753788710740172, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5047860145568848, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.7018691301345825, + "num_tokens": 394513361.0, + "step": 15256 + }, + { + "epoch": 1.675488688776631, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.519644260406494, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7067460417747498, + "num_tokens": 394539867.0, + "step": 15257 + }, + { + "epoch": 1.6755985064792445, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3999221324920654, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7014918327331543, + "num_tokens": 394571484.0, + "step": 15258 + }, + { + "epoch": 1.675708324181858, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6715633869171143, + "learning_rate": 1e-06, + "loss": 0.8788, + "mean_token_accuracy": 0.7378364205360413, + "num_tokens": 394592336.0, + "step": 15259 + }, + { + "epoch": 1.6758181418844718, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.692547559738159, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7007355093955994, + "num_tokens": 394617743.0, + "step": 15260 + }, + { + "epoch": 1.6759279595870855, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.704263210296631, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7148312926292419, + "num_tokens": 394642515.0, + "step": 15261 + }, + { + "epoch": 1.676037777289699, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6957666873931885, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.6998489499092102, + "num_tokens": 394665608.0, + "step": 15262 + }, + { + "epoch": 1.6761475949923128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6559360027313232, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7112302184104919, + "num_tokens": 394690613.0, + "step": 15263 + }, + { + "epoch": 1.6762574126949263, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4371588230133057, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6875361204147339, + "num_tokens": 394720380.0, + "step": 15264 + }, + { + "epoch": 1.67636723039754, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.540616989135742, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.699583888053894, + "num_tokens": 394750186.0, + "step": 15265 + }, + { + "epoch": 1.6764770481001539, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.48239803314209, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7042838931083679, + "num_tokens": 394780990.0, + "step": 15266 + }, + { + "epoch": 1.6765868658027674, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7654201984405518, + "learning_rate": 1e-06, + "loss": 0.9173, + "mean_token_accuracy": 0.7294551134109497, + "num_tokens": 394802607.0, + "step": 15267 + }, + { + "epoch": 1.676696683505381, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7356278896331787, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7136946320533752, + "num_tokens": 394825914.0, + "step": 15268 + }, + { + "epoch": 1.6768065012079947, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4045279026031494, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7159830331802368, + "num_tokens": 394855125.0, + "step": 15269 + }, + { + "epoch": 1.6769163189106084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.692473888397217, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6844412684440613, + "num_tokens": 394883159.0, + "step": 15270 + }, + { + "epoch": 1.6770261366132222, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5362374782562256, + "learning_rate": 1e-06, + "loss": 1.1057, + "mean_token_accuracy": 0.6899703145027161, + "num_tokens": 394909074.0, + "step": 15271 + }, + { + "epoch": 1.6771359543158357, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6473779678344727, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7039094567298889, + "num_tokens": 394932896.0, + "step": 15272 + }, + { + "epoch": 1.6772457720184493, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7313201427459717, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6921442747116089, + "num_tokens": 394957311.0, + "step": 15273 + }, + { + "epoch": 1.677355589721063, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8065407276153564, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6913111805915833, + "num_tokens": 394981625.0, + "step": 15274 + }, + { + "epoch": 1.6774654074236768, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.762735366821289, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7026636004447937, + "num_tokens": 395011025.0, + "step": 15275 + }, + { + "epoch": 1.6775752251262903, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8415422439575195, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7044025659561157, + "num_tokens": 395031310.0, + "step": 15276 + }, + { + "epoch": 1.677685042828904, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2627549171447754, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6926891207695007, + "num_tokens": 395064109.0, + "step": 15277 + }, + { + "epoch": 1.6777948605315176, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5727341175079346, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.695810854434967, + "num_tokens": 395092223.0, + "step": 15278 + }, + { + "epoch": 1.6779046782341314, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.463747978210449, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7055749893188477, + "num_tokens": 395121767.0, + "step": 15279 + }, + { + "epoch": 1.6780144959367451, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.619018077850342, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7270209789276123, + "num_tokens": 395145604.0, + "step": 15280 + }, + { + "epoch": 1.6781243136393587, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.579834222793579, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7059134244918823, + "num_tokens": 395172840.0, + "step": 15281 + }, + { + "epoch": 1.6782341313419722, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9061596393585205, + "learning_rate": 1e-06, + "loss": 0.8322, + "mean_token_accuracy": 0.7581879496574402, + "num_tokens": 395192119.0, + "step": 15282 + }, + { + "epoch": 1.678343949044586, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4617018699645996, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6932803988456726, + "num_tokens": 395220787.0, + "step": 15283 + }, + { + "epoch": 1.6784537667471997, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6504199504852295, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6828292012214661, + "num_tokens": 395246050.0, + "step": 15284 + }, + { + "epoch": 1.6785635844498135, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7352054119110107, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7133444547653198, + "num_tokens": 395267040.0, + "step": 15285 + }, + { + "epoch": 1.678673402152427, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.471104383468628, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7009637355804443, + "num_tokens": 395293994.0, + "step": 15286 + }, + { + "epoch": 1.6787832198550405, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.724970817565918, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.709648609161377, + "num_tokens": 395319169.0, + "step": 15287 + }, + { + "epoch": 1.6788930375576543, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.775615930557251, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.700325071811676, + "num_tokens": 395341689.0, + "step": 15288 + }, + { + "epoch": 1.679002855260268, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.585606336593628, + "learning_rate": 1e-06, + "loss": 1.0839, + "mean_token_accuracy": 0.6884270906448364, + "num_tokens": 395369341.0, + "step": 15289 + }, + { + "epoch": 1.6791126729628816, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7570574283599854, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7048683166503906, + "num_tokens": 395391825.0, + "step": 15290 + }, + { + "epoch": 1.6792224906654951, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.622830390930176, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7022053003311157, + "num_tokens": 395416109.0, + "step": 15291 + }, + { + "epoch": 1.6793323083681089, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6647818088531494, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6892426609992981, + "num_tokens": 395440664.0, + "step": 15292 + }, + { + "epoch": 1.6794421260707226, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6770455837249756, + "learning_rate": 1e-06, + "loss": 0.9234, + "mean_token_accuracy": 0.7283822894096375, + "num_tokens": 395460977.0, + "step": 15293 + }, + { + "epoch": 1.6795519437733364, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8080837726593018, + "learning_rate": 1e-06, + "loss": 1.122, + "mean_token_accuracy": 0.6822906732559204, + "num_tokens": 395486277.0, + "step": 15294 + }, + { + "epoch": 1.67966176147595, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8002877235412598, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7096861600875854, + "num_tokens": 395511382.0, + "step": 15295 + }, + { + "epoch": 1.6797715791785635, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6194510459899902, + "learning_rate": 1e-06, + "loss": 0.927, + "mean_token_accuracy": 0.7252806425094604, + "num_tokens": 395535975.0, + "step": 15296 + }, + { + "epoch": 1.6798813968811772, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5784566402435303, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6953129768371582, + "num_tokens": 395561926.0, + "step": 15297 + }, + { + "epoch": 1.679991214583791, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5273332595825195, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6823439598083496, + "num_tokens": 395589193.0, + "step": 15298 + }, + { + "epoch": 1.6801010322864047, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3241002559661865, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6954538822174072, + "num_tokens": 395622093.0, + "step": 15299 + }, + { + "epoch": 1.6802108499890183, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6369729042053223, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7238848209381104, + "num_tokens": 395649011.0, + "step": 15300 + }, + { + "epoch": 1.6803206676916318, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6425700187683105, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7146884202957153, + "num_tokens": 395673215.0, + "step": 15301 + }, + { + "epoch": 1.6804304853942456, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.558561325073242, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6928613185882568, + "num_tokens": 395699330.0, + "step": 15302 + }, + { + "epoch": 1.6805403030968593, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6889054775238037, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7145259380340576, + "num_tokens": 395723199.0, + "step": 15303 + }, + { + "epoch": 1.6806501207994728, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6272714138031006, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.6975337862968445, + "num_tokens": 395748689.0, + "step": 15304 + }, + { + "epoch": 1.6807599385020864, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7486393451690674, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7275311946868896, + "num_tokens": 395775038.0, + "step": 15305 + }, + { + "epoch": 1.6808697562047001, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.466470956802368, + "learning_rate": 1e-06, + "loss": 0.906, + "mean_token_accuracy": 0.7347927689552307, + "num_tokens": 395803491.0, + "step": 15306 + }, + { + "epoch": 1.680979573907314, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.42633318901062, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7194117903709412, + "num_tokens": 395829483.0, + "step": 15307 + }, + { + "epoch": 1.6810893916099277, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7921090126037598, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7046854496002197, + "num_tokens": 395849479.0, + "step": 15308 + }, + { + "epoch": 1.6811992093125412, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5201990604400635, + "learning_rate": 1e-06, + "loss": 1.1235, + "mean_token_accuracy": 0.6799207925796509, + "num_tokens": 395877741.0, + "step": 15309 + }, + { + "epoch": 1.6813090270151547, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6278796195983887, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7173097729682922, + "num_tokens": 395900303.0, + "step": 15310 + }, + { + "epoch": 1.6814188447177685, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.483487129211426, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7042665481567383, + "num_tokens": 395928543.0, + "step": 15311 + }, + { + "epoch": 1.6815286624203822, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.650878429412842, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7144132852554321, + "num_tokens": 395954915.0, + "step": 15312 + }, + { + "epoch": 1.681638480122996, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7113850116729736, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7342020273208618, + "num_tokens": 395977319.0, + "step": 15313 + }, + { + "epoch": 1.6817482978256095, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.626085042953491, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7022746801376343, + "num_tokens": 396000471.0, + "step": 15314 + }, + { + "epoch": 1.681858115528223, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.400156021118164, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.7004626393318176, + "num_tokens": 396033624.0, + "step": 15315 + }, + { + "epoch": 1.6819679332308368, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.620210647583008, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7061915397644043, + "num_tokens": 396060174.0, + "step": 15316 + }, + { + "epoch": 1.6820777509334506, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.797891855239868, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6980962753295898, + "num_tokens": 396081819.0, + "step": 15317 + }, + { + "epoch": 1.6821875686360641, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.51689076423645, + "learning_rate": 1e-06, + "loss": 1.163, + "mean_token_accuracy": 0.6669135093688965, + "num_tokens": 396112089.0, + "step": 15318 + }, + { + "epoch": 1.6822973863386776, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8571107387542725, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7150046825408936, + "num_tokens": 396134208.0, + "step": 15319 + }, + { + "epoch": 1.6824072040412914, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6321420669555664, + "learning_rate": 1e-06, + "loss": 1.1268, + "mean_token_accuracy": 0.672875165939331, + "num_tokens": 396165284.0, + "step": 15320 + }, + { + "epoch": 1.6825170217439052, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8005752563476562, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7034690976142883, + "num_tokens": 396185911.0, + "step": 15321 + }, + { + "epoch": 1.682626839446519, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7619147300720215, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7083379030227661, + "num_tokens": 396208581.0, + "step": 15322 + }, + { + "epoch": 1.6827366571491325, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4628498554229736, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.7252782583236694, + "num_tokens": 396235996.0, + "step": 15323 + }, + { + "epoch": 1.682846474851746, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.625248670578003, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6949988603591919, + "num_tokens": 396259865.0, + "step": 15324 + }, + { + "epoch": 1.6829562925543597, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4202282428741455, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6999074220657349, + "num_tokens": 396288236.0, + "step": 15325 + }, + { + "epoch": 1.6830661102569735, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.822981834411621, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7150667905807495, + "num_tokens": 396310576.0, + "step": 15326 + }, + { + "epoch": 1.683175927959587, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.677316904067993, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7038468718528748, + "num_tokens": 396334790.0, + "step": 15327 + }, + { + "epoch": 1.6832857456622008, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4569931030273438, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6986852884292603, + "num_tokens": 396362040.0, + "step": 15328 + }, + { + "epoch": 1.6833955633648143, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.781536817550659, + "learning_rate": 1e-06, + "loss": 1.1152, + "mean_token_accuracy": 0.6795020699501038, + "num_tokens": 396385734.0, + "step": 15329 + }, + { + "epoch": 1.683505381067428, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9274868965148926, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7107366323471069, + "num_tokens": 396410754.0, + "step": 15330 + }, + { + "epoch": 1.6836151987700418, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5444259643554688, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6837717294692993, + "num_tokens": 396440831.0, + "step": 15331 + }, + { + "epoch": 1.6837250164726554, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6506803035736084, + "learning_rate": 1e-06, + "loss": 1.0745, + "mean_token_accuracy": 0.6839022040367126, + "num_tokens": 396466952.0, + "step": 15332 + }, + { + "epoch": 1.683834834175269, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5112550258636475, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7084269523620605, + "num_tokens": 396496122.0, + "step": 15333 + }, + { + "epoch": 1.6839446518778827, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5474441051483154, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7058241367340088, + "num_tokens": 396518979.0, + "step": 15334 + }, + { + "epoch": 1.6840544695804964, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5150675773620605, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7015995979309082, + "num_tokens": 396549076.0, + "step": 15335 + }, + { + "epoch": 1.6841642872831102, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.473592519760132, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6978864669799805, + "num_tokens": 396579410.0, + "step": 15336 + }, + { + "epoch": 1.6842741049857237, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.577249526977539, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7103033065795898, + "num_tokens": 396605417.0, + "step": 15337 + }, + { + "epoch": 1.6843839226883373, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6265628337860107, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7095506191253662, + "num_tokens": 396629173.0, + "step": 15338 + }, + { + "epoch": 1.684493740390951, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 4.078454494476318, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.7000184655189514, + "num_tokens": 396657129.0, + "step": 15339 + }, + { + "epoch": 1.6846035580935648, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.865999221801758, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7239648699760437, + "num_tokens": 396678250.0, + "step": 15340 + }, + { + "epoch": 1.6847133757961783, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6276376247406006, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7222320437431335, + "num_tokens": 396702668.0, + "step": 15341 + }, + { + "epoch": 1.684823193498792, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.614057779312134, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7129297852516174, + "num_tokens": 396727824.0, + "step": 15342 + }, + { + "epoch": 1.6849330112014056, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.347034215927124, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7040871381759644, + "num_tokens": 396756753.0, + "step": 15343 + }, + { + "epoch": 1.6850428289040194, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.692880630493164, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7074128985404968, + "num_tokens": 396780860.0, + "step": 15344 + }, + { + "epoch": 1.685152646606633, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7707700729370117, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6902850866317749, + "num_tokens": 396804809.0, + "step": 15345 + }, + { + "epoch": 1.6852624643092466, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8684916496276855, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.7265583276748657, + "num_tokens": 396827140.0, + "step": 15346 + }, + { + "epoch": 1.6853722820118602, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.684285879135132, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7021328210830688, + "num_tokens": 396851950.0, + "step": 15347 + }, + { + "epoch": 1.685482099714474, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.577378034591675, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7202926278114319, + "num_tokens": 396876087.0, + "step": 15348 + }, + { + "epoch": 1.6855919174170877, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3767950534820557, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6971954107284546, + "num_tokens": 396905161.0, + "step": 15349 + }, + { + "epoch": 1.6857017351197014, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.740802526473999, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7067931890487671, + "num_tokens": 396928353.0, + "step": 15350 + }, + { + "epoch": 1.685811552822315, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6035985946655273, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7111306190490723, + "num_tokens": 396952381.0, + "step": 15351 + }, + { + "epoch": 1.6859213705249285, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.570890188217163, + "learning_rate": 1e-06, + "loss": 1.1249, + "mean_token_accuracy": 0.6762370467185974, + "num_tokens": 396981463.0, + "step": 15352 + }, + { + "epoch": 1.6860311882275423, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.81447696685791, + "learning_rate": 1e-06, + "loss": 1.1379, + "mean_token_accuracy": 0.6783449649810791, + "num_tokens": 397008751.0, + "step": 15353 + }, + { + "epoch": 1.686141005930156, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7453219890594482, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.707685112953186, + "num_tokens": 397030367.0, + "step": 15354 + }, + { + "epoch": 1.6862508236327696, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3767924308776855, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.703328549861908, + "num_tokens": 397063919.0, + "step": 15355 + }, + { + "epoch": 1.686360641335383, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7812018394470215, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7099262475967407, + "num_tokens": 397090232.0, + "step": 15356 + }, + { + "epoch": 1.6864704590379969, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6946942806243896, + "learning_rate": 1e-06, + "loss": 0.938, + "mean_token_accuracy": 0.7188488841056824, + "num_tokens": 397111520.0, + "step": 15357 + }, + { + "epoch": 1.6865802767406106, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.384836435317993, + "learning_rate": 1e-06, + "loss": 1.0999, + "mean_token_accuracy": 0.6896803379058838, + "num_tokens": 397144790.0, + "step": 15358 + }, + { + "epoch": 1.6866900944432244, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4733715057373047, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.7045893669128418, + "num_tokens": 397171378.0, + "step": 15359 + }, + { + "epoch": 1.686799912145838, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6233887672424316, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7122603058815002, + "num_tokens": 397192336.0, + "step": 15360 + }, + { + "epoch": 1.6869097298484514, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6018006801605225, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7277751564979553, + "num_tokens": 397217011.0, + "step": 15361 + }, + { + "epoch": 1.6870195475510652, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.862363576889038, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6958687901496887, + "num_tokens": 397236937.0, + "step": 15362 + }, + { + "epoch": 1.687129365253679, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6177988052368164, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7004088163375854, + "num_tokens": 397262375.0, + "step": 15363 + }, + { + "epoch": 1.6872391829562927, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.47304368019104, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7174095511436462, + "num_tokens": 397287748.0, + "step": 15364 + }, + { + "epoch": 1.6873490006589063, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.287050485610962, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6886391639709473, + "num_tokens": 397318339.0, + "step": 15365 + }, + { + "epoch": 1.6874588183615198, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.836418390274048, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7263637781143188, + "num_tokens": 397339436.0, + "step": 15366 + }, + { + "epoch": 1.6875686360641335, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5251824855804443, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7183907628059387, + "num_tokens": 397364289.0, + "step": 15367 + }, + { + "epoch": 1.6876784537667473, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.550269842147827, + "learning_rate": 1e-06, + "loss": 1.1274, + "mean_token_accuracy": 0.6795666217803955, + "num_tokens": 397390891.0, + "step": 15368 + }, + { + "epoch": 1.6877882714693608, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5433285236358643, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7117234468460083, + "num_tokens": 397415729.0, + "step": 15369 + }, + { + "epoch": 1.6878980891719744, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5583086013793945, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.701105535030365, + "num_tokens": 397444613.0, + "step": 15370 + }, + { + "epoch": 1.6880079068745881, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5055086612701416, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.7090823650360107, + "num_tokens": 397470442.0, + "step": 15371 + }, + { + "epoch": 1.6881177245772019, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5519063472747803, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7192173004150391, + "num_tokens": 397493162.0, + "step": 15372 + }, + { + "epoch": 1.6882275422798156, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.827730178833008, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6891621947288513, + "num_tokens": 397527873.0, + "step": 15373 + }, + { + "epoch": 1.6883373599824292, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5765676498413086, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.7225469946861267, + "num_tokens": 397551717.0, + "step": 15374 + }, + { + "epoch": 1.6884471776850427, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6532211303710938, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.710791826248169, + "num_tokens": 397576725.0, + "step": 15375 + }, + { + "epoch": 1.6885569953876565, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.668070077896118, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7092974185943604, + "num_tokens": 397597859.0, + "step": 15376 + }, + { + "epoch": 1.6886668130902702, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7159125804901123, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.7026394605636597, + "num_tokens": 397620983.0, + "step": 15377 + }, + { + "epoch": 1.6887766307928838, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3062520027160645, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7133133411407471, + "num_tokens": 397654202.0, + "step": 15378 + }, + { + "epoch": 1.6888864484954975, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.336560010910034, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7207899689674377, + "num_tokens": 397684205.0, + "step": 15379 + }, + { + "epoch": 1.688996266198111, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.482330322265625, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.709877073764801, + "num_tokens": 397709024.0, + "step": 15380 + }, + { + "epoch": 1.6891060839007248, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.497933864593506, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7213661074638367, + "num_tokens": 397734698.0, + "step": 15381 + }, + { + "epoch": 1.6892159016033386, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 80.54350280761719, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.710795521736145, + "num_tokens": 397757587.0, + "step": 15382 + }, + { + "epoch": 1.689325719305952, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0953032970428467, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7225322723388672, + "num_tokens": 397776130.0, + "step": 15383 + }, + { + "epoch": 1.6894355370085656, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.654500961303711, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.709046483039856, + "num_tokens": 397801847.0, + "step": 15384 + }, + { + "epoch": 1.6895453547111794, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.518800973892212, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7167348861694336, + "num_tokens": 397830107.0, + "step": 15385 + }, + { + "epoch": 1.6896551724137931, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0668399333953857, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6953728199005127, + "num_tokens": 397856490.0, + "step": 15386 + }, + { + "epoch": 1.689764990116407, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.563870429992676, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6976544857025146, + "num_tokens": 397886156.0, + "step": 15387 + }, + { + "epoch": 1.6898748078190204, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.624748706817627, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7231736779212952, + "num_tokens": 397910565.0, + "step": 15388 + }, + { + "epoch": 1.689984625521634, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8143131732940674, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7281118631362915, + "num_tokens": 397931254.0, + "step": 15389 + }, + { + "epoch": 1.6900944432242477, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4898014068603516, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7068353891372681, + "num_tokens": 397960044.0, + "step": 15390 + }, + { + "epoch": 1.6902042609268615, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.536550760269165, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6917137503623962, + "num_tokens": 397988078.0, + "step": 15391 + }, + { + "epoch": 1.690314078629475, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5573465824127197, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6901067495346069, + "num_tokens": 398014185.0, + "step": 15392 + }, + { + "epoch": 1.6904238963320888, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.45535945892334, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7063106894493103, + "num_tokens": 398041080.0, + "step": 15393 + }, + { + "epoch": 1.6905337140347023, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5635108947753906, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7125030755996704, + "num_tokens": 398066396.0, + "step": 15394 + }, + { + "epoch": 1.690643531737316, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.82515811920166, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7087501883506775, + "num_tokens": 398091231.0, + "step": 15395 + }, + { + "epoch": 1.6907533494399298, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.821683406829834, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7075578570365906, + "num_tokens": 398113988.0, + "step": 15396 + }, + { + "epoch": 1.6908631671425434, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6076793670654297, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6919755935668945, + "num_tokens": 398139097.0, + "step": 15397 + }, + { + "epoch": 1.690972984845157, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5511345863342285, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7181769609451294, + "num_tokens": 398166310.0, + "step": 15398 + }, + { + "epoch": 1.6910828025477707, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6733760833740234, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7188339233398438, + "num_tokens": 398188479.0, + "step": 15399 + }, + { + "epoch": 1.6911926202503844, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4905574321746826, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7097076773643494, + "num_tokens": 398214429.0, + "step": 15400 + }, + { + "epoch": 1.6913024379529982, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.390235185623169, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6995812654495239, + "num_tokens": 398246406.0, + "step": 15401 + }, + { + "epoch": 1.6914122556556117, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.471351146697998, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7034800052642822, + "num_tokens": 398273894.0, + "step": 15402 + }, + { + "epoch": 1.6915220733582252, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.794114828109741, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7101070880889893, + "num_tokens": 398295936.0, + "step": 15403 + }, + { + "epoch": 1.691631891060839, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7636165618896484, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.699371337890625, + "num_tokens": 398322470.0, + "step": 15404 + }, + { + "epoch": 1.6917417087634528, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.63431978225708, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6901394724845886, + "num_tokens": 398347501.0, + "step": 15405 + }, + { + "epoch": 1.6918515264660663, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.42126202583313, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.7011992335319519, + "num_tokens": 398377375.0, + "step": 15406 + }, + { + "epoch": 1.6919613441686798, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7331151962280273, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7154819965362549, + "num_tokens": 398397998.0, + "step": 15407 + }, + { + "epoch": 1.6920711618712936, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.217428207397461, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.6957917213439941, + "num_tokens": 398433069.0, + "step": 15408 + }, + { + "epoch": 1.6921809795739073, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0085203647613525, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.717832088470459, + "num_tokens": 398455663.0, + "step": 15409 + }, + { + "epoch": 1.692290797276521, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8145029544830322, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7057950496673584, + "num_tokens": 398478737.0, + "step": 15410 + }, + { + "epoch": 1.6924006149791346, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4467084407806396, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7144982814788818, + "num_tokens": 398506106.0, + "step": 15411 + }, + { + "epoch": 1.6925104326817482, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7259109020233154, + "learning_rate": 1e-06, + "loss": 0.8976, + "mean_token_accuracy": 0.7359262704849243, + "num_tokens": 398526706.0, + "step": 15412 + }, + { + "epoch": 1.692620250384362, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3840551376342773, + "learning_rate": 1e-06, + "loss": 1.1159, + "mean_token_accuracy": 0.6780753135681152, + "num_tokens": 398558623.0, + "step": 15413 + }, + { + "epoch": 1.6927300680869757, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.765225410461426, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7206965684890747, + "num_tokens": 398580334.0, + "step": 15414 + }, + { + "epoch": 1.6928398857895894, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8596079349517822, + "learning_rate": 1e-06, + "loss": 0.8928, + "mean_token_accuracy": 0.7397656440734863, + "num_tokens": 398601379.0, + "step": 15415 + }, + { + "epoch": 1.692949703492203, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.725841760635376, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.7278356552124023, + "num_tokens": 398629504.0, + "step": 15416 + }, + { + "epoch": 1.6930595211948165, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8140017986297607, + "learning_rate": 1e-06, + "loss": 1.098, + "mean_token_accuracy": 0.6806539297103882, + "num_tokens": 398655967.0, + "step": 15417 + }, + { + "epoch": 1.6931693388974303, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7286415100097656, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7007877826690674, + "num_tokens": 398677510.0, + "step": 15418 + }, + { + "epoch": 1.693279156600044, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6148691177368164, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6987901926040649, + "num_tokens": 398705020.0, + "step": 15419 + }, + { + "epoch": 1.6933889743026576, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9672892093658447, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7054128050804138, + "num_tokens": 398730620.0, + "step": 15420 + }, + { + "epoch": 1.693498792005271, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6442930698394775, + "learning_rate": 1e-06, + "loss": 0.9134, + "mean_token_accuracy": 0.7307363152503967, + "num_tokens": 398752590.0, + "step": 15421 + }, + { + "epoch": 1.6936086097078848, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 16.6875, + "learning_rate": 1e-06, + "loss": 0.9034, + "mean_token_accuracy": 0.7322227358818054, + "num_tokens": 398775873.0, + "step": 15422 + }, + { + "epoch": 1.6937184274104986, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7304654121398926, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7140015363693237, + "num_tokens": 398798566.0, + "step": 15423 + }, + { + "epoch": 1.6938282451131124, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6901817321777344, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7030123472213745, + "num_tokens": 398822146.0, + "step": 15424 + }, + { + "epoch": 1.693938062815726, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5215604305267334, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.6841336488723755, + "num_tokens": 398851007.0, + "step": 15425 + }, + { + "epoch": 1.6940478805183394, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8201210498809814, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7124828100204468, + "num_tokens": 398874032.0, + "step": 15426 + }, + { + "epoch": 1.6941576982209532, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.325436592102051, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7026008367538452, + "num_tokens": 398907392.0, + "step": 15427 + }, + { + "epoch": 1.694267515923567, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6717689037323, + "learning_rate": 1e-06, + "loss": 1.1028, + "mean_token_accuracy": 0.692111611366272, + "num_tokens": 398932488.0, + "step": 15428 + }, + { + "epoch": 1.6943773336261807, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.745607376098633, + "learning_rate": 1e-06, + "loss": 0.9203, + "mean_token_accuracy": 0.7278692126274109, + "num_tokens": 398955318.0, + "step": 15429 + }, + { + "epoch": 1.6944871513287942, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.138040781021118, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7256249189376831, + "num_tokens": 398973510.0, + "step": 15430 + }, + { + "epoch": 1.6945969690314078, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.563716411590576, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7179957628250122, + "num_tokens": 398998123.0, + "step": 15431 + }, + { + "epoch": 1.6947067867340215, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7870655059814453, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7150975465774536, + "num_tokens": 399021861.0, + "step": 15432 + }, + { + "epoch": 1.6948166044366353, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9084601402282715, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7222325801849365, + "num_tokens": 399042557.0, + "step": 15433 + }, + { + "epoch": 1.6949264221392488, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4129812717437744, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7179455757141113, + "num_tokens": 399072586.0, + "step": 15434 + }, + { + "epoch": 1.6950362398418624, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5720205307006836, + "learning_rate": 1e-06, + "loss": 1.1039, + "mean_token_accuracy": 0.6868361234664917, + "num_tokens": 399099061.0, + "step": 15435 + }, + { + "epoch": 1.6951460575444761, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.026705741882324, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7069186568260193, + "num_tokens": 399117783.0, + "step": 15436 + }, + { + "epoch": 1.6952558752470899, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0416152477264404, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7078371047973633, + "num_tokens": 399141358.0, + "step": 15437 + }, + { + "epoch": 1.6953656929497036, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.751551866531372, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6833392381668091, + "num_tokens": 399169370.0, + "step": 15438 + }, + { + "epoch": 1.6954755106523172, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.512148380279541, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7128050923347473, + "num_tokens": 399197959.0, + "step": 15439 + }, + { + "epoch": 1.6955853283549307, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.911017894744873, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.7031794786453247, + "num_tokens": 399220990.0, + "step": 15440 + }, + { + "epoch": 1.6956951460575445, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.538881540298462, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.69507896900177, + "num_tokens": 399246665.0, + "step": 15441 + }, + { + "epoch": 1.6958049637601582, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4019572734832764, + "learning_rate": 1e-06, + "loss": 1.1084, + "mean_token_accuracy": 0.6812715530395508, + "num_tokens": 399276337.0, + "step": 15442 + }, + { + "epoch": 1.6959147814627717, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.695526123046875, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7163140773773193, + "num_tokens": 399298346.0, + "step": 15443 + }, + { + "epoch": 1.6960245991653855, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.793085813522339, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.697659432888031, + "num_tokens": 399323963.0, + "step": 15444 + }, + { + "epoch": 1.696134416867999, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.246022939682007, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7101933360099792, + "num_tokens": 399344041.0, + "step": 15445 + }, + { + "epoch": 1.6962442345706128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8095407485961914, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7029843330383301, + "num_tokens": 399365154.0, + "step": 15446 + }, + { + "epoch": 1.6963540522732266, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6141626834869385, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7138248085975647, + "num_tokens": 399390134.0, + "step": 15447 + }, + { + "epoch": 1.69646386997584, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5457634925842285, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7095052003860474, + "num_tokens": 399414294.0, + "step": 15448 + }, + { + "epoch": 1.6965736876784536, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 16.697294235229492, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.6799725294113159, + "num_tokens": 399442632.0, + "step": 15449 + }, + { + "epoch": 1.6966835053810674, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6081161499023438, + "learning_rate": 1e-06, + "loss": 1.1091, + "mean_token_accuracy": 0.6745064854621887, + "num_tokens": 399471270.0, + "step": 15450 + }, + { + "epoch": 1.6967933230836811, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.792672634124756, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7105216979980469, + "num_tokens": 399495752.0, + "step": 15451 + }, + { + "epoch": 1.696903140786295, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4651012420654297, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.6973849534988403, + "num_tokens": 399523581.0, + "step": 15452 + }, + { + "epoch": 1.6970129584889084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.470615863800049, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6948161125183105, + "num_tokens": 399550092.0, + "step": 15453 + }, + { + "epoch": 1.697122776191522, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.277068853378296, + "learning_rate": 1e-06, + "loss": 0.9326, + "mean_token_accuracy": 0.7238207459449768, + "num_tokens": 399579278.0, + "step": 15454 + }, + { + "epoch": 1.6972325938941357, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2372031211853027, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7084327340126038, + "num_tokens": 399610000.0, + "step": 15455 + }, + { + "epoch": 1.6973424115967495, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1279780864715576, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7242201566696167, + "num_tokens": 399635600.0, + "step": 15456 + }, + { + "epoch": 1.697452229299363, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5493950843811035, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.7059570550918579, + "num_tokens": 399660664.0, + "step": 15457 + }, + { + "epoch": 1.6975620470019768, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5407724380493164, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6920155882835388, + "num_tokens": 399687216.0, + "step": 15458 + }, + { + "epoch": 1.6976718647045903, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5940659046173096, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6876298189163208, + "num_tokens": 399711844.0, + "step": 15459 + }, + { + "epoch": 1.697781682407204, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4044837951660156, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6811444163322449, + "num_tokens": 399743060.0, + "step": 15460 + }, + { + "epoch": 1.6978915001098178, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6817305088043213, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.693534791469574, + "num_tokens": 399769595.0, + "step": 15461 + }, + { + "epoch": 1.6980013178124314, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0074145793914795, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7038505673408508, + "num_tokens": 399794162.0, + "step": 15462 + }, + { + "epoch": 1.698111135515045, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6677024364471436, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6861188411712646, + "num_tokens": 399820012.0, + "step": 15463 + }, + { + "epoch": 1.6982209532176586, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2002134323120117, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6986625790596008, + "num_tokens": 399846566.0, + "step": 15464 + }, + { + "epoch": 1.6983307709202724, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.533330202102661, + "learning_rate": 1e-06, + "loss": 0.9051, + "mean_token_accuracy": 0.7287206649780273, + "num_tokens": 399872193.0, + "step": 15465 + }, + { + "epoch": 1.6984405886228862, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.525487184524536, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.701943039894104, + "num_tokens": 399897730.0, + "step": 15466 + }, + { + "epoch": 1.6985504063254997, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.528088331222534, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6941428780555725, + "num_tokens": 399926421.0, + "step": 15467 + }, + { + "epoch": 1.6986602240281132, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.798879623413086, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.709406852722168, + "num_tokens": 399946819.0, + "step": 15468 + }, + { + "epoch": 1.698770041730727, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0412485599517822, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7138198614120483, + "num_tokens": 399970656.0, + "step": 15469 + }, + { + "epoch": 1.6988798594333407, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.383794069290161, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7077075242996216, + "num_tokens": 399998962.0, + "step": 15470 + }, + { + "epoch": 1.6989896771359543, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.396721124649048, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7090807557106018, + "num_tokens": 400029421.0, + "step": 15471 + }, + { + "epoch": 1.6990994948385678, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8213820457458496, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6916214823722839, + "num_tokens": 400053607.0, + "step": 15472 + }, + { + "epoch": 1.6992093125411816, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5261738300323486, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7174521684646606, + "num_tokens": 400080343.0, + "step": 15473 + }, + { + "epoch": 1.6993191302437953, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.492215871810913, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6936027407646179, + "num_tokens": 400109965.0, + "step": 15474 + }, + { + "epoch": 1.699428947946409, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.673051595687866, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7114739418029785, + "num_tokens": 400132559.0, + "step": 15475 + }, + { + "epoch": 1.6995387656490226, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3889880180358887, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7119133472442627, + "num_tokens": 400159520.0, + "step": 15476 + }, + { + "epoch": 1.6996485833516362, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.706221580505371, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6975898146629333, + "num_tokens": 400182574.0, + "step": 15477 + }, + { + "epoch": 1.69975840105425, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4385907649993896, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6959179639816284, + "num_tokens": 400212425.0, + "step": 15478 + }, + { + "epoch": 1.6998682187568637, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7301995754241943, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7174085378646851, + "num_tokens": 400234508.0, + "step": 15479 + }, + { + "epoch": 1.6999780364594774, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4315974712371826, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7140446901321411, + "num_tokens": 400261336.0, + "step": 15480 + }, + { + "epoch": 1.700087854162091, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6910617351531982, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7046923637390137, + "num_tokens": 400284616.0, + "step": 15481 + }, + { + "epoch": 1.7001976718647045, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.907194137573242, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.6938485503196716, + "num_tokens": 400311134.0, + "step": 15482 + }, + { + "epoch": 1.7003074895673183, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7900359630584717, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7139061689376831, + "num_tokens": 400333231.0, + "step": 15483 + }, + { + "epoch": 1.700417307269932, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.425340175628662, + "learning_rate": 1e-06, + "loss": 1.091, + "mean_token_accuracy": 0.6809878945350647, + "num_tokens": 400366946.0, + "step": 15484 + }, + { + "epoch": 1.7005271249725455, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6207311153411865, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7101833820343018, + "num_tokens": 400391020.0, + "step": 15485 + }, + { + "epoch": 1.700636942675159, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4270901679992676, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6985164284706116, + "num_tokens": 400420854.0, + "step": 15486 + }, + { + "epoch": 1.7007467603777728, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.547382116317749, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7133890986442566, + "num_tokens": 400445400.0, + "step": 15487 + }, + { + "epoch": 1.7008565780803866, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5711445808410645, + "learning_rate": 1e-06, + "loss": 1.0883, + "mean_token_accuracy": 0.6902470588684082, + "num_tokens": 400472258.0, + "step": 15488 + }, + { + "epoch": 1.7009663957830004, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.714747428894043, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7141755819320679, + "num_tokens": 400495777.0, + "step": 15489 + }, + { + "epoch": 1.7010762134856139, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8187496662139893, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.696264386177063, + "num_tokens": 400523789.0, + "step": 15490 + }, + { + "epoch": 1.7011860311882274, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.839276075363159, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7164808511734009, + "num_tokens": 400548564.0, + "step": 15491 + }, + { + "epoch": 1.7012958488908412, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6160056591033936, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6946811676025391, + "num_tokens": 400572430.0, + "step": 15492 + }, + { + "epoch": 1.701405666593455, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.516718626022339, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7082428932189941, + "num_tokens": 400600268.0, + "step": 15493 + }, + { + "epoch": 1.7015154842960687, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.739680290222168, + "learning_rate": 1e-06, + "loss": 0.935, + "mean_token_accuracy": 0.7247247695922852, + "num_tokens": 400622347.0, + "step": 15494 + }, + { + "epoch": 1.7016253019986822, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.493922233581543, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7182400226593018, + "num_tokens": 400640295.0, + "step": 15495 + }, + { + "epoch": 1.7017351197012958, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0461559295654297, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7051934003829956, + "num_tokens": 400657890.0, + "step": 15496 + }, + { + "epoch": 1.7018449374039095, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6132616996765137, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.7077049016952515, + "num_tokens": 400682174.0, + "step": 15497 + }, + { + "epoch": 1.7019547551065233, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6870431900024414, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7117801308631897, + "num_tokens": 400707258.0, + "step": 15498 + }, + { + "epoch": 1.7020645728091368, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4866104125976562, + "learning_rate": 1e-06, + "loss": 1.0965, + "mean_token_accuracy": 0.6800503730773926, + "num_tokens": 400736267.0, + "step": 15499 + }, + { + "epoch": 1.7021743905117503, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8563120365142822, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7035517692565918, + "num_tokens": 400758113.0, + "step": 15500 + }, + { + "epoch": 1.702284208214364, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9084742069244385, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7220285534858704, + "num_tokens": 400779398.0, + "step": 15501 + }, + { + "epoch": 1.7023940259169779, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.988107919692993, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7157611846923828, + "num_tokens": 400803902.0, + "step": 15502 + }, + { + "epoch": 1.7025038436195916, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6951706409454346, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7051408290863037, + "num_tokens": 400828366.0, + "step": 15503 + }, + { + "epoch": 1.7026136613222052, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5488860607147217, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.6946857571601868, + "num_tokens": 400857298.0, + "step": 15504 + }, + { + "epoch": 1.7027234790248187, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8322665691375732, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7132615447044373, + "num_tokens": 400878272.0, + "step": 15505 + }, + { + "epoch": 1.7028332967274324, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6480374336242676, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.7035834789276123, + "num_tokens": 400904791.0, + "step": 15506 + }, + { + "epoch": 1.7029431144300462, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4992563724517822, + "learning_rate": 1e-06, + "loss": 0.9143, + "mean_token_accuracy": 0.7290487885475159, + "num_tokens": 400931359.0, + "step": 15507 + }, + { + "epoch": 1.7030529321326597, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.154538631439209, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6950192451477051, + "num_tokens": 400968802.0, + "step": 15508 + }, + { + "epoch": 1.7031627498352735, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.610809564590454, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7005253434181213, + "num_tokens": 400993412.0, + "step": 15509 + }, + { + "epoch": 1.703272567537887, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.742279529571533, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7048977613449097, + "num_tokens": 401016039.0, + "step": 15510 + }, + { + "epoch": 1.7033823852405008, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6196534633636475, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7147679328918457, + "num_tokens": 401037545.0, + "step": 15511 + }, + { + "epoch": 1.7034922029431145, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9872353076934814, + "learning_rate": 1e-06, + "loss": 1.1502, + "mean_token_accuracy": 0.6684414148330688, + "num_tokens": 401064993.0, + "step": 15512 + }, + { + "epoch": 1.703602020645728, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3834869861602783, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.7142741680145264, + "num_tokens": 401091084.0, + "step": 15513 + }, + { + "epoch": 1.7037118383483416, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3233625888824463, + "learning_rate": 1e-06, + "loss": 1.1185, + "mean_token_accuracy": 0.6825544238090515, + "num_tokens": 401122728.0, + "step": 15514 + }, + { + "epoch": 1.7038216560509554, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8937437534332275, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7031419277191162, + "num_tokens": 401142675.0, + "step": 15515 + }, + { + "epoch": 1.7039314737535691, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.538552761077881, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6926395297050476, + "num_tokens": 401170929.0, + "step": 15516 + }, + { + "epoch": 1.7040412914561829, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7079241275787354, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6860722899436951, + "num_tokens": 401198278.0, + "step": 15517 + }, + { + "epoch": 1.7041511091587964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.411125659942627, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7080520391464233, + "num_tokens": 401225270.0, + "step": 15518 + }, + { + "epoch": 1.70426092686141, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.470431089401245, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7262310981750488, + "num_tokens": 401249881.0, + "step": 15519 + }, + { + "epoch": 1.7043707445640237, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6848247051239014, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7109590768814087, + "num_tokens": 401272146.0, + "step": 15520 + }, + { + "epoch": 1.7044805622666375, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9397897720336914, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7099912166595459, + "num_tokens": 401293937.0, + "step": 15521 + }, + { + "epoch": 1.704590379969251, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6044507026672363, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7238725423812866, + "num_tokens": 401317738.0, + "step": 15522 + }, + { + "epoch": 1.7047001976718648, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.525451898574829, + "learning_rate": 1e-06, + "loss": 1.1233, + "mean_token_accuracy": 0.6724132299423218, + "num_tokens": 401348387.0, + "step": 15523 + }, + { + "epoch": 1.7048100153744783, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.648698329925537, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7065381407737732, + "num_tokens": 401381340.0, + "step": 15524 + }, + { + "epoch": 1.704919833077092, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.746826171875, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7091255784034729, + "num_tokens": 401402595.0, + "step": 15525 + }, + { + "epoch": 1.7050296507797058, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.569254159927368, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7183454036712646, + "num_tokens": 401425200.0, + "step": 15526 + }, + { + "epoch": 1.7051394684823193, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7291858196258545, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6921025514602661, + "num_tokens": 401448836.0, + "step": 15527 + }, + { + "epoch": 1.7052492861849329, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.789776086807251, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7131811380386353, + "num_tokens": 401470107.0, + "step": 15528 + }, + { + "epoch": 1.7053591038875466, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.512977361679077, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6918585300445557, + "num_tokens": 401497024.0, + "step": 15529 + }, + { + "epoch": 1.7054689215901604, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3555290699005127, + "learning_rate": 1e-06, + "loss": 1.1443, + "mean_token_accuracy": 0.6673814654350281, + "num_tokens": 401529644.0, + "step": 15530 + }, + { + "epoch": 1.7055787392927741, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.618549108505249, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7025622725486755, + "num_tokens": 401553693.0, + "step": 15531 + }, + { + "epoch": 1.7056885569953877, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6157844066619873, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7026432156562805, + "num_tokens": 401577197.0, + "step": 15532 + }, + { + "epoch": 1.7057983746980012, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.388331651687622, + "learning_rate": 1e-06, + "loss": 1.0974, + "mean_token_accuracy": 0.6825733184814453, + "num_tokens": 401608064.0, + "step": 15533 + }, + { + "epoch": 1.705908192400615, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8187365531921387, + "learning_rate": 1e-06, + "loss": 0.9271, + "mean_token_accuracy": 0.727165699005127, + "num_tokens": 401628407.0, + "step": 15534 + }, + { + "epoch": 1.7060180101032287, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.471681594848633, + "learning_rate": 1e-06, + "loss": 1.1409, + "mean_token_accuracy": 0.6755051612854004, + "num_tokens": 401656515.0, + "step": 15535 + }, + { + "epoch": 1.7061278278058423, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.626211404800415, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7095028162002563, + "num_tokens": 401681436.0, + "step": 15536 + }, + { + "epoch": 1.7062376455084558, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.665510892868042, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7035031318664551, + "num_tokens": 401706904.0, + "step": 15537 + }, + { + "epoch": 1.7063474632110696, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6282007694244385, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7160787582397461, + "num_tokens": 401730977.0, + "step": 15538 + }, + { + "epoch": 1.7064572809136833, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.480358123779297, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.718710720539093, + "num_tokens": 401757877.0, + "step": 15539 + }, + { + "epoch": 1.706567098616297, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.636932849884033, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7124688625335693, + "num_tokens": 401784118.0, + "step": 15540 + }, + { + "epoch": 1.7066769163189106, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.558436870574951, + "learning_rate": 1e-06, + "loss": 0.8763, + "mean_token_accuracy": 0.7415035963058472, + "num_tokens": 401809143.0, + "step": 15541 + }, + { + "epoch": 1.7067867340215241, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7779808044433594, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.7008399367332458, + "num_tokens": 401830948.0, + "step": 15542 + }, + { + "epoch": 1.706896551724138, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7096877098083496, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7164902687072754, + "num_tokens": 401853041.0, + "step": 15543 + }, + { + "epoch": 1.7070063694267517, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.444493293762207, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6968541145324707, + "num_tokens": 401883021.0, + "step": 15544 + }, + { + "epoch": 1.7071161871293654, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.66336989402771, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.701374888420105, + "num_tokens": 401905419.0, + "step": 15545 + }, + { + "epoch": 1.707226004831979, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.48207688331604, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6900410652160645, + "num_tokens": 401934862.0, + "step": 15546 + }, + { + "epoch": 1.7073358225345925, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3874011039733887, + "learning_rate": 1e-06, + "loss": 1.1021, + "mean_token_accuracy": 0.6788779497146606, + "num_tokens": 401965412.0, + "step": 15547 + }, + { + "epoch": 1.7074456402372062, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5925018787384033, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6903078556060791, + "num_tokens": 401990219.0, + "step": 15548 + }, + { + "epoch": 1.70755545793982, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8101766109466553, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7111315727233887, + "num_tokens": 402010596.0, + "step": 15549 + }, + { + "epoch": 1.7076652756424335, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5312514305114746, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7103526592254639, + "num_tokens": 402035555.0, + "step": 15550 + }, + { + "epoch": 1.707775093345047, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.594364643096924, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6925688982009888, + "num_tokens": 402064140.0, + "step": 15551 + }, + { + "epoch": 1.7078849110476608, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.833961248397827, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7089858055114746, + "num_tokens": 402085461.0, + "step": 15552 + }, + { + "epoch": 1.7079947287502746, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5985958576202393, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.7132570743560791, + "num_tokens": 402113389.0, + "step": 15553 + }, + { + "epoch": 1.7081045464528883, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6772027015686035, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.705368161201477, + "num_tokens": 402134794.0, + "step": 15554 + }, + { + "epoch": 1.7082143641555019, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4755163192749023, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6897668838500977, + "num_tokens": 402162204.0, + "step": 15555 + }, + { + "epoch": 1.7083241818581154, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8962113857269287, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7183560132980347, + "num_tokens": 402180576.0, + "step": 15556 + }, + { + "epoch": 1.7084339995607292, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.627241611480713, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.726604700088501, + "num_tokens": 402202089.0, + "step": 15557 + }, + { + "epoch": 1.708543817263343, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8205132484436035, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7260018587112427, + "num_tokens": 402221884.0, + "step": 15558 + }, + { + "epoch": 1.7086536349659565, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7711405754089355, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7035340070724487, + "num_tokens": 402243592.0, + "step": 15559 + }, + { + "epoch": 1.7087634526685702, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9530467987060547, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6893330216407776, + "num_tokens": 402268787.0, + "step": 15560 + }, + { + "epoch": 1.7088732703711838, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.399583578109741, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.69675213098526, + "num_tokens": 402299378.0, + "step": 15561 + }, + { + "epoch": 1.7089830880737975, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.49691104888916, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.713651180267334, + "num_tokens": 402324054.0, + "step": 15562 + }, + { + "epoch": 1.7090929057764113, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5284900665283203, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7137826681137085, + "num_tokens": 402346913.0, + "step": 15563 + }, + { + "epoch": 1.7092027234790248, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.530565023422241, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6952387094497681, + "num_tokens": 402373163.0, + "step": 15564 + }, + { + "epoch": 1.7093125411816383, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5428669452667236, + "learning_rate": 1e-06, + "loss": 0.939, + "mean_token_accuracy": 0.7299027442932129, + "num_tokens": 402396775.0, + "step": 15565 + }, + { + "epoch": 1.709422358884252, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.684356689453125, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7076764106750488, + "num_tokens": 402426353.0, + "step": 15566 + }, + { + "epoch": 1.7095321765868658, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3239312171936035, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6813598871231079, + "num_tokens": 402457270.0, + "step": 15567 + }, + { + "epoch": 1.7096419942894796, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.464451551437378, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7020465135574341, + "num_tokens": 402483166.0, + "step": 15568 + }, + { + "epoch": 1.7097518119920931, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5384578704833984, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7185089588165283, + "num_tokens": 402510246.0, + "step": 15569 + }, + { + "epoch": 1.7098616296947067, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 4.346603870391846, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7147444486618042, + "num_tokens": 402540081.0, + "step": 15570 + }, + { + "epoch": 1.7099714473973204, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4643993377685547, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.687779426574707, + "num_tokens": 402570877.0, + "step": 15571 + }, + { + "epoch": 1.7100812650999342, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6571366786956787, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7183791399002075, + "num_tokens": 402594434.0, + "step": 15572 + }, + { + "epoch": 1.7101910828025477, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0806565284729004, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7137859463691711, + "num_tokens": 402618572.0, + "step": 15573 + }, + { + "epoch": 1.7103009005051615, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8238320350646973, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.70103919506073, + "num_tokens": 402639406.0, + "step": 15574 + }, + { + "epoch": 1.710410718207775, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9997193813323975, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.6952393054962158, + "num_tokens": 402658305.0, + "step": 15575 + }, + { + "epoch": 1.7105205359103888, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.491598606109619, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.6830980181694031, + "num_tokens": 402690186.0, + "step": 15576 + }, + { + "epoch": 1.7106303536130025, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.440899610519409, + "learning_rate": 1e-06, + "loss": 1.143, + "mean_token_accuracy": 0.6741035580635071, + "num_tokens": 402722651.0, + "step": 15577 + }, + { + "epoch": 1.710740171315616, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9934568405151367, + "learning_rate": 1e-06, + "loss": 0.9294, + "mean_token_accuracy": 0.7262082099914551, + "num_tokens": 402746767.0, + "step": 15578 + }, + { + "epoch": 1.7108499890182296, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4087984561920166, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7136507034301758, + "num_tokens": 402777349.0, + "step": 15579 + }, + { + "epoch": 1.7109598067208434, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.408217668533325, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.7018738985061646, + "num_tokens": 402806854.0, + "step": 15580 + }, + { + "epoch": 1.7110696244234571, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.4562535285949707, + "learning_rate": 1e-06, + "loss": 1.1001, + "mean_token_accuracy": 0.691840648651123, + "num_tokens": 402836077.0, + "step": 15581 + }, + { + "epoch": 1.7111794421260709, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5907809734344482, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.6918745040893555, + "num_tokens": 402862507.0, + "step": 15582 + }, + { + "epoch": 1.7112892598286844, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3533217906951904, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7075469493865967, + "num_tokens": 402895496.0, + "step": 15583 + }, + { + "epoch": 1.711399077531298, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.591238498687744, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.702750563621521, + "num_tokens": 402922000.0, + "step": 15584 + }, + { + "epoch": 1.7115088952339117, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5808374881744385, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6991356611251831, + "num_tokens": 402946708.0, + "step": 15585 + }, + { + "epoch": 1.7116187129365255, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.449523687362671, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6994495391845703, + "num_tokens": 402974797.0, + "step": 15586 + }, + { + "epoch": 1.711728530639139, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5862042903900146, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7117854356765747, + "num_tokens": 402999664.0, + "step": 15587 + }, + { + "epoch": 1.7118383483417525, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4089415073394775, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6850122213363647, + "num_tokens": 403031461.0, + "step": 15588 + }, + { + "epoch": 1.7119481660443663, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4852607250213623, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7104362845420837, + "num_tokens": 403058704.0, + "step": 15589 + }, + { + "epoch": 1.71205798374698, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4457459449768066, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7214658260345459, + "num_tokens": 403089446.0, + "step": 15590 + }, + { + "epoch": 1.7121678014495938, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6917879581451416, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6822625398635864, + "num_tokens": 403116757.0, + "step": 15591 + }, + { + "epoch": 1.7122776191522073, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6595115661621094, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7130728960037231, + "num_tokens": 403138993.0, + "step": 15592 + }, + { + "epoch": 1.7123874368548209, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6964406967163086, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7011271119117737, + "num_tokens": 403162916.0, + "step": 15593 + }, + { + "epoch": 1.7124972545574346, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3810272216796875, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7089002728462219, + "num_tokens": 403191582.0, + "step": 15594 + }, + { + "epoch": 1.7126070722600484, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.384788751602173, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6928882002830505, + "num_tokens": 403218747.0, + "step": 15595 + }, + { + "epoch": 1.7127168899626621, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.576927423477173, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.690058171749115, + "num_tokens": 403247716.0, + "step": 15596 + }, + { + "epoch": 1.7128267076652757, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.425586462020874, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7143995761871338, + "num_tokens": 403277059.0, + "step": 15597 + }, + { + "epoch": 1.7129365253678892, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.643860101699829, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6913629174232483, + "num_tokens": 403304477.0, + "step": 15598 + }, + { + "epoch": 1.713046343070503, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.752628803253174, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.7187975645065308, + "num_tokens": 403327045.0, + "step": 15599 + }, + { + "epoch": 1.7131561607731167, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8618078231811523, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7199655771255493, + "num_tokens": 403346736.0, + "step": 15600 + }, + { + "epoch": 1.7132659784757303, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.630593776702881, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.707938015460968, + "num_tokens": 403372578.0, + "step": 15601 + }, + { + "epoch": 1.7133757961783438, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7551963329315186, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7013331651687622, + "num_tokens": 403393365.0, + "step": 15602 + }, + { + "epoch": 1.7134856138809575, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7957587242126465, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.6995404958724976, + "num_tokens": 403419727.0, + "step": 15603 + }, + { + "epoch": 1.7135954315835713, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.391345262527466, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7074483633041382, + "num_tokens": 403447141.0, + "step": 15604 + }, + { + "epoch": 1.713705249286185, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6277706623077393, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7007076740264893, + "num_tokens": 403471657.0, + "step": 15605 + }, + { + "epoch": 1.7138150669887986, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6593544483184814, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7081035375595093, + "num_tokens": 403496854.0, + "step": 15606 + }, + { + "epoch": 1.7139248846914121, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4924476146698, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7030746936798096, + "num_tokens": 403524503.0, + "step": 15607 + }, + { + "epoch": 1.7140347023940259, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4207651615142822, + "learning_rate": 1e-06, + "loss": 1.0956, + "mean_token_accuracy": 0.6891818046569824, + "num_tokens": 403554962.0, + "step": 15608 + }, + { + "epoch": 1.7141445200966396, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2546916007995605, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6994086503982544, + "num_tokens": 403587714.0, + "step": 15609 + }, + { + "epoch": 1.7142543377992534, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.614790678024292, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.6869027018547058, + "num_tokens": 403616592.0, + "step": 15610 + }, + { + "epoch": 1.714364155501867, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5638229846954346, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7101175785064697, + "num_tokens": 403641132.0, + "step": 15611 + }, + { + "epoch": 1.7144739732044805, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.748300552368164, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7042034864425659, + "num_tokens": 403673395.0, + "step": 15612 + }, + { + "epoch": 1.7145837909070942, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6690866947174072, + "learning_rate": 1e-06, + "loss": 0.8903, + "mean_token_accuracy": 0.7358505129814148, + "num_tokens": 403695074.0, + "step": 15613 + }, + { + "epoch": 1.714693608609708, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8056416511535645, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.6986023187637329, + "num_tokens": 403722761.0, + "step": 15614 + }, + { + "epoch": 1.7148034263123215, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1052989959716797, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6956418752670288, + "num_tokens": 403743307.0, + "step": 15615 + }, + { + "epoch": 1.714913244014935, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7117669582366943, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7024700045585632, + "num_tokens": 403767386.0, + "step": 15616 + }, + { + "epoch": 1.7150230617175488, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.513516664505005, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7141969799995422, + "num_tokens": 403795522.0, + "step": 15617 + }, + { + "epoch": 1.7151328794201626, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.603663206100464, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7033336758613586, + "num_tokens": 403821048.0, + "step": 15618 + }, + { + "epoch": 1.7152426971227763, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3796908855438232, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7109211683273315, + "num_tokens": 403854081.0, + "step": 15619 + }, + { + "epoch": 1.7153525148253899, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2549750804901123, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7140672206878662, + "num_tokens": 403887105.0, + "step": 15620 + }, + { + "epoch": 1.7154623325280034, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.681640625, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.695852518081665, + "num_tokens": 403911293.0, + "step": 15621 + }, + { + "epoch": 1.7155721502306172, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6861565113067627, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7056115865707397, + "num_tokens": 403932423.0, + "step": 15622 + }, + { + "epoch": 1.715681967933231, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.392939805984497, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.695552408695221, + "num_tokens": 403964248.0, + "step": 15623 + }, + { + "epoch": 1.7157917856358444, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.456195116043091, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7226804494857788, + "num_tokens": 403990120.0, + "step": 15624 + }, + { + "epoch": 1.7159016033384582, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8158092498779297, + "learning_rate": 1e-06, + "loss": 0.9259, + "mean_token_accuracy": 0.7243226170539856, + "num_tokens": 404009131.0, + "step": 15625 + }, + { + "epoch": 1.7160114210410717, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.459641933441162, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6972904801368713, + "num_tokens": 404039592.0, + "step": 15626 + }, + { + "epoch": 1.7161212387436855, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.647656202316284, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7134570479393005, + "num_tokens": 404061611.0, + "step": 15627 + }, + { + "epoch": 1.7162310564462993, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4958584308624268, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.6912944316864014, + "num_tokens": 404089891.0, + "step": 15628 + }, + { + "epoch": 1.7163408741489128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.358381748199463, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6944462060928345, + "num_tokens": 404118032.0, + "step": 15629 + }, + { + "epoch": 1.7164506918515263, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.977076530456543, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7046874761581421, + "num_tokens": 404139583.0, + "step": 15630 + }, + { + "epoch": 1.71656050955414, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.813356876373291, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.699648380279541, + "num_tokens": 404162484.0, + "step": 15631 + }, + { + "epoch": 1.7166703272567538, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6924803256988525, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.723253071308136, + "num_tokens": 404185439.0, + "step": 15632 + }, + { + "epoch": 1.7167801449593676, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.837630271911621, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7128745913505554, + "num_tokens": 404209797.0, + "step": 15633 + }, + { + "epoch": 1.7168899626619811, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.790520191192627, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7300596237182617, + "num_tokens": 404230140.0, + "step": 15634 + }, + { + "epoch": 1.7169997803645947, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.2776870727539062, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6964353322982788, + "num_tokens": 404254200.0, + "step": 15635 + }, + { + "epoch": 1.7171095980672084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6504642963409424, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7071727514266968, + "num_tokens": 404278519.0, + "step": 15636 + }, + { + "epoch": 1.7172194157698222, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.643261671066284, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.693825900554657, + "num_tokens": 404303770.0, + "step": 15637 + }, + { + "epoch": 1.7173292334724357, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8090124130249023, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7175806760787964, + "num_tokens": 404325138.0, + "step": 15638 + }, + { + "epoch": 1.7174390511750495, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6880083084106445, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.689103364944458, + "num_tokens": 404349688.0, + "step": 15639 + }, + { + "epoch": 1.717548868877663, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3168447017669678, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.700019359588623, + "num_tokens": 404382347.0, + "step": 15640 + }, + { + "epoch": 1.7176586865802768, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.405306577682495, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7047596573829651, + "num_tokens": 404409840.0, + "step": 15641 + }, + { + "epoch": 1.7177685042828905, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.8859238624572754, + "learning_rate": 1e-06, + "loss": 1.1385, + "mean_token_accuracy": 0.677091658115387, + "num_tokens": 404441008.0, + "step": 15642 + }, + { + "epoch": 1.717878321985504, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4305574893951416, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.695008397102356, + "num_tokens": 404468403.0, + "step": 15643 + }, + { + "epoch": 1.7179881396881176, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4004831314086914, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.706893265247345, + "num_tokens": 404497085.0, + "step": 15644 + }, + { + "epoch": 1.7180979573907313, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.230405330657959, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6901117563247681, + "num_tokens": 404529952.0, + "step": 15645 + }, + { + "epoch": 1.718207775093345, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.597430467605591, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6885067224502563, + "num_tokens": 404559855.0, + "step": 15646 + }, + { + "epoch": 1.7183175927959589, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.510618209838867, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.6997193098068237, + "num_tokens": 404585447.0, + "step": 15647 + }, + { + "epoch": 1.7184274104985724, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6016831398010254, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6982439756393433, + "num_tokens": 404615258.0, + "step": 15648 + }, + { + "epoch": 1.718537228201186, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6639575958251953, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6974042654037476, + "num_tokens": 404640487.0, + "step": 15649 + }, + { + "epoch": 1.7186470459037997, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5084052085876465, + "learning_rate": 1e-06, + "loss": 1.1005, + "mean_token_accuracy": 0.6880668997764587, + "num_tokens": 404669984.0, + "step": 15650 + }, + { + "epoch": 1.7187568636064134, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4842543601989746, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6972565054893494, + "num_tokens": 404695795.0, + "step": 15651 + }, + { + "epoch": 1.718866681309027, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8609979152679443, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7198883295059204, + "num_tokens": 404716863.0, + "step": 15652 + }, + { + "epoch": 1.7189764990116405, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.027649402618408, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7148153185844421, + "num_tokens": 404733635.0, + "step": 15653 + }, + { + "epoch": 1.7190863167142543, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.973194122314453, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.718186616897583, + "num_tokens": 404754534.0, + "step": 15654 + }, + { + "epoch": 1.719196134416868, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7485179901123047, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6853134632110596, + "num_tokens": 404786352.0, + "step": 15655 + }, + { + "epoch": 1.7193059521194818, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.663292407989502, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7013917565345764, + "num_tokens": 404812670.0, + "step": 15656 + }, + { + "epoch": 1.7194157698220953, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.331117868423462, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7151561379432678, + "num_tokens": 404833971.0, + "step": 15657 + }, + { + "epoch": 1.7195255875247089, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.347231149673462, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.699594259262085, + "num_tokens": 404865009.0, + "step": 15658 + }, + { + "epoch": 1.7196354052273226, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7461681365966797, + "learning_rate": 1e-06, + "loss": 1.1292, + "mean_token_accuracy": 0.6849989891052246, + "num_tokens": 404891573.0, + "step": 15659 + }, + { + "epoch": 1.7197452229299364, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.542531967163086, + "learning_rate": 1e-06, + "loss": 1.1376, + "mean_token_accuracy": 0.6973956823348999, + "num_tokens": 404918890.0, + "step": 15660 + }, + { + "epoch": 1.7198550406325501, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4788706302642822, + "learning_rate": 1e-06, + "loss": 0.9217, + "mean_token_accuracy": 0.7334750890731812, + "num_tokens": 404943132.0, + "step": 15661 + }, + { + "epoch": 1.7199648583351637, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5076231956481934, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6977761387825012, + "num_tokens": 404970183.0, + "step": 15662 + }, + { + "epoch": 1.7200746760377772, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.495016098022461, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.7010077238082886, + "num_tokens": 404996090.0, + "step": 15663 + }, + { + "epoch": 1.720184493740391, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3329343795776367, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7078948020935059, + "num_tokens": 405024775.0, + "step": 15664 + }, + { + "epoch": 1.7202943114430047, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4019060134887695, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6909662485122681, + "num_tokens": 405056691.0, + "step": 15665 + }, + { + "epoch": 1.7204041291456182, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5697343349456787, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6921656131744385, + "num_tokens": 405082271.0, + "step": 15666 + }, + { + "epoch": 1.7205139468482318, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.80898118019104, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7030697464942932, + "num_tokens": 405104542.0, + "step": 15667 + }, + { + "epoch": 1.7206237645508455, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5156517028808594, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7118908166885376, + "num_tokens": 405135485.0, + "step": 15668 + }, + { + "epoch": 1.7207335822534593, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5331876277923584, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7124112844467163, + "num_tokens": 405159707.0, + "step": 15669 + }, + { + "epoch": 1.720843399956073, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.677832841873169, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7049442529678345, + "num_tokens": 405185350.0, + "step": 15670 + }, + { + "epoch": 1.7209532176586866, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4334943294525146, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7194274663925171, + "num_tokens": 405214522.0, + "step": 15671 + }, + { + "epoch": 1.7210630353613001, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8808655738830566, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7109030485153198, + "num_tokens": 405236811.0, + "step": 15672 + }, + { + "epoch": 1.7211728530639139, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.238560914993286, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7323402166366577, + "num_tokens": 405265540.0, + "step": 15673 + }, + { + "epoch": 1.7212826707665276, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.589186906814575, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6935649514198303, + "num_tokens": 405295474.0, + "step": 15674 + }, + { + "epoch": 1.7213924884691414, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.716487169265747, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7085762619972229, + "num_tokens": 405321196.0, + "step": 15675 + }, + { + "epoch": 1.721502306171755, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3217153549194336, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6889596581459045, + "num_tokens": 405351944.0, + "step": 15676 + }, + { + "epoch": 1.7216121238743685, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6948416233062744, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6963201761245728, + "num_tokens": 405378336.0, + "step": 15677 + }, + { + "epoch": 1.7217219415769822, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4346485137939453, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7222801446914673, + "num_tokens": 405404115.0, + "step": 15678 + }, + { + "epoch": 1.721831759279596, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6467456817626953, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6974993944168091, + "num_tokens": 405427676.0, + "step": 15679 + }, + { + "epoch": 1.7219415769822095, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4326000213623047, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6848131418228149, + "num_tokens": 405456316.0, + "step": 15680 + }, + { + "epoch": 1.722051394684823, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8500404357910156, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.697638750076294, + "num_tokens": 405478827.0, + "step": 15681 + }, + { + "epoch": 1.7221612123874368, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6610124111175537, + "learning_rate": 1e-06, + "loss": 0.9234, + "mean_token_accuracy": 0.7268861532211304, + "num_tokens": 405503775.0, + "step": 15682 + }, + { + "epoch": 1.7222710300900506, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5628843307495117, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6950921416282654, + "num_tokens": 405530284.0, + "step": 15683 + }, + { + "epoch": 1.7223808477926643, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5386576652526855, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6941890716552734, + "num_tokens": 405560487.0, + "step": 15684 + }, + { + "epoch": 1.7224906654952779, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6552746295928955, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.715342104434967, + "num_tokens": 405586029.0, + "step": 15685 + }, + { + "epoch": 1.7226004831978914, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4744980335235596, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7063426971435547, + "num_tokens": 405614303.0, + "step": 15686 + }, + { + "epoch": 1.7227103009005051, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6606216430664062, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7063891887664795, + "num_tokens": 405637574.0, + "step": 15687 + }, + { + "epoch": 1.722820118603119, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.453317642211914, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6820254921913147, + "num_tokens": 405664647.0, + "step": 15688 + }, + { + "epoch": 1.7229299363057324, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2417678833007812, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7167983651161194, + "num_tokens": 405695245.0, + "step": 15689 + }, + { + "epoch": 1.7230397540083462, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.458164930343628, + "learning_rate": 1e-06, + "loss": 0.9411, + "mean_token_accuracy": 0.7273210287094116, + "num_tokens": 405721043.0, + "step": 15690 + }, + { + "epoch": 1.7231495717109597, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4882731437683105, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6917816996574402, + "num_tokens": 405748441.0, + "step": 15691 + }, + { + "epoch": 1.7232593894135735, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.795106887817383, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.6990836262702942, + "num_tokens": 405770211.0, + "step": 15692 + }, + { + "epoch": 1.7233692071161872, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4387097358703613, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7207786440849304, + "num_tokens": 405799085.0, + "step": 15693 + }, + { + "epoch": 1.7234790248188008, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.505162477493286, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7195459604263306, + "num_tokens": 405825810.0, + "step": 15694 + }, + { + "epoch": 1.7235888425214143, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.584754228591919, + "learning_rate": 1e-06, + "loss": 1.1211, + "mean_token_accuracy": 0.6865416765213013, + "num_tokens": 405853812.0, + "step": 15695 + }, + { + "epoch": 1.723698660224028, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7979695796966553, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7156275510787964, + "num_tokens": 405879039.0, + "step": 15696 + }, + { + "epoch": 1.7238084779266418, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.501676559448242, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.6976154446601868, + "num_tokens": 405908301.0, + "step": 15697 + }, + { + "epoch": 1.7239182956292556, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.1876022815704346, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7214560508728027, + "num_tokens": 405928513.0, + "step": 15698 + }, + { + "epoch": 1.7240281133318691, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.436816930770874, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6952553987503052, + "num_tokens": 405958633.0, + "step": 15699 + }, + { + "epoch": 1.7241379310344827, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.56355357170105, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7169854044914246, + "num_tokens": 405983518.0, + "step": 15700 + }, + { + "epoch": 1.7242477487370964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4342403411865234, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7128098011016846, + "num_tokens": 406009479.0, + "step": 15701 + }, + { + "epoch": 1.7243575664397102, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9351091384887695, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7074145078659058, + "num_tokens": 406030749.0, + "step": 15702 + }, + { + "epoch": 1.7244673841423237, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6933200359344482, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.7040985822677612, + "num_tokens": 406056447.0, + "step": 15703 + }, + { + "epoch": 1.7245772018449375, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.980001926422119, + "learning_rate": 1e-06, + "loss": 0.8786, + "mean_token_accuracy": 0.739463210105896, + "num_tokens": 406074410.0, + "step": 15704 + }, + { + "epoch": 1.724687019547551, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7593469619750977, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7264394164085388, + "num_tokens": 406096397.0, + "step": 15705 + }, + { + "epoch": 1.7247968372501647, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6815502643585205, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6877191066741943, + "num_tokens": 406123869.0, + "step": 15706 + }, + { + "epoch": 1.7249066549527785, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.729621171951294, + "learning_rate": 1e-06, + "loss": 0.9073, + "mean_token_accuracy": 0.7355421781539917, + "num_tokens": 406147337.0, + "step": 15707 + }, + { + "epoch": 1.725016472655392, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.60014271736145, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6986736059188843, + "num_tokens": 406171769.0, + "step": 15708 + }, + { + "epoch": 1.7251262903580056, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5472371578216553, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6986503601074219, + "num_tokens": 406196379.0, + "step": 15709 + }, + { + "epoch": 1.7252361080606193, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.557068347930908, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7035134434700012, + "num_tokens": 406226674.0, + "step": 15710 + }, + { + "epoch": 1.725345925763233, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.482926607131958, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7167482972145081, + "num_tokens": 406251785.0, + "step": 15711 + }, + { + "epoch": 1.7254557434658468, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7326130867004395, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.7239024639129639, + "num_tokens": 406272446.0, + "step": 15712 + }, + { + "epoch": 1.7255655611684604, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.649521589279175, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7122324705123901, + "num_tokens": 406297309.0, + "step": 15713 + }, + { + "epoch": 1.725675378871074, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5540900230407715, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.703599750995636, + "num_tokens": 406322257.0, + "step": 15714 + }, + { + "epoch": 1.7257851965736877, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.671477794647217, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7033440470695496, + "num_tokens": 406346570.0, + "step": 15715 + }, + { + "epoch": 1.7258950142763014, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 9.721525192260742, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7107603549957275, + "num_tokens": 406372128.0, + "step": 15716 + }, + { + "epoch": 1.726004831978915, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.135005235671997, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.7265335321426392, + "num_tokens": 406393200.0, + "step": 15717 + }, + { + "epoch": 1.7261146496815285, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5370779037475586, + "learning_rate": 1e-06, + "loss": 1.1212, + "mean_token_accuracy": 0.6771774888038635, + "num_tokens": 406421258.0, + "step": 15718 + }, + { + "epoch": 1.7262244673841423, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3083279132843018, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7246240973472595, + "num_tokens": 406454628.0, + "step": 15719 + }, + { + "epoch": 1.726334285086756, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4552717208862305, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7214208841323853, + "num_tokens": 406484194.0, + "step": 15720 + }, + { + "epoch": 1.7264441027893698, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5547587871551514, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.7023613452911377, + "num_tokens": 406508641.0, + "step": 15721 + }, + { + "epoch": 1.7265539204919833, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4272899627685547, + "learning_rate": 1e-06, + "loss": 1.0847, + "mean_token_accuracy": 0.6843367218971252, + "num_tokens": 406535886.0, + "step": 15722 + }, + { + "epoch": 1.7266637381945968, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9438953399658203, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.7015740871429443, + "num_tokens": 406557711.0, + "step": 15723 + }, + { + "epoch": 1.7267735558972106, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.753538131713867, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7136960029602051, + "num_tokens": 406579582.0, + "step": 15724 + }, + { + "epoch": 1.7268833735998244, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7382102012634277, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7234334349632263, + "num_tokens": 406603065.0, + "step": 15725 + }, + { + "epoch": 1.7269931913024381, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6713767051696777, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.7008728981018066, + "num_tokens": 406627249.0, + "step": 15726 + }, + { + "epoch": 1.7271030090050516, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.570991277694702, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7286115884780884, + "num_tokens": 406651523.0, + "step": 15727 + }, + { + "epoch": 1.7272128267076652, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.509080410003662, + "learning_rate": 1e-06, + "loss": 1.0909, + "mean_token_accuracy": 0.6844757795333862, + "num_tokens": 406677929.0, + "step": 15728 + }, + { + "epoch": 1.727322644410279, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 16.835500717163086, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7010927200317383, + "num_tokens": 406703411.0, + "step": 15729 + }, + { + "epoch": 1.7274324621128927, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7278201580047607, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.7002487778663635, + "num_tokens": 406727723.0, + "step": 15730 + }, + { + "epoch": 1.7275422798155062, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5590410232543945, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6908243298530579, + "num_tokens": 406755501.0, + "step": 15731 + }, + { + "epoch": 1.7276520975181198, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0760881900787354, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7177963852882385, + "num_tokens": 406775289.0, + "step": 15732 + }, + { + "epoch": 1.7277619152207335, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5612576007843018, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7162125706672668, + "num_tokens": 406804107.0, + "step": 15733 + }, + { + "epoch": 1.7278717329233473, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.485072612762451, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7035343647003174, + "num_tokens": 406832425.0, + "step": 15734 + }, + { + "epoch": 1.727981550625961, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6388351917266846, + "learning_rate": 1e-06, + "loss": 0.8546, + "mean_token_accuracy": 0.7457050681114197, + "num_tokens": 406855526.0, + "step": 15735 + }, + { + "epoch": 1.7280913683285746, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.919048547744751, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7078575491905212, + "num_tokens": 406876907.0, + "step": 15736 + }, + { + "epoch": 1.728201186031188, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.034353017807007, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.6868616342544556, + "num_tokens": 406904760.0, + "step": 15737 + }, + { + "epoch": 1.7283110037338019, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.726016044616699, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7125526070594788, + "num_tokens": 406928549.0, + "step": 15738 + }, + { + "epoch": 1.7284208214364156, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.83439564704895, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.700605571269989, + "num_tokens": 406949708.0, + "step": 15739 + }, + { + "epoch": 1.7285306391390292, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4853601455688477, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.7004708051681519, + "num_tokens": 406977852.0, + "step": 15740 + }, + { + "epoch": 1.728640456841643, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5861310958862305, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.6962980031967163, + "num_tokens": 407004183.0, + "step": 15741 + }, + { + "epoch": 1.7287502745442564, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4827468395233154, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7134166955947876, + "num_tokens": 407029641.0, + "step": 15742 + }, + { + "epoch": 1.7288600922468702, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2701475620269775, + "learning_rate": 1e-06, + "loss": 1.1229, + "mean_token_accuracy": 0.6813082695007324, + "num_tokens": 407061311.0, + "step": 15743 + }, + { + "epoch": 1.728969909949484, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6972761154174805, + "learning_rate": 1e-06, + "loss": 0.9405, + "mean_token_accuracy": 0.7174738049507141, + "num_tokens": 407083842.0, + "step": 15744 + }, + { + "epoch": 1.7290797276520975, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5708136558532715, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7000560760498047, + "num_tokens": 407106403.0, + "step": 15745 + }, + { + "epoch": 1.729189545354711, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7819180488586426, + "learning_rate": 1e-06, + "loss": 0.9233, + "mean_token_accuracy": 0.7256520986557007, + "num_tokens": 407127611.0, + "step": 15746 + }, + { + "epoch": 1.7292993630573248, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5248262882232666, + "learning_rate": 1e-06, + "loss": 1.0779, + "mean_token_accuracy": 0.6937832832336426, + "num_tokens": 407155998.0, + "step": 15747 + }, + { + "epoch": 1.7294091807599385, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5066843032836914, + "learning_rate": 1e-06, + "loss": 0.885, + "mean_token_accuracy": 0.7425447702407837, + "num_tokens": 407180192.0, + "step": 15748 + }, + { + "epoch": 1.7295189984625523, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5742173194885254, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.6928030252456665, + "num_tokens": 407207098.0, + "step": 15749 + }, + { + "epoch": 1.7296288161651658, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.640289306640625, + "learning_rate": 1e-06, + "loss": 0.9209, + "mean_token_accuracy": 0.7291110754013062, + "num_tokens": 407229106.0, + "step": 15750 + }, + { + "epoch": 1.7297386338677794, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.685065269470215, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7211949825286865, + "num_tokens": 407251733.0, + "step": 15751 + }, + { + "epoch": 1.7298484515703931, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6257224082946777, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6940982937812805, + "num_tokens": 407276925.0, + "step": 15752 + }, + { + "epoch": 1.7299582692730069, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.589956521987915, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7016350030899048, + "num_tokens": 407301318.0, + "step": 15753 + }, + { + "epoch": 1.7300680869756204, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.813875436782837, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7189031839370728, + "num_tokens": 407323652.0, + "step": 15754 + }, + { + "epoch": 1.7301779046782342, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4810633659362793, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7075085639953613, + "num_tokens": 407353689.0, + "step": 15755 + }, + { + "epoch": 1.7302877223808477, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6064341068267822, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6914031505584717, + "num_tokens": 407378942.0, + "step": 15756 + }, + { + "epoch": 1.7303975400834615, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5921480655670166, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7097468376159668, + "num_tokens": 407404492.0, + "step": 15757 + }, + { + "epoch": 1.7305073577860752, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.594619035720825, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7081952095031738, + "num_tokens": 407429503.0, + "step": 15758 + }, + { + "epoch": 1.7306171754886888, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3610317707061768, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7104986906051636, + "num_tokens": 407459862.0, + "step": 15759 + }, + { + "epoch": 1.7307269931913023, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.762836456298828, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7160723209381104, + "num_tokens": 407485344.0, + "step": 15760 + }, + { + "epoch": 1.730836810893916, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.651709794998169, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7128248810768127, + "num_tokens": 407508639.0, + "step": 15761 + }, + { + "epoch": 1.7309466285965298, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5472261905670166, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7036681771278381, + "num_tokens": 407535977.0, + "step": 15762 + }, + { + "epoch": 1.7310564462991436, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.71490478515625, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7232394814491272, + "num_tokens": 407557839.0, + "step": 15763 + }, + { + "epoch": 1.731166264001757, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8315579891204834, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.71593177318573, + "num_tokens": 407577992.0, + "step": 15764 + }, + { + "epoch": 1.7312760817043706, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6651928424835205, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7243704199790955, + "num_tokens": 407602088.0, + "step": 15765 + }, + { + "epoch": 1.7313858994069844, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.895848274230957, + "learning_rate": 1e-06, + "loss": 0.9302, + "mean_token_accuracy": 0.7237407565116882, + "num_tokens": 407624483.0, + "step": 15766 + }, + { + "epoch": 1.7314957171095982, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.687666893005371, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6866739988327026, + "num_tokens": 407648718.0, + "step": 15767 + }, + { + "epoch": 1.7316055348122117, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6748640537261963, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7077950239181519, + "num_tokens": 407671524.0, + "step": 15768 + }, + { + "epoch": 1.7317153525148252, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5590453147888184, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.7023929357528687, + "num_tokens": 407696736.0, + "step": 15769 + }, + { + "epoch": 1.731825170217439, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.970848560333252, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7215139865875244, + "num_tokens": 407719242.0, + "step": 15770 + }, + { + "epoch": 1.7319349879200527, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2126407623291016, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.71397864818573, + "num_tokens": 407741647.0, + "step": 15771 + }, + { + "epoch": 1.7320448056226665, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6567773818969727, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7157410979270935, + "num_tokens": 407766431.0, + "step": 15772 + }, + { + "epoch": 1.73215462332528, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4671027660369873, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7070854902267456, + "num_tokens": 407791725.0, + "step": 15773 + }, + { + "epoch": 1.7322644410278936, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.82305645942688, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6860513687133789, + "num_tokens": 407814167.0, + "step": 15774 + }, + { + "epoch": 1.7323742587305073, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4916300773620605, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6860249042510986, + "num_tokens": 407843181.0, + "step": 15775 + }, + { + "epoch": 1.732484076433121, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.650214433670044, + "learning_rate": 1e-06, + "loss": 1.1127, + "mean_token_accuracy": 0.6839221119880676, + "num_tokens": 407868658.0, + "step": 15776 + }, + { + "epoch": 1.7325938941357348, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6113765239715576, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7220897674560547, + "num_tokens": 407895419.0, + "step": 15777 + }, + { + "epoch": 1.7327037118383484, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6077444553375244, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7168152332305908, + "num_tokens": 407920037.0, + "step": 15778 + }, + { + "epoch": 1.732813529540962, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.704145908355713, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7033243775367737, + "num_tokens": 407943819.0, + "step": 15779 + }, + { + "epoch": 1.7329233472435757, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.479642391204834, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6979830265045166, + "num_tokens": 407974168.0, + "step": 15780 + }, + { + "epoch": 1.7330331649461894, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.660766124725342, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7146322727203369, + "num_tokens": 408000107.0, + "step": 15781 + }, + { + "epoch": 1.733142982648803, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.441830635070801, + "learning_rate": 1e-06, + "loss": 1.1357, + "mean_token_accuracy": 0.6705383658409119, + "num_tokens": 408029346.0, + "step": 15782 + }, + { + "epoch": 1.7332528003514165, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.647808790206909, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6927961707115173, + "num_tokens": 408054076.0, + "step": 15783 + }, + { + "epoch": 1.7333626180540302, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.388866901397705, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6920614242553711, + "num_tokens": 408087889.0, + "step": 15784 + }, + { + "epoch": 1.733472435756644, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3626489639282227, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7026638984680176, + "num_tokens": 408115327.0, + "step": 15785 + }, + { + "epoch": 1.7335822534592578, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6933958530426025, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6835373640060425, + "num_tokens": 408139826.0, + "step": 15786 + }, + { + "epoch": 1.7336920711618713, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.744021415710449, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7016414403915405, + "num_tokens": 408163102.0, + "step": 15787 + }, + { + "epoch": 1.7338018888644848, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8276872634887695, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7037865519523621, + "num_tokens": 408186929.0, + "step": 15788 + }, + { + "epoch": 1.7339117065670986, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.688065767288208, + "learning_rate": 1e-06, + "loss": 1.1369, + "mean_token_accuracy": 0.6726722717285156, + "num_tokens": 408213628.0, + "step": 15789 + }, + { + "epoch": 1.7340215242697123, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.539306163787842, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6863097548484802, + "num_tokens": 408244207.0, + "step": 15790 + }, + { + "epoch": 1.734131341972326, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3484809398651123, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7162455320358276, + "num_tokens": 408273952.0, + "step": 15791 + }, + { + "epoch": 1.7342411596749396, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6015405654907227, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7082191705703735, + "num_tokens": 408300411.0, + "step": 15792 + }, + { + "epoch": 1.7343509773775532, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.984142780303955, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7225933074951172, + "num_tokens": 408321815.0, + "step": 15793 + }, + { + "epoch": 1.734460795080167, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.838636875152588, + "learning_rate": 1e-06, + "loss": 0.8992, + "mean_token_accuracy": 0.7325915694236755, + "num_tokens": 408343248.0, + "step": 15794 + }, + { + "epoch": 1.7345706127827807, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7260935306549072, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6912710070610046, + "num_tokens": 408368877.0, + "step": 15795 + }, + { + "epoch": 1.7346804304853942, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6577517986297607, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6859630346298218, + "num_tokens": 408394331.0, + "step": 15796 + }, + { + "epoch": 1.7347902481880078, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4697835445404053, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6884240508079529, + "num_tokens": 408419574.0, + "step": 15797 + }, + { + "epoch": 1.7349000658906215, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.402902126312256, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7061601281166077, + "num_tokens": 408446745.0, + "step": 15798 + }, + { + "epoch": 1.7350098835932353, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5908987522125244, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7001065015792847, + "num_tokens": 408470346.0, + "step": 15799 + }, + { + "epoch": 1.735119701295849, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.59135365486145, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6997959613800049, + "num_tokens": 408497927.0, + "step": 15800 + }, + { + "epoch": 1.7352295189984626, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6629326343536377, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6940593719482422, + "num_tokens": 408523528.0, + "step": 15801 + }, + { + "epoch": 1.735339336701076, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2078349590301514, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6966807246208191, + "num_tokens": 408557516.0, + "step": 15802 + }, + { + "epoch": 1.7354491544036899, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5618274211883545, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6975568532943726, + "num_tokens": 408582193.0, + "step": 15803 + }, + { + "epoch": 1.7355589721063036, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8960728645324707, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7156111001968384, + "num_tokens": 408602622.0, + "step": 15804 + }, + { + "epoch": 1.7356687898089171, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8118374347686768, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7093154191970825, + "num_tokens": 408624316.0, + "step": 15805 + }, + { + "epoch": 1.735778607511531, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6425070762634277, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7120887041091919, + "num_tokens": 408652235.0, + "step": 15806 + }, + { + "epoch": 1.7358884252141444, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.653434991836548, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.7014434933662415, + "num_tokens": 408677377.0, + "step": 15807 + }, + { + "epoch": 1.7359982429167582, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6466054916381836, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6997930407524109, + "num_tokens": 408703453.0, + "step": 15808 + }, + { + "epoch": 1.736108060619372, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.699652910232544, + "learning_rate": 1e-06, + "loss": 0.9233, + "mean_token_accuracy": 0.729259729385376, + "num_tokens": 408727630.0, + "step": 15809 + }, + { + "epoch": 1.7362178783219855, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.68672251701355, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6975547075271606, + "num_tokens": 408750923.0, + "step": 15810 + }, + { + "epoch": 1.736327696024599, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.508657217025757, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6992947459220886, + "num_tokens": 408778970.0, + "step": 15811 + }, + { + "epoch": 1.7364375137272128, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.71846604347229, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.730392575263977, + "num_tokens": 408803405.0, + "step": 15812 + }, + { + "epoch": 1.7365473314298265, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6303532123565674, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6873096227645874, + "num_tokens": 408828951.0, + "step": 15813 + }, + { + "epoch": 1.7366571491324403, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3743762969970703, + "learning_rate": 1e-06, + "loss": 1.1142, + "mean_token_accuracy": 0.6681681275367737, + "num_tokens": 408864782.0, + "step": 15814 + }, + { + "epoch": 1.7367669668350538, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.884413957595825, + "learning_rate": 1e-06, + "loss": 0.9176, + "mean_token_accuracy": 0.7227669954299927, + "num_tokens": 408883987.0, + "step": 15815 + }, + { + "epoch": 1.7368767845376674, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5205628871917725, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.6928038001060486, + "num_tokens": 408911114.0, + "step": 15816 + }, + { + "epoch": 1.7369866022402811, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.822239875793457, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7219396829605103, + "num_tokens": 408932732.0, + "step": 15817 + }, + { + "epoch": 1.7370964199428949, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.864396333694458, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.726120114326477, + "num_tokens": 408956139.0, + "step": 15818 + }, + { + "epoch": 1.7372062376455084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.729820966720581, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7161526083946228, + "num_tokens": 408979450.0, + "step": 15819 + }, + { + "epoch": 1.7373160553481222, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7248950004577637, + "learning_rate": 1e-06, + "loss": 0.9456, + "mean_token_accuracy": 0.7237576246261597, + "num_tokens": 409002261.0, + "step": 15820 + }, + { + "epoch": 1.7374258730507357, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.689610481262207, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6979283094406128, + "num_tokens": 409031473.0, + "step": 15821 + }, + { + "epoch": 1.7375356907533495, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.746107339859009, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7075014114379883, + "num_tokens": 409053824.0, + "step": 15822 + }, + { + "epoch": 1.7376455084559632, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5967464447021484, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6948037147521973, + "num_tokens": 409081507.0, + "step": 15823 + }, + { + "epoch": 1.7377553261585768, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7437963485717773, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.732919454574585, + "num_tokens": 409104613.0, + "step": 15824 + }, + { + "epoch": 1.7378651438611903, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6623950004577637, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7014380693435669, + "num_tokens": 409131953.0, + "step": 15825 + }, + { + "epoch": 1.737974961563804, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.695495367050171, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6981320381164551, + "num_tokens": 409156621.0, + "step": 15826 + }, + { + "epoch": 1.7380847792664178, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.552093505859375, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.6962103843688965, + "num_tokens": 409183001.0, + "step": 15827 + }, + { + "epoch": 1.7381945969690316, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7339394092559814, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7213322520256042, + "num_tokens": 409203539.0, + "step": 15828 + }, + { + "epoch": 1.738304414671645, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5985612869262695, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7105511426925659, + "num_tokens": 409230142.0, + "step": 15829 + }, + { + "epoch": 1.7384142323742586, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4938149452209473, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.690648078918457, + "num_tokens": 409255909.0, + "step": 15830 + }, + { + "epoch": 1.7385240500768724, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4329075813293457, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7182588577270508, + "num_tokens": 409282733.0, + "step": 15831 + }, + { + "epoch": 1.7386338677794861, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4960970878601074, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7169669270515442, + "num_tokens": 409309821.0, + "step": 15832 + }, + { + "epoch": 1.7387436854820997, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6664793491363525, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7321710586547852, + "num_tokens": 409334055.0, + "step": 15833 + }, + { + "epoch": 1.7388535031847132, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8289482593536377, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.7059455513954163, + "num_tokens": 409354915.0, + "step": 15834 + }, + { + "epoch": 1.738963320887327, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.230764627456665, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7316184043884277, + "num_tokens": 409372334.0, + "step": 15835 + }, + { + "epoch": 1.7390731385899407, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.551191806793213, + "learning_rate": 1e-06, + "loss": 1.1087, + "mean_token_accuracy": 0.682623028755188, + "num_tokens": 409401713.0, + "step": 15836 + }, + { + "epoch": 1.7391829562925545, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5183072090148926, + "learning_rate": 1e-06, + "loss": 1.1026, + "mean_token_accuracy": 0.6822574734687805, + "num_tokens": 409431433.0, + "step": 15837 + }, + { + "epoch": 1.739292773995168, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.642486095428467, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.708389401435852, + "num_tokens": 409454487.0, + "step": 15838 + }, + { + "epoch": 1.7394025916977816, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5130014419555664, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7118884325027466, + "num_tokens": 409479958.0, + "step": 15839 + }, + { + "epoch": 1.7395124094003953, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.447727918624878, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.709610104560852, + "num_tokens": 409509442.0, + "step": 15840 + }, + { + "epoch": 1.739622227103009, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8483190536499023, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6930755972862244, + "num_tokens": 409530593.0, + "step": 15841 + }, + { + "epoch": 1.7397320448056228, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5603981018066406, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7030548453330994, + "num_tokens": 409563789.0, + "step": 15842 + }, + { + "epoch": 1.7398418625082364, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.341308116912842, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6948211789131165, + "num_tokens": 409592783.0, + "step": 15843 + }, + { + "epoch": 1.73995168021085, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5765998363494873, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7029507756233215, + "num_tokens": 409617309.0, + "step": 15844 + }, + { + "epoch": 1.7400614979134637, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.641756296157837, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6979460120201111, + "num_tokens": 409642941.0, + "step": 15845 + }, + { + "epoch": 1.7401713156160774, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4615232944488525, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7153780460357666, + "num_tokens": 409669861.0, + "step": 15846 + }, + { + "epoch": 1.740281133318691, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7602531909942627, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.693598747253418, + "num_tokens": 409691814.0, + "step": 15847 + }, + { + "epoch": 1.7403909510213045, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8873789310455322, + "learning_rate": 1e-06, + "loss": 1.0736, + "mean_token_accuracy": 0.6972069144248962, + "num_tokens": 409717653.0, + "step": 15848 + }, + { + "epoch": 1.7405007687239182, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.737971782684326, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7081735730171204, + "num_tokens": 409738514.0, + "step": 15849 + }, + { + "epoch": 1.740610586426532, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7880096435546875, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.691916286945343, + "num_tokens": 409762917.0, + "step": 15850 + }, + { + "epoch": 1.7407204041291457, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.636551856994629, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7135602235794067, + "num_tokens": 409786764.0, + "step": 15851 + }, + { + "epoch": 1.7408302218317593, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5482096672058105, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6989174485206604, + "num_tokens": 409814012.0, + "step": 15852 + }, + { + "epoch": 1.7409400395343728, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.473680257797241, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.6990669965744019, + "num_tokens": 409841290.0, + "step": 15853 + }, + { + "epoch": 1.7410498572369866, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5931880474090576, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.7010695338249207, + "num_tokens": 409865997.0, + "step": 15854 + }, + { + "epoch": 1.7411596749396003, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6445696353912354, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6884483695030212, + "num_tokens": 409888915.0, + "step": 15855 + }, + { + "epoch": 1.741269492642214, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2070536613464355, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6855226755142212, + "num_tokens": 409922378.0, + "step": 15856 + }, + { + "epoch": 1.7413793103448276, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5271248817443848, + "learning_rate": 1e-06, + "loss": 0.8818, + "mean_token_accuracy": 0.7390308380126953, + "num_tokens": 409945678.0, + "step": 15857 + }, + { + "epoch": 1.7414891280474412, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.670305013656616, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.6990318298339844, + "num_tokens": 409968835.0, + "step": 15858 + }, + { + "epoch": 1.741598945750055, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.448636054992676, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6981306076049805, + "num_tokens": 409997469.0, + "step": 15859 + }, + { + "epoch": 1.7417087634526687, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7696146965026855, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6929692625999451, + "num_tokens": 410022375.0, + "step": 15860 + }, + { + "epoch": 1.7418185811552822, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3757262229919434, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.6792722940444946, + "num_tokens": 410054020.0, + "step": 15861 + }, + { + "epoch": 1.7419283988578957, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 9.769468307495117, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6931535601615906, + "num_tokens": 410079572.0, + "step": 15862 + }, + { + "epoch": 1.7420382165605095, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.525118589401245, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7162474989891052, + "num_tokens": 410106514.0, + "step": 15863 + }, + { + "epoch": 1.7421480342631233, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.642350912094116, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7015289068222046, + "num_tokens": 410132343.0, + "step": 15864 + }, + { + "epoch": 1.742257851965737, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.548851251602173, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7074532508850098, + "num_tokens": 410161866.0, + "step": 15865 + }, + { + "epoch": 1.7423676696683505, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.612098217010498, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6811431050300598, + "num_tokens": 410187236.0, + "step": 15866 + }, + { + "epoch": 1.742477487370964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4613263607025146, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7037643194198608, + "num_tokens": 410216432.0, + "step": 15867 + }, + { + "epoch": 1.7425873050735778, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3782224655151367, + "learning_rate": 1e-06, + "loss": 0.9761, + "mean_token_accuracy": 0.71347975730896, + "num_tokens": 410244252.0, + "step": 15868 + }, + { + "epoch": 1.7426971227761916, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4889752864837646, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6886788010597229, + "num_tokens": 410271135.0, + "step": 15869 + }, + { + "epoch": 1.7428069404788051, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5966134071350098, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7064104676246643, + "num_tokens": 410296705.0, + "step": 15870 + }, + { + "epoch": 1.742916758181419, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.579871892929077, + "learning_rate": 1e-06, + "loss": 1.1115, + "mean_token_accuracy": 0.6744881272315979, + "num_tokens": 410324785.0, + "step": 15871 + }, + { + "epoch": 1.7430265758840324, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4678213596343994, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7159934639930725, + "num_tokens": 410350793.0, + "step": 15872 + }, + { + "epoch": 1.7431363935866462, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4276602268218994, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7089133858680725, + "num_tokens": 410380193.0, + "step": 15873 + }, + { + "epoch": 1.74324621128926, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.620408535003662, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.7303762435913086, + "num_tokens": 410403374.0, + "step": 15874 + }, + { + "epoch": 1.7433560289918735, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4963905811309814, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6931208372116089, + "num_tokens": 410431523.0, + "step": 15875 + }, + { + "epoch": 1.743465846694487, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.617114305496216, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7236891984939575, + "num_tokens": 410455780.0, + "step": 15876 + }, + { + "epoch": 1.7435756643971008, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5296459197998047, + "learning_rate": 1e-06, + "loss": 1.1123, + "mean_token_accuracy": 0.6869741082191467, + "num_tokens": 410485812.0, + "step": 15877 + }, + { + "epoch": 1.7436854820997145, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8145885467529297, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6975780129432678, + "num_tokens": 410507164.0, + "step": 15878 + }, + { + "epoch": 1.7437952998023283, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.598200798034668, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.716222882270813, + "num_tokens": 410534324.0, + "step": 15879 + }, + { + "epoch": 1.7439051175049418, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.682142734527588, + "learning_rate": 1e-06, + "loss": 1.1347, + "mean_token_accuracy": 0.670904815196991, + "num_tokens": 410564035.0, + "step": 15880 + }, + { + "epoch": 1.7440149352075554, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9960427284240723, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7144800424575806, + "num_tokens": 410582722.0, + "step": 15881 + }, + { + "epoch": 1.744124752910169, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5791172981262207, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7210353016853333, + "num_tokens": 410606232.0, + "step": 15882 + }, + { + "epoch": 1.7442345706127829, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5860981941223145, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6932371854782104, + "num_tokens": 410631951.0, + "step": 15883 + }, + { + "epoch": 1.7443443883153964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.411438465118408, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.735716700553894, + "num_tokens": 410660421.0, + "step": 15884 + }, + { + "epoch": 1.7444542060180102, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5999162197113037, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6914269924163818, + "num_tokens": 410688148.0, + "step": 15885 + }, + { + "epoch": 1.7445640237206237, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.2257936000823975, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6907512545585632, + "num_tokens": 410721790.0, + "step": 15886 + }, + { + "epoch": 1.7446738414232374, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.960355758666992, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7161950469017029, + "num_tokens": 410740257.0, + "step": 15887 + }, + { + "epoch": 1.7447836591258512, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5566811561584473, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7199786901473999, + "num_tokens": 410763734.0, + "step": 15888 + }, + { + "epoch": 1.7448934768284647, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7768285274505615, + "learning_rate": 1e-06, + "loss": 0.9532, + "mean_token_accuracy": 0.7166315913200378, + "num_tokens": 410785032.0, + "step": 15889 + }, + { + "epoch": 1.7450032945310783, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7102901935577393, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6774494647979736, + "num_tokens": 410810970.0, + "step": 15890 + }, + { + "epoch": 1.745113112233692, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.510964870452881, + "learning_rate": 1e-06, + "loss": 1.1164, + "mean_token_accuracy": 0.6797290444374084, + "num_tokens": 410841451.0, + "step": 15891 + }, + { + "epoch": 1.7452229299363058, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.626077890396118, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7067763209342957, + "num_tokens": 410865315.0, + "step": 15892 + }, + { + "epoch": 1.7453327476389195, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.319230079650879, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.6970101594924927, + "num_tokens": 410889008.0, + "step": 15893 + }, + { + "epoch": 1.745442565341533, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.755826234817505, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6903032064437866, + "num_tokens": 410912808.0, + "step": 15894 + }, + { + "epoch": 1.7455523830441466, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.524282693862915, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7040306329727173, + "num_tokens": 410940907.0, + "step": 15895 + }, + { + "epoch": 1.7456622007467604, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.651884078979492, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7034413814544678, + "num_tokens": 410964706.0, + "step": 15896 + }, + { + "epoch": 1.7457720184493741, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5862321853637695, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6966270804405212, + "num_tokens": 410992747.0, + "step": 15897 + }, + { + "epoch": 1.7458818361519877, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.626152992248535, + "learning_rate": 1e-06, + "loss": 0.8802, + "mean_token_accuracy": 0.7397804260253906, + "num_tokens": 411016194.0, + "step": 15898 + }, + { + "epoch": 1.7459916538546012, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.779667854309082, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6899871826171875, + "num_tokens": 411037537.0, + "step": 15899 + }, + { + "epoch": 1.746101471557215, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.679144859313965, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7155515551567078, + "num_tokens": 411061701.0, + "step": 15900 + }, + { + "epoch": 1.7462112892598287, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.616046667098999, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7107175588607788, + "num_tokens": 411086094.0, + "step": 15901 + }, + { + "epoch": 1.7463211069624425, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.789100170135498, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7003011703491211, + "num_tokens": 411107452.0, + "step": 15902 + }, + { + "epoch": 1.746430924665056, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6925883293151855, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.723564088344574, + "num_tokens": 411129754.0, + "step": 15903 + }, + { + "epoch": 1.7465407423676695, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.818633794784546, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6955938935279846, + "num_tokens": 411155350.0, + "step": 15904 + }, + { + "epoch": 1.7466505600702833, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.334282636642456, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.707747757434845, + "num_tokens": 411183490.0, + "step": 15905 + }, + { + "epoch": 1.746760377772897, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.585876703262329, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6949072480201721, + "num_tokens": 411210915.0, + "step": 15906 + }, + { + "epoch": 1.7468701954755108, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7129933834075928, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7124800682067871, + "num_tokens": 411237205.0, + "step": 15907 + }, + { + "epoch": 1.7469800131781243, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5370402336120605, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6925355195999146, + "num_tokens": 411264664.0, + "step": 15908 + }, + { + "epoch": 1.7470898308807379, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0824501514434814, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7041038274765015, + "num_tokens": 411291043.0, + "step": 15909 + }, + { + "epoch": 1.7471996485833516, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9195477962493896, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6916906833648682, + "num_tokens": 411313227.0, + "step": 15910 + }, + { + "epoch": 1.7473094662859654, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 7.97407865524292, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6829431056976318, + "num_tokens": 411339882.0, + "step": 15911 + }, + { + "epoch": 1.747419283988579, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7300727367401123, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.701371431350708, + "num_tokens": 411366020.0, + "step": 15912 + }, + { + "epoch": 1.7475291016911925, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.672302484512329, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7144694924354553, + "num_tokens": 411388987.0, + "step": 15913 + }, + { + "epoch": 1.7476389193938062, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4398882389068604, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7218226194381714, + "num_tokens": 411417028.0, + "step": 15914 + }, + { + "epoch": 1.74774873709642, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.801870584487915, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.6972126960754395, + "num_tokens": 411439589.0, + "step": 15915 + }, + { + "epoch": 1.7478585547990337, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8825411796569824, + "learning_rate": 1e-06, + "loss": 1.1211, + "mean_token_accuracy": 0.6850427389144897, + "num_tokens": 411461548.0, + "step": 15916 + }, + { + "epoch": 1.7479683725016473, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3041064739227295, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7173383235931396, + "num_tokens": 411489361.0, + "step": 15917 + }, + { + "epoch": 1.7480781902042608, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.481901168823242, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.698230504989624, + "num_tokens": 411519407.0, + "step": 15918 + }, + { + "epoch": 1.7481880079068746, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4135780334472656, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7113094329833984, + "num_tokens": 411547927.0, + "step": 15919 + }, + { + "epoch": 1.7482978256094883, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5014560222625732, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6983642578125, + "num_tokens": 411575453.0, + "step": 15920 + }, + { + "epoch": 1.7484076433121019, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 7.869171619415283, + "learning_rate": 1e-06, + "loss": 0.9785, + "mean_token_accuracy": 0.716910719871521, + "num_tokens": 411603169.0, + "step": 15921 + }, + { + "epoch": 1.7485174610147156, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.517545700073242, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6913224458694458, + "num_tokens": 411635310.0, + "step": 15922 + }, + { + "epoch": 1.7486272787173291, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.373490571975708, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7240517139434814, + "num_tokens": 411666693.0, + "step": 15923 + }, + { + "epoch": 1.748737096419943, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.623861312866211, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6957782506942749, + "num_tokens": 411694091.0, + "step": 15924 + }, + { + "epoch": 1.7488469141225567, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8139750957489014, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7057830095291138, + "num_tokens": 411715521.0, + "step": 15925 + }, + { + "epoch": 1.7489567318251702, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.365030288696289, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7095746994018555, + "num_tokens": 411744107.0, + "step": 15926 + }, + { + "epoch": 1.7490665495277837, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.481051206588745, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6978042125701904, + "num_tokens": 411774228.0, + "step": 15927 + }, + { + "epoch": 1.7491763672303975, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.0051729679107666, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7147569060325623, + "num_tokens": 411791279.0, + "step": 15928 + }, + { + "epoch": 1.7492861849330112, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.166437864303589, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7150434851646423, + "num_tokens": 411813971.0, + "step": 15929 + }, + { + "epoch": 1.749396002635625, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7307159900665283, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7079476118087769, + "num_tokens": 411836624.0, + "step": 15930 + }, + { + "epoch": 1.7495058203382385, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6132547855377197, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.7026985287666321, + "num_tokens": 411862879.0, + "step": 15931 + }, + { + "epoch": 1.749615638040852, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.612144708633423, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7090440392494202, + "num_tokens": 411890777.0, + "step": 15932 + }, + { + "epoch": 1.7497254557434658, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.21042799949646, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6949033141136169, + "num_tokens": 411929648.0, + "step": 15933 + }, + { + "epoch": 1.7498352734460796, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.240978956222534, + "learning_rate": 1e-06, + "loss": 0.9147, + "mean_token_accuracy": 0.7282711267471313, + "num_tokens": 411959716.0, + "step": 15934 + }, + { + "epoch": 1.7499450911486931, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6015164852142334, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7071548700332642, + "num_tokens": 411984294.0, + "step": 15935 + }, + { + "epoch": 1.7500549088513069, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5109734535217285, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7038827538490295, + "num_tokens": 412008278.0, + "step": 15936 + }, + { + "epoch": 1.7501647265539204, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.513165235519409, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.719474196434021, + "num_tokens": 412035117.0, + "step": 15937 + }, + { + "epoch": 1.7502745442565342, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.510639190673828, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7075111269950867, + "num_tokens": 412059205.0, + "step": 15938 + }, + { + "epoch": 1.750384361959148, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.755707025527954, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6946645379066467, + "num_tokens": 412082773.0, + "step": 15939 + }, + { + "epoch": 1.7504941796617615, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.533508539199829, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7033956050872803, + "num_tokens": 412111838.0, + "step": 15940 + }, + { + "epoch": 1.750603997364375, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.933729410171509, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7168853282928467, + "num_tokens": 412131623.0, + "step": 15941 + }, + { + "epoch": 1.7507138150669888, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.494868278503418, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7034791707992554, + "num_tokens": 412161242.0, + "step": 15942 + }, + { + "epoch": 1.7508236327696025, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.906198024749756, + "learning_rate": 1e-06, + "loss": 0.9261, + "mean_token_accuracy": 0.7232560515403748, + "num_tokens": 412182544.0, + "step": 15943 + }, + { + "epoch": 1.7509334504722163, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.687417507171631, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.6953447461128235, + "num_tokens": 412207490.0, + "step": 15944 + }, + { + "epoch": 1.7510432681748298, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5309271812438965, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6951347589492798, + "num_tokens": 412235966.0, + "step": 15945 + }, + { + "epoch": 1.7511530858774433, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7992608547210693, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.713257908821106, + "num_tokens": 412259956.0, + "step": 15946 + }, + { + "epoch": 1.751262903580057, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4772884845733643, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6846448183059692, + "num_tokens": 412289410.0, + "step": 15947 + }, + { + "epoch": 1.7513727212826709, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.565666913986206, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7057372331619263, + "num_tokens": 412316136.0, + "step": 15948 + }, + { + "epoch": 1.7514825389852844, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.413240909576416, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7117704153060913, + "num_tokens": 412344879.0, + "step": 15949 + }, + { + "epoch": 1.7515923566878981, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4595799446105957, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7089781165122986, + "num_tokens": 412371632.0, + "step": 15950 + }, + { + "epoch": 1.7517021743905117, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.315493583679199, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7178758382797241, + "num_tokens": 412400581.0, + "step": 15951 + }, + { + "epoch": 1.7518119920931254, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.69415283203125, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6851833462715149, + "num_tokens": 412425480.0, + "step": 15952 + }, + { + "epoch": 1.7519218097957392, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.754546642303467, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7092750072479248, + "num_tokens": 412448255.0, + "step": 15953 + }, + { + "epoch": 1.7520316274983527, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.71826434135437, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.70465487241745, + "num_tokens": 412471255.0, + "step": 15954 + }, + { + "epoch": 1.7521414452009663, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.7837939262390137, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.7275891304016113, + "num_tokens": 412491773.0, + "step": 15955 + }, + { + "epoch": 1.75225126290358, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.564602851867676, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.7061309218406677, + "num_tokens": 412521308.0, + "step": 15956 + }, + { + "epoch": 1.7523610806061938, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6981940269470215, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6995688676834106, + "num_tokens": 412547588.0, + "step": 15957 + }, + { + "epoch": 1.7524708983088075, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5115127563476562, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6892513036727905, + "num_tokens": 412576696.0, + "step": 15958 + }, + { + "epoch": 1.752580716011421, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.801755905151367, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.6951450109481812, + "num_tokens": 412598512.0, + "step": 15959 + }, + { + "epoch": 1.7526905337140346, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.646573543548584, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7039721012115479, + "num_tokens": 412624152.0, + "step": 15960 + }, + { + "epoch": 1.7528003514166484, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.450998306274414, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.725678563117981, + "num_tokens": 412649925.0, + "step": 15961 + }, + { + "epoch": 1.7529101691192621, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.129643201828003, + "learning_rate": 1e-06, + "loss": 1.0847, + "mean_token_accuracy": 0.6913264393806458, + "num_tokens": 412689788.0, + "step": 15962 + }, + { + "epoch": 1.7530199868218757, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6325275897979736, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.7267730832099915, + "num_tokens": 412712857.0, + "step": 15963 + }, + { + "epoch": 1.7531298045244892, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5924904346466064, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6927194595336914, + "num_tokens": 412739070.0, + "step": 15964 + }, + { + "epoch": 1.753239622227103, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.341888189315796, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7100105285644531, + "num_tokens": 412768299.0, + "step": 15965 + }, + { + "epoch": 1.7533494399297167, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 3.2724647521972656, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7023077607154846, + "num_tokens": 412790910.0, + "step": 15966 + }, + { + "epoch": 1.7534592576323305, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.613705635070801, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7092636823654175, + "num_tokens": 412816326.0, + "step": 15967 + }, + { + "epoch": 1.753569075334944, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5591235160827637, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7237867116928101, + "num_tokens": 412842083.0, + "step": 15968 + }, + { + "epoch": 1.7536788930375575, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.839542865753174, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6978256106376648, + "num_tokens": 412862592.0, + "step": 15969 + }, + { + "epoch": 1.7537887107401713, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9213814735412598, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7084919214248657, + "num_tokens": 412886549.0, + "step": 15970 + }, + { + "epoch": 1.753898528442785, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4897000789642334, + "learning_rate": 1e-06, + "loss": 1.0935, + "mean_token_accuracy": 0.6799659729003906, + "num_tokens": 412912929.0, + "step": 15971 + }, + { + "epoch": 1.7540083461453988, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6944079399108887, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7081055641174316, + "num_tokens": 412934306.0, + "step": 15972 + }, + { + "epoch": 1.7541181638480123, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.430800199508667, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.708210289478302, + "num_tokens": 412963595.0, + "step": 15973 + }, + { + "epoch": 1.7542279815506259, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.346113920211792, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6973936557769775, + "num_tokens": 412999685.0, + "step": 15974 + }, + { + "epoch": 1.7543377992532396, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3648386001586914, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7000889778137207, + "num_tokens": 413030652.0, + "step": 15975 + }, + { + "epoch": 1.7544476169558534, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.673461437225342, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6915175914764404, + "num_tokens": 413058995.0, + "step": 15976 + }, + { + "epoch": 1.754557434658467, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.923593759536743, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.6985475420951843, + "num_tokens": 413081066.0, + "step": 15977 + }, + { + "epoch": 1.7546672523610805, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5461673736572266, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7086818814277649, + "num_tokens": 413109710.0, + "step": 15978 + }, + { + "epoch": 1.7547770700636942, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5653984546661377, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7167664766311646, + "num_tokens": 413133397.0, + "step": 15979 + }, + { + "epoch": 1.754886887766308, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3650598526000977, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6975345611572266, + "num_tokens": 413165579.0, + "step": 15980 + }, + { + "epoch": 1.7549967054689217, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.773972988128662, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7205525636672974, + "num_tokens": 413190182.0, + "step": 15981 + }, + { + "epoch": 1.7551065231715353, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5581600666046143, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7031184434890747, + "num_tokens": 413214874.0, + "step": 15982 + }, + { + "epoch": 1.7552163408741488, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.366272449493408, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7035054564476013, + "num_tokens": 413244392.0, + "step": 15983 + }, + { + "epoch": 1.7553261585767626, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.389469861984253, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7033541202545166, + "num_tokens": 413276386.0, + "step": 15984 + }, + { + "epoch": 1.7554359762793763, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4083359241485596, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7093438506126404, + "num_tokens": 413304406.0, + "step": 15985 + }, + { + "epoch": 1.7555457939819898, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5270462036132812, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7113688588142395, + "num_tokens": 413329957.0, + "step": 15986 + }, + { + "epoch": 1.7556556116846036, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.574167251586914, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7054818272590637, + "num_tokens": 413355676.0, + "step": 15987 + }, + { + "epoch": 1.7557654293872171, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.568302631378174, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.7119591236114502, + "num_tokens": 413377392.0, + "step": 15988 + }, + { + "epoch": 1.755875247089831, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.9573631286621094, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7289873361587524, + "num_tokens": 413398830.0, + "step": 15989 + }, + { + "epoch": 1.7559850647924446, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3964836597442627, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6898403167724609, + "num_tokens": 413429855.0, + "step": 15990 + }, + { + "epoch": 1.7560948824950582, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4413630962371826, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7115216255187988, + "num_tokens": 413456455.0, + "step": 15991 + }, + { + "epoch": 1.7562047001976717, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.74094820022583, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7032350301742554, + "num_tokens": 413479740.0, + "step": 15992 + }, + { + "epoch": 1.7563145179002855, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6970863342285156, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6898139715194702, + "num_tokens": 413505148.0, + "step": 15993 + }, + { + "epoch": 1.7564243356028992, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6988725662231445, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.707295835018158, + "num_tokens": 413530087.0, + "step": 15994 + }, + { + "epoch": 1.756534153305513, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8986897468566895, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7225973606109619, + "num_tokens": 413556001.0, + "step": 15995 + }, + { + "epoch": 1.7566439710081265, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.756000518798828, + "learning_rate": 1e-06, + "loss": 0.8998, + "mean_token_accuracy": 0.7335495352745056, + "num_tokens": 413578697.0, + "step": 15996 + }, + { + "epoch": 1.75675378871074, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5847091674804688, + "learning_rate": 1e-06, + "loss": 1.0929, + "mean_token_accuracy": 0.6887439489364624, + "num_tokens": 413607179.0, + "step": 15997 + }, + { + "epoch": 1.7568636064133538, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.55352783203125, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6986945271492004, + "num_tokens": 413635705.0, + "step": 15998 + }, + { + "epoch": 1.7569734241159676, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5872771739959717, + "learning_rate": 1e-06, + "loss": 0.9225, + "mean_token_accuracy": 0.7324666380882263, + "num_tokens": 413658062.0, + "step": 15999 + }, + { + "epoch": 1.757083241818581, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.649432420730591, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7222554683685303, + "num_tokens": 413679936.0, + "step": 16000 + }, + { + "epoch": 1.7571930595211949, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.783834934234619, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7024462819099426, + "num_tokens": 413700924.0, + "step": 16001 + }, + { + "epoch": 1.7573028772238084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.918825149536133, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7154313325881958, + "num_tokens": 413720728.0, + "step": 16002 + }, + { + "epoch": 1.7574126949264222, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4799182415008545, + "learning_rate": 1e-06, + "loss": 0.9526, + "mean_token_accuracy": 0.7227138876914978, + "num_tokens": 413747592.0, + "step": 16003 + }, + { + "epoch": 1.757522512629036, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3844423294067383, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7086669206619263, + "num_tokens": 413779379.0, + "step": 16004 + }, + { + "epoch": 1.7576323303316495, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.792943000793457, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6992281675338745, + "num_tokens": 413801675.0, + "step": 16005 + }, + { + "epoch": 1.757742148034263, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5193710327148438, + "learning_rate": 1e-06, + "loss": 1.1014, + "mean_token_accuracy": 0.6914419531822205, + "num_tokens": 413828275.0, + "step": 16006 + }, + { + "epoch": 1.7578519657368767, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4341940879821777, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7088820338249207, + "num_tokens": 413859114.0, + "step": 16007 + }, + { + "epoch": 1.7579617834394905, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4509761333465576, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7204003930091858, + "num_tokens": 413886825.0, + "step": 16008 + }, + { + "epoch": 1.7580716011421043, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2674190998077393, + "learning_rate": 1e-06, + "loss": 1.1645, + "mean_token_accuracy": 0.672078013420105, + "num_tokens": 413925084.0, + "step": 16009 + }, + { + "epoch": 1.7581814188447178, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.719785451889038, + "learning_rate": 1e-06, + "loss": 1.0995, + "mean_token_accuracy": 0.6810593008995056, + "num_tokens": 413952206.0, + "step": 16010 + }, + { + "epoch": 1.7582912365473313, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7027764320373535, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.691404402256012, + "num_tokens": 413975722.0, + "step": 16011 + }, + { + "epoch": 1.758401054249945, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5689423084259033, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7229055166244507, + "num_tokens": 414000120.0, + "step": 16012 + }, + { + "epoch": 1.7585108719525588, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7807857990264893, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7068964242935181, + "num_tokens": 414022521.0, + "step": 16013 + }, + { + "epoch": 1.7586206896551724, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6003000736236572, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6913490891456604, + "num_tokens": 414047699.0, + "step": 16014 + }, + { + "epoch": 1.758730507357786, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7568609714508057, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7035782933235168, + "num_tokens": 414071095.0, + "step": 16015 + }, + { + "epoch": 1.7588403250603997, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.585303544998169, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7234305739402771, + "num_tokens": 414096434.0, + "step": 16016 + }, + { + "epoch": 1.7589501427630134, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.681880474090576, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7080234289169312, + "num_tokens": 414120608.0, + "step": 16017 + }, + { + "epoch": 1.7590599604656272, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6419384479522705, + "learning_rate": 1e-06, + "loss": 0.8795, + "mean_token_accuracy": 0.747675895690918, + "num_tokens": 414142862.0, + "step": 16018 + }, + { + "epoch": 1.7591697781682407, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4170796871185303, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6883271932601929, + "num_tokens": 414173434.0, + "step": 16019 + }, + { + "epoch": 1.7592795958708543, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.86686635017395, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7211612462997437, + "num_tokens": 414199609.0, + "step": 16020 + }, + { + "epoch": 1.759389413573468, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.639699935913086, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6999527812004089, + "num_tokens": 414225107.0, + "step": 16021 + }, + { + "epoch": 1.7594992312760818, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.324585199356079, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7251421809196472, + "num_tokens": 414251708.0, + "step": 16022 + }, + { + "epoch": 1.7596090489786955, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.446371078491211, + "learning_rate": 1e-06, + "loss": 0.919, + "mean_token_accuracy": 0.7378506660461426, + "num_tokens": 414279664.0, + "step": 16023 + }, + { + "epoch": 1.759718866681309, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4228718280792236, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6868218183517456, + "num_tokens": 414309844.0, + "step": 16024 + }, + { + "epoch": 1.7598286843839226, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.614152193069458, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7060413360595703, + "num_tokens": 414333773.0, + "step": 16025 + }, + { + "epoch": 1.7599385020865363, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.412672281265259, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7027823328971863, + "num_tokens": 414360364.0, + "step": 16026 + }, + { + "epoch": 1.76004831978915, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5638415813446045, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7322959899902344, + "num_tokens": 414387365.0, + "step": 16027 + }, + { + "epoch": 1.7601581374917636, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.618065595626831, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7156673669815063, + "num_tokens": 414413067.0, + "step": 16028 + }, + { + "epoch": 1.7602679551943772, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6266062259674072, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6955534815788269, + "num_tokens": 414440864.0, + "step": 16029 + }, + { + "epoch": 1.760377772896991, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.601583242416382, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6977521777153015, + "num_tokens": 414467877.0, + "step": 16030 + }, + { + "epoch": 1.7604875905996047, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5440895557403564, + "learning_rate": 1e-06, + "loss": 0.9401, + "mean_token_accuracy": 0.7239879369735718, + "num_tokens": 414492795.0, + "step": 16031 + }, + { + "epoch": 1.7605974083022184, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.557215929031372, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7172300219535828, + "num_tokens": 414519885.0, + "step": 16032 + }, + { + "epoch": 1.760707226004832, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3446037769317627, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7112279534339905, + "num_tokens": 414549805.0, + "step": 16033 + }, + { + "epoch": 1.7608170437074455, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4595768451690674, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7168399095535278, + "num_tokens": 414576865.0, + "step": 16034 + }, + { + "epoch": 1.7609268614100593, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5099048614501953, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6981422901153564, + "num_tokens": 414605167.0, + "step": 16035 + }, + { + "epoch": 1.761036679112673, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.837998151779175, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7109791040420532, + "num_tokens": 414626986.0, + "step": 16036 + }, + { + "epoch": 1.7611464968152868, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.515596866607666, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.695292592048645, + "num_tokens": 414655905.0, + "step": 16037 + }, + { + "epoch": 1.7612563145179003, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5533416271209717, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.696621298789978, + "num_tokens": 414682696.0, + "step": 16038 + }, + { + "epoch": 1.7613661322205139, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5362274646759033, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7068538665771484, + "num_tokens": 414709072.0, + "step": 16039 + }, + { + "epoch": 1.7614759499231276, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.577829360961914, + "learning_rate": 1e-06, + "loss": 0.9453, + "mean_token_accuracy": 0.7253546714782715, + "num_tokens": 414734163.0, + "step": 16040 + }, + { + "epoch": 1.7615857676257414, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.754683256149292, + "learning_rate": 1e-06, + "loss": 0.9297, + "mean_token_accuracy": 0.7269999980926514, + "num_tokens": 414757528.0, + "step": 16041 + }, + { + "epoch": 1.761695585328355, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.695688247680664, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.700245201587677, + "num_tokens": 414781026.0, + "step": 16042 + }, + { + "epoch": 1.7618054030309684, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.330073118209839, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7078279256820679, + "num_tokens": 414811374.0, + "step": 16043 + }, + { + "epoch": 1.7619152207335822, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6269969940185547, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7087469100952148, + "num_tokens": 414834105.0, + "step": 16044 + }, + { + "epoch": 1.762025038436196, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.357119560241699, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6930685043334961, + "num_tokens": 414863277.0, + "step": 16045 + }, + { + "epoch": 1.7621348561388097, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6976077556610107, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7169101238250732, + "num_tokens": 414888365.0, + "step": 16046 + }, + { + "epoch": 1.7622446738414232, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.441622257232666, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7169505953788757, + "num_tokens": 414914796.0, + "step": 16047 + }, + { + "epoch": 1.7623544915440368, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4507546424865723, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6966505646705627, + "num_tokens": 414946541.0, + "step": 16048 + }, + { + "epoch": 1.7624643092466505, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8792812824249268, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7063358426094055, + "num_tokens": 414970385.0, + "step": 16049 + }, + { + "epoch": 1.7625741269492643, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.463132381439209, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7022484540939331, + "num_tokens": 414997241.0, + "step": 16050 + }, + { + "epoch": 1.7626839446518778, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6205968856811523, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7054960131645203, + "num_tokens": 415025244.0, + "step": 16051 + }, + { + "epoch": 1.7627937623544916, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.3856160640716553, + "learning_rate": 1e-06, + "loss": 0.9237, + "mean_token_accuracy": 0.7316299676895142, + "num_tokens": 415051255.0, + "step": 16052 + }, + { + "epoch": 1.7629035800571051, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.377065896987915, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7214029431343079, + "num_tokens": 415080247.0, + "step": 16053 + }, + { + "epoch": 1.7630133977597189, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4143869876861572, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.6951147317886353, + "num_tokens": 415109155.0, + "step": 16054 + }, + { + "epoch": 1.7631232154623326, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5975184440612793, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7109652757644653, + "num_tokens": 415133469.0, + "step": 16055 + }, + { + "epoch": 1.7632330331649462, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7517776489257812, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6966572999954224, + "num_tokens": 415155715.0, + "step": 16056 + }, + { + "epoch": 1.7633428508675597, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.1604156494140625, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6982478499412537, + "num_tokens": 415192246.0, + "step": 16057 + }, + { + "epoch": 1.7634526685701735, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.6534457206726074, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.705571174621582, + "num_tokens": 415219216.0, + "step": 16058 + }, + { + "epoch": 1.7635624862727872, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6085469722747803, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7106038331985474, + "num_tokens": 415245103.0, + "step": 16059 + }, + { + "epoch": 1.763672303975401, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 7.898812770843506, + "learning_rate": 1e-06, + "loss": 1.1111, + "mean_token_accuracy": 0.6856324672698975, + "num_tokens": 415276220.0, + "step": 16060 + }, + { + "epoch": 1.7637821216780145, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.612276077270508, + "learning_rate": 1e-06, + "loss": 0.9404, + "mean_token_accuracy": 0.7183999419212341, + "num_tokens": 415303037.0, + "step": 16061 + }, + { + "epoch": 1.763891939380628, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9547617435455322, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7059407830238342, + "num_tokens": 415324617.0, + "step": 16062 + }, + { + "epoch": 1.7640017570832418, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7816786766052246, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.70367032289505, + "num_tokens": 415349364.0, + "step": 16063 + }, + { + "epoch": 1.7641115747858556, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.634155035018921, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6945428848266602, + "num_tokens": 415376114.0, + "step": 16064 + }, + { + "epoch": 1.764221392488469, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5125582218170166, + "learning_rate": 1e-06, + "loss": 1.1637, + "mean_token_accuracy": 0.6686691641807556, + "num_tokens": 415406244.0, + "step": 16065 + }, + { + "epoch": 1.7643312101910829, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.651986837387085, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6922675371170044, + "num_tokens": 415428018.0, + "step": 16066 + }, + { + "epoch": 1.7644410278936964, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5203471183776855, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.6901835799217224, + "num_tokens": 415455075.0, + "step": 16067 + }, + { + "epoch": 1.7645508455963101, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4359312057495117, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6903582811355591, + "num_tokens": 415485792.0, + "step": 16068 + }, + { + "epoch": 1.764660663298924, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6587698459625244, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6907505989074707, + "num_tokens": 415514560.0, + "step": 16069 + }, + { + "epoch": 1.7647704810015374, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.842012405395508, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7086067199707031, + "num_tokens": 415537026.0, + "step": 16070 + }, + { + "epoch": 1.764880298704151, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.4843287467956543, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6875410676002502, + "num_tokens": 415565677.0, + "step": 16071 + }, + { + "epoch": 1.7649901164067647, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.542407989501953, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.698491096496582, + "num_tokens": 415594587.0, + "step": 16072 + }, + { + "epoch": 1.7650999341093785, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6674511432647705, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7007005214691162, + "num_tokens": 415622755.0, + "step": 16073 + }, + { + "epoch": 1.7652097518119922, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.366180419921875, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7072468996047974, + "num_tokens": 415652436.0, + "step": 16074 + }, + { + "epoch": 1.7653195695146058, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.826434850692749, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6992696523666382, + "num_tokens": 415676475.0, + "step": 16075 + }, + { + "epoch": 1.7654293872172193, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.5392813682556152, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.6948807835578918, + "num_tokens": 415703020.0, + "step": 16076 + }, + { + "epoch": 1.765539204919833, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.565297842025757, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.695401132106781, + "num_tokens": 415729866.0, + "step": 16077 + }, + { + "epoch": 1.7656490226224468, + "ewc_loss": 9.5367431640625e-06, + "grad_norm": 2.8957695960998535, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7124764919281006, + "num_tokens": 415752142.0, + "step": 16078 + }, + { + "epoch": 1.7657588403250604, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5103225708007812, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6953069567680359, + "num_tokens": 415780469.0, + "step": 16079 + }, + { + "epoch": 1.765868658027674, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 7.902066707611084, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6929999589920044, + "num_tokens": 415809635.0, + "step": 16080 + }, + { + "epoch": 1.7659784757302877, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7381279468536377, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7004618644714355, + "num_tokens": 415832723.0, + "step": 16081 + }, + { + "epoch": 1.7660882934329014, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.689513683319092, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6904020309448242, + "num_tokens": 415857939.0, + "step": 16082 + }, + { + "epoch": 1.7661981111355152, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.637688398361206, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7036818861961365, + "num_tokens": 415881968.0, + "step": 16083 + }, + { + "epoch": 1.7663079288381287, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3894636631011963, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6989517211914062, + "num_tokens": 415911357.0, + "step": 16084 + }, + { + "epoch": 1.7664177465407422, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4319140911102295, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7171626091003418, + "num_tokens": 415936779.0, + "step": 16085 + }, + { + "epoch": 1.766527564243356, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6026296615600586, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6877855062484741, + "num_tokens": 415964168.0, + "step": 16086 + }, + { + "epoch": 1.7666373819459698, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6318888664245605, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7152705788612366, + "num_tokens": 415988411.0, + "step": 16087 + }, + { + "epoch": 1.7667471996485835, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3078458309173584, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6861401200294495, + "num_tokens": 416018363.0, + "step": 16088 + }, + { + "epoch": 1.766857017351197, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5535876750946045, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7194470763206482, + "num_tokens": 416046232.0, + "step": 16089 + }, + { + "epoch": 1.7669668350538106, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.431880474090576, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7226758599281311, + "num_tokens": 416073759.0, + "step": 16090 + }, + { + "epoch": 1.7670766527564243, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.536048650741577, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.7018325328826904, + "num_tokens": 416098163.0, + "step": 16091 + }, + { + "epoch": 1.767186470459038, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.297489881515503, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.7009465098381042, + "num_tokens": 416126423.0, + "step": 16092 + }, + { + "epoch": 1.7672962881616516, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.1536810398101807, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7265990376472473, + "num_tokens": 416147882.0, + "step": 16093 + }, + { + "epoch": 1.7674061058642652, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.739128828048706, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7088766098022461, + "num_tokens": 416169273.0, + "step": 16094 + }, + { + "epoch": 1.767515923566879, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7714693546295166, + "learning_rate": 1e-06, + "loss": 1.0725, + "mean_token_accuracy": 0.6876745223999023, + "num_tokens": 416192765.0, + "step": 16095 + }, + { + "epoch": 1.7676257412694927, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.59541916847229, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7110821008682251, + "num_tokens": 416216952.0, + "step": 16096 + }, + { + "epoch": 1.7677355589721064, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.544887065887451, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7161466479301453, + "num_tokens": 416243681.0, + "step": 16097 + }, + { + "epoch": 1.76784537667472, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4745259284973145, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7174409627914429, + "num_tokens": 416270792.0, + "step": 16098 + }, + { + "epoch": 1.7679551943773335, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.527968168258667, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6973900198936462, + "num_tokens": 416298995.0, + "step": 16099 + }, + { + "epoch": 1.7680650120799473, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6445858478546143, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7137660980224609, + "num_tokens": 416324419.0, + "step": 16100 + }, + { + "epoch": 1.768174829782561, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8423900604248047, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7132183909416199, + "num_tokens": 416347089.0, + "step": 16101 + }, + { + "epoch": 1.7682846474851748, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6666057109832764, + "learning_rate": 1e-06, + "loss": 1.1061, + "mean_token_accuracy": 0.6803460121154785, + "num_tokens": 416372595.0, + "step": 16102 + }, + { + "epoch": 1.7683944651877883, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7938191890716553, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7113279700279236, + "num_tokens": 416393401.0, + "step": 16103 + }, + { + "epoch": 1.7685042828904018, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7656028270721436, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7144691944122314, + "num_tokens": 416416802.0, + "step": 16104 + }, + { + "epoch": 1.7686141005930156, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5448708534240723, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6934303045272827, + "num_tokens": 416446841.0, + "step": 16105 + }, + { + "epoch": 1.7687239182956294, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.615960121154785, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7250735759735107, + "num_tokens": 416471615.0, + "step": 16106 + }, + { + "epoch": 1.768833735998243, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4514882564544678, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.694801390171051, + "num_tokens": 416500878.0, + "step": 16107 + }, + { + "epoch": 1.7689435537008564, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.691065788269043, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.717133641242981, + "num_tokens": 416524480.0, + "step": 16108 + }, + { + "epoch": 1.7690533714034702, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.645806074142456, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7189745903015137, + "num_tokens": 416549422.0, + "step": 16109 + }, + { + "epoch": 1.769163189106084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9444940090179443, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7211905121803284, + "num_tokens": 416574267.0, + "step": 16110 + }, + { + "epoch": 1.7692730068086977, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.49713397026062, + "learning_rate": 1e-06, + "loss": 1.082, + "mean_token_accuracy": 0.6935827136039734, + "num_tokens": 416602495.0, + "step": 16111 + }, + { + "epoch": 1.7693828245113112, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4318690299987793, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6958581209182739, + "num_tokens": 416631999.0, + "step": 16112 + }, + { + "epoch": 1.7694926422139248, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.389488458633423, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6950525641441345, + "num_tokens": 416664289.0, + "step": 16113 + }, + { + "epoch": 1.7696024599165385, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.826603412628174, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.702601432800293, + "num_tokens": 416686554.0, + "step": 16114 + }, + { + "epoch": 1.7697122776191523, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.563633918762207, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6995548009872437, + "num_tokens": 416715570.0, + "step": 16115 + }, + { + "epoch": 1.7698220953217658, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4750964641571045, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6933172941207886, + "num_tokens": 416743489.0, + "step": 16116 + }, + { + "epoch": 1.7699319130243796, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4650886058807373, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7080862522125244, + "num_tokens": 416771487.0, + "step": 16117 + }, + { + "epoch": 1.7700417307269931, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5156352519989014, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7041324377059937, + "num_tokens": 416801416.0, + "step": 16118 + }, + { + "epoch": 1.7701515484296069, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.747096061706543, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7165725231170654, + "num_tokens": 416823188.0, + "step": 16119 + }, + { + "epoch": 1.7702613661322206, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6892993450164795, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7063074111938477, + "num_tokens": 416849104.0, + "step": 16120 + }, + { + "epoch": 1.7703711838348342, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5376765727996826, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6905447244644165, + "num_tokens": 416876435.0, + "step": 16121 + }, + { + "epoch": 1.7704810015374477, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3390164375305176, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7035340070724487, + "num_tokens": 416910832.0, + "step": 16122 + }, + { + "epoch": 1.7705908192400615, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6298978328704834, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6931449174880981, + "num_tokens": 416939844.0, + "step": 16123 + }, + { + "epoch": 1.7707006369426752, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.490131378173828, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7182155251502991, + "num_tokens": 416966003.0, + "step": 16124 + }, + { + "epoch": 1.770810454645289, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8019444942474365, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7097310423851013, + "num_tokens": 416985736.0, + "step": 16125 + }, + { + "epoch": 1.7709202723479025, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8283329010009766, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7069208025932312, + "num_tokens": 417007408.0, + "step": 16126 + }, + { + "epoch": 1.771030090050516, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.472414016723633, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7219001054763794, + "num_tokens": 417039428.0, + "step": 16127 + }, + { + "epoch": 1.7711399077531298, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5454084873199463, + "learning_rate": 1e-06, + "loss": 0.8821, + "mean_token_accuracy": 0.7315072417259216, + "num_tokens": 417062279.0, + "step": 16128 + }, + { + "epoch": 1.7712497254557436, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6106767654418945, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7153738737106323, + "num_tokens": 417086262.0, + "step": 16129 + }, + { + "epoch": 1.771359543158357, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5358784198760986, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6955274343490601, + "num_tokens": 417113676.0, + "step": 16130 + }, + { + "epoch": 1.7714693608609708, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.545380115509033, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.707222580909729, + "num_tokens": 417137542.0, + "step": 16131 + }, + { + "epoch": 1.7715791785635844, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.767690896987915, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7042717337608337, + "num_tokens": 417160678.0, + "step": 16132 + }, + { + "epoch": 1.7716889962661981, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.790057420730591, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6945233345031738, + "num_tokens": 417183766.0, + "step": 16133 + }, + { + "epoch": 1.771798813968812, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.722050428390503, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6838936805725098, + "num_tokens": 417207715.0, + "step": 16134 + }, + { + "epoch": 1.7719086316714254, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6613495349884033, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7129197716712952, + "num_tokens": 417232832.0, + "step": 16135 + }, + { + "epoch": 1.772018449374039, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4431803226470947, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.705803632736206, + "num_tokens": 417262091.0, + "step": 16136 + }, + { + "epoch": 1.7721282670766527, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5228378772735596, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6971795558929443, + "num_tokens": 417293057.0, + "step": 16137 + }, + { + "epoch": 1.7722380847792665, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.649005174636841, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7091763019561768, + "num_tokens": 417317893.0, + "step": 16138 + }, + { + "epoch": 1.7723479024818802, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.489144802093506, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7165958881378174, + "num_tokens": 417345068.0, + "step": 16139 + }, + { + "epoch": 1.7724577201844938, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.532635450363159, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7167848348617554, + "num_tokens": 417372931.0, + "step": 16140 + }, + { + "epoch": 1.7725675378871073, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.71639347076416, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6987988948822021, + "num_tokens": 417395646.0, + "step": 16141 + }, + { + "epoch": 1.772677355589721, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4827632904052734, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6949994564056396, + "num_tokens": 417426697.0, + "step": 16142 + }, + { + "epoch": 1.7727871732923348, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7961959838867188, + "learning_rate": 1e-06, + "loss": 0.9186, + "mean_token_accuracy": 0.7311558723449707, + "num_tokens": 417448738.0, + "step": 16143 + }, + { + "epoch": 1.7728969909949484, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.765302896499634, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7155517935752869, + "num_tokens": 417472669.0, + "step": 16144 + }, + { + "epoch": 1.7730068086975619, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3767783641815186, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7136446237564087, + "num_tokens": 417500389.0, + "step": 16145 + }, + { + "epoch": 1.7731166264001756, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4985766410827637, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6901187300682068, + "num_tokens": 417529568.0, + "step": 16146 + }, + { + "epoch": 1.7732264441027894, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.522137403488159, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7083323001861572, + "num_tokens": 417554658.0, + "step": 16147 + }, + { + "epoch": 1.7733362618054032, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.451092481613159, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7003874778747559, + "num_tokens": 417587294.0, + "step": 16148 + }, + { + "epoch": 1.7734460795080167, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6407477855682373, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7076725959777832, + "num_tokens": 417612958.0, + "step": 16149 + }, + { + "epoch": 1.7735558972106302, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7810285091400146, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7150393724441528, + "num_tokens": 417634148.0, + "step": 16150 + }, + { + "epoch": 1.773665714913244, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.515509605407715, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.6983504891395569, + "num_tokens": 417659456.0, + "step": 16151 + }, + { + "epoch": 1.7737755326158577, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4137301445007324, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6906754970550537, + "num_tokens": 417687557.0, + "step": 16152 + }, + { + "epoch": 1.7738853503184715, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.437987804412842, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7011440396308899, + "num_tokens": 417712667.0, + "step": 16153 + }, + { + "epoch": 1.773995168021085, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.661456346511841, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7238869071006775, + "num_tokens": 417735849.0, + "step": 16154 + }, + { + "epoch": 1.7741049857236986, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4743852615356445, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.688928484916687, + "num_tokens": 417765359.0, + "step": 16155 + }, + { + "epoch": 1.7742148034263123, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.312469720840454, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6893582344055176, + "num_tokens": 417797089.0, + "step": 16156 + }, + { + "epoch": 1.774324621128926, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6342625617980957, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7101894617080688, + "num_tokens": 417820062.0, + "step": 16157 + }, + { + "epoch": 1.7744344388315396, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5734808444976807, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.684115469455719, + "num_tokens": 417848752.0, + "step": 16158 + }, + { + "epoch": 1.7745442565341532, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2341408729553223, + "learning_rate": 1e-06, + "loss": 1.1488, + "mean_token_accuracy": 0.672677755355835, + "num_tokens": 417881739.0, + "step": 16159 + }, + { + "epoch": 1.774654074236767, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.398986339569092, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7040408849716187, + "num_tokens": 417913173.0, + "step": 16160 + }, + { + "epoch": 1.7747638919393807, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.504127264022827, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7073690891265869, + "num_tokens": 417939617.0, + "step": 16161 + }, + { + "epoch": 1.7748737096419944, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.563961982727051, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7109699845314026, + "num_tokens": 417963686.0, + "step": 16162 + }, + { + "epoch": 1.774983527344608, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4417924880981445, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6964637041091919, + "num_tokens": 417992361.0, + "step": 16163 + }, + { + "epoch": 1.7750933450472215, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5327670574188232, + "learning_rate": 1e-06, + "loss": 0.9209, + "mean_token_accuracy": 0.7292908430099487, + "num_tokens": 418017981.0, + "step": 16164 + }, + { + "epoch": 1.7752031627498353, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6540892124176025, + "learning_rate": 1e-06, + "loss": 0.9143, + "mean_token_accuracy": 0.7230079770088196, + "num_tokens": 418043341.0, + "step": 16165 + }, + { + "epoch": 1.775312980452449, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5824568271636963, + "learning_rate": 1e-06, + "loss": 1.0994, + "mean_token_accuracy": 0.6815654039382935, + "num_tokens": 418069128.0, + "step": 16166 + }, + { + "epoch": 1.7754227981550625, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.918107748031616, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7151198387145996, + "num_tokens": 418090271.0, + "step": 16167 + }, + { + "epoch": 1.7755326158576763, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8107335567474365, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7195441722869873, + "num_tokens": 418113119.0, + "step": 16168 + }, + { + "epoch": 1.7756424335602898, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.569105625152588, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7114257216453552, + "num_tokens": 418138183.0, + "step": 16169 + }, + { + "epoch": 1.7757522512629036, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3821280002593994, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6847994327545166, + "num_tokens": 418171349.0, + "step": 16170 + }, + { + "epoch": 1.7758620689655173, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.30183744430542, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7064656019210815, + "num_tokens": 418201351.0, + "step": 16171 + }, + { + "epoch": 1.7759718866681309, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5109691619873047, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7225558757781982, + "num_tokens": 418226757.0, + "step": 16172 + }, + { + "epoch": 1.7760817043707444, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7249884605407715, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7103400826454163, + "num_tokens": 418250369.0, + "step": 16173 + }, + { + "epoch": 1.7761915220733582, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.898846387863159, + "learning_rate": 1e-06, + "loss": 0.9105, + "mean_token_accuracy": 0.7395029067993164, + "num_tokens": 418271077.0, + "step": 16174 + }, + { + "epoch": 1.776301339775972, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8330671787261963, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.722801685333252, + "num_tokens": 418292164.0, + "step": 16175 + }, + { + "epoch": 1.7764111574785857, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7337167263031006, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7160705924034119, + "num_tokens": 418315319.0, + "step": 16176 + }, + { + "epoch": 1.7765209751811992, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.711763381958008, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.688849151134491, + "num_tokens": 418339650.0, + "step": 16177 + }, + { + "epoch": 1.7766307928838128, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.368255376815796, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7066057324409485, + "num_tokens": 418370376.0, + "step": 16178 + }, + { + "epoch": 1.7767406105864265, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5319676399230957, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6856175661087036, + "num_tokens": 418395633.0, + "step": 16179 + }, + { + "epoch": 1.7768504282890403, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.826077699661255, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7097460031509399, + "num_tokens": 418417308.0, + "step": 16180 + }, + { + "epoch": 1.7769602459916538, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6484735012054443, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7106175422668457, + "num_tokens": 418441138.0, + "step": 16181 + }, + { + "epoch": 1.7770700636942676, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.313046932220459, + "learning_rate": 1e-06, + "loss": 1.1134, + "mean_token_accuracy": 0.6878013610839844, + "num_tokens": 418470919.0, + "step": 16182 + }, + { + "epoch": 1.777179881396881, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.672390937805176, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7005103230476379, + "num_tokens": 418493758.0, + "step": 16183 + }, + { + "epoch": 1.7772896990994949, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5082669258117676, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6892530918121338, + "num_tokens": 418520258.0, + "step": 16184 + }, + { + "epoch": 1.7773995168021086, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.63604736328125, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7239737510681152, + "num_tokens": 418544191.0, + "step": 16185 + }, + { + "epoch": 1.7775093345047221, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7878427505493164, + "learning_rate": 1e-06, + "loss": 0.8517, + "mean_token_accuracy": 0.7469297647476196, + "num_tokens": 418564807.0, + "step": 16186 + }, + { + "epoch": 1.7776191522073357, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4549009799957275, + "learning_rate": 1e-06, + "loss": 0.9019, + "mean_token_accuracy": 0.7256403565406799, + "num_tokens": 418590708.0, + "step": 16187 + }, + { + "epoch": 1.7777289699099494, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.521374464035034, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6971130967140198, + "num_tokens": 418617068.0, + "step": 16188 + }, + { + "epoch": 1.7778387876125632, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5777652263641357, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7085311412811279, + "num_tokens": 418640811.0, + "step": 16189 + }, + { + "epoch": 1.777948605315177, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5821001529693604, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7065993547439575, + "num_tokens": 418667841.0, + "step": 16190 + }, + { + "epoch": 1.7780584230177905, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2984564304351807, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7057385444641113, + "num_tokens": 418696645.0, + "step": 16191 + }, + { + "epoch": 1.778168240720404, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3730804920196533, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7149269580841064, + "num_tokens": 418726463.0, + "step": 16192 + }, + { + "epoch": 1.7782780584230178, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3615636825561523, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.699152946472168, + "num_tokens": 418755149.0, + "step": 16193 + }, + { + "epoch": 1.7783878761256315, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5838749408721924, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7124111652374268, + "num_tokens": 418778806.0, + "step": 16194 + }, + { + "epoch": 1.778497693828245, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8196663856506348, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7225675582885742, + "num_tokens": 418799077.0, + "step": 16195 + }, + { + "epoch": 1.7786075115308586, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7523610591888428, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7196817398071289, + "num_tokens": 418823886.0, + "step": 16196 + }, + { + "epoch": 1.7787173292334724, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5835938453674316, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6988998651504517, + "num_tokens": 418850668.0, + "step": 16197 + }, + { + "epoch": 1.7788271469360861, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5709354877471924, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.709374725818634, + "num_tokens": 418876540.0, + "step": 16198 + }, + { + "epoch": 1.7789369646386999, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.928011655807495, + "learning_rate": 1e-06, + "loss": 0.8579, + "mean_token_accuracy": 0.742964506149292, + "num_tokens": 418894376.0, + "step": 16199 + }, + { + "epoch": 1.7790467823413134, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.578913927078247, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6922376751899719, + "num_tokens": 418919237.0, + "step": 16200 + }, + { + "epoch": 1.779156600043927, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.297546148300171, + "learning_rate": 1e-06, + "loss": 1.1106, + "mean_token_accuracy": 0.6751612424850464, + "num_tokens": 418954808.0, + "step": 16201 + }, + { + "epoch": 1.7792664177465407, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.603473424911499, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7006113529205322, + "num_tokens": 418979222.0, + "step": 16202 + }, + { + "epoch": 1.7793762354491545, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4622607231140137, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7198989391326904, + "num_tokens": 419005298.0, + "step": 16203 + }, + { + "epoch": 1.7794860531517682, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5490095615386963, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.6985781192779541, + "num_tokens": 419031203.0, + "step": 16204 + }, + { + "epoch": 1.7795958708543818, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.511317491531372, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7237330079078674, + "num_tokens": 419056050.0, + "step": 16205 + }, + { + "epoch": 1.7797056885569953, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6005473136901855, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.7001115083694458, + "num_tokens": 419085664.0, + "step": 16206 + }, + { + "epoch": 1.779815506259609, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.017674446105957, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.705409586429596, + "num_tokens": 419107035.0, + "step": 16207 + }, + { + "epoch": 1.7799253239622228, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.669489860534668, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7049425840377808, + "num_tokens": 419134962.0, + "step": 16208 + }, + { + "epoch": 1.7800351416648363, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.870298385620117, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7077975273132324, + "num_tokens": 419156469.0, + "step": 16209 + }, + { + "epoch": 1.7801449593674499, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.814903974533081, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.698309063911438, + "num_tokens": 419180412.0, + "step": 16210 + }, + { + "epoch": 1.7802547770700636, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3852319717407227, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7051303386688232, + "num_tokens": 419209102.0, + "step": 16211 + }, + { + "epoch": 1.7803645947726774, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.000145435333252, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7171194553375244, + "num_tokens": 419230360.0, + "step": 16212 + }, + { + "epoch": 1.7804744124752911, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5012125968933105, + "learning_rate": 1e-06, + "loss": 1.167, + "mean_token_accuracy": 0.6630575656890869, + "num_tokens": 419263585.0, + "step": 16213 + }, + { + "epoch": 1.7805842301779047, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.81593918800354, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.6980305314064026, + "num_tokens": 419290289.0, + "step": 16214 + }, + { + "epoch": 1.7806940478805182, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5767879486083984, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.693464994430542, + "num_tokens": 419315191.0, + "step": 16215 + }, + { + "epoch": 1.780803865583132, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4109835624694824, + "learning_rate": 1e-06, + "loss": 1.136, + "mean_token_accuracy": 0.6732851266860962, + "num_tokens": 419345979.0, + "step": 16216 + }, + { + "epoch": 1.7809136832857457, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7048847675323486, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.7279304265975952, + "num_tokens": 419369204.0, + "step": 16217 + }, + { + "epoch": 1.7810235009883595, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.616696834564209, + "learning_rate": 1e-06, + "loss": 0.9313, + "mean_token_accuracy": 0.7259696125984192, + "num_tokens": 419393122.0, + "step": 16218 + }, + { + "epoch": 1.781133318690973, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5658555030822754, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6989820003509521, + "num_tokens": 419420255.0, + "step": 16219 + }, + { + "epoch": 1.7812431363935866, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.548832654953003, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7049746513366699, + "num_tokens": 419445421.0, + "step": 16220 + }, + { + "epoch": 1.7813529540962003, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.628845453262329, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.685592770576477, + "num_tokens": 419473130.0, + "step": 16221 + }, + { + "epoch": 1.781462771798814, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8602538108825684, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.6997994184494019, + "num_tokens": 419494795.0, + "step": 16222 + }, + { + "epoch": 1.7815725895014276, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8193814754486084, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6939451694488525, + "num_tokens": 419519442.0, + "step": 16223 + }, + { + "epoch": 1.7816824072040411, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.721301555633545, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.6995066404342651, + "num_tokens": 419543835.0, + "step": 16224 + }, + { + "epoch": 1.781792224906655, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8030922412872314, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7053941488265991, + "num_tokens": 419565539.0, + "step": 16225 + }, + { + "epoch": 1.7819020426092687, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.638732671737671, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.708882749080658, + "num_tokens": 419589785.0, + "step": 16226 + }, + { + "epoch": 1.7820118603118824, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.522207736968994, + "learning_rate": 1e-06, + "loss": 0.8814, + "mean_token_accuracy": 0.7402111291885376, + "num_tokens": 419614859.0, + "step": 16227 + }, + { + "epoch": 1.782121678014496, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.284050464630127, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7221872806549072, + "num_tokens": 419645316.0, + "step": 16228 + }, + { + "epoch": 1.7822314957171095, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6826999187469482, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7214055061340332, + "num_tokens": 419669390.0, + "step": 16229 + }, + { + "epoch": 1.7823413134197232, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8823792934417725, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6908924579620361, + "num_tokens": 419692156.0, + "step": 16230 + }, + { + "epoch": 1.782451131122337, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.679166316986084, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7113985419273376, + "num_tokens": 419718398.0, + "step": 16231 + }, + { + "epoch": 1.7825609488249505, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5042266845703125, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7145172357559204, + "num_tokens": 419743950.0, + "step": 16232 + }, + { + "epoch": 1.7826707665275643, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8151443004608154, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7265753149986267, + "num_tokens": 419764640.0, + "step": 16233 + }, + { + "epoch": 1.7827805842301778, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.627504825592041, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7082778811454773, + "num_tokens": 419792119.0, + "step": 16234 + }, + { + "epoch": 1.7828904019327916, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.33655047416687, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7123335599899292, + "num_tokens": 419820532.0, + "step": 16235 + }, + { + "epoch": 1.7830002196354053, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.0776028633117676, + "learning_rate": 1e-06, + "loss": 0.9177, + "mean_token_accuracy": 0.7318416237831116, + "num_tokens": 419845548.0, + "step": 16236 + }, + { + "epoch": 1.7831100373380189, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7292258739471436, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7102323770523071, + "num_tokens": 419867079.0, + "step": 16237 + }, + { + "epoch": 1.7832198550406324, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6425700187683105, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7123093008995056, + "num_tokens": 419891368.0, + "step": 16238 + }, + { + "epoch": 1.7833296727432462, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.715594530105591, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.7011127471923828, + "num_tokens": 419915401.0, + "step": 16239 + }, + { + "epoch": 1.78343949044586, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5268218517303467, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.704571008682251, + "num_tokens": 419941102.0, + "step": 16240 + }, + { + "epoch": 1.7835493081484737, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.0361294746398926, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7134314775466919, + "num_tokens": 419960710.0, + "step": 16241 + }, + { + "epoch": 1.7836591258510872, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.762962579727173, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7179346084594727, + "num_tokens": 419979796.0, + "step": 16242 + }, + { + "epoch": 1.7837689435537007, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9755921363830566, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7119992971420288, + "num_tokens": 419999905.0, + "step": 16243 + }, + { + "epoch": 1.7838787612563145, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7882261276245117, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7012888193130493, + "num_tokens": 420029631.0, + "step": 16244 + }, + { + "epoch": 1.7839885789589283, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7397987842559814, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7149444818496704, + "num_tokens": 420051474.0, + "step": 16245 + }, + { + "epoch": 1.7840983966615418, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.7318203449249268, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7110142707824707, + "num_tokens": 420071650.0, + "step": 16246 + }, + { + "epoch": 1.7842082143641556, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8236801624298096, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.7052766680717468, + "num_tokens": 420094751.0, + "step": 16247 + }, + { + "epoch": 1.784318032066769, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.529782295227051, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.711028516292572, + "num_tokens": 420121656.0, + "step": 16248 + }, + { + "epoch": 1.7844278497693828, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6907362937927246, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.719097375869751, + "num_tokens": 420146275.0, + "step": 16249 + }, + { + "epoch": 1.7845376674719966, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5480666160583496, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6897838711738586, + "num_tokens": 420173068.0, + "step": 16250 + }, + { + "epoch": 1.7846474851746101, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3526220321655273, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7041229605674744, + "num_tokens": 420202313.0, + "step": 16251 + }, + { + "epoch": 1.7847573028772237, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6755433082580566, + "learning_rate": 1e-06, + "loss": 0.9111, + "mean_token_accuracy": 0.7334722280502319, + "num_tokens": 420225258.0, + "step": 16252 + }, + { + "epoch": 1.7848671205798374, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.402151107788086, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7068630456924438, + "num_tokens": 420254837.0, + "step": 16253 + }, + { + "epoch": 1.7849769382824512, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3514833450317383, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.686745285987854, + "num_tokens": 420284882.0, + "step": 16254 + }, + { + "epoch": 1.785086755985065, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5434582233428955, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.7019036412239075, + "num_tokens": 420312526.0, + "step": 16255 + }, + { + "epoch": 1.7851965736876785, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.45352840423584, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.6984549760818481, + "num_tokens": 420341823.0, + "step": 16256 + }, + { + "epoch": 1.785306391390292, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.333148241043091, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7087392807006836, + "num_tokens": 420372663.0, + "step": 16257 + }, + { + "epoch": 1.7854162090929058, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.914799213409424, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6970630884170532, + "num_tokens": 420396983.0, + "step": 16258 + }, + { + "epoch": 1.7855260267955195, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5189402103424072, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7084767818450928, + "num_tokens": 420421503.0, + "step": 16259 + }, + { + "epoch": 1.785635844498133, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.499056339263916, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6860909461975098, + "num_tokens": 420451332.0, + "step": 16260 + }, + { + "epoch": 1.7857456622007466, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.405949354171753, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7028788328170776, + "num_tokens": 420481896.0, + "step": 16261 + }, + { + "epoch": 1.7858554799033604, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6587440967559814, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7162479162216187, + "num_tokens": 420507325.0, + "step": 16262 + }, + { + "epoch": 1.7859652976059741, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.692420244216919, + "learning_rate": 1e-06, + "loss": 0.8719, + "mean_token_accuracy": 0.7390186786651611, + "num_tokens": 420532423.0, + "step": 16263 + }, + { + "epoch": 1.7860751153085879, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4834229946136475, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7061289548873901, + "num_tokens": 420558188.0, + "step": 16264 + }, + { + "epoch": 1.7861849330112014, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.498222589492798, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7166362404823303, + "num_tokens": 420583440.0, + "step": 16265 + }, + { + "epoch": 1.786294750713815, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.503965377807617, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7156703472137451, + "num_tokens": 420608649.0, + "step": 16266 + }, + { + "epoch": 1.7864045684164287, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.2046120166778564, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7112457752227783, + "num_tokens": 420627120.0, + "step": 16267 + }, + { + "epoch": 1.7865143861190425, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4132273197174072, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.6975253820419312, + "num_tokens": 420656474.0, + "step": 16268 + }, + { + "epoch": 1.7866242038216562, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.443934679031372, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7211792469024658, + "num_tokens": 420683699.0, + "step": 16269 + }, + { + "epoch": 1.7867340215242697, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5539581775665283, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.7001898288726807, + "num_tokens": 420708766.0, + "step": 16270 + }, + { + "epoch": 1.7868438392268833, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.444486379623413, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6836087107658386, + "num_tokens": 420741260.0, + "step": 16271 + }, + { + "epoch": 1.786953656929497, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.422344446182251, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.704810619354248, + "num_tokens": 420768544.0, + "step": 16272 + }, + { + "epoch": 1.7870634746321108, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6308958530426025, + "learning_rate": 1e-06, + "loss": 0.9382, + "mean_token_accuracy": 0.7261002063751221, + "num_tokens": 420790272.0, + "step": 16273 + }, + { + "epoch": 1.7871732923347243, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.484175205230713, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6967366337776184, + "num_tokens": 420818546.0, + "step": 16274 + }, + { + "epoch": 1.7872831100373379, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.873347520828247, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6876582503318787, + "num_tokens": 420848155.0, + "step": 16275 + }, + { + "epoch": 1.7873929277399516, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.577568531036377, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7117602825164795, + "num_tokens": 420873991.0, + "step": 16276 + }, + { + "epoch": 1.7875027454425654, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.2912776470184326, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7084172964096069, + "num_tokens": 420899750.0, + "step": 16277 + }, + { + "epoch": 1.7876125631451791, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.568728446960449, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.702831506729126, + "num_tokens": 420926433.0, + "step": 16278 + }, + { + "epoch": 1.7877223808477927, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.833850145339966, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7181709408760071, + "num_tokens": 420951268.0, + "step": 16279 + }, + { + "epoch": 1.7878321985504062, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3214714527130127, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7034308910369873, + "num_tokens": 420981931.0, + "step": 16280 + }, + { + "epoch": 1.78794201625302, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.703117847442627, + "learning_rate": 1e-06, + "loss": 1.1187, + "mean_token_accuracy": 0.6761562824249268, + "num_tokens": 421007442.0, + "step": 16281 + }, + { + "epoch": 1.7880518339556337, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4783880710601807, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6989344358444214, + "num_tokens": 421034920.0, + "step": 16282 + }, + { + "epoch": 1.7881616516582475, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5701894760131836, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.688022255897522, + "num_tokens": 421066890.0, + "step": 16283 + }, + { + "epoch": 1.788271469360861, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.69558048248291, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6872882843017578, + "num_tokens": 421090461.0, + "step": 16284 + }, + { + "epoch": 1.7883812870634745, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.794804096221924, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7081785202026367, + "num_tokens": 421112929.0, + "step": 16285 + }, + { + "epoch": 1.7884911047660883, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.886650800704956, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7105339765548706, + "num_tokens": 421137933.0, + "step": 16286 + }, + { + "epoch": 1.788600922468702, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4576587677001953, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7097024917602539, + "num_tokens": 421165253.0, + "step": 16287 + }, + { + "epoch": 1.7887107401713156, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5756149291992188, + "learning_rate": 1e-06, + "loss": 0.8979, + "mean_token_accuracy": 0.7327526807785034, + "num_tokens": 421187552.0, + "step": 16288 + }, + { + "epoch": 1.7888205578739291, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4445464611053467, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7149296402931213, + "num_tokens": 421213194.0, + "step": 16289 + }, + { + "epoch": 1.7889303755765429, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.677612781524658, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.7054346799850464, + "num_tokens": 421236247.0, + "step": 16290 + }, + { + "epoch": 1.7890401932791566, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9630346298217773, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7106562852859497, + "num_tokens": 421257100.0, + "step": 16291 + }, + { + "epoch": 1.7891500109817704, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6081039905548096, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7167474031448364, + "num_tokens": 421280175.0, + "step": 16292 + }, + { + "epoch": 1.789259828684384, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7941880226135254, + "learning_rate": 1e-06, + "loss": 0.8809, + "mean_token_accuracy": 0.7371585369110107, + "num_tokens": 421300730.0, + "step": 16293 + }, + { + "epoch": 1.7893696463869975, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6801609992980957, + "learning_rate": 1e-06, + "loss": 0.9237, + "mean_token_accuracy": 0.7283749580383301, + "num_tokens": 421323715.0, + "step": 16294 + }, + { + "epoch": 1.7894794640896112, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.70048451423645, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6960443258285522, + "num_tokens": 421346426.0, + "step": 16295 + }, + { + "epoch": 1.789589281792225, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7409796714782715, + "learning_rate": 1e-06, + "loss": 1.1286, + "mean_token_accuracy": 0.6771162748336792, + "num_tokens": 421371371.0, + "step": 16296 + }, + { + "epoch": 1.7896990994948385, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.664978504180908, + "learning_rate": 1e-06, + "loss": 1.0909, + "mean_token_accuracy": 0.6913763284683228, + "num_tokens": 421396099.0, + "step": 16297 + }, + { + "epoch": 1.7898089171974523, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.451202392578125, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7097623348236084, + "num_tokens": 421424550.0, + "step": 16298 + }, + { + "epoch": 1.7899187349000658, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6507158279418945, + "learning_rate": 1e-06, + "loss": 0.9299, + "mean_token_accuracy": 0.7249928712844849, + "num_tokens": 421446253.0, + "step": 16299 + }, + { + "epoch": 1.7900285526026796, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6692955493927, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.7001198530197144, + "num_tokens": 421480936.0, + "step": 16300 + }, + { + "epoch": 1.7901383703052933, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.5950162410736084, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7067722678184509, + "num_tokens": 421501546.0, + "step": 16301 + }, + { + "epoch": 1.7902481880079069, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6758620738983154, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7046941518783569, + "num_tokens": 421523198.0, + "step": 16302 + }, + { + "epoch": 1.7903580057105204, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.777460813522339, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7098721265792847, + "num_tokens": 421549623.0, + "step": 16303 + }, + { + "epoch": 1.7904678234131342, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6360180377960205, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6855193972587585, + "num_tokens": 421574848.0, + "step": 16304 + }, + { + "epoch": 1.790577641115748, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.770981788635254, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7070698738098145, + "num_tokens": 421596570.0, + "step": 16305 + }, + { + "epoch": 1.7906874588183617, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3771629333496094, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.697575569152832, + "num_tokens": 421626890.0, + "step": 16306 + }, + { + "epoch": 1.7907972765209752, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5324020385742188, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7109542489051819, + "num_tokens": 421652410.0, + "step": 16307 + }, + { + "epoch": 1.7909070942235887, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.605105400085449, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6979305148124695, + "num_tokens": 421677817.0, + "step": 16308 + }, + { + "epoch": 1.7910169119262025, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5942070484161377, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7186183929443359, + "num_tokens": 421701895.0, + "step": 16309 + }, + { + "epoch": 1.7911267296288162, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5400826930999756, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7088430523872375, + "num_tokens": 421727887.0, + "step": 16310 + }, + { + "epoch": 1.7912365473314298, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.572530508041382, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7054013609886169, + "num_tokens": 421754232.0, + "step": 16311 + }, + { + "epoch": 1.7913463650340435, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9374566078186035, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.6981540322303772, + "num_tokens": 421774594.0, + "step": 16312 + }, + { + "epoch": 1.791456182736657, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.713002920150757, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6898560523986816, + "num_tokens": 421799392.0, + "step": 16313 + }, + { + "epoch": 1.7915660004392708, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5713613033294678, + "learning_rate": 1e-06, + "loss": 1.0606, + "mean_token_accuracy": 0.6919600367546082, + "num_tokens": 421827295.0, + "step": 16314 + }, + { + "epoch": 1.7916758181418846, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.715507745742798, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7012051343917847, + "num_tokens": 421853525.0, + "step": 16315 + }, + { + "epoch": 1.7917856358444981, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5772202014923096, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7206919193267822, + "num_tokens": 421880922.0, + "step": 16316 + }, + { + "epoch": 1.7918954535471117, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.661402940750122, + "learning_rate": 1e-06, + "loss": 0.9508, + "mean_token_accuracy": 0.7178150415420532, + "num_tokens": 421905972.0, + "step": 16317 + }, + { + "epoch": 1.7920052712497254, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.788034439086914, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7053951025009155, + "num_tokens": 421928724.0, + "step": 16318 + }, + { + "epoch": 1.7921150889523392, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.364785671234131, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7015167474746704, + "num_tokens": 421956659.0, + "step": 16319 + }, + { + "epoch": 1.792224906654953, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5627923011779785, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6980017423629761, + "num_tokens": 421979564.0, + "step": 16320 + }, + { + "epoch": 1.7923347243575665, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6512467861175537, + "learning_rate": 1e-06, + "loss": 0.8766, + "mean_token_accuracy": 0.741990327835083, + "num_tokens": 422000886.0, + "step": 16321 + }, + { + "epoch": 1.79244454206018, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6001803874969482, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6933191418647766, + "num_tokens": 422027894.0, + "step": 16322 + }, + { + "epoch": 1.7925543597627938, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7163941860198975, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7052594423294067, + "num_tokens": 422050366.0, + "step": 16323 + }, + { + "epoch": 1.7926641774654075, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.447075366973877, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.710176408290863, + "num_tokens": 422078367.0, + "step": 16324 + }, + { + "epoch": 1.792773995168021, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7111027240753174, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7165588140487671, + "num_tokens": 422102979.0, + "step": 16325 + }, + { + "epoch": 1.7928838128706346, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7031099796295166, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7105805277824402, + "num_tokens": 422127190.0, + "step": 16326 + }, + { + "epoch": 1.7929936305732483, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.99460506439209, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6951441764831543, + "num_tokens": 422151054.0, + "step": 16327 + }, + { + "epoch": 1.793103448275862, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4446232318878174, + "learning_rate": 1e-06, + "loss": 0.9186, + "mean_token_accuracy": 0.7251193523406982, + "num_tokens": 422176163.0, + "step": 16328 + }, + { + "epoch": 1.7932132659784759, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.602560520172119, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7018368244171143, + "num_tokens": 422200960.0, + "step": 16329 + }, + { + "epoch": 1.7933230836810894, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3351120948791504, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7168539762496948, + "num_tokens": 422229630.0, + "step": 16330 + }, + { + "epoch": 1.793432901383703, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3898239135742188, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6977015733718872, + "num_tokens": 422259552.0, + "step": 16331 + }, + { + "epoch": 1.7935427190863167, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.363466739654541, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7015962600708008, + "num_tokens": 422289751.0, + "step": 16332 + }, + { + "epoch": 1.7936525367889304, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.577017068862915, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7141867876052856, + "num_tokens": 422314633.0, + "step": 16333 + }, + { + "epoch": 1.7937623544915442, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.44858455657959, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7140344381332397, + "num_tokens": 422341077.0, + "step": 16334 + }, + { + "epoch": 1.7938721721941577, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5278050899505615, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.683885931968689, + "num_tokens": 422369625.0, + "step": 16335 + }, + { + "epoch": 1.7939819898967713, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6268608570098877, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7108862400054932, + "num_tokens": 422394613.0, + "step": 16336 + }, + { + "epoch": 1.794091807599385, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.701918840408325, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.7083907127380371, + "num_tokens": 422417982.0, + "step": 16337 + }, + { + "epoch": 1.7942016253019988, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.496246337890625, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7208631634712219, + "num_tokens": 422443533.0, + "step": 16338 + }, + { + "epoch": 1.7943114430046123, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.65230393409729, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6924155950546265, + "num_tokens": 422468024.0, + "step": 16339 + }, + { + "epoch": 1.7944212607072259, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.808204412460327, + "learning_rate": 1e-06, + "loss": 1.0559, + "mean_token_accuracy": 0.6975746750831604, + "num_tokens": 422490614.0, + "step": 16340 + }, + { + "epoch": 1.7945310784098396, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.801466464996338, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7149884104728699, + "num_tokens": 422512313.0, + "step": 16341 + }, + { + "epoch": 1.7946408961124534, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5250461101531982, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7157315015792847, + "num_tokens": 422537722.0, + "step": 16342 + }, + { + "epoch": 1.7947507138150671, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.440847396850586, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7087973356246948, + "num_tokens": 422567382.0, + "step": 16343 + }, + { + "epoch": 1.7948605315176807, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.425105094909668, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.689539909362793, + "num_tokens": 422595383.0, + "step": 16344 + }, + { + "epoch": 1.7949703492202942, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.471956491470337, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.702672004699707, + "num_tokens": 422623755.0, + "step": 16345 + }, + { + "epoch": 1.795080166922908, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.643538475036621, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7008505463600159, + "num_tokens": 422649779.0, + "step": 16346 + }, + { + "epoch": 1.7951899846255217, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.723280906677246, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7139025330543518, + "num_tokens": 422671435.0, + "step": 16347 + }, + { + "epoch": 1.7952998023281352, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.439993143081665, + "learning_rate": 1e-06, + "loss": 0.9079, + "mean_token_accuracy": 0.7275567054748535, + "num_tokens": 422696878.0, + "step": 16348 + }, + { + "epoch": 1.795409620030749, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5518081188201904, + "learning_rate": 1e-06, + "loss": 0.9395, + "mean_token_accuracy": 0.7316042184829712, + "num_tokens": 422720433.0, + "step": 16349 + }, + { + "epoch": 1.7955194377333625, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.702364683151245, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.6879947185516357, + "num_tokens": 422743653.0, + "step": 16350 + }, + { + "epoch": 1.7956292554359763, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.736618757247925, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6975135803222656, + "num_tokens": 422771067.0, + "step": 16351 + }, + { + "epoch": 1.79573907313859, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.002443552017212, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6858054399490356, + "num_tokens": 422798121.0, + "step": 16352 + }, + { + "epoch": 1.7958488908412036, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7275514602661133, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.6991535425186157, + "num_tokens": 422821499.0, + "step": 16353 + }, + { + "epoch": 1.7959587085438171, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6178855895996094, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7230283617973328, + "num_tokens": 422843980.0, + "step": 16354 + }, + { + "epoch": 1.7960685262464309, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6393275260925293, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7119686603546143, + "num_tokens": 422866031.0, + "step": 16355 + }, + { + "epoch": 1.7961783439490446, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.369440793991089, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7073001861572266, + "num_tokens": 422900563.0, + "step": 16356 + }, + { + "epoch": 1.7962881616516584, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.556762218475342, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7007559537887573, + "num_tokens": 422926656.0, + "step": 16357 + }, + { + "epoch": 1.796397979354272, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.545863389968872, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7101683616638184, + "num_tokens": 422952485.0, + "step": 16358 + }, + { + "epoch": 1.7965077970568855, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.444711208343506, + "learning_rate": 1e-06, + "loss": 1.1037, + "mean_token_accuracy": 0.6855602264404297, + "num_tokens": 422984735.0, + "step": 16359 + }, + { + "epoch": 1.7966176147594992, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.600118398666382, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7169256806373596, + "num_tokens": 423008382.0, + "step": 16360 + }, + { + "epoch": 1.796727432462113, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8408257961273193, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7216377854347229, + "num_tokens": 423028419.0, + "step": 16361 + }, + { + "epoch": 1.7968372501647265, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0284230709075928, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7078537940979004, + "num_tokens": 423053152.0, + "step": 16362 + }, + { + "epoch": 1.7969470678673403, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4709761142730713, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7206547260284424, + "num_tokens": 423079331.0, + "step": 16363 + }, + { + "epoch": 1.7970568855699538, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7356536388397217, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.7000711560249329, + "num_tokens": 423102215.0, + "step": 16364 + }, + { + "epoch": 1.7971667032725676, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3071322441101074, + "learning_rate": 1e-06, + "loss": 0.8858, + "mean_token_accuracy": 0.7358471155166626, + "num_tokens": 423131041.0, + "step": 16365 + }, + { + "epoch": 1.7972765209751813, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4966213703155518, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7140889167785645, + "num_tokens": 423158313.0, + "step": 16366 + }, + { + "epoch": 1.7973863386777948, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8773622512817383, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7084847688674927, + "num_tokens": 423177086.0, + "step": 16367 + }, + { + "epoch": 1.7974961563804084, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6462411880493164, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.706475019454956, + "num_tokens": 423202341.0, + "step": 16368 + }, + { + "epoch": 1.7976059740830221, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4036548137664795, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6916286945343018, + "num_tokens": 423232975.0, + "step": 16369 + }, + { + "epoch": 1.797715791785636, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.712143659591675, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6990293860435486, + "num_tokens": 423255271.0, + "step": 16370 + }, + { + "epoch": 1.7978256094882497, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5725789070129395, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6969176530838013, + "num_tokens": 423279840.0, + "step": 16371 + }, + { + "epoch": 1.7979354271908632, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.71451735496521, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.714788019657135, + "num_tokens": 423301839.0, + "step": 16372 + }, + { + "epoch": 1.7980452448934767, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.673564910888672, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6961973905563354, + "num_tokens": 423325332.0, + "step": 16373 + }, + { + "epoch": 1.7981550625960905, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4111733436584473, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7150194644927979, + "num_tokens": 423354366.0, + "step": 16374 + }, + { + "epoch": 1.7982648802987042, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.685034990310669, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.6926704049110413, + "num_tokens": 423378460.0, + "step": 16375 + }, + { + "epoch": 1.7983746980013178, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.440929651260376, + "learning_rate": 1e-06, + "loss": 1.1223, + "mean_token_accuracy": 0.6933777332305908, + "num_tokens": 423409450.0, + "step": 16376 + }, + { + "epoch": 1.7984845157039313, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.719053268432617, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.7044836282730103, + "num_tokens": 423432078.0, + "step": 16377 + }, + { + "epoch": 1.798594333406545, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3324108123779297, + "learning_rate": 1e-06, + "loss": 1.0876, + "mean_token_accuracy": 0.6838580965995789, + "num_tokens": 423463737.0, + "step": 16378 + }, + { + "epoch": 1.7987041511091588, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.505535840988159, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.722511351108551, + "num_tokens": 423490202.0, + "step": 16379 + }, + { + "epoch": 1.7988139688117726, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8152966499328613, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7139899730682373, + "num_tokens": 423518171.0, + "step": 16380 + }, + { + "epoch": 1.7989237865143861, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4437167644500732, + "learning_rate": 1e-06, + "loss": 1.0672, + "mean_token_accuracy": 0.6909061670303345, + "num_tokens": 423549214.0, + "step": 16381 + }, + { + "epoch": 1.7990336042169996, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.338688850402832, + "learning_rate": 1e-06, + "loss": 1.1069, + "mean_token_accuracy": 0.6802942752838135, + "num_tokens": 423583822.0, + "step": 16382 + }, + { + "epoch": 1.7991434219196134, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5454466342926025, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7042669653892517, + "num_tokens": 423611450.0, + "step": 16383 + }, + { + "epoch": 1.7992532396222272, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4193196296691895, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7140095233917236, + "num_tokens": 423639246.0, + "step": 16384 + }, + { + "epoch": 1.799363057324841, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.546396255493164, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7048129439353943, + "num_tokens": 423663237.0, + "step": 16385 + }, + { + "epoch": 1.7994728750274545, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.592315435409546, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.7024093866348267, + "num_tokens": 423682575.0, + "step": 16386 + }, + { + "epoch": 1.799582692730068, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5502469539642334, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6976314187049866, + "num_tokens": 423709095.0, + "step": 16387 + }, + { + "epoch": 1.7996925104326817, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8295018672943115, + "learning_rate": 1e-06, + "loss": 0.9318, + "mean_token_accuracy": 0.7317447662353516, + "num_tokens": 423729133.0, + "step": 16388 + }, + { + "epoch": 1.7998023281352955, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.530651330947876, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7224447131156921, + "num_tokens": 423756251.0, + "step": 16389 + }, + { + "epoch": 1.799912145837909, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.592172622680664, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6911212801933289, + "num_tokens": 423780483.0, + "step": 16390 + }, + { + "epoch": 1.8000219635405226, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.519080877304077, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7081232070922852, + "num_tokens": 423807886.0, + "step": 16391 + }, + { + "epoch": 1.8001317812431363, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7089946269989014, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7218055129051208, + "num_tokens": 423830403.0, + "step": 16392 + }, + { + "epoch": 1.80024159894575, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3544788360595703, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.691300630569458, + "num_tokens": 423862554.0, + "step": 16393 + }, + { + "epoch": 1.8003514166483638, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6178317070007324, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7165861129760742, + "num_tokens": 423889404.0, + "step": 16394 + }, + { + "epoch": 1.8004612343509774, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.236339569091797, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7009645700454712, + "num_tokens": 423920984.0, + "step": 16395 + }, + { + "epoch": 1.800571052053591, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.655855417251587, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6895612478256226, + "num_tokens": 423944673.0, + "step": 16396 + }, + { + "epoch": 1.8006808697562047, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5782291889190674, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.698532223701477, + "num_tokens": 423971785.0, + "step": 16397 + }, + { + "epoch": 1.8007906874588184, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.109081506729126, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.722133219242096, + "num_tokens": 423993324.0, + "step": 16398 + }, + { + "epoch": 1.8009005051614322, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.921741008758545, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.7081499099731445, + "num_tokens": 424022594.0, + "step": 16399 + }, + { + "epoch": 1.8010103228640457, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.509566307067871, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7003611326217651, + "num_tokens": 424050715.0, + "step": 16400 + }, + { + "epoch": 1.8011201405666593, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.620621919631958, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6951759457588196, + "num_tokens": 424075604.0, + "step": 16401 + }, + { + "epoch": 1.801229958269273, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9933059215545654, + "learning_rate": 1e-06, + "loss": 0.8757, + "mean_token_accuracy": 0.7387118935585022, + "num_tokens": 424093160.0, + "step": 16402 + }, + { + "epoch": 1.8013397759718868, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4366726875305176, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6853408217430115, + "num_tokens": 424121916.0, + "step": 16403 + }, + { + "epoch": 1.8014495936745003, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.518096923828125, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7132912874221802, + "num_tokens": 424146915.0, + "step": 16404 + }, + { + "epoch": 1.8015594113771138, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7078170776367188, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6889734268188477, + "num_tokens": 424174791.0, + "step": 16405 + }, + { + "epoch": 1.8016692290797276, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.311662197113037, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7004960775375366, + "num_tokens": 424206433.0, + "step": 16406 + }, + { + "epoch": 1.8017790467823414, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6017322540283203, + "learning_rate": 1e-06, + "loss": 0.8485, + "mean_token_accuracy": 0.7483294010162354, + "num_tokens": 424229767.0, + "step": 16407 + }, + { + "epoch": 1.801888864484955, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.502261161804199, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7070496082305908, + "num_tokens": 424258047.0, + "step": 16408 + }, + { + "epoch": 1.8019986821875686, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6471195220947266, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6909637451171875, + "num_tokens": 424281964.0, + "step": 16409 + }, + { + "epoch": 1.8021084998901822, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4459264278411865, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.7053136229515076, + "num_tokens": 424309618.0, + "step": 16410 + }, + { + "epoch": 1.802218317592796, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.526871681213379, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7111167907714844, + "num_tokens": 424336400.0, + "step": 16411 + }, + { + "epoch": 1.8023281352954097, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.820075750350952, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7087575793266296, + "num_tokens": 424358957.0, + "step": 16412 + }, + { + "epoch": 1.8024379529980232, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.443976640701294, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.6994175910949707, + "num_tokens": 424390698.0, + "step": 16413 + }, + { + "epoch": 1.802547770700637, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9647490978240967, + "learning_rate": 1e-06, + "loss": 0.9407, + "mean_token_accuracy": 0.7193795442581177, + "num_tokens": 424408802.0, + "step": 16414 + }, + { + "epoch": 1.8026575884032505, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.937345266342163, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.707974374294281, + "num_tokens": 424432093.0, + "step": 16415 + }, + { + "epoch": 1.8027674061058643, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.677783727645874, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7096314430236816, + "num_tokens": 424457333.0, + "step": 16416 + }, + { + "epoch": 1.802877223808478, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7775728702545166, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7126388549804688, + "num_tokens": 424479372.0, + "step": 16417 + }, + { + "epoch": 1.8029870415110916, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.521942615509033, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.6871930360794067, + "num_tokens": 424506414.0, + "step": 16418 + }, + { + "epoch": 1.803096859213705, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.693493127822876, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6831478476524353, + "num_tokens": 424530747.0, + "step": 16419 + }, + { + "epoch": 1.8032066769163189, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.621190309524536, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7068676352500916, + "num_tokens": 424554944.0, + "step": 16420 + }, + { + "epoch": 1.8033164946189326, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6115286350250244, + "learning_rate": 1e-06, + "loss": 0.9066, + "mean_token_accuracy": 0.7370241284370422, + "num_tokens": 424577934.0, + "step": 16421 + }, + { + "epoch": 1.8034263123215464, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6913530826568604, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6977360844612122, + "num_tokens": 424600956.0, + "step": 16422 + }, + { + "epoch": 1.80353613002416, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.63370680809021, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7135850191116333, + "num_tokens": 424625842.0, + "step": 16423 + }, + { + "epoch": 1.8036459477267734, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5502841472625732, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7112641930580139, + "num_tokens": 424652210.0, + "step": 16424 + }, + { + "epoch": 1.8037557654293872, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.72208309173584, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7167917490005493, + "num_tokens": 424676026.0, + "step": 16425 + }, + { + "epoch": 1.803865583132001, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0223641395568848, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.6969178915023804, + "num_tokens": 424697535.0, + "step": 16426 + }, + { + "epoch": 1.8039754008346145, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5292842388153076, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7180250287055969, + "num_tokens": 424725156.0, + "step": 16427 + }, + { + "epoch": 1.8040852185372283, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.935182809829712, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7038832902908325, + "num_tokens": 424746333.0, + "step": 16428 + }, + { + "epoch": 1.8041950362398418, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5231564044952393, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7205743789672852, + "num_tokens": 424771557.0, + "step": 16429 + }, + { + "epoch": 1.8043048539424555, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7154998779296875, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.6967852115631104, + "num_tokens": 424792970.0, + "step": 16430 + }, + { + "epoch": 1.8044146716450693, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7482426166534424, + "learning_rate": 1e-06, + "loss": 0.8504, + "mean_token_accuracy": 0.7456492185592651, + "num_tokens": 424812348.0, + "step": 16431 + }, + { + "epoch": 1.8045244893476828, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4456353187561035, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6933538317680359, + "num_tokens": 424840825.0, + "step": 16432 + }, + { + "epoch": 1.8046343070502964, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.619372606277466, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7279220819473267, + "num_tokens": 424864673.0, + "step": 16433 + }, + { + "epoch": 1.8047441247529101, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.2549421787261963, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7164759635925293, + "num_tokens": 424885996.0, + "step": 16434 + }, + { + "epoch": 1.8048539424555239, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.728639602661133, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.6986374855041504, + "num_tokens": 424913553.0, + "step": 16435 + }, + { + "epoch": 1.8049637601581376, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.765944480895996, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6963806748390198, + "num_tokens": 424936463.0, + "step": 16436 + }, + { + "epoch": 1.8050735778607512, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4309728145599365, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6890228986740112, + "num_tokens": 424967622.0, + "step": 16437 + }, + { + "epoch": 1.8051833955633647, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5801923274993896, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6940895318984985, + "num_tokens": 424996346.0, + "step": 16438 + }, + { + "epoch": 1.8052932132659785, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.903999090194702, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7140459418296814, + "num_tokens": 425016697.0, + "step": 16439 + }, + { + "epoch": 1.8054030309685922, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6592671871185303, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7081912159919739, + "num_tokens": 425040663.0, + "step": 16440 + }, + { + "epoch": 1.8055128486712058, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4770724773406982, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7234311699867249, + "num_tokens": 425066619.0, + "step": 16441 + }, + { + "epoch": 1.8056226663738193, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3988940715789795, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6990433931350708, + "num_tokens": 425095651.0, + "step": 16442 + }, + { + "epoch": 1.805732484076433, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.301492214202881, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6922124624252319, + "num_tokens": 425129103.0, + "step": 16443 + }, + { + "epoch": 1.8058423017790468, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.015533685684204, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7058929204940796, + "num_tokens": 425158032.0, + "step": 16444 + }, + { + "epoch": 1.8059521194816606, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7868494987487793, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7010982632637024, + "num_tokens": 425179784.0, + "step": 16445 + }, + { + "epoch": 1.806061937184274, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.574984550476074, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7207196950912476, + "num_tokens": 425204646.0, + "step": 16446 + }, + { + "epoch": 1.8061717548868876, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8477394580841064, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6875836849212646, + "num_tokens": 425230154.0, + "step": 16447 + }, + { + "epoch": 1.8062815725895014, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5582282543182373, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7162628173828125, + "num_tokens": 425252949.0, + "step": 16448 + }, + { + "epoch": 1.8063913902921152, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.639437675476074, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7057724595069885, + "num_tokens": 425279039.0, + "step": 16449 + }, + { + "epoch": 1.806501207994729, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.261854410171509, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6992579698562622, + "num_tokens": 425314120.0, + "step": 16450 + }, + { + "epoch": 1.8066110256973424, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1480448246002197, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.7053945064544678, + "num_tokens": 425341235.0, + "step": 16451 + }, + { + "epoch": 1.806720843399956, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7296323776245117, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7295002937316895, + "num_tokens": 425362936.0, + "step": 16452 + }, + { + "epoch": 1.8068306611025697, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5821893215179443, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7009581327438354, + "num_tokens": 425387980.0, + "step": 16453 + }, + { + "epoch": 1.8069404788051835, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4775919914245605, + "learning_rate": 1e-06, + "loss": 0.9785, + "mean_token_accuracy": 0.7167640924453735, + "num_tokens": 425415000.0, + "step": 16454 + }, + { + "epoch": 1.807050296507797, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6542344093322754, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7106912136077881, + "num_tokens": 425438014.0, + "step": 16455 + }, + { + "epoch": 1.8071601142104106, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4986274242401123, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6965394616127014, + "num_tokens": 425465809.0, + "step": 16456 + }, + { + "epoch": 1.8072699319130243, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4143776893615723, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.6920709609985352, + "num_tokens": 425494739.0, + "step": 16457 + }, + { + "epoch": 1.807379749615638, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.621668815612793, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7035273313522339, + "num_tokens": 425519222.0, + "step": 16458 + }, + { + "epoch": 1.8074895673182518, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.959411382675171, + "learning_rate": 1e-06, + "loss": 1.0965, + "mean_token_accuracy": 0.6855064630508423, + "num_tokens": 425546860.0, + "step": 16459 + }, + { + "epoch": 1.8075993850208654, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3572804927825928, + "learning_rate": 1e-06, + "loss": 1.1346, + "mean_token_accuracy": 0.6774815320968628, + "num_tokens": 425578098.0, + "step": 16460 + }, + { + "epoch": 1.807709202723479, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0342938899993896, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6992182731628418, + "num_tokens": 425603109.0, + "step": 16461 + }, + { + "epoch": 1.8078190204260927, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9230728149414062, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6892164945602417, + "num_tokens": 425624711.0, + "step": 16462 + }, + { + "epoch": 1.8079288381287064, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6134114265441895, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6987770795822144, + "num_tokens": 425651333.0, + "step": 16463 + }, + { + "epoch": 1.8080386558313202, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.638667583465576, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.7038942575454712, + "num_tokens": 425681369.0, + "step": 16464 + }, + { + "epoch": 1.8081484735339337, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.770190477371216, + "learning_rate": 1e-06, + "loss": 0.9037, + "mean_token_accuracy": 0.7370991110801697, + "num_tokens": 425705748.0, + "step": 16465 + }, + { + "epoch": 1.8082582912365472, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4444618225097656, + "learning_rate": 1e-06, + "loss": 0.9127, + "mean_token_accuracy": 0.7342263460159302, + "num_tokens": 425730636.0, + "step": 16466 + }, + { + "epoch": 1.808368108939161, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3816511631011963, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.695372998714447, + "num_tokens": 425758932.0, + "step": 16467 + }, + { + "epoch": 1.8084779266417748, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8439838886260986, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7139296531677246, + "num_tokens": 425778446.0, + "step": 16468 + }, + { + "epoch": 1.8085877443443883, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.42324161529541, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7154524326324463, + "num_tokens": 425807780.0, + "step": 16469 + }, + { + "epoch": 1.8086975620470018, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6005961894989014, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6839602589607239, + "num_tokens": 425834352.0, + "step": 16470 + }, + { + "epoch": 1.8088073797496156, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3611433506011963, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.6945390701293945, + "num_tokens": 425864174.0, + "step": 16471 + }, + { + "epoch": 1.8089171974522293, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3348124027252197, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.7129594087600708, + "num_tokens": 425894335.0, + "step": 16472 + }, + { + "epoch": 1.809027015154843, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.423306703567505, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6938087940216064, + "num_tokens": 425925002.0, + "step": 16473 + }, + { + "epoch": 1.8091368328574566, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.572629451751709, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6966521739959717, + "num_tokens": 425950776.0, + "step": 16474 + }, + { + "epoch": 1.8092466505600702, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4936866760253906, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6994878649711609, + "num_tokens": 425977702.0, + "step": 16475 + }, + { + "epoch": 1.809356468262684, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8227107524871826, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.727553129196167, + "num_tokens": 425999777.0, + "step": 16476 + }, + { + "epoch": 1.8094662859652977, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.661515474319458, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6938732266426086, + "num_tokens": 426023601.0, + "step": 16477 + }, + { + "epoch": 1.8095761036679112, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4657890796661377, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7096343040466309, + "num_tokens": 426052195.0, + "step": 16478 + }, + { + "epoch": 1.809685921370525, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6537725925445557, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.702526330947876, + "num_tokens": 426076543.0, + "step": 16479 + }, + { + "epoch": 1.8097957390731385, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6771128177642822, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.6959748268127441, + "num_tokens": 426102646.0, + "step": 16480 + }, + { + "epoch": 1.8099055567757523, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.493368148803711, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7140305042266846, + "num_tokens": 426128498.0, + "step": 16481 + }, + { + "epoch": 1.810015374478366, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9833974838256836, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7081392407417297, + "num_tokens": 426152242.0, + "step": 16482 + }, + { + "epoch": 1.8101251921809796, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5500776767730713, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7059522271156311, + "num_tokens": 426179491.0, + "step": 16483 + }, + { + "epoch": 1.810235009883593, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.699326515197754, + "learning_rate": 1e-06, + "loss": 1.1509, + "mean_token_accuracy": 0.6698735952377319, + "num_tokens": 426210144.0, + "step": 16484 + }, + { + "epoch": 1.8103448275862069, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4197847843170166, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6913274526596069, + "num_tokens": 426240115.0, + "step": 16485 + }, + { + "epoch": 1.8104546452888206, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5784411430358887, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7048747539520264, + "num_tokens": 426264571.0, + "step": 16486 + }, + { + "epoch": 1.8105644629914344, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.291025400161743, + "learning_rate": 1e-06, + "loss": 1.0937, + "mean_token_accuracy": 0.6850661039352417, + "num_tokens": 426299083.0, + "step": 16487 + }, + { + "epoch": 1.810674280694048, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4475719928741455, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7046328186988831, + "num_tokens": 426324702.0, + "step": 16488 + }, + { + "epoch": 1.8107840983966614, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7711503505706787, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.716614842414856, + "num_tokens": 426346607.0, + "step": 16489 + }, + { + "epoch": 1.8108939160992752, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.589235305786133, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6923534274101257, + "num_tokens": 426370961.0, + "step": 16490 + }, + { + "epoch": 1.811003733801889, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4826292991638184, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7085409164428711, + "num_tokens": 426401716.0, + "step": 16491 + }, + { + "epoch": 1.8111135515045025, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8576853275299072, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.700127363204956, + "num_tokens": 426423232.0, + "step": 16492 + }, + { + "epoch": 1.8112233692071162, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.895007610321045, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7053373456001282, + "num_tokens": 426444908.0, + "step": 16493 + }, + { + "epoch": 1.8113331869097298, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9152050018310547, + "learning_rate": 1e-06, + "loss": 0.9045, + "mean_token_accuracy": 0.7331498265266418, + "num_tokens": 426464873.0, + "step": 16494 + }, + { + "epoch": 1.8114430046123435, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.729234218597412, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.6989463567733765, + "num_tokens": 426493448.0, + "step": 16495 + }, + { + "epoch": 1.8115528223149573, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.798036575317383, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7079890966415405, + "num_tokens": 426514810.0, + "step": 16496 + }, + { + "epoch": 1.8116626400175708, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7650322914123535, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7240986227989197, + "num_tokens": 426535324.0, + "step": 16497 + }, + { + "epoch": 1.8117724577201844, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4605634212493896, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7086237668991089, + "num_tokens": 426563419.0, + "step": 16498 + }, + { + "epoch": 1.8118822754227981, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5600249767303467, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7172684073448181, + "num_tokens": 426589380.0, + "step": 16499 + }, + { + "epoch": 1.8119920931254119, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.712160587310791, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7113372087478638, + "num_tokens": 426613698.0, + "step": 16500 + }, + { + "epoch": 1.8121019108280256, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2796409130096436, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6779390573501587, + "num_tokens": 426652593.0, + "step": 16501 + }, + { + "epoch": 1.8122117285306392, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3836915493011475, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.6909343004226685, + "num_tokens": 426682565.0, + "step": 16502 + }, + { + "epoch": 1.8123215462332527, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.255350112915039, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6993266344070435, + "num_tokens": 426718171.0, + "step": 16503 + }, + { + "epoch": 1.8124313639358665, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.201885938644409, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7162882089614868, + "num_tokens": 426751561.0, + "step": 16504 + }, + { + "epoch": 1.8125411816384802, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.579460382461548, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6888318061828613, + "num_tokens": 426776114.0, + "step": 16505 + }, + { + "epoch": 1.8126509993410937, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3444154262542725, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.6991591453552246, + "num_tokens": 426805410.0, + "step": 16506 + }, + { + "epoch": 1.8127608170437073, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5444140434265137, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6883140802383423, + "num_tokens": 426831192.0, + "step": 16507 + }, + { + "epoch": 1.812870634746321, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.473814010620117, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6882129907608032, + "num_tokens": 426858715.0, + "step": 16508 + }, + { + "epoch": 1.8129804524489348, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4411351680755615, + "learning_rate": 1e-06, + "loss": 1.0797, + "mean_token_accuracy": 0.6833958625793457, + "num_tokens": 426888925.0, + "step": 16509 + }, + { + "epoch": 1.8130902701515486, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7424280643463135, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.707250714302063, + "num_tokens": 426911423.0, + "step": 16510 + }, + { + "epoch": 1.813200087854162, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.680107831954956, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6889925003051758, + "num_tokens": 426935761.0, + "step": 16511 + }, + { + "epoch": 1.8133099055567756, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.564124345779419, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7136445045471191, + "num_tokens": 426963565.0, + "step": 16512 + }, + { + "epoch": 1.8134197232593894, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.465254545211792, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7238925695419312, + "num_tokens": 426988193.0, + "step": 16513 + }, + { + "epoch": 1.8135295409620031, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.561971426010132, + "learning_rate": 1e-06, + "loss": 1.1568, + "mean_token_accuracy": 0.6737635135650635, + "num_tokens": 427022948.0, + "step": 16514 + }, + { + "epoch": 1.813639358664617, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.54079532623291, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7069480419158936, + "num_tokens": 427052540.0, + "step": 16515 + }, + { + "epoch": 1.8137491763672304, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6727516651153564, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7043176293373108, + "num_tokens": 427075944.0, + "step": 16516 + }, + { + "epoch": 1.813858994069844, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2292416095733643, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7104663848876953, + "num_tokens": 427108854.0, + "step": 16517 + }, + { + "epoch": 1.8139688117724577, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.659043312072754, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7028985619544983, + "num_tokens": 427131274.0, + "step": 16518 + }, + { + "epoch": 1.8140786294750715, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5950207710266113, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7021799683570862, + "num_tokens": 427155856.0, + "step": 16519 + }, + { + "epoch": 1.814188447177685, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.49066162109375, + "learning_rate": 1e-06, + "loss": 1.1596, + "mean_token_accuracy": 0.6693305373191833, + "num_tokens": 427185417.0, + "step": 16520 + }, + { + "epoch": 1.8142982648802986, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.526944875717163, + "learning_rate": 1e-06, + "loss": 1.115, + "mean_token_accuracy": 0.6835036277770996, + "num_tokens": 427212866.0, + "step": 16521 + }, + { + "epoch": 1.8144080825829123, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4704833030700684, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.693793535232544, + "num_tokens": 427238942.0, + "step": 16522 + }, + { + "epoch": 1.814517900285526, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.549921751022339, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7063787579536438, + "num_tokens": 427264887.0, + "step": 16523 + }, + { + "epoch": 1.8146277179881398, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.5703353881835938, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.720702052116394, + "num_tokens": 427292572.0, + "step": 16524 + }, + { + "epoch": 1.8147375356907534, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6780333518981934, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6956249475479126, + "num_tokens": 427316787.0, + "step": 16525 + }, + { + "epoch": 1.814847353393367, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4532907009124756, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6784070730209351, + "num_tokens": 427348035.0, + "step": 16526 + }, + { + "epoch": 1.8149571710959806, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.436582088470459, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.707687258720398, + "num_tokens": 427373501.0, + "step": 16527 + }, + { + "epoch": 1.8150669887985944, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9027676582336426, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.719818651676178, + "num_tokens": 427392578.0, + "step": 16528 + }, + { + "epoch": 1.815176806501208, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.416308641433716, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.7065642476081848, + "num_tokens": 427420076.0, + "step": 16529 + }, + { + "epoch": 1.8152866242038217, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4896442890167236, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6922248005867004, + "num_tokens": 427447000.0, + "step": 16530 + }, + { + "epoch": 1.8153964419064352, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7942864894866943, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6978319883346558, + "num_tokens": 427467724.0, + "step": 16531 + }, + { + "epoch": 1.815506259609049, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.489501953125, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7164287567138672, + "num_tokens": 427498695.0, + "step": 16532 + }, + { + "epoch": 1.8156160773116627, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.655184745788574, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7056962251663208, + "num_tokens": 427524312.0, + "step": 16533 + }, + { + "epoch": 1.8157258950142763, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6800708770751953, + "learning_rate": 1e-06, + "loss": 1.1081, + "mean_token_accuracy": 0.6803828477859497, + "num_tokens": 427552420.0, + "step": 16534 + }, + { + "epoch": 1.8158357127168898, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6173501014709473, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7097434997558594, + "num_tokens": 427577265.0, + "step": 16535 + }, + { + "epoch": 1.8159455304195036, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3218488693237305, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7178770303726196, + "num_tokens": 427607241.0, + "step": 16536 + }, + { + "epoch": 1.8160553481221173, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8954224586486816, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7142969369888306, + "num_tokens": 427626819.0, + "step": 16537 + }, + { + "epoch": 1.816165165824731, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.401573896408081, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7152401208877563, + "num_tokens": 427656171.0, + "step": 16538 + }, + { + "epoch": 1.8162749835273446, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.512430429458618, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6849484443664551, + "num_tokens": 427683797.0, + "step": 16539 + }, + { + "epoch": 1.8163848012299582, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3901824951171875, + "learning_rate": 1e-06, + "loss": 0.9318, + "mean_token_accuracy": 0.7251611948013306, + "num_tokens": 427710480.0, + "step": 16540 + }, + { + "epoch": 1.816494618932572, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.779205083847046, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7074559926986694, + "num_tokens": 427736371.0, + "step": 16541 + }, + { + "epoch": 1.8166044366351857, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.829817533493042, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6921457052230835, + "num_tokens": 427758803.0, + "step": 16542 + }, + { + "epoch": 1.8167142543377992, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.704418659210205, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.715499758720398, + "num_tokens": 427779095.0, + "step": 16543 + }, + { + "epoch": 1.816824072040413, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8020660877227783, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6936556100845337, + "num_tokens": 427799114.0, + "step": 16544 + }, + { + "epoch": 1.8169338897430265, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.610905647277832, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.6810476779937744, + "num_tokens": 427826069.0, + "step": 16545 + }, + { + "epoch": 1.8170437074456403, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7559502124786377, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7065072059631348, + "num_tokens": 427849250.0, + "step": 16546 + }, + { + "epoch": 1.817153525148254, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.601167678833008, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.695284366607666, + "num_tokens": 427880158.0, + "step": 16547 + }, + { + "epoch": 1.8172633428508675, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.391174554824829, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7103248834609985, + "num_tokens": 427911055.0, + "step": 16548 + }, + { + "epoch": 1.817373160553481, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.464689254760742, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6896326541900635, + "num_tokens": 427941798.0, + "step": 16549 + }, + { + "epoch": 1.8174829782560948, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.612278461456299, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7164990901947021, + "num_tokens": 427965018.0, + "step": 16550 + }, + { + "epoch": 1.8175927959587086, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7056477069854736, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7061381340026855, + "num_tokens": 427986831.0, + "step": 16551 + }, + { + "epoch": 1.8177026136613224, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5261011123657227, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7277302145957947, + "num_tokens": 428011835.0, + "step": 16552 + }, + { + "epoch": 1.8178124313639359, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.254987955093384, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7126061916351318, + "num_tokens": 428030219.0, + "step": 16553 + }, + { + "epoch": 1.8179222490665494, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.918462038040161, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7084386348724365, + "num_tokens": 428050693.0, + "step": 16554 + }, + { + "epoch": 1.8180320667691632, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.646789789199829, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7146418690681458, + "num_tokens": 428075890.0, + "step": 16555 + }, + { + "epoch": 1.818141884471777, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5338051319122314, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.6974375247955322, + "num_tokens": 428104492.0, + "step": 16556 + }, + { + "epoch": 1.8182517021743905, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.6980066299438477, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.709362268447876, + "num_tokens": 428125142.0, + "step": 16557 + }, + { + "epoch": 1.818361519877004, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.63371205329895, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7034679651260376, + "num_tokens": 428149052.0, + "step": 16558 + }, + { + "epoch": 1.8184713375796178, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.487428903579712, + "learning_rate": 1e-06, + "loss": 0.8965, + "mean_token_accuracy": 0.7350314855575562, + "num_tokens": 428175138.0, + "step": 16559 + }, + { + "epoch": 1.8185811552822315, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5238590240478516, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7073702812194824, + "num_tokens": 428201616.0, + "step": 16560 + }, + { + "epoch": 1.8186909729848453, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5101518630981445, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7121214866638184, + "num_tokens": 428229204.0, + "step": 16561 + }, + { + "epoch": 1.8188007906874588, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6170711517333984, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7032743096351624, + "num_tokens": 428252398.0, + "step": 16562 + }, + { + "epoch": 1.8189106083900723, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4335458278656006, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7236568927764893, + "num_tokens": 428280091.0, + "step": 16563 + }, + { + "epoch": 1.819020426092686, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.920637369155884, + "learning_rate": 1e-06, + "loss": 0.9131, + "mean_token_accuracy": 0.7285542488098145, + "num_tokens": 428303463.0, + "step": 16564 + }, + { + "epoch": 1.8191302437952999, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7097063064575195, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7075872421264648, + "num_tokens": 428326495.0, + "step": 16565 + }, + { + "epoch": 1.8192400614979136, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5948565006256104, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6832156777381897, + "num_tokens": 428356987.0, + "step": 16566 + }, + { + "epoch": 1.8193498792005272, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.6732966899871826, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.69904625415802, + "num_tokens": 428382039.0, + "step": 16567 + }, + { + "epoch": 1.8194596969031407, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7782037258148193, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7097055912017822, + "num_tokens": 428407816.0, + "step": 16568 + }, + { + "epoch": 1.8195695146057544, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.627309560775757, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7040339708328247, + "num_tokens": 428434878.0, + "step": 16569 + }, + { + "epoch": 1.8196793323083682, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.479860544204712, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.699138879776001, + "num_tokens": 428461640.0, + "step": 16570 + }, + { + "epoch": 1.8197891500109817, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4762699604034424, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.7006503343582153, + "num_tokens": 428486922.0, + "step": 16571 + }, + { + "epoch": 1.8198989677135953, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8007354736328125, + "learning_rate": 1e-06, + "loss": 0.8897, + "mean_token_accuracy": 0.7344447374343872, + "num_tokens": 428509178.0, + "step": 16572 + }, + { + "epoch": 1.820008785416209, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.3601911067962646, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.719666063785553, + "num_tokens": 428526688.0, + "step": 16573 + }, + { + "epoch": 1.8201186031188228, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.608018398284912, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6950315237045288, + "num_tokens": 428552595.0, + "step": 16574 + }, + { + "epoch": 1.8202284208214365, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.514080047607422, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7162576913833618, + "num_tokens": 428580590.0, + "step": 16575 + }, + { + "epoch": 1.82033823852405, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3802318572998047, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7173334360122681, + "num_tokens": 428610668.0, + "step": 16576 + }, + { + "epoch": 1.8204480562266636, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7343432903289795, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7151708602905273, + "num_tokens": 428630137.0, + "step": 16577 + }, + { + "epoch": 1.8205578739292774, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5362226963043213, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7173916101455688, + "num_tokens": 428656268.0, + "step": 16578 + }, + { + "epoch": 1.8206676916318911, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4985852241516113, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7140194177627563, + "num_tokens": 428684183.0, + "step": 16579 + }, + { + "epoch": 1.8207775093345049, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6616432666778564, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.726222038269043, + "num_tokens": 428708305.0, + "step": 16580 + }, + { + "epoch": 1.8208873270371184, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.598879098892212, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7158203721046448, + "num_tokens": 428732776.0, + "step": 16581 + }, + { + "epoch": 1.820997144739732, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.302427053451538, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6996606588363647, + "num_tokens": 428764808.0, + "step": 16582 + }, + { + "epoch": 1.8211069624423457, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.9007785320281982, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7032681703567505, + "num_tokens": 428784878.0, + "step": 16583 + }, + { + "epoch": 1.8212167801449595, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4139535427093506, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.727047324180603, + "num_tokens": 428815897.0, + "step": 16584 + }, + { + "epoch": 1.821326597847573, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8793582916259766, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.7178213596343994, + "num_tokens": 428837386.0, + "step": 16585 + }, + { + "epoch": 1.8214364155501865, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3549439907073975, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7130011916160583, + "num_tokens": 428863800.0, + "step": 16586 + }, + { + "epoch": 1.8215462332528003, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7824654579162598, + "learning_rate": 1e-06, + "loss": 0.88, + "mean_token_accuracy": 0.7400014400482178, + "num_tokens": 428884082.0, + "step": 16587 + }, + { + "epoch": 1.821656050955414, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.952954053878784, + "learning_rate": 1e-06, + "loss": 0.9113, + "mean_token_accuracy": 0.7263213396072388, + "num_tokens": 428904607.0, + "step": 16588 + }, + { + "epoch": 1.8217658686580278, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.591719627380371, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7037664651870728, + "num_tokens": 428930390.0, + "step": 16589 + }, + { + "epoch": 1.8218756863606413, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.711028575897217, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7035607695579529, + "num_tokens": 428955621.0, + "step": 16590 + }, + { + "epoch": 1.8219855040632549, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.425384759902954, + "learning_rate": 1e-06, + "loss": 0.9429, + "mean_token_accuracy": 0.7230412364006042, + "num_tokens": 428986531.0, + "step": 16591 + }, + { + "epoch": 1.8220953217658686, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.608733654022217, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7234885692596436, + "num_tokens": 429011547.0, + "step": 16592 + }, + { + "epoch": 1.8222051394684824, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4476754665374756, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7085331678390503, + "num_tokens": 429037503.0, + "step": 16593 + }, + { + "epoch": 1.822314957171096, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.628065586090088, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.69930499792099, + "num_tokens": 429063438.0, + "step": 16594 + }, + { + "epoch": 1.8224247748737097, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5331146717071533, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7127201557159424, + "num_tokens": 429089960.0, + "step": 16595 + }, + { + "epoch": 1.8225345925763232, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.693847894668579, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.704677939414978, + "num_tokens": 429112940.0, + "step": 16596 + }, + { + "epoch": 1.822644410278937, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4820122718811035, + "learning_rate": 1e-06, + "loss": 0.9147, + "mean_token_accuracy": 0.7280571460723877, + "num_tokens": 429139904.0, + "step": 16597 + }, + { + "epoch": 1.8227542279815507, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.397568941116333, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.7006914615631104, + "num_tokens": 429172666.0, + "step": 16598 + }, + { + "epoch": 1.8228640456841643, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.723501205444336, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7132608890533447, + "num_tokens": 429195114.0, + "step": 16599 + }, + { + "epoch": 1.8229738633867778, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4922869205474854, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.6879264116287231, + "num_tokens": 429222275.0, + "step": 16600 + }, + { + "epoch": 1.8230836810893916, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6699464321136475, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6892402768135071, + "num_tokens": 429248468.0, + "step": 16601 + }, + { + "epoch": 1.8231934987920053, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5402941703796387, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.6988035440444946, + "num_tokens": 429276517.0, + "step": 16602 + }, + { + "epoch": 1.823303316494619, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.481001615524292, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6961462497711182, + "num_tokens": 429305965.0, + "step": 16603 + }, + { + "epoch": 1.8234131341972326, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5959692001342773, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7001870274543762, + "num_tokens": 429329134.0, + "step": 16604 + }, + { + "epoch": 1.8235229518998461, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.420308828353882, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6978130340576172, + "num_tokens": 429358836.0, + "step": 16605 + }, + { + "epoch": 1.82363276960246, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7358779907226562, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.7218689918518066, + "num_tokens": 429380611.0, + "step": 16606 + }, + { + "epoch": 1.8237425873050737, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5807430744171143, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6888874769210815, + "num_tokens": 429407180.0, + "step": 16607 + }, + { + "epoch": 1.8238524050076872, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9490487575531006, + "learning_rate": 1e-06, + "loss": 1.107, + "mean_token_accuracy": 0.6831186413764954, + "num_tokens": 429435290.0, + "step": 16608 + }, + { + "epoch": 1.823962222710301, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.690349578857422, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7047441601753235, + "num_tokens": 429460877.0, + "step": 16609 + }, + { + "epoch": 1.8240720404129145, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.489546775817871, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7106993794441223, + "num_tokens": 429490257.0, + "step": 16610 + }, + { + "epoch": 1.8241818581155282, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3586723804473877, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6868895292282104, + "num_tokens": 429518656.0, + "step": 16611 + }, + { + "epoch": 1.824291675818142, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.2668237686157227, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6783887147903442, + "num_tokens": 429551292.0, + "step": 16612 + }, + { + "epoch": 1.8244014935207555, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3746705055236816, + "learning_rate": 1e-06, + "loss": 1.1205, + "mean_token_accuracy": 0.6857520341873169, + "num_tokens": 429581395.0, + "step": 16613 + }, + { + "epoch": 1.824511311223369, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.376246690750122, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7102770805358887, + "num_tokens": 429610545.0, + "step": 16614 + }, + { + "epoch": 1.8246211289259828, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3989293575286865, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6897047758102417, + "num_tokens": 429640094.0, + "step": 16615 + }, + { + "epoch": 1.8247309466285966, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.563230276107788, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.681682825088501, + "num_tokens": 429664718.0, + "step": 16616 + }, + { + "epoch": 1.8248407643312103, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3783884048461914, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7100995182991028, + "num_tokens": 429693462.0, + "step": 16617 + }, + { + "epoch": 1.8249505820338239, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6170527935028076, + "learning_rate": 1e-06, + "loss": 1.1019, + "mean_token_accuracy": 0.6818121075630188, + "num_tokens": 429722076.0, + "step": 16618 + }, + { + "epoch": 1.8250603997364374, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5119409561157227, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6910765171051025, + "num_tokens": 429751375.0, + "step": 16619 + }, + { + "epoch": 1.8251702174390512, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.8019638061523438, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7062896490097046, + "num_tokens": 429773201.0, + "step": 16620 + }, + { + "epoch": 1.825280035141665, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4336204528808594, + "learning_rate": 1e-06, + "loss": 0.9309, + "mean_token_accuracy": 0.7316361665725708, + "num_tokens": 429801196.0, + "step": 16621 + }, + { + "epoch": 1.8253898528442785, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6557610034942627, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7076331973075867, + "num_tokens": 429823930.0, + "step": 16622 + }, + { + "epoch": 1.825499670546892, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4942874908447266, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7016774415969849, + "num_tokens": 429852022.0, + "step": 16623 + }, + { + "epoch": 1.8256094882495058, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.432795524597168, + "learning_rate": 1e-06, + "loss": 1.0754, + "mean_token_accuracy": 0.6862703561782837, + "num_tokens": 429881626.0, + "step": 16624 + }, + { + "epoch": 1.8257193059521195, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.503934383392334, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7044916749000549, + "num_tokens": 429910784.0, + "step": 16625 + }, + { + "epoch": 1.8258291236547333, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3624377250671387, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7101640701293945, + "num_tokens": 429940358.0, + "step": 16626 + }, + { + "epoch": 1.8259389413573468, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.403245210647583, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.6843591928482056, + "num_tokens": 429969579.0, + "step": 16627 + }, + { + "epoch": 1.8260487590599603, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.4349632263183594, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.695478618144989, + "num_tokens": 429996881.0, + "step": 16628 + }, + { + "epoch": 1.826158576762574, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 3.0341427326202393, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.711345911026001, + "num_tokens": 430022276.0, + "step": 16629 + }, + { + "epoch": 1.8262683944651878, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7073981761932373, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7097066640853882, + "num_tokens": 430047514.0, + "step": 16630 + }, + { + "epoch": 1.8263782121678016, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8150062561035156, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7053514719009399, + "num_tokens": 430068455.0, + "step": 16631 + }, + { + "epoch": 1.8264880298704151, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.745615243911743, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7017651796340942, + "num_tokens": 430091890.0, + "step": 16632 + }, + { + "epoch": 1.8265978475730287, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.494844436645508, + "learning_rate": 1e-06, + "loss": 1.1165, + "mean_token_accuracy": 0.6766074895858765, + "num_tokens": 430122910.0, + "step": 16633 + }, + { + "epoch": 1.8267076652756424, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.575263261795044, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.698157548904419, + "num_tokens": 430148544.0, + "step": 16634 + }, + { + "epoch": 1.8268174829782562, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.510831117630005, + "learning_rate": 1e-06, + "loss": 1.1241, + "mean_token_accuracy": 0.6861253976821899, + "num_tokens": 430177519.0, + "step": 16635 + }, + { + "epoch": 1.8269273006808697, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7001943588256836, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7047839164733887, + "num_tokens": 430202491.0, + "step": 16636 + }, + { + "epoch": 1.8270371183834833, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5836992263793945, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7136407494544983, + "num_tokens": 430228359.0, + "step": 16637 + }, + { + "epoch": 1.827146936086097, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.087970495223999, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7088571786880493, + "num_tokens": 430246665.0, + "step": 16638 + }, + { + "epoch": 1.8272567537887108, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.55480694770813, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.7106912136077881, + "num_tokens": 430271257.0, + "step": 16639 + }, + { + "epoch": 1.8273665714913245, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.479011058807373, + "learning_rate": 1e-06, + "loss": 0.9125, + "mean_token_accuracy": 0.7316752672195435, + "num_tokens": 430295702.0, + "step": 16640 + }, + { + "epoch": 1.827476389193938, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5847859382629395, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7142363786697388, + "num_tokens": 430319395.0, + "step": 16641 + }, + { + "epoch": 1.8275862068965516, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7021186351776123, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.698655366897583, + "num_tokens": 430344876.0, + "step": 16642 + }, + { + "epoch": 1.8276960245991654, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.624687433242798, + "learning_rate": 1e-06, + "loss": 0.9327, + "mean_token_accuracy": 0.7236788868904114, + "num_tokens": 430369863.0, + "step": 16643 + }, + { + "epoch": 1.8278058423017791, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.578582525253296, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6973428726196289, + "num_tokens": 430394954.0, + "step": 16644 + }, + { + "epoch": 1.8279156600043929, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.3651957511901855, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6821821928024292, + "num_tokens": 430426202.0, + "step": 16645 + }, + { + "epoch": 1.8280254777070064, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.6064538955688477, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6940317749977112, + "num_tokens": 430452728.0, + "step": 16646 + }, + { + "epoch": 1.82813529540962, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6702396869659424, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7085452079772949, + "num_tokens": 430476884.0, + "step": 16647 + }, + { + "epoch": 1.8282451131122337, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5889647006988525, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7088101506233215, + "num_tokens": 430501811.0, + "step": 16648 + }, + { + "epoch": 1.8283549308148475, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.564591646194458, + "learning_rate": 1e-06, + "loss": 0.9402, + "mean_token_accuracy": 0.7205241918563843, + "num_tokens": 430525781.0, + "step": 16649 + }, + { + "epoch": 1.828464748517461, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.394940137863159, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7164819240570068, + "num_tokens": 430556236.0, + "step": 16650 + }, + { + "epoch": 1.8285745662200745, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.0157058238983154, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7209774255752563, + "num_tokens": 430575557.0, + "step": 16651 + }, + { + "epoch": 1.8286843839226883, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3978805541992188, + "learning_rate": 1e-06, + "loss": 0.936, + "mean_token_accuracy": 0.7268825769424438, + "num_tokens": 430602998.0, + "step": 16652 + }, + { + "epoch": 1.828794201625302, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7547616958618164, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7108040452003479, + "num_tokens": 430626485.0, + "step": 16653 + }, + { + "epoch": 1.8289040193279158, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5238900184631348, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7094553709030151, + "num_tokens": 430651993.0, + "step": 16654 + }, + { + "epoch": 1.8290138370305293, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.579561233520508, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.6960806846618652, + "num_tokens": 430677563.0, + "step": 16655 + }, + { + "epoch": 1.8291236547331429, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5639984607696533, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6856702566146851, + "num_tokens": 430705592.0, + "step": 16656 + }, + { + "epoch": 1.8292334724357566, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 3.015657663345337, + "learning_rate": 1e-06, + "loss": 1.101, + "mean_token_accuracy": 0.706496000289917, + "num_tokens": 430725170.0, + "step": 16657 + }, + { + "epoch": 1.8293432901383704, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.551490306854248, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6916695237159729, + "num_tokens": 430751860.0, + "step": 16658 + }, + { + "epoch": 1.829453107840984, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4104270935058594, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6941961050033569, + "num_tokens": 430781725.0, + "step": 16659 + }, + { + "epoch": 1.8295629255435977, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.462773323059082, + "learning_rate": 1e-06, + "loss": 0.9328, + "mean_token_accuracy": 0.7292200922966003, + "num_tokens": 430807699.0, + "step": 16660 + }, + { + "epoch": 1.8296727432462112, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6883773803710938, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7113679647445679, + "num_tokens": 430832221.0, + "step": 16661 + }, + { + "epoch": 1.829782560948825, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.58809494972229, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7076027989387512, + "num_tokens": 430856288.0, + "step": 16662 + }, + { + "epoch": 1.8298923786514387, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6200737953186035, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.6996481418609619, + "num_tokens": 430878692.0, + "step": 16663 + }, + { + "epoch": 1.8300021963540523, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5556602478027344, + "learning_rate": 1e-06, + "loss": 0.9397, + "mean_token_accuracy": 0.7207057476043701, + "num_tokens": 430902390.0, + "step": 16664 + }, + { + "epoch": 1.8301120140566658, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5001614093780518, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7015088200569153, + "num_tokens": 430927656.0, + "step": 16665 + }, + { + "epoch": 1.8302218317592795, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3733160495758057, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7100347876548767, + "num_tokens": 430955307.0, + "step": 16666 + }, + { + "epoch": 1.8303316494618933, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.490288257598877, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7044486403465271, + "num_tokens": 430984328.0, + "step": 16667 + }, + { + "epoch": 1.830441467164507, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.245495319366455, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6930772066116333, + "num_tokens": 431019994.0, + "step": 16668 + }, + { + "epoch": 1.8305512848671206, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5853500366210938, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7226620316505432, + "num_tokens": 431042177.0, + "step": 16669 + }, + { + "epoch": 1.8306611025697341, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3870208263397217, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7034280300140381, + "num_tokens": 431068939.0, + "step": 16670 + }, + { + "epoch": 1.830770920272348, + "ewc_loss": 9.59634780883789e-06, + "grad_norm": 2.7838003635406494, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.6920167207717896, + "num_tokens": 431089876.0, + "step": 16671 + }, + { + "epoch": 1.8308807379749616, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.521885633468628, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6827132701873779, + "num_tokens": 431114712.0, + "step": 16672 + }, + { + "epoch": 1.8309905556775752, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.729276657104492, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7187334895133972, + "num_tokens": 431136945.0, + "step": 16673 + }, + { + "epoch": 1.831100373380189, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6689703464508057, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7065668702125549, + "num_tokens": 431159197.0, + "step": 16674 + }, + { + "epoch": 1.8312101910828025, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7389214038848877, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6951169967651367, + "num_tokens": 431183436.0, + "step": 16675 + }, + { + "epoch": 1.8313200087854162, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5704808235168457, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7034115791320801, + "num_tokens": 431208000.0, + "step": 16676 + }, + { + "epoch": 1.83142982648803, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.917661666870117, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7031790018081665, + "num_tokens": 431229066.0, + "step": 16677 + }, + { + "epoch": 1.8315396441906435, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6403427124023438, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7172116637229919, + "num_tokens": 431253228.0, + "step": 16678 + }, + { + "epoch": 1.831649461893257, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5890376567840576, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7010300159454346, + "num_tokens": 431277606.0, + "step": 16679 + }, + { + "epoch": 1.8317592795958708, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.418687105178833, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6851907968521118, + "num_tokens": 431306796.0, + "step": 16680 + }, + { + "epoch": 1.8318690972984846, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4650800228118896, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6993088722229004, + "num_tokens": 431332376.0, + "step": 16681 + }, + { + "epoch": 1.8319789150010983, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2680130004882812, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7095773220062256, + "num_tokens": 431361858.0, + "step": 16682 + }, + { + "epoch": 1.8320887327037119, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6827292442321777, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.7082287073135376, + "num_tokens": 431385909.0, + "step": 16683 + }, + { + "epoch": 1.8321985504063254, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.721625566482544, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6844249367713928, + "num_tokens": 431411587.0, + "step": 16684 + }, + { + "epoch": 1.8323083681089392, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.546703577041626, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7055969834327698, + "num_tokens": 431438484.0, + "step": 16685 + }, + { + "epoch": 1.832418185811553, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.580528974533081, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.687890887260437, + "num_tokens": 431463481.0, + "step": 16686 + }, + { + "epoch": 1.8325280035141664, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5926437377929688, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7068967819213867, + "num_tokens": 431491588.0, + "step": 16687 + }, + { + "epoch": 1.83263782121678, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 5.645234107971191, + "learning_rate": 1e-06, + "loss": 0.9159, + "mean_token_accuracy": 0.7323286533355713, + "num_tokens": 431517339.0, + "step": 16688 + }, + { + "epoch": 1.8327476389193937, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6113929748535156, + "learning_rate": 1e-06, + "loss": 0.9033, + "mean_token_accuracy": 0.7369473576545715, + "num_tokens": 431543952.0, + "step": 16689 + }, + { + "epoch": 1.8328574566220075, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7884483337402344, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.701340913772583, + "num_tokens": 431566628.0, + "step": 16690 + }, + { + "epoch": 1.8329672743246213, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5682778358459473, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6966445446014404, + "num_tokens": 431595692.0, + "step": 16691 + }, + { + "epoch": 1.8330770920272348, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.741304397583008, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6979779005050659, + "num_tokens": 431617405.0, + "step": 16692 + }, + { + "epoch": 1.8331869097298483, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9294662475585938, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7109459638595581, + "num_tokens": 431639808.0, + "step": 16693 + }, + { + "epoch": 1.833296727432462, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.571727752685547, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.711756706237793, + "num_tokens": 431665541.0, + "step": 16694 + }, + { + "epoch": 1.8334065451350758, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.795243501663208, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7275465130805969, + "num_tokens": 431686833.0, + "step": 16695 + }, + { + "epoch": 1.8335163628376896, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8054041862487793, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7008732557296753, + "num_tokens": 431707512.0, + "step": 16696 + }, + { + "epoch": 1.8336261805403031, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.545248031616211, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7078596949577332, + "num_tokens": 431734053.0, + "step": 16697 + }, + { + "epoch": 1.8337359982429167, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.395822763442993, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6982222199440002, + "num_tokens": 431762374.0, + "step": 16698 + }, + { + "epoch": 1.8338458159455304, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4236903190612793, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6973546743392944, + "num_tokens": 431791432.0, + "step": 16699 + }, + { + "epoch": 1.8339556336481442, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5379812717437744, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7029867172241211, + "num_tokens": 431815212.0, + "step": 16700 + }, + { + "epoch": 1.8340654513507577, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5933401584625244, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7275602221488953, + "num_tokens": 431840246.0, + "step": 16701 + }, + { + "epoch": 1.8341752690533712, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.7187705039978027, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6961795091629028, + "num_tokens": 431865239.0, + "step": 16702 + }, + { + "epoch": 1.834285086755985, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.640958070755005, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.6979369521141052, + "num_tokens": 431889686.0, + "step": 16703 + }, + { + "epoch": 1.8343949044585988, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.4536163806915283, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6981112957000732, + "num_tokens": 431916534.0, + "step": 16704 + }, + { + "epoch": 1.8345047221612125, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.665179491043091, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.700567901134491, + "num_tokens": 431939591.0, + "step": 16705 + }, + { + "epoch": 1.834614539863826, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5294017791748047, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.720183253288269, + "num_tokens": 431962665.0, + "step": 16706 + }, + { + "epoch": 1.8347243575664396, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6199729442596436, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.7102732062339783, + "num_tokens": 431986591.0, + "step": 16707 + }, + { + "epoch": 1.8348341752690533, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5104572772979736, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7211735248565674, + "num_tokens": 432012179.0, + "step": 16708 + }, + { + "epoch": 1.834943992971667, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.791595935821533, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7049357295036316, + "num_tokens": 432035559.0, + "step": 16709 + }, + { + "epoch": 1.8350538106742806, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6926562786102295, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7189961075782776, + "num_tokens": 432058218.0, + "step": 16710 + }, + { + "epoch": 1.8351636283768944, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.670729637145996, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7169625759124756, + "num_tokens": 432087561.0, + "step": 16711 + }, + { + "epoch": 1.835273446079508, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.488781452178955, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7202142477035522, + "num_tokens": 432115675.0, + "step": 16712 + }, + { + "epoch": 1.8353832637821217, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.478370189666748, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.690798819065094, + "num_tokens": 432143933.0, + "step": 16713 + }, + { + "epoch": 1.8354930814847354, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.2935352325439453, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.6952501535415649, + "num_tokens": 432174272.0, + "step": 16714 + }, + { + "epoch": 1.835602899187349, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6977458000183105, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6913877725601196, + "num_tokens": 432196508.0, + "step": 16715 + }, + { + "epoch": 1.8357127168899625, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.715001106262207, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7029286026954651, + "num_tokens": 432219663.0, + "step": 16716 + }, + { + "epoch": 1.8358225345925763, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.403811454772949, + "learning_rate": 1e-06, + "loss": 1.1835, + "mean_token_accuracy": 0.6682676076889038, + "num_tokens": 432251645.0, + "step": 16717 + }, + { + "epoch": 1.83593235229519, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.330606460571289, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7019293308258057, + "num_tokens": 432279998.0, + "step": 16718 + }, + { + "epoch": 1.8360421699978038, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6130149364471436, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6971752047538757, + "num_tokens": 432309369.0, + "step": 16719 + }, + { + "epoch": 1.8361519877004173, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5854403972625732, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7231537103652954, + "num_tokens": 432335263.0, + "step": 16720 + }, + { + "epoch": 1.8362618054030309, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.869990348815918, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6997543573379517, + "num_tokens": 432362386.0, + "step": 16721 + }, + { + "epoch": 1.8363716231056446, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6042041778564453, + "learning_rate": 1e-06, + "loss": 1.1369, + "mean_token_accuracy": 0.6810185313224792, + "num_tokens": 432388428.0, + "step": 16722 + }, + { + "epoch": 1.8364814408082584, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.570612668991089, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.6907875537872314, + "num_tokens": 432415030.0, + "step": 16723 + }, + { + "epoch": 1.836591258510872, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.474215030670166, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7061713933944702, + "num_tokens": 432438623.0, + "step": 16724 + }, + { + "epoch": 1.8367010762134857, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.78877854347229, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7240524291992188, + "num_tokens": 432458752.0, + "step": 16725 + }, + { + "epoch": 1.8368108939160992, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6045985221862793, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6993166208267212, + "num_tokens": 432482944.0, + "step": 16726 + }, + { + "epoch": 1.836920711618713, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 16.675554275512695, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7028687596321106, + "num_tokens": 432510600.0, + "step": 16727 + }, + { + "epoch": 1.8370305293213267, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.884633779525757, + "learning_rate": 1e-06, + "loss": 0.9212, + "mean_token_accuracy": 0.728890061378479, + "num_tokens": 432532220.0, + "step": 16728 + }, + { + "epoch": 1.8371403470239402, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7861642837524414, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7131115198135376, + "num_tokens": 432556467.0, + "step": 16729 + }, + { + "epoch": 1.8372501647265538, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4129908084869385, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6904362440109253, + "num_tokens": 432588323.0, + "step": 16730 + }, + { + "epoch": 1.8373599824291675, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6378016471862793, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6915632486343384, + "num_tokens": 432610998.0, + "step": 16731 + }, + { + "epoch": 1.8374698001317813, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4183387756347656, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6983690857887268, + "num_tokens": 432639635.0, + "step": 16732 + }, + { + "epoch": 1.837579617834395, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6168737411499023, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7266719937324524, + "num_tokens": 432665412.0, + "step": 16733 + }, + { + "epoch": 1.8376894355370086, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.908905029296875, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.7179116010665894, + "num_tokens": 432684610.0, + "step": 16734 + }, + { + "epoch": 1.8377992532396221, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.721782922744751, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.7000374794006348, + "num_tokens": 432707798.0, + "step": 16735 + }, + { + "epoch": 1.8379090709422359, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.649867296218872, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7152351140975952, + "num_tokens": 432731428.0, + "step": 16736 + }, + { + "epoch": 1.8380188886448496, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.866633892059326, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6976837515830994, + "num_tokens": 432755040.0, + "step": 16737 + }, + { + "epoch": 1.8381287063474632, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9494118690490723, + "learning_rate": 1e-06, + "loss": 0.9161, + "mean_token_accuracy": 0.7291849851608276, + "num_tokens": 432775161.0, + "step": 16738 + }, + { + "epoch": 1.838238524050077, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.782179117202759, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7125263214111328, + "num_tokens": 432797686.0, + "step": 16739 + }, + { + "epoch": 1.8383483417526905, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.384732723236084, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6908097267150879, + "num_tokens": 432831722.0, + "step": 16740 + }, + { + "epoch": 1.8384581594553042, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.63106632232666, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7069528102874756, + "num_tokens": 432857199.0, + "step": 16741 + }, + { + "epoch": 1.838567977157918, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8413941860198975, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6924651861190796, + "num_tokens": 432884640.0, + "step": 16742 + }, + { + "epoch": 1.8386777948605315, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8637890815734863, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.7042340040206909, + "num_tokens": 432906955.0, + "step": 16743 + }, + { + "epoch": 1.838787612563145, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4349253177642822, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7101192474365234, + "num_tokens": 432936136.0, + "step": 16744 + }, + { + "epoch": 1.8388974302657588, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.587588310241699, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7039819955825806, + "num_tokens": 432962710.0, + "step": 16745 + }, + { + "epoch": 1.8390072479683726, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6987860202789307, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7093235850334167, + "num_tokens": 432985909.0, + "step": 16746 + }, + { + "epoch": 1.8391170656709863, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.728513240814209, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7176803350448608, + "num_tokens": 433005395.0, + "step": 16747 + }, + { + "epoch": 1.8392268833735999, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5922956466674805, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7222506999969482, + "num_tokens": 433031676.0, + "step": 16748 + }, + { + "epoch": 1.8393367010762134, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.51210355758667, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.7045122981071472, + "num_tokens": 433058830.0, + "step": 16749 + }, + { + "epoch": 1.8394465187788271, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8710060119628906, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7150634527206421, + "num_tokens": 433082843.0, + "step": 16750 + }, + { + "epoch": 1.839556336481441, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.9308974742889404, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7143234014511108, + "num_tokens": 433102919.0, + "step": 16751 + }, + { + "epoch": 1.8396661541840544, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.986159563064575, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7245905995368958, + "num_tokens": 433120615.0, + "step": 16752 + }, + { + "epoch": 1.839775971886668, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.733886480331421, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.7056344151496887, + "num_tokens": 433144967.0, + "step": 16753 + }, + { + "epoch": 1.8398857895892817, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7526233196258545, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.7041815519332886, + "num_tokens": 433167433.0, + "step": 16754 + }, + { + "epoch": 1.8399956072918955, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.567927837371826, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.7020719647407532, + "num_tokens": 433194566.0, + "step": 16755 + }, + { + "epoch": 1.8401054249945092, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.367584466934204, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7118883728981018, + "num_tokens": 433222826.0, + "step": 16756 + }, + { + "epoch": 1.8402152426971228, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 16.699932098388672, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7063910365104675, + "num_tokens": 433246856.0, + "step": 16757 + }, + { + "epoch": 1.8403250603997363, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.699655294418335, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7186641693115234, + "num_tokens": 433270364.0, + "step": 16758 + }, + { + "epoch": 1.84043487810235, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5708119869232178, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6886987686157227, + "num_tokens": 433298091.0, + "step": 16759 + }, + { + "epoch": 1.8405446958049638, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3812148571014404, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7295349836349487, + "num_tokens": 433326190.0, + "step": 16760 + }, + { + "epoch": 1.8406545135075776, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2976291179656982, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7101908922195435, + "num_tokens": 433356604.0, + "step": 16761 + }, + { + "epoch": 1.8407643312101911, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4598562717437744, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6969809532165527, + "num_tokens": 433383447.0, + "step": 16762 + }, + { + "epoch": 1.8408741489128047, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4616518020629883, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6815946698188782, + "num_tokens": 433411016.0, + "step": 16763 + }, + { + "epoch": 1.8409839666154184, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4773507118225098, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6968198418617249, + "num_tokens": 433437278.0, + "step": 16764 + }, + { + "epoch": 1.8410937843180322, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6871824264526367, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7180706262588501, + "num_tokens": 433458780.0, + "step": 16765 + }, + { + "epoch": 1.8412036020206457, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5931105613708496, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6916601657867432, + "num_tokens": 433483918.0, + "step": 16766 + }, + { + "epoch": 1.8413134197232592, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8307693004608154, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7093420624732971, + "num_tokens": 433505207.0, + "step": 16767 + }, + { + "epoch": 1.841423237425873, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6315457820892334, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.732255756855011, + "num_tokens": 433527139.0, + "step": 16768 + }, + { + "epoch": 1.8415330551284868, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5778512954711914, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.7033272981643677, + "num_tokens": 433552161.0, + "step": 16769 + }, + { + "epoch": 1.8416428728311005, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6645615100860596, + "learning_rate": 1e-06, + "loss": 1.1136, + "mean_token_accuracy": 0.6811209917068481, + "num_tokens": 433578670.0, + "step": 16770 + }, + { + "epoch": 1.841752690533714, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6827588081359863, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7004674673080444, + "num_tokens": 433599905.0, + "step": 16771 + }, + { + "epoch": 1.8418625082363276, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4533374309539795, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6881955862045288, + "num_tokens": 433628044.0, + "step": 16772 + }, + { + "epoch": 1.8419723259389413, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.53244686126709, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7054795026779175, + "num_tokens": 433654532.0, + "step": 16773 + }, + { + "epoch": 1.842082143641555, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.319535732269287, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7120637893676758, + "num_tokens": 433685377.0, + "step": 16774 + }, + { + "epoch": 1.8421919613441686, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6306049823760986, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.7031626105308533, + "num_tokens": 433708832.0, + "step": 16775 + }, + { + "epoch": 1.8423017790467824, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6378378868103027, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.7330735325813293, + "num_tokens": 433733598.0, + "step": 16776 + }, + { + "epoch": 1.842411596749396, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5748720169067383, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7232030630111694, + "num_tokens": 433757291.0, + "step": 16777 + }, + { + "epoch": 1.8425214144520097, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.533750057220459, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.7186806201934814, + "num_tokens": 433781578.0, + "step": 16778 + }, + { + "epoch": 1.8426312321546234, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6309151649475098, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6917999386787415, + "num_tokens": 433807409.0, + "step": 16779 + }, + { + "epoch": 1.842741049857237, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.417689561843872, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7051515579223633, + "num_tokens": 433836391.0, + "step": 16780 + }, + { + "epoch": 1.8428508675598505, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.407116651535034, + "learning_rate": 1e-06, + "loss": 1.1149, + "mean_token_accuracy": 0.681788980960846, + "num_tokens": 433869430.0, + "step": 16781 + }, + { + "epoch": 1.8429606852624643, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3645262718200684, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7095023393630981, + "num_tokens": 433899256.0, + "step": 16782 + }, + { + "epoch": 1.843070502965078, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6883347034454346, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.6985363960266113, + "num_tokens": 433922331.0, + "step": 16783 + }, + { + "epoch": 1.8431803206676918, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.259931802749634, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.7042636871337891, + "num_tokens": 433952389.0, + "step": 16784 + }, + { + "epoch": 1.8432901383703053, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8115670680999756, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7183359861373901, + "num_tokens": 433972878.0, + "step": 16785 + }, + { + "epoch": 1.8433999560729188, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.794783592224121, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7154442071914673, + "num_tokens": 433996169.0, + "step": 16786 + }, + { + "epoch": 1.8435097737755326, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3894293308258057, + "learning_rate": 1e-06, + "loss": 0.9191, + "mean_token_accuracy": 0.7317544221878052, + "num_tokens": 434023402.0, + "step": 16787 + }, + { + "epoch": 1.8436195914781464, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.704333782196045, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7130504846572876, + "num_tokens": 434048443.0, + "step": 16788 + }, + { + "epoch": 1.84372940918076, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6903343200683594, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7204632759094238, + "num_tokens": 434074861.0, + "step": 16789 + }, + { + "epoch": 1.8438392268833736, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3733317852020264, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6919181942939758, + "num_tokens": 434105980.0, + "step": 16790 + }, + { + "epoch": 1.8439490445859872, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6196224689483643, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7084164619445801, + "num_tokens": 434129951.0, + "step": 16791 + }, + { + "epoch": 1.844058862288601, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.803118944168091, + "learning_rate": 1e-06, + "loss": 0.9197, + "mean_token_accuracy": 0.726443886756897, + "num_tokens": 434151178.0, + "step": 16792 + }, + { + "epoch": 1.8441686799912147, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5916874408721924, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7183105945587158, + "num_tokens": 434179016.0, + "step": 16793 + }, + { + "epoch": 1.8442784976938282, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.2454628944396973, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.6972943544387817, + "num_tokens": 434198588.0, + "step": 16794 + }, + { + "epoch": 1.8443883153964418, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.780740261077881, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7268000841140747, + "num_tokens": 434220242.0, + "step": 16795 + }, + { + "epoch": 1.8444981330990555, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5179619789123535, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7022722363471985, + "num_tokens": 434246939.0, + "step": 16796 + }, + { + "epoch": 1.8446079508016693, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.439840316772461, + "learning_rate": 1e-06, + "loss": 1.1097, + "mean_token_accuracy": 0.6810817122459412, + "num_tokens": 434275946.0, + "step": 16797 + }, + { + "epoch": 1.844717768504283, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7716448307037354, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6993383765220642, + "num_tokens": 434299740.0, + "step": 16798 + }, + { + "epoch": 1.8448275862068966, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3848798274993896, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6852713823318481, + "num_tokens": 434330720.0, + "step": 16799 + }, + { + "epoch": 1.84493740390951, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.329409599304199, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7024262547492981, + "num_tokens": 434359474.0, + "step": 16800 + }, + { + "epoch": 1.8450472216121239, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.991943597793579, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7149896025657654, + "num_tokens": 434383772.0, + "step": 16801 + }, + { + "epoch": 1.8451570393147376, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.381391763687134, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7066992521286011, + "num_tokens": 434412512.0, + "step": 16802 + }, + { + "epoch": 1.8452668570173512, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5669028759002686, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7071583271026611, + "num_tokens": 434435345.0, + "step": 16803 + }, + { + "epoch": 1.8453766747199647, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6396570205688477, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7028440237045288, + "num_tokens": 434459575.0, + "step": 16804 + }, + { + "epoch": 1.8454864924225785, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5368216037750244, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6976315975189209, + "num_tokens": 434487834.0, + "step": 16805 + }, + { + "epoch": 1.8455963101251922, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8115298748016357, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7103995084762573, + "num_tokens": 434508043.0, + "step": 16806 + }, + { + "epoch": 1.845706127827806, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5203566551208496, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7213835716247559, + "num_tokens": 434531909.0, + "step": 16807 + }, + { + "epoch": 1.8458159455304195, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.627427101135254, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7041195631027222, + "num_tokens": 434556812.0, + "step": 16808 + }, + { + "epoch": 1.845925763233033, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.657761573791504, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7143397331237793, + "num_tokens": 434585807.0, + "step": 16809 + }, + { + "epoch": 1.8460355809356468, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3234753608703613, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.6986993551254272, + "num_tokens": 434619430.0, + "step": 16810 + }, + { + "epoch": 1.8461453986382605, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.540292739868164, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7006771564483643, + "num_tokens": 434644649.0, + "step": 16811 + }, + { + "epoch": 1.8462552163408743, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3763768672943115, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7101967334747314, + "num_tokens": 434675536.0, + "step": 16812 + }, + { + "epoch": 1.8463650340434878, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.618583917617798, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7026920914649963, + "num_tokens": 434702111.0, + "step": 16813 + }, + { + "epoch": 1.8464748517461014, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4874823093414307, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7253382802009583, + "num_tokens": 434727742.0, + "step": 16814 + }, + { + "epoch": 1.8465846694487151, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8498716354370117, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7100602388381958, + "num_tokens": 434753236.0, + "step": 16815 + }, + { + "epoch": 1.846694487151329, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8669137954711914, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6966714859008789, + "num_tokens": 434780497.0, + "step": 16816 + }, + { + "epoch": 1.8468043048539424, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9416539669036865, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7240480184555054, + "num_tokens": 434801305.0, + "step": 16817 + }, + { + "epoch": 1.846914122556556, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.677621841430664, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7216312885284424, + "num_tokens": 434825040.0, + "step": 16818 + }, + { + "epoch": 1.8470239402591697, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.818523406982422, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.696186900138855, + "num_tokens": 434848252.0, + "step": 16819 + }, + { + "epoch": 1.8471337579617835, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.637270927429199, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6996273994445801, + "num_tokens": 434874096.0, + "step": 16820 + }, + { + "epoch": 1.8472435756643972, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7577970027923584, + "learning_rate": 1e-06, + "loss": 0.9134, + "mean_token_accuracy": 0.7298038005828857, + "num_tokens": 434894708.0, + "step": 16821 + }, + { + "epoch": 1.8473533933670108, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.431166172027588, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7039902210235596, + "num_tokens": 434923205.0, + "step": 16822 + }, + { + "epoch": 1.8474632110696243, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.734921455383301, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7255934476852417, + "num_tokens": 434944794.0, + "step": 16823 + }, + { + "epoch": 1.847573028772238, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4745304584503174, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.6971023082733154, + "num_tokens": 434969538.0, + "step": 16824 + }, + { + "epoch": 1.8476828464748518, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.71743106842041, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7007741332054138, + "num_tokens": 434991412.0, + "step": 16825 + }, + { + "epoch": 1.8477926641774656, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.931689739227295, + "learning_rate": 1e-06, + "loss": 1.1094, + "mean_token_accuracy": 0.6856662034988403, + "num_tokens": 435021016.0, + "step": 16826 + }, + { + "epoch": 1.847902481880079, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6652843952178955, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.695344090461731, + "num_tokens": 435044767.0, + "step": 16827 + }, + { + "epoch": 1.8480122995826926, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6999545097351074, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.7125651836395264, + "num_tokens": 435068460.0, + "step": 16828 + }, + { + "epoch": 1.8481221172853064, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5749995708465576, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7068343758583069, + "num_tokens": 435092176.0, + "step": 16829 + }, + { + "epoch": 1.8482319349879202, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4349710941314697, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6968024969100952, + "num_tokens": 435122503.0, + "step": 16830 + }, + { + "epoch": 1.8483417526905337, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5353190898895264, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7089073657989502, + "num_tokens": 435147829.0, + "step": 16831 + }, + { + "epoch": 1.8484515703931472, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6169371604919434, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6923254728317261, + "num_tokens": 435170486.0, + "step": 16832 + }, + { + "epoch": 1.848561388095761, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.50549578666687, + "learning_rate": 1e-06, + "loss": 1.111, + "mean_token_accuracy": 0.7022172212600708, + "num_tokens": 435198505.0, + "step": 16833 + }, + { + "epoch": 1.8486712057983747, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.779625654220581, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.7036876082420349, + "num_tokens": 435220991.0, + "step": 16834 + }, + { + "epoch": 1.8487810235009885, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.57254958152771, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.7009605169296265, + "num_tokens": 435247749.0, + "step": 16835 + }, + { + "epoch": 1.848890841203602, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.026101589202881, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.699947714805603, + "num_tokens": 435266982.0, + "step": 16836 + }, + { + "epoch": 1.8490006589062156, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9443199634552, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7271382808685303, + "num_tokens": 435286124.0, + "step": 16837 + }, + { + "epoch": 1.8491104766088293, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.654707908630371, + "learning_rate": 1e-06, + "loss": 0.9248, + "mean_token_accuracy": 0.7313181757926941, + "num_tokens": 435308497.0, + "step": 16838 + }, + { + "epoch": 1.849220294311443, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5299277305603027, + "learning_rate": 1e-06, + "loss": 1.1287, + "mean_token_accuracy": 0.6719577312469482, + "num_tokens": 435337722.0, + "step": 16839 + }, + { + "epoch": 1.8493301120140566, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8268778324127197, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.7226082682609558, + "num_tokens": 435358826.0, + "step": 16840 + }, + { + "epoch": 1.8494399297166704, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7593772411346436, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6911106109619141, + "num_tokens": 435383016.0, + "step": 16841 + }, + { + "epoch": 1.849549747419284, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.593364953994751, + "learning_rate": 1e-06, + "loss": 1.1144, + "mean_token_accuracy": 0.691291332244873, + "num_tokens": 435411040.0, + "step": 16842 + }, + { + "epoch": 1.8496595651218977, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.510575771331787, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7180607318878174, + "num_tokens": 435437714.0, + "step": 16843 + }, + { + "epoch": 1.8497693828245114, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.52734375, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7067753672599792, + "num_tokens": 435463892.0, + "step": 16844 + }, + { + "epoch": 1.849879200527125, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.33178448677063, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6908231973648071, + "num_tokens": 435494700.0, + "step": 16845 + }, + { + "epoch": 1.8499890182297385, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8865485191345215, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7130938768386841, + "num_tokens": 435514755.0, + "step": 16846 + }, + { + "epoch": 1.8500988359323522, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.816326379776001, + "learning_rate": 1e-06, + "loss": 0.9123, + "mean_token_accuracy": 0.731338381767273, + "num_tokens": 435534376.0, + "step": 16847 + }, + { + "epoch": 1.850208653634966, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.740441083908081, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7011991739273071, + "num_tokens": 435557910.0, + "step": 16848 + }, + { + "epoch": 1.8503184713375798, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3788249492645264, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6897050142288208, + "num_tokens": 435593600.0, + "step": 16849 + }, + { + "epoch": 1.8504282890401933, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8699145317077637, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6854934692382812, + "num_tokens": 435615993.0, + "step": 16850 + }, + { + "epoch": 1.8505381067428068, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5598361492156982, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.700875997543335, + "num_tokens": 435640678.0, + "step": 16851 + }, + { + "epoch": 1.8506479244454206, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.597226858139038, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6926300525665283, + "num_tokens": 435666368.0, + "step": 16852 + }, + { + "epoch": 1.8507577421480343, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 80.54882049560547, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7159510254859924, + "num_tokens": 435686611.0, + "step": 16853 + }, + { + "epoch": 1.8508675598506479, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.580512046813965, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7050575017929077, + "num_tokens": 435715453.0, + "step": 16854 + }, + { + "epoch": 1.8509773775532616, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.962336540222168, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6856259107589722, + "num_tokens": 435739985.0, + "step": 16855 + }, + { + "epoch": 1.8510871952558752, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6378862857818604, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.6965664029121399, + "num_tokens": 435767221.0, + "step": 16856 + }, + { + "epoch": 1.851197012958489, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9028289318084717, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7287750244140625, + "num_tokens": 435790554.0, + "step": 16857 + }, + { + "epoch": 1.8513068306611027, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5055830478668213, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7138763070106506, + "num_tokens": 435817734.0, + "step": 16858 + }, + { + "epoch": 1.8514166483637162, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.0862183570861816, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6920585632324219, + "num_tokens": 435845409.0, + "step": 16859 + }, + { + "epoch": 1.8515264660663298, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 16.757070541381836, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.71123206615448, + "num_tokens": 435867146.0, + "step": 16860 + }, + { + "epoch": 1.8516362837689435, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6177961826324463, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.700947642326355, + "num_tokens": 435896313.0, + "step": 16861 + }, + { + "epoch": 1.8517461014715573, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.1510260105133057, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.7010778784751892, + "num_tokens": 435916873.0, + "step": 16862 + }, + { + "epoch": 1.851855919174171, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6061460971832275, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.7228423357009888, + "num_tokens": 435940352.0, + "step": 16863 + }, + { + "epoch": 1.8519657368767846, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.577155113220215, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7163820862770081, + "num_tokens": 435965409.0, + "step": 16864 + }, + { + "epoch": 1.852075554579398, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6042428016662598, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7212030291557312, + "num_tokens": 435989123.0, + "step": 16865 + }, + { + "epoch": 1.8521853722820119, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.87929105758667, + "learning_rate": 1e-06, + "loss": 0.9471, + "mean_token_accuracy": 0.7229211330413818, + "num_tokens": 436010896.0, + "step": 16866 + }, + { + "epoch": 1.8522951899846256, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6299123764038086, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6957564949989319, + "num_tokens": 436037438.0, + "step": 16867 + }, + { + "epoch": 1.8524050076872391, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.0154459476470947, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.6978914737701416, + "num_tokens": 436056645.0, + "step": 16868 + }, + { + "epoch": 1.8525148253898527, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.1045174598693848, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6887892484664917, + "num_tokens": 436084261.0, + "step": 16869 + }, + { + "epoch": 1.8526246430924664, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4776220321655273, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6960017681121826, + "num_tokens": 436115170.0, + "step": 16870 + }, + { + "epoch": 1.8527344607950802, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8792765140533447, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7097672820091248, + "num_tokens": 436138017.0, + "step": 16871 + }, + { + "epoch": 1.852844278497694, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6584432125091553, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6931203603744507, + "num_tokens": 436164826.0, + "step": 16872 + }, + { + "epoch": 1.8529540962003075, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6983189582824707, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.724011242389679, + "num_tokens": 436186950.0, + "step": 16873 + }, + { + "epoch": 1.853063913902921, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3722140789031982, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6883723735809326, + "num_tokens": 436218554.0, + "step": 16874 + }, + { + "epoch": 1.8531737316055348, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.5489745140075684, + "learning_rate": 1e-06, + "loss": 1.1595, + "mean_token_accuracy": 0.6738190650939941, + "num_tokens": 436245732.0, + "step": 16875 + }, + { + "epoch": 1.8532835493081485, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.286163806915283, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7178943753242493, + "num_tokens": 436275907.0, + "step": 16876 + }, + { + "epoch": 1.8533933670107623, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6597278118133545, + "learning_rate": 1e-06, + "loss": 0.9046, + "mean_token_accuracy": 0.7344143390655518, + "num_tokens": 436298559.0, + "step": 16877 + }, + { + "epoch": 1.8535031847133758, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.413822889328003, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6943709254264832, + "num_tokens": 436328711.0, + "step": 16878 + }, + { + "epoch": 1.8536130024159894, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.757840156555176, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7087414264678955, + "num_tokens": 436351002.0, + "step": 16879 + }, + { + "epoch": 1.8537228201186031, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.714022397994995, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6937177777290344, + "num_tokens": 436376844.0, + "step": 16880 + }, + { + "epoch": 1.8538326378212169, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.56207537651062, + "learning_rate": 1e-06, + "loss": 1.0916, + "mean_token_accuracy": 0.6853254437446594, + "num_tokens": 436403455.0, + "step": 16881 + }, + { + "epoch": 1.8539424555238304, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6037681102752686, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.7016354203224182, + "num_tokens": 436427524.0, + "step": 16882 + }, + { + "epoch": 1.854052273226444, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5873658657073975, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7146511673927307, + "num_tokens": 436454649.0, + "step": 16883 + }, + { + "epoch": 1.8541620909290577, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.6475491523742676, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7218012809753418, + "num_tokens": 436477195.0, + "step": 16884 + }, + { + "epoch": 1.8542719086316715, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3588497638702393, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.7073624134063721, + "num_tokens": 436505904.0, + "step": 16885 + }, + { + "epoch": 1.8543817263342852, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 4.410009860992432, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7042763233184814, + "num_tokens": 436531585.0, + "step": 16886 + }, + { + "epoch": 1.8544915440368988, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4599153995513916, + "learning_rate": 1e-06, + "loss": 1.1056, + "mean_token_accuracy": 0.6846268177032471, + "num_tokens": 436563798.0, + "step": 16887 + }, + { + "epoch": 1.8546013617395123, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4300320148468018, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.7042233347892761, + "num_tokens": 436592580.0, + "step": 16888 + }, + { + "epoch": 1.854711179442126, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.863818883895874, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7200090885162354, + "num_tokens": 436615814.0, + "step": 16889 + }, + { + "epoch": 1.8548209971447398, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5088438987731934, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7103214263916016, + "num_tokens": 436642617.0, + "step": 16890 + }, + { + "epoch": 1.8549308148473536, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.962352752685547, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6948621869087219, + "num_tokens": 436666971.0, + "step": 16891 + }, + { + "epoch": 1.855040632549967, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4550647735595703, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6947341561317444, + "num_tokens": 436695065.0, + "step": 16892 + }, + { + "epoch": 1.8551504502525806, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.107989549636841, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6984683871269226, + "num_tokens": 436724658.0, + "step": 16893 + }, + { + "epoch": 1.8552602679551944, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5312960147857666, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7028397917747498, + "num_tokens": 436750097.0, + "step": 16894 + }, + { + "epoch": 1.8553700856578081, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5873420238494873, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.714561939239502, + "num_tokens": 436774023.0, + "step": 16895 + }, + { + "epoch": 1.8554799033604217, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.603050947189331, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.705197274684906, + "num_tokens": 436800885.0, + "step": 16896 + }, + { + "epoch": 1.8555897210630352, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8075451850891113, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.696792721748352, + "num_tokens": 436824212.0, + "step": 16897 + }, + { + "epoch": 1.855699538765649, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.606912612915039, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7232856750488281, + "num_tokens": 436843050.0, + "step": 16898 + }, + { + "epoch": 1.8558093564682627, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.3136708736419678, + "learning_rate": 1e-06, + "loss": 1.1059, + "mean_token_accuracy": 0.6868158578872681, + "num_tokens": 436880926.0, + "step": 16899 + }, + { + "epoch": 1.8559191741708765, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.558582067489624, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7021079063415527, + "num_tokens": 436908418.0, + "step": 16900 + }, + { + "epoch": 1.85602899187349, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.66583514213562, + "learning_rate": 1e-06, + "loss": 0.924, + "mean_token_accuracy": 0.7243140935897827, + "num_tokens": 436933281.0, + "step": 16901 + }, + { + "epoch": 1.8561388095761036, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.409457206726074, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7069318294525146, + "num_tokens": 436959755.0, + "step": 16902 + }, + { + "epoch": 1.8562486272787173, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.558957576751709, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7364856004714966, + "num_tokens": 436983183.0, + "step": 16903 + }, + { + "epoch": 1.856358444981331, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.8940610885620117, + "learning_rate": 1e-06, + "loss": 0.8517, + "mean_token_accuracy": 0.7494571208953857, + "num_tokens": 437000635.0, + "step": 16904 + }, + { + "epoch": 1.8564682626839446, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5890953540802, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6977924108505249, + "num_tokens": 437025503.0, + "step": 16905 + }, + { + "epoch": 1.8565780803865584, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6820266246795654, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7099226713180542, + "num_tokens": 437051630.0, + "step": 16906 + }, + { + "epoch": 1.856687898089172, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5609214305877686, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7290015816688538, + "num_tokens": 437076770.0, + "step": 16907 + }, + { + "epoch": 1.8567977157917857, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.045626401901245, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7132842540740967, + "num_tokens": 437108400.0, + "step": 16908 + }, + { + "epoch": 1.8569075334943994, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2981135845184326, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6965346336364746, + "num_tokens": 437140555.0, + "step": 16909 + }, + { + "epoch": 1.857017351197013, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7227942943573, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7107136249542236, + "num_tokens": 437162231.0, + "step": 16910 + }, + { + "epoch": 1.8571271688996265, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.522731065750122, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7024426460266113, + "num_tokens": 437188757.0, + "step": 16911 + }, + { + "epoch": 1.8572369866022402, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.37066912651062, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7292371392250061, + "num_tokens": 437215715.0, + "step": 16912 + }, + { + "epoch": 1.857346804304854, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.513371706008911, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7107876539230347, + "num_tokens": 437240293.0, + "step": 16913 + }, + { + "epoch": 1.8574566220074678, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.405353307723999, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.714817464351654, + "num_tokens": 437271806.0, + "step": 16914 + }, + { + "epoch": 1.8575664397100813, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.777085781097412, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7221054434776306, + "num_tokens": 437295395.0, + "step": 16915 + }, + { + "epoch": 1.8576762574126948, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7985281944274902, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7038525342941284, + "num_tokens": 437316129.0, + "step": 16916 + }, + { + "epoch": 1.8577860751153086, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.49187970161438, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6878504753112793, + "num_tokens": 437342790.0, + "step": 16917 + }, + { + "epoch": 1.8578958928179223, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.379335880279541, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6862987279891968, + "num_tokens": 437372755.0, + "step": 16918 + }, + { + "epoch": 1.8580057105205359, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2808589935302734, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7009762525558472, + "num_tokens": 437406550.0, + "step": 16919 + }, + { + "epoch": 1.8581155282231496, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.986453056335449, + "learning_rate": 1e-06, + "loss": 0.8989, + "mean_token_accuracy": 0.7340644598007202, + "num_tokens": 437428676.0, + "step": 16920 + }, + { + "epoch": 1.8582253459257632, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4936506748199463, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6896870732307434, + "num_tokens": 437455018.0, + "step": 16921 + }, + { + "epoch": 1.858335163628377, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.754793643951416, + "learning_rate": 1e-06, + "loss": 1.1175, + "mean_token_accuracy": 0.6920943260192871, + "num_tokens": 437479847.0, + "step": 16922 + }, + { + "epoch": 1.8584449813309907, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5506296157836914, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.695972740650177, + "num_tokens": 437507113.0, + "step": 16923 + }, + { + "epoch": 1.8585547990336042, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.775066614151001, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6936338543891907, + "num_tokens": 437529316.0, + "step": 16924 + }, + { + "epoch": 1.8586646167362177, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3409481048583984, + "learning_rate": 1e-06, + "loss": 1.1007, + "mean_token_accuracy": 0.6780986785888672, + "num_tokens": 437560522.0, + "step": 16925 + }, + { + "epoch": 1.8587744344388315, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4481360912323, + "learning_rate": 1e-06, + "loss": 1.1009, + "mean_token_accuracy": 0.6830514669418335, + "num_tokens": 437590821.0, + "step": 16926 + }, + { + "epoch": 1.8588842521414453, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.551490068435669, + "learning_rate": 1e-06, + "loss": 1.1515, + "mean_token_accuracy": 0.6882007122039795, + "num_tokens": 437619416.0, + "step": 16927 + }, + { + "epoch": 1.858994069844059, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.759833574295044, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.6862889528274536, + "num_tokens": 437643616.0, + "step": 16928 + }, + { + "epoch": 1.8591038875466726, + "ewc_loss": 9.655952453613281e-06, + "grad_norm": 2.394538402557373, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7160575985908508, + "num_tokens": 437670853.0, + "step": 16929 + }, + { + "epoch": 1.859213705249286, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.674062490463257, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.714820384979248, + "num_tokens": 437692695.0, + "step": 16930 + }, + { + "epoch": 1.8593235229518998, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.606123924255371, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7117269039154053, + "num_tokens": 437719371.0, + "step": 16931 + }, + { + "epoch": 1.8594333406545136, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6782445907592773, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.718211829662323, + "num_tokens": 437743999.0, + "step": 16932 + }, + { + "epoch": 1.8595431583571271, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.728137731552124, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.6955057382583618, + "num_tokens": 437766725.0, + "step": 16933 + }, + { + "epoch": 1.8596529760597407, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5894062519073486, + "learning_rate": 1e-06, + "loss": 0.8929, + "mean_token_accuracy": 0.7322913408279419, + "num_tokens": 437790488.0, + "step": 16934 + }, + { + "epoch": 1.8597627937623544, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5894978046417236, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.6990251541137695, + "num_tokens": 437814868.0, + "step": 16935 + }, + { + "epoch": 1.8598726114649682, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9138877391815186, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7107832431793213, + "num_tokens": 437834285.0, + "step": 16936 + }, + { + "epoch": 1.859982429167582, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3965022563934326, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6856026649475098, + "num_tokens": 437862652.0, + "step": 16937 + }, + { + "epoch": 1.8600922468701955, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.549987316131592, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7053113579750061, + "num_tokens": 437886820.0, + "step": 16938 + }, + { + "epoch": 1.860202064572809, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4986467361450195, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7130457758903503, + "num_tokens": 437919256.0, + "step": 16939 + }, + { + "epoch": 1.8603118822754228, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.464111328125, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.6880775690078735, + "num_tokens": 437947549.0, + "step": 16940 + }, + { + "epoch": 1.8604216999780365, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6000404357910156, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6921216249465942, + "num_tokens": 437972843.0, + "step": 16941 + }, + { + "epoch": 1.8605315176806503, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.571043014526367, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7124202251434326, + "num_tokens": 437998600.0, + "step": 16942 + }, + { + "epoch": 1.8606413353832638, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6031572818756104, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7085956335067749, + "num_tokens": 438028102.0, + "step": 16943 + }, + { + "epoch": 1.8607511530858774, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6549103260040283, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.6922316551208496, + "num_tokens": 438052822.0, + "step": 16944 + }, + { + "epoch": 1.860860970788491, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.476207971572876, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.6966229677200317, + "num_tokens": 438082844.0, + "step": 16945 + }, + { + "epoch": 1.8609707884911049, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.630401134490967, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7043461799621582, + "num_tokens": 438109890.0, + "step": 16946 + }, + { + "epoch": 1.8610806061937184, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.405557632446289, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.7032334208488464, + "num_tokens": 438140895.0, + "step": 16947 + }, + { + "epoch": 1.861190423896332, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5802619457244873, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7197798490524292, + "num_tokens": 438165806.0, + "step": 16948 + }, + { + "epoch": 1.8613002415989457, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8406982421875, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7200334072113037, + "num_tokens": 438190317.0, + "step": 16949 + }, + { + "epoch": 1.8614100593015594, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4946558475494385, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7030317783355713, + "num_tokens": 438215540.0, + "step": 16950 + }, + { + "epoch": 1.8615198770041732, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.670700788497925, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7169285416603088, + "num_tokens": 438239349.0, + "step": 16951 + }, + { + "epoch": 1.8616296947067867, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.343379259109497, + "learning_rate": 1e-06, + "loss": 1.1021, + "mean_token_accuracy": 0.689709484577179, + "num_tokens": 438272407.0, + "step": 16952 + }, + { + "epoch": 1.8617395124094003, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8853633403778076, + "learning_rate": 1e-06, + "loss": 0.9132, + "mean_token_accuracy": 0.7379822731018066, + "num_tokens": 438292060.0, + "step": 16953 + }, + { + "epoch": 1.861849330112014, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3250112533569336, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6948214769363403, + "num_tokens": 438323163.0, + "step": 16954 + }, + { + "epoch": 1.8619591478146278, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4772415161132812, + "learning_rate": 1e-06, + "loss": 1.1121, + "mean_token_accuracy": 0.6867282390594482, + "num_tokens": 438349393.0, + "step": 16955 + }, + { + "epoch": 1.8620689655172413, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4941322803497314, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7090544700622559, + "num_tokens": 438374889.0, + "step": 16956 + }, + { + "epoch": 1.862178783219855, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.363267421722412, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6939277052879333, + "num_tokens": 438404116.0, + "step": 16957 + }, + { + "epoch": 1.8622886009224686, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5519094467163086, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.703833818435669, + "num_tokens": 438428559.0, + "step": 16958 + }, + { + "epoch": 1.8623984186250824, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7458760738372803, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6926215887069702, + "num_tokens": 438452106.0, + "step": 16959 + }, + { + "epoch": 1.8625082363276961, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.587897539138794, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.7027679681777954, + "num_tokens": 438476888.0, + "step": 16960 + }, + { + "epoch": 1.8626180540303097, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.272707462310791, + "learning_rate": 1e-06, + "loss": 1.094, + "mean_token_accuracy": 0.6826304793357849, + "num_tokens": 438509799.0, + "step": 16961 + }, + { + "epoch": 1.8627278717329232, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.679405927658081, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7062884569168091, + "num_tokens": 438533819.0, + "step": 16962 + }, + { + "epoch": 1.862837689435537, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6219098567962646, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6912252902984619, + "num_tokens": 438559396.0, + "step": 16963 + }, + { + "epoch": 1.8629475071381507, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6015963554382324, + "learning_rate": 1e-06, + "loss": 0.8821, + "mean_token_accuracy": 0.7382776737213135, + "num_tokens": 438581493.0, + "step": 16964 + }, + { + "epoch": 1.8630573248407645, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.721674680709839, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7263966798782349, + "num_tokens": 438603969.0, + "step": 16965 + }, + { + "epoch": 1.863167142543378, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.5097908973693848, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7213889360427856, + "num_tokens": 438629259.0, + "step": 16966 + }, + { + "epoch": 1.8632769602459915, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4215989112854004, + "learning_rate": 1e-06, + "loss": 1.1369, + "mean_token_accuracy": 0.6729351282119751, + "num_tokens": 438659759.0, + "step": 16967 + }, + { + "epoch": 1.8633867779486053, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2968227863311768, + "learning_rate": 1e-06, + "loss": 1.1058, + "mean_token_accuracy": 0.6826001405715942, + "num_tokens": 438693993.0, + "step": 16968 + }, + { + "epoch": 1.863496595651219, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9201557636260986, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7185423970222473, + "num_tokens": 438717877.0, + "step": 16969 + }, + { + "epoch": 1.8636064133538326, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6245458126068115, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7068910002708435, + "num_tokens": 438741555.0, + "step": 16970 + }, + { + "epoch": 1.8637162310564463, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3883261680603027, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6946654319763184, + "num_tokens": 438772483.0, + "step": 16971 + }, + { + "epoch": 1.8638260487590599, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.67922043800354, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7219324707984924, + "num_tokens": 438796946.0, + "step": 16972 + }, + { + "epoch": 1.8639358664616736, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.236159563064575, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6949613094329834, + "num_tokens": 438828630.0, + "step": 16973 + }, + { + "epoch": 1.8640456841642874, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6532816886901855, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7098265886306763, + "num_tokens": 438854638.0, + "step": 16974 + }, + { + "epoch": 1.864155501866901, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6148455142974854, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7173446416854858, + "num_tokens": 438880277.0, + "step": 16975 + }, + { + "epoch": 1.8642653195695145, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.250194787979126, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6946957111358643, + "num_tokens": 438913785.0, + "step": 16976 + }, + { + "epoch": 1.8643751372721282, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8531930446624756, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7175400257110596, + "num_tokens": 438935297.0, + "step": 16977 + }, + { + "epoch": 1.864484954974742, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5878255367279053, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7082115411758423, + "num_tokens": 438960935.0, + "step": 16978 + }, + { + "epoch": 1.8645947726773557, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5401813983917236, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7116482257843018, + "num_tokens": 438985536.0, + "step": 16979 + }, + { + "epoch": 1.8647045903799693, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5280094146728516, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6820467710494995, + "num_tokens": 439014239.0, + "step": 16980 + }, + { + "epoch": 1.8648144080825828, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7443161010742188, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7073369026184082, + "num_tokens": 439039532.0, + "step": 16981 + }, + { + "epoch": 1.8649242257851966, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.33829665184021, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.705486536026001, + "num_tokens": 439070929.0, + "step": 16982 + }, + { + "epoch": 1.8650340434878103, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6576662063598633, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7093716859817505, + "num_tokens": 439093825.0, + "step": 16983 + }, + { + "epoch": 1.8651438611904239, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6375668048858643, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6998482346534729, + "num_tokens": 439117753.0, + "step": 16984 + }, + { + "epoch": 1.8652536788930374, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.431246757507324, + "learning_rate": 1e-06, + "loss": 0.8968, + "mean_token_accuracy": 0.7374452352523804, + "num_tokens": 439144930.0, + "step": 16985 + }, + { + "epoch": 1.8653634965956511, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.922614097595215, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.705375075340271, + "num_tokens": 439165820.0, + "step": 16986 + }, + { + "epoch": 1.865473314298265, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.767582416534424, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7243001461029053, + "num_tokens": 439188198.0, + "step": 16987 + }, + { + "epoch": 1.8655831320008787, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4273104667663574, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7202067375183105, + "num_tokens": 439215580.0, + "step": 16988 + }, + { + "epoch": 1.8656929497034922, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7066471576690674, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7127532958984375, + "num_tokens": 439238300.0, + "step": 16989 + }, + { + "epoch": 1.8658027674061057, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.588383674621582, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6952712535858154, + "num_tokens": 439263492.0, + "step": 16990 + }, + { + "epoch": 1.8659125851087195, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.539454460144043, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7016822099685669, + "num_tokens": 439290625.0, + "step": 16991 + }, + { + "epoch": 1.8660224028113332, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8450090885162354, + "learning_rate": 1e-06, + "loss": 1.1111, + "mean_token_accuracy": 0.6809325218200684, + "num_tokens": 439316988.0, + "step": 16992 + }, + { + "epoch": 1.866132220513947, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.004892110824585, + "learning_rate": 1e-06, + "loss": 0.9775, + "mean_token_accuracy": 0.7164576053619385, + "num_tokens": 439338441.0, + "step": 16993 + }, + { + "epoch": 1.8662420382165605, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.906160593032837, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7166658639907837, + "num_tokens": 439359117.0, + "step": 16994 + }, + { + "epoch": 1.866351855919174, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6670589447021484, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6995152831077576, + "num_tokens": 439383259.0, + "step": 16995 + }, + { + "epoch": 1.8664616736217878, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.596345901489258, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7020581364631653, + "num_tokens": 439407786.0, + "step": 16996 + }, + { + "epoch": 1.8665714913244016, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4621829986572266, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6971110105514526, + "num_tokens": 439437130.0, + "step": 16997 + }, + { + "epoch": 1.8666813090270151, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.507143497467041, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7245830297470093, + "num_tokens": 439465442.0, + "step": 16998 + }, + { + "epoch": 1.8667911267296287, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4542462825775146, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7002105712890625, + "num_tokens": 439494648.0, + "step": 16999 + }, + { + "epoch": 1.8669009444322424, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.751077175140381, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7103683352470398, + "num_tokens": 439520459.0, + "step": 17000 + }, + { + "epoch": 1.8670107621348562, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.677070140838623, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.69197678565979, + "num_tokens": 439546501.0, + "step": 17001 + }, + { + "epoch": 1.86712057983747, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4355721473693848, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7098485231399536, + "num_tokens": 439576103.0, + "step": 17002 + }, + { + "epoch": 1.8672303975400835, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1441874504089355, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.7018674612045288, + "num_tokens": 439597541.0, + "step": 17003 + }, + { + "epoch": 1.867340215242697, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.458369016647339, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7141450047492981, + "num_tokens": 439627409.0, + "step": 17004 + }, + { + "epoch": 1.8674500329453108, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8504600524902344, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.6984797716140747, + "num_tokens": 439649990.0, + "step": 17005 + }, + { + "epoch": 1.8675598506479245, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.574556589126587, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7136210799217224, + "num_tokens": 439676100.0, + "step": 17006 + }, + { + "epoch": 1.8676696683505383, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.099095106124878, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6929147243499756, + "num_tokens": 439701954.0, + "step": 17007 + }, + { + "epoch": 1.8677794860531518, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5798661708831787, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7055190205574036, + "num_tokens": 439729123.0, + "step": 17008 + }, + { + "epoch": 1.8678893037557653, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.499551773071289, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.6871621608734131, + "num_tokens": 439755615.0, + "step": 17009 + }, + { + "epoch": 1.867999121458379, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3947031497955322, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6950489282608032, + "num_tokens": 439786232.0, + "step": 17010 + }, + { + "epoch": 1.8681089391609929, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5682694911956787, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.71159827709198, + "num_tokens": 439810890.0, + "step": 17011 + }, + { + "epoch": 1.8682187568636064, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.718573808670044, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7026548385620117, + "num_tokens": 439833482.0, + "step": 17012 + }, + { + "epoch": 1.86832857456622, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.391521692276001, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7135220170021057, + "num_tokens": 439859092.0, + "step": 17013 + }, + { + "epoch": 1.8684383922688337, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4769694805145264, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7067310810089111, + "num_tokens": 439886906.0, + "step": 17014 + }, + { + "epoch": 1.8685482099714474, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.497556447982788, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.693628191947937, + "num_tokens": 439912722.0, + "step": 17015 + }, + { + "epoch": 1.8686580276740612, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.611362934112549, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.719517707824707, + "num_tokens": 439936368.0, + "step": 17016 + }, + { + "epoch": 1.8687678453766747, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.808948278427124, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7077057361602783, + "num_tokens": 439961316.0, + "step": 17017 + }, + { + "epoch": 1.8688776630792883, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.483438491821289, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6854532957077026, + "num_tokens": 439989532.0, + "step": 17018 + }, + { + "epoch": 1.868987480781902, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.504138708114624, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.703417181968689, + "num_tokens": 440016373.0, + "step": 17019 + }, + { + "epoch": 1.8690972984845158, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4849843978881836, + "learning_rate": 1e-06, + "loss": 0.923, + "mean_token_accuracy": 0.7244691848754883, + "num_tokens": 440041652.0, + "step": 17020 + }, + { + "epoch": 1.8692071161871293, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 8.010173797607422, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7008824348449707, + "num_tokens": 440065463.0, + "step": 17021 + }, + { + "epoch": 1.869316933889743, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.627307653427124, + "learning_rate": 1e-06, + "loss": 1.0904, + "mean_token_accuracy": 0.6907472610473633, + "num_tokens": 440092608.0, + "step": 17022 + }, + { + "epoch": 1.8694267515923566, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.352428913116455, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7036267518997192, + "num_tokens": 440122046.0, + "step": 17023 + }, + { + "epoch": 1.8695365692949704, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.328476905822754, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7145854830741882, + "num_tokens": 440152237.0, + "step": 17024 + }, + { + "epoch": 1.8696463869975841, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4495677947998047, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6917203664779663, + "num_tokens": 440179011.0, + "step": 17025 + }, + { + "epoch": 1.8697562047001977, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5386364459991455, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7170112133026123, + "num_tokens": 440202428.0, + "step": 17026 + }, + { + "epoch": 1.8698660224028112, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.721581220626831, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6883930563926697, + "num_tokens": 440228411.0, + "step": 17027 + }, + { + "epoch": 1.869975840105425, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6440999507904053, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.7123634219169617, + "num_tokens": 440253868.0, + "step": 17028 + }, + { + "epoch": 1.8700856578080387, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5952491760253906, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7143563032150269, + "num_tokens": 440276775.0, + "step": 17029 + }, + { + "epoch": 1.8701954755106525, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.739353656768799, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6989251375198364, + "num_tokens": 440304245.0, + "step": 17030 + }, + { + "epoch": 1.870305293213266, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5644147396087646, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7124967575073242, + "num_tokens": 440328836.0, + "step": 17031 + }, + { + "epoch": 1.8704151109158795, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7613070011138916, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7115985155105591, + "num_tokens": 440350971.0, + "step": 17032 + }, + { + "epoch": 1.8705249286184933, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6791398525238037, + "learning_rate": 1e-06, + "loss": 1.1586, + "mean_token_accuracy": 0.662889301776886, + "num_tokens": 440377766.0, + "step": 17033 + }, + { + "epoch": 1.870634746321107, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.449101209640503, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7246767282485962, + "num_tokens": 440407060.0, + "step": 17034 + }, + { + "epoch": 1.8707445640237206, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5195794105529785, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6903601884841919, + "num_tokens": 440433321.0, + "step": 17035 + }, + { + "epoch": 1.8708543817263343, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3767106533050537, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7097971439361572, + "num_tokens": 440461191.0, + "step": 17036 + }, + { + "epoch": 1.8709641994289479, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.645573377609253, + "learning_rate": 1e-06, + "loss": 1.1129, + "mean_token_accuracy": 0.6829168200492859, + "num_tokens": 440485817.0, + "step": 17037 + }, + { + "epoch": 1.8710740171315616, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3434929847717285, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7031072378158569, + "num_tokens": 440512880.0, + "step": 17038 + }, + { + "epoch": 1.8711838348341754, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 7.861570835113525, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.714276134967804, + "num_tokens": 440543790.0, + "step": 17039 + }, + { + "epoch": 1.871293652536789, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6995890140533447, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6973260641098022, + "num_tokens": 440568853.0, + "step": 17040 + }, + { + "epoch": 1.8714034702394025, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.402690887451172, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6889854669570923, + "num_tokens": 440599282.0, + "step": 17041 + }, + { + "epoch": 1.8715132879420162, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4643189907073975, + "learning_rate": 1e-06, + "loss": 0.8814, + "mean_token_accuracy": 0.7350984811782837, + "num_tokens": 440629140.0, + "step": 17042 + }, + { + "epoch": 1.87162310564463, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.930196762084961, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7086915373802185, + "num_tokens": 440653455.0, + "step": 17043 + }, + { + "epoch": 1.8717329233472437, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3711915016174316, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.710594117641449, + "num_tokens": 440680585.0, + "step": 17044 + }, + { + "epoch": 1.8718427410498573, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.757747173309326, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7145462036132812, + "num_tokens": 440703091.0, + "step": 17045 + }, + { + "epoch": 1.8719525587524708, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4637045860290527, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7099465131759644, + "num_tokens": 440729516.0, + "step": 17046 + }, + { + "epoch": 1.8720623764550846, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3512625694274902, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7136126756668091, + "num_tokens": 440764333.0, + "step": 17047 + }, + { + "epoch": 1.8721721941576983, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5898141860961914, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6915207505226135, + "num_tokens": 440794112.0, + "step": 17048 + }, + { + "epoch": 1.8722820118603118, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7349283695220947, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.724091649055481, + "num_tokens": 440816068.0, + "step": 17049 + }, + { + "epoch": 1.8723918295629254, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6667304039001465, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7184165120124817, + "num_tokens": 440839606.0, + "step": 17050 + }, + { + "epoch": 1.8725016472655391, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5708744525909424, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6924922466278076, + "num_tokens": 440867155.0, + "step": 17051 + }, + { + "epoch": 1.872611464968153, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.422617197036743, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.7047548294067383, + "num_tokens": 440892677.0, + "step": 17052 + }, + { + "epoch": 1.8727212826707667, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.654344081878662, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6958951354026794, + "num_tokens": 440920143.0, + "step": 17053 + }, + { + "epoch": 1.8728311003733802, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.846935272216797, + "learning_rate": 1e-06, + "loss": 0.8285, + "mean_token_accuracy": 0.7530354261398315, + "num_tokens": 440940554.0, + "step": 17054 + }, + { + "epoch": 1.8729409180759937, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.340886354446411, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7192200422286987, + "num_tokens": 440971347.0, + "step": 17055 + }, + { + "epoch": 1.8730507357786075, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.59456467628479, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6996610760688782, + "num_tokens": 440997428.0, + "step": 17056 + }, + { + "epoch": 1.8731605534812212, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7721521854400635, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7151174545288086, + "num_tokens": 441020541.0, + "step": 17057 + }, + { + "epoch": 1.873270371183835, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6128880977630615, + "learning_rate": 1e-06, + "loss": 0.8671, + "mean_token_accuracy": 0.7456352114677429, + "num_tokens": 441044714.0, + "step": 17058 + }, + { + "epoch": 1.8733801888864485, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.587655544281006, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7191363573074341, + "num_tokens": 441068532.0, + "step": 17059 + }, + { + "epoch": 1.873490006589062, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.741199016571045, + "learning_rate": 1e-06, + "loss": 0.9004, + "mean_token_accuracy": 0.7348915338516235, + "num_tokens": 441091202.0, + "step": 17060 + }, + { + "epoch": 1.8735998242916758, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0082623958587646, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.6996245980262756, + "num_tokens": 441112498.0, + "step": 17061 + }, + { + "epoch": 1.8737096419942896, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.496734380722046, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6972187161445618, + "num_tokens": 441142109.0, + "step": 17062 + }, + { + "epoch": 1.8738194596969031, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1448497772216797, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6891352534294128, + "num_tokens": 441164003.0, + "step": 17063 + }, + { + "epoch": 1.8739292773995166, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.015505313873291, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6800153851509094, + "num_tokens": 441189905.0, + "step": 17064 + }, + { + "epoch": 1.8740390951021304, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6908349990844727, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7087115049362183, + "num_tokens": 441215805.0, + "step": 17065 + }, + { + "epoch": 1.8741489128047442, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.947995662689209, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6963515877723694, + "num_tokens": 441246601.0, + "step": 17066 + }, + { + "epoch": 1.874258730507358, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5927090644836426, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.715509295463562, + "num_tokens": 441274797.0, + "step": 17067 + }, + { + "epoch": 1.8743685482099715, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.579153299331665, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6939878463745117, + "num_tokens": 441301370.0, + "step": 17068 + }, + { + "epoch": 1.874478365912585, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.852587938308716, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7107072472572327, + "num_tokens": 441323660.0, + "step": 17069 + }, + { + "epoch": 1.8745881836151987, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4826033115386963, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6965116262435913, + "num_tokens": 441352209.0, + "step": 17070 + }, + { + "epoch": 1.8746980013178125, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.481238842010498, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7057534456253052, + "num_tokens": 441383721.0, + "step": 17071 + }, + { + "epoch": 1.8748078190204263, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.630204677581787, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6967459321022034, + "num_tokens": 441409299.0, + "step": 17072 + }, + { + "epoch": 1.8749176367230398, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6049797534942627, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.704555869102478, + "num_tokens": 441434302.0, + "step": 17073 + }, + { + "epoch": 1.8750274544256533, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4500811100006104, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7086524963378906, + "num_tokens": 441461110.0, + "step": 17074 + }, + { + "epoch": 1.875137272128267, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2510433197021484, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7053614258766174, + "num_tokens": 441493487.0, + "step": 17075 + }, + { + "epoch": 1.8752470898308808, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.445050001144409, + "learning_rate": 1e-06, + "loss": 0.9182, + "mean_token_accuracy": 0.731086254119873, + "num_tokens": 441518779.0, + "step": 17076 + }, + { + "epoch": 1.8753569075334944, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.639146089553833, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.7291555404663086, + "num_tokens": 441542646.0, + "step": 17077 + }, + { + "epoch": 1.875466725236108, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2299904823303223, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7109044790267944, + "num_tokens": 441573331.0, + "step": 17078 + }, + { + "epoch": 1.8755765429387217, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.452751636505127, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7195686101913452, + "num_tokens": 441602259.0, + "step": 17079 + }, + { + "epoch": 1.8756863606413354, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.46128249168396, + "learning_rate": 1e-06, + "loss": 1.1243, + "mean_token_accuracy": 0.6826690435409546, + "num_tokens": 441634513.0, + "step": 17080 + }, + { + "epoch": 1.8757961783439492, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7727229595184326, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7144074440002441, + "num_tokens": 441657125.0, + "step": 17081 + }, + { + "epoch": 1.8759059960465627, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.571125030517578, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6982409954071045, + "num_tokens": 441683874.0, + "step": 17082 + }, + { + "epoch": 1.8760158137491763, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.588758945465088, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6842596530914307, + "num_tokens": 441710732.0, + "step": 17083 + }, + { + "epoch": 1.87612563145179, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.546835422515869, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7189419269561768, + "num_tokens": 441736131.0, + "step": 17084 + }, + { + "epoch": 1.8762354491544038, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.783477783203125, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7015269994735718, + "num_tokens": 441758858.0, + "step": 17085 + }, + { + "epoch": 1.8763452668570173, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8832972049713135, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7161670923233032, + "num_tokens": 441782241.0, + "step": 17086 + }, + { + "epoch": 1.876455084559631, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6302599906921387, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7166421413421631, + "num_tokens": 441805904.0, + "step": 17087 + }, + { + "epoch": 1.8765649022622446, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.942936420440674, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7162107825279236, + "num_tokens": 441826314.0, + "step": 17088 + }, + { + "epoch": 1.8766747199648584, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.717440128326416, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7051428556442261, + "num_tokens": 441852105.0, + "step": 17089 + }, + { + "epoch": 1.876784537667472, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6248068809509277, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6987099647521973, + "num_tokens": 441882518.0, + "step": 17090 + }, + { + "epoch": 1.8768943553700856, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.924314022064209, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6922416687011719, + "num_tokens": 441905581.0, + "step": 17091 + }, + { + "epoch": 1.8770041730726992, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1567635536193848, + "learning_rate": 1e-06, + "loss": 0.899, + "mean_token_accuracy": 0.7311403751373291, + "num_tokens": 441921785.0, + "step": 17092 + }, + { + "epoch": 1.877113990775313, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7281033992767334, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7218096256256104, + "num_tokens": 441944956.0, + "step": 17093 + }, + { + "epoch": 1.8772238084779267, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7615013122558594, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7112707495689392, + "num_tokens": 441967212.0, + "step": 17094 + }, + { + "epoch": 1.8773336261805404, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.414315700531006, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7008342742919922, + "num_tokens": 441997960.0, + "step": 17095 + }, + { + "epoch": 1.877443443883154, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.3892016410827637, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6980769038200378, + "num_tokens": 442022187.0, + "step": 17096 + }, + { + "epoch": 1.8775532615857675, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6556785106658936, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6865295171737671, + "num_tokens": 442050708.0, + "step": 17097 + }, + { + "epoch": 1.8776630792883813, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4994747638702393, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7304554581642151, + "num_tokens": 442075492.0, + "step": 17098 + }, + { + "epoch": 1.877772896990995, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.576962471008301, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6858745217323303, + "num_tokens": 442102548.0, + "step": 17099 + }, + { + "epoch": 1.8778827146936086, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6030211448669434, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6957137584686279, + "num_tokens": 442130805.0, + "step": 17100 + }, + { + "epoch": 1.8779925323962223, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.683124303817749, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6992254257202148, + "num_tokens": 442154279.0, + "step": 17101 + }, + { + "epoch": 1.8781023500988359, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.439289093017578, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7007913589477539, + "num_tokens": 442182025.0, + "step": 17102 + }, + { + "epoch": 1.8782121678014496, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5573136806488037, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7185582518577576, + "num_tokens": 442205503.0, + "step": 17103 + }, + { + "epoch": 1.8783219855040634, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.609159469604492, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7187330722808838, + "num_tokens": 442228249.0, + "step": 17104 + }, + { + "epoch": 1.878431803206677, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7388691902160645, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7020300626754761, + "num_tokens": 442249551.0, + "step": 17105 + }, + { + "epoch": 1.8785416209092904, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.808985471725464, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7189688682556152, + "num_tokens": 442271194.0, + "step": 17106 + }, + { + "epoch": 1.8786514386119042, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.416382074356079, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.709534227848053, + "num_tokens": 442300042.0, + "step": 17107 + }, + { + "epoch": 1.878761256314518, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5056748390197754, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6906290054321289, + "num_tokens": 442328394.0, + "step": 17108 + }, + { + "epoch": 1.8788710740171317, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2699737548828125, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6865326166152954, + "num_tokens": 442360783.0, + "step": 17109 + }, + { + "epoch": 1.8789808917197452, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6848526000976562, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7052656412124634, + "num_tokens": 442385450.0, + "step": 17110 + }, + { + "epoch": 1.8790907094223588, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5279791355133057, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6998056769371033, + "num_tokens": 442410895.0, + "step": 17111 + }, + { + "epoch": 1.8792005271249725, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.561418056488037, + "learning_rate": 1e-06, + "loss": 0.9323, + "mean_token_accuracy": 0.72540682554245, + "num_tokens": 442433978.0, + "step": 17112 + }, + { + "epoch": 1.8793103448275863, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.575025796890259, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.7026328444480896, + "num_tokens": 442459487.0, + "step": 17113 + }, + { + "epoch": 1.8794201625301998, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5852627754211426, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7034023404121399, + "num_tokens": 442488748.0, + "step": 17114 + }, + { + "epoch": 1.8795299802328134, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5097928047180176, + "learning_rate": 1e-06, + "loss": 0.9397, + "mean_token_accuracy": 0.7263602614402771, + "num_tokens": 442513277.0, + "step": 17115 + }, + { + "epoch": 1.8796397979354271, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6680500507354736, + "learning_rate": 1e-06, + "loss": 0.9016, + "mean_token_accuracy": 0.7318377494812012, + "num_tokens": 442535248.0, + "step": 17116 + }, + { + "epoch": 1.8797496156380409, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.343672037124634, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7044211626052856, + "num_tokens": 442565238.0, + "step": 17117 + }, + { + "epoch": 1.8798594333406546, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5329864025115967, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7178683876991272, + "num_tokens": 442592048.0, + "step": 17118 + }, + { + "epoch": 1.8799692510432682, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.636610984802246, + "learning_rate": 1e-06, + "loss": 1.0915, + "mean_token_accuracy": 0.6942733526229858, + "num_tokens": 442619496.0, + "step": 17119 + }, + { + "epoch": 1.8800790687458817, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5967838764190674, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7005167603492737, + "num_tokens": 442644476.0, + "step": 17120 + }, + { + "epoch": 1.8801888864484955, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7514703273773193, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.69227534532547, + "num_tokens": 442670967.0, + "step": 17121 + }, + { + "epoch": 1.8802987041511092, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.718956708908081, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7058089971542358, + "num_tokens": 442694055.0, + "step": 17122 + }, + { + "epoch": 1.880408521853723, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5110127925872803, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6842636466026306, + "num_tokens": 442719608.0, + "step": 17123 + }, + { + "epoch": 1.8805183395563365, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.43204927444458, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7039963006973267, + "num_tokens": 442745522.0, + "step": 17124 + }, + { + "epoch": 1.88062815725895, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.259958028793335, + "learning_rate": 1e-06, + "loss": 0.9167, + "mean_token_accuracy": 0.7305688261985779, + "num_tokens": 442774483.0, + "step": 17125 + }, + { + "epoch": 1.8807379749615638, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6632049083709717, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.707865834236145, + "num_tokens": 442799053.0, + "step": 17126 + }, + { + "epoch": 1.8808477926641776, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.085430383682251, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6880915760993958, + "num_tokens": 442830422.0, + "step": 17127 + }, + { + "epoch": 1.880957610366791, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4825448989868164, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6916947364807129, + "num_tokens": 442858114.0, + "step": 17128 + }, + { + "epoch": 1.8810674280694046, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 6.586851596832275, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6839932203292847, + "num_tokens": 442885656.0, + "step": 17129 + }, + { + "epoch": 1.8811772457720184, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4966888427734375, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.7229257822036743, + "num_tokens": 442915149.0, + "step": 17130 + }, + { + "epoch": 1.8812870634746321, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6843583583831787, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7057094573974609, + "num_tokens": 442940307.0, + "step": 17131 + }, + { + "epoch": 1.881396881177246, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2710890769958496, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7003498077392578, + "num_tokens": 442971327.0, + "step": 17132 + }, + { + "epoch": 1.8815066988798594, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.693577527999878, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7086492776870728, + "num_tokens": 442993456.0, + "step": 17133 + }, + { + "epoch": 1.881616516582473, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.478827476501465, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6963168978691101, + "num_tokens": 443022173.0, + "step": 17134 + }, + { + "epoch": 1.8817263342850867, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4810380935668945, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7192438840866089, + "num_tokens": 443048031.0, + "step": 17135 + }, + { + "epoch": 1.8818361519877005, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.811220407485962, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7061750888824463, + "num_tokens": 443068998.0, + "step": 17136 + }, + { + "epoch": 1.881945969690314, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7307074069976807, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6948009729385376, + "num_tokens": 443090374.0, + "step": 17137 + }, + { + "epoch": 1.8820557873929278, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4842543601989746, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7168364524841309, + "num_tokens": 443116477.0, + "step": 17138 + }, + { + "epoch": 1.8821656050955413, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5505051612854004, + "learning_rate": 1e-06, + "loss": 0.9139, + "mean_token_accuracy": 0.7293094396591187, + "num_tokens": 443142580.0, + "step": 17139 + }, + { + "epoch": 1.882275422798155, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.793494462966919, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7042132616043091, + "num_tokens": 443165982.0, + "step": 17140 + }, + { + "epoch": 1.8823852405007688, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3172266483306885, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7140077352523804, + "num_tokens": 443197490.0, + "step": 17141 + }, + { + "epoch": 1.8824950582033824, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8442609310150146, + "learning_rate": 1e-06, + "loss": 0.8935, + "mean_token_accuracy": 0.7357997894287109, + "num_tokens": 443216578.0, + "step": 17142 + }, + { + "epoch": 1.882604875905996, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.303325653076172, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7195178270339966, + "num_tokens": 443246303.0, + "step": 17143 + }, + { + "epoch": 1.8827146936086097, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6383581161499023, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.702467679977417, + "num_tokens": 443272511.0, + "step": 17144 + }, + { + "epoch": 1.8828245113112234, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5547537803649902, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6995570063591003, + "num_tokens": 443297898.0, + "step": 17145 + }, + { + "epoch": 1.8829343290138372, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.835963487625122, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7120676040649414, + "num_tokens": 443320756.0, + "step": 17146 + }, + { + "epoch": 1.8830441467164507, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6774325370788574, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7006999254226685, + "num_tokens": 443344590.0, + "step": 17147 + }, + { + "epoch": 1.8831539644190642, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1955604553222656, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.6935505867004395, + "num_tokens": 443372794.0, + "step": 17148 + }, + { + "epoch": 1.883263782121678, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.465179920196533, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6903244256973267, + "num_tokens": 443401270.0, + "step": 17149 + }, + { + "epoch": 1.8833735998242918, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4207539558410645, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.7003725171089172, + "num_tokens": 443432763.0, + "step": 17150 + }, + { + "epoch": 1.8834834175269053, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.409210205078125, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.691317081451416, + "num_tokens": 443462146.0, + "step": 17151 + }, + { + "epoch": 1.883593235229519, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4078667163848877, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7248210906982422, + "num_tokens": 443489265.0, + "step": 17152 + }, + { + "epoch": 1.8837030529321326, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3344838619232178, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6987552642822266, + "num_tokens": 443521043.0, + "step": 17153 + }, + { + "epoch": 1.8838128706347463, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.414027690887451, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.706829309463501, + "num_tokens": 443548868.0, + "step": 17154 + }, + { + "epoch": 1.88392268833736, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5104763507843018, + "learning_rate": 1e-06, + "loss": 0.8532, + "mean_token_accuracy": 0.7464011907577515, + "num_tokens": 443572380.0, + "step": 17155 + }, + { + "epoch": 1.8840325060399736, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5929510593414307, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7094607353210449, + "num_tokens": 443595791.0, + "step": 17156 + }, + { + "epoch": 1.8841423237425872, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4032492637634277, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.6960760354995728, + "num_tokens": 443623415.0, + "step": 17157 + }, + { + "epoch": 1.884252141445201, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.415316343307495, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7009735107421875, + "num_tokens": 443653345.0, + "step": 17158 + }, + { + "epoch": 1.8843619591478147, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.395193576812744, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7189887762069702, + "num_tokens": 443681577.0, + "step": 17159 + }, + { + "epoch": 1.8844717768504284, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5431392192840576, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7089153528213501, + "num_tokens": 443707466.0, + "step": 17160 + }, + { + "epoch": 1.884581594553042, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.790398597717285, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7043611407279968, + "num_tokens": 443730996.0, + "step": 17161 + }, + { + "epoch": 1.8846914122556555, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6871752738952637, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6957327723503113, + "num_tokens": 443754314.0, + "step": 17162 + }, + { + "epoch": 1.8848012299582693, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7897818088531494, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7156293988227844, + "num_tokens": 443775221.0, + "step": 17163 + }, + { + "epoch": 1.884911047660883, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.650803327560425, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6826868057250977, + "num_tokens": 443799455.0, + "step": 17164 + }, + { + "epoch": 1.8850208653634966, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.677917718887329, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6913260817527771, + "num_tokens": 443823852.0, + "step": 17165 + }, + { + "epoch": 1.88513068306611, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6328375339508057, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.6837360262870789, + "num_tokens": 443850175.0, + "step": 17166 + }, + { + "epoch": 1.8852405007687238, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6066317558288574, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.7199878692626953, + "num_tokens": 443874583.0, + "step": 17167 + }, + { + "epoch": 1.8853503184713376, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.800070285797119, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7206318378448486, + "num_tokens": 443894730.0, + "step": 17168 + }, + { + "epoch": 1.8854601361739514, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8715553283691406, + "learning_rate": 1e-06, + "loss": 0.9014, + "mean_token_accuracy": 0.7319910526275635, + "num_tokens": 443914815.0, + "step": 17169 + }, + { + "epoch": 1.885569953876565, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.335313558578491, + "learning_rate": 1e-06, + "loss": 1.1094, + "mean_token_accuracy": 0.682709813117981, + "num_tokens": 443946817.0, + "step": 17170 + }, + { + "epoch": 1.8856797715791784, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.556478977203369, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.6981512308120728, + "num_tokens": 443972576.0, + "step": 17171 + }, + { + "epoch": 1.8857895892817922, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5696310997009277, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7145463228225708, + "num_tokens": 443997637.0, + "step": 17172 + }, + { + "epoch": 1.885899406984406, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0633764266967773, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7108356952667236, + "num_tokens": 444016602.0, + "step": 17173 + }, + { + "epoch": 1.8860092246870197, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5630404949188232, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.7130469083786011, + "num_tokens": 444042596.0, + "step": 17174 + }, + { + "epoch": 1.8861190423896332, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2666330337524414, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.725499153137207, + "num_tokens": 444071027.0, + "step": 17175 + }, + { + "epoch": 1.8862288600922468, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4183785915374756, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6939612627029419, + "num_tokens": 444097511.0, + "step": 17176 + }, + { + "epoch": 1.8863386777948605, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.479170799255371, + "learning_rate": 1e-06, + "loss": 1.0969, + "mean_token_accuracy": 0.6975617408752441, + "num_tokens": 444123105.0, + "step": 17177 + }, + { + "epoch": 1.8864484954974743, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.361553192138672, + "learning_rate": 1e-06, + "loss": 1.1077, + "mean_token_accuracy": 0.6860621571540833, + "num_tokens": 444152880.0, + "step": 17178 + }, + { + "epoch": 1.8865583132000878, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6604981422424316, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6942302584648132, + "num_tokens": 444177159.0, + "step": 17179 + }, + { + "epoch": 1.8866681309027014, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.407017230987549, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7163997292518616, + "num_tokens": 444204765.0, + "step": 17180 + }, + { + "epoch": 1.8867779486053151, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 8.73437786102295, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6951109170913696, + "num_tokens": 444240019.0, + "step": 17181 + }, + { + "epoch": 1.8868877663079289, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.540583848953247, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7128437757492065, + "num_tokens": 444265462.0, + "step": 17182 + }, + { + "epoch": 1.8869975840105426, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.544034957885742, + "learning_rate": 1e-06, + "loss": 0.9315, + "mean_token_accuracy": 0.7325326204299927, + "num_tokens": 444290496.0, + "step": 17183 + }, + { + "epoch": 1.8871074017131562, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4909918308258057, + "learning_rate": 1e-06, + "loss": 0.9462, + "mean_token_accuracy": 0.7203992605209351, + "num_tokens": 444316974.0, + "step": 17184 + }, + { + "epoch": 1.8872172194157697, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3507637977600098, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7048808336257935, + "num_tokens": 444347182.0, + "step": 17185 + }, + { + "epoch": 1.8873270371183835, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.509922504425049, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6958262920379639, + "num_tokens": 444373363.0, + "step": 17186 + }, + { + "epoch": 1.8874368548209972, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.964108943939209, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6957036852836609, + "num_tokens": 444395418.0, + "step": 17187 + }, + { + "epoch": 1.887546672523611, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5030577182769775, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.729931116104126, + "num_tokens": 444423375.0, + "step": 17188 + }, + { + "epoch": 1.8876564902262245, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7687106132507324, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7130390405654907, + "num_tokens": 444444920.0, + "step": 17189 + }, + { + "epoch": 1.887766307928838, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4355998039245605, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.696134090423584, + "num_tokens": 444474529.0, + "step": 17190 + }, + { + "epoch": 1.8878761256314518, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.823118209838867, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6944737434387207, + "num_tokens": 444500093.0, + "step": 17191 + }, + { + "epoch": 1.8879859433340656, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5815296173095703, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6929696798324585, + "num_tokens": 444527583.0, + "step": 17192 + }, + { + "epoch": 1.888095761036679, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4730312824249268, + "learning_rate": 1e-06, + "loss": 0.9297, + "mean_token_accuracy": 0.7332319617271423, + "num_tokens": 444554468.0, + "step": 17193 + }, + { + "epoch": 1.8882055787392926, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.731531858444214, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6957454681396484, + "num_tokens": 444576911.0, + "step": 17194 + }, + { + "epoch": 1.8883153964419064, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.388791799545288, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6940968036651611, + "num_tokens": 444606474.0, + "step": 17195 + }, + { + "epoch": 1.8884252141445201, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4707934856414795, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6826759576797485, + "num_tokens": 444634777.0, + "step": 17196 + }, + { + "epoch": 1.888535031847134, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5335259437561035, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.703854501247406, + "num_tokens": 444659687.0, + "step": 17197 + }, + { + "epoch": 1.8886448495497474, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6813342571258545, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7210254073143005, + "num_tokens": 444684014.0, + "step": 17198 + }, + { + "epoch": 1.888754667252361, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8367903232574463, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6906149387359619, + "num_tokens": 444706289.0, + "step": 17199 + }, + { + "epoch": 1.8888644849549747, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7920095920562744, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.707014799118042, + "num_tokens": 444727952.0, + "step": 17200 + }, + { + "epoch": 1.8889743026575885, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1511762142181396, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.707330584526062, + "num_tokens": 444757089.0, + "step": 17201 + }, + { + "epoch": 1.889084120360202, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3138177394866943, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6881025433540344, + "num_tokens": 444787948.0, + "step": 17202 + }, + { + "epoch": 1.8891939380628158, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.223945379257202, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6783406138420105, + "num_tokens": 444821121.0, + "step": 17203 + }, + { + "epoch": 1.8893037557654293, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9014148712158203, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7050309181213379, + "num_tokens": 444843954.0, + "step": 17204 + }, + { + "epoch": 1.889413573468043, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.779339551925659, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6915069818496704, + "num_tokens": 444865200.0, + "step": 17205 + }, + { + "epoch": 1.8895233911706568, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.644132375717163, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7094326615333557, + "num_tokens": 444888083.0, + "step": 17206 + }, + { + "epoch": 1.8896332088732704, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.598914384841919, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7257062196731567, + "num_tokens": 444912017.0, + "step": 17207 + }, + { + "epoch": 1.8897430265758839, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7788257598876953, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7211564183235168, + "num_tokens": 444934303.0, + "step": 17208 + }, + { + "epoch": 1.8898528442784976, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.785919427871704, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7112407684326172, + "num_tokens": 444956792.0, + "step": 17209 + }, + { + "epoch": 1.8899626619811114, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8226852416992188, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7106220722198486, + "num_tokens": 444978512.0, + "step": 17210 + }, + { + "epoch": 1.8900724796837252, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.440034866333008, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7034071683883667, + "num_tokens": 445007847.0, + "step": 17211 + }, + { + "epoch": 1.8901822973863387, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4598817825317383, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7015964984893799, + "num_tokens": 445034565.0, + "step": 17212 + }, + { + "epoch": 1.8902921150889522, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2858052253723145, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6920242309570312, + "num_tokens": 445066671.0, + "step": 17213 + }, + { + "epoch": 1.890401932791566, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.493149757385254, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6932108998298645, + "num_tokens": 445100563.0, + "step": 17214 + }, + { + "epoch": 1.8905117504941797, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.35072922706604, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7157341837882996, + "num_tokens": 445129719.0, + "step": 17215 + }, + { + "epoch": 1.8906215681967933, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.24774169921875, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6961663961410522, + "num_tokens": 445152273.0, + "step": 17216 + }, + { + "epoch": 1.890731385899407, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5491113662719727, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7162644863128662, + "num_tokens": 445176937.0, + "step": 17217 + }, + { + "epoch": 1.8908412036020206, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.408480644226074, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6854666471481323, + "num_tokens": 445207189.0, + "step": 17218 + }, + { + "epoch": 1.8909510213046343, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6256706714630127, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7056546211242676, + "num_tokens": 445229436.0, + "step": 17219 + }, + { + "epoch": 1.891060839007248, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6564719676971436, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7074494361877441, + "num_tokens": 445257816.0, + "step": 17220 + }, + { + "epoch": 1.8911706567098616, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.584139108657837, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6996034979820251, + "num_tokens": 445282658.0, + "step": 17221 + }, + { + "epoch": 1.8912804744124752, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4396371841430664, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6962885856628418, + "num_tokens": 445309648.0, + "step": 17222 + }, + { + "epoch": 1.891390292115089, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5763208866119385, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7063935995101929, + "num_tokens": 445336103.0, + "step": 17223 + }, + { + "epoch": 1.8915001098177027, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5303757190704346, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7045942544937134, + "num_tokens": 445366873.0, + "step": 17224 + }, + { + "epoch": 1.8916099275203164, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6147661209106445, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.6872866749763489, + "num_tokens": 445394047.0, + "step": 17225 + }, + { + "epoch": 1.89171974522293, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.808140993118286, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.699301540851593, + "num_tokens": 445418118.0, + "step": 17226 + }, + { + "epoch": 1.8918295629255435, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6271941661834717, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7040833234786987, + "num_tokens": 445446647.0, + "step": 17227 + }, + { + "epoch": 1.8919393806281573, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.85566782951355, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7124435305595398, + "num_tokens": 445470464.0, + "step": 17228 + }, + { + "epoch": 1.892049198330771, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8188493251800537, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7243900299072266, + "num_tokens": 445492287.0, + "step": 17229 + }, + { + "epoch": 1.8921590160333845, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7406911849975586, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.698766827583313, + "num_tokens": 445516898.0, + "step": 17230 + }, + { + "epoch": 1.892268833735998, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6332972049713135, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6985248327255249, + "num_tokens": 445541226.0, + "step": 17231 + }, + { + "epoch": 1.8923786514386118, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7116763591766357, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7143571972846985, + "num_tokens": 445564130.0, + "step": 17232 + }, + { + "epoch": 1.8924884691412256, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.662095069885254, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.716090977191925, + "num_tokens": 445589156.0, + "step": 17233 + }, + { + "epoch": 1.8925982868438394, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6447033882141113, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.704062819480896, + "num_tokens": 445613269.0, + "step": 17234 + }, + { + "epoch": 1.8927081045464529, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8169000148773193, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6928406953811646, + "num_tokens": 445642926.0, + "step": 17235 + }, + { + "epoch": 1.8928179222490664, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.636197328567505, + "learning_rate": 1e-06, + "loss": 0.8868, + "mean_token_accuracy": 0.7349435091018677, + "num_tokens": 445663012.0, + "step": 17236 + }, + { + "epoch": 1.8929277399516802, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4627835750579834, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.709975004196167, + "num_tokens": 445690161.0, + "step": 17237 + }, + { + "epoch": 1.893037557654294, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.554136276245117, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.6938557624816895, + "num_tokens": 445716987.0, + "step": 17238 + }, + { + "epoch": 1.8931473753569077, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.570502996444702, + "learning_rate": 1e-06, + "loss": 0.9079, + "mean_token_accuracy": 0.7299116253852844, + "num_tokens": 445741219.0, + "step": 17239 + }, + { + "epoch": 1.8932571930595212, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.677727222442627, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.6922317743301392, + "num_tokens": 445766150.0, + "step": 17240 + }, + { + "epoch": 1.8933670107621348, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4665441513061523, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6954926252365112, + "num_tokens": 445793915.0, + "step": 17241 + }, + { + "epoch": 1.8934768284647485, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4844744205474854, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6952344179153442, + "num_tokens": 445819999.0, + "step": 17242 + }, + { + "epoch": 1.8935866461673623, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.450869560241699, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.7004890441894531, + "num_tokens": 445849330.0, + "step": 17243 + }, + { + "epoch": 1.8936964638699758, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.521204948425293, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6908761262893677, + "num_tokens": 445876150.0, + "step": 17244 + }, + { + "epoch": 1.8938062815725893, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.845651626586914, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.7046884894371033, + "num_tokens": 445906024.0, + "step": 17245 + }, + { + "epoch": 1.893916099275203, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4555914402008057, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6889873147010803, + "num_tokens": 445934276.0, + "step": 17246 + }, + { + "epoch": 1.8940259169778169, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6471195220947266, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6920058727264404, + "num_tokens": 445961624.0, + "step": 17247 + }, + { + "epoch": 1.8941357346804306, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.398275136947632, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7102465629577637, + "num_tokens": 445991363.0, + "step": 17248 + }, + { + "epoch": 1.8942455523830442, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.694563627243042, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6970298290252686, + "num_tokens": 446014534.0, + "step": 17249 + }, + { + "epoch": 1.8943553700856577, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8198304176330566, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7212173938751221, + "num_tokens": 446036606.0, + "step": 17250 + }, + { + "epoch": 1.8944651877882714, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8013088703155518, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7093702554702759, + "num_tokens": 446056241.0, + "step": 17251 + }, + { + "epoch": 1.8945750054908852, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.066951274871826, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7106342315673828, + "num_tokens": 446073179.0, + "step": 17252 + }, + { + "epoch": 1.894684823193499, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6770410537719727, + "learning_rate": 1e-06, + "loss": 1.093, + "mean_token_accuracy": 0.6909177303314209, + "num_tokens": 446102627.0, + "step": 17253 + }, + { + "epoch": 1.8947946408961125, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.59721040725708, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7101786136627197, + "num_tokens": 446129432.0, + "step": 17254 + }, + { + "epoch": 1.894904458598726, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7357940673828125, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.7183800935745239, + "num_tokens": 446151081.0, + "step": 17255 + }, + { + "epoch": 1.8950142763013398, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.722599744796753, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6976790428161621, + "num_tokens": 446176354.0, + "step": 17256 + }, + { + "epoch": 1.8951240940039535, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.597125291824341, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7055634260177612, + "num_tokens": 446202852.0, + "step": 17257 + }, + { + "epoch": 1.895233911706567, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.010136365890503, + "learning_rate": 1e-06, + "loss": 1.1093, + "mean_token_accuracy": 0.6781076192855835, + "num_tokens": 446229433.0, + "step": 17258 + }, + { + "epoch": 1.8953437294091806, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7287375926971436, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7203145623207092, + "num_tokens": 446250044.0, + "step": 17259 + }, + { + "epoch": 1.8954535471117944, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7728230953216553, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6948485374450684, + "num_tokens": 446274982.0, + "step": 17260 + }, + { + "epoch": 1.8955633648144081, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.935831308364868, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7142208814620972, + "num_tokens": 446296475.0, + "step": 17261 + }, + { + "epoch": 1.8956731825170219, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.370439291000366, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6972087025642395, + "num_tokens": 446329801.0, + "step": 17262 + }, + { + "epoch": 1.8957830002196354, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5943095684051514, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.6897393465042114, + "num_tokens": 446354878.0, + "step": 17263 + }, + { + "epoch": 1.895892817922249, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3306310176849365, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7058089971542358, + "num_tokens": 446388505.0, + "step": 17264 + }, + { + "epoch": 1.8960026356248627, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.760300636291504, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6951909065246582, + "num_tokens": 446409303.0, + "step": 17265 + }, + { + "epoch": 1.8961124533274765, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4032905101776123, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.6999808549880981, + "num_tokens": 446437388.0, + "step": 17266 + }, + { + "epoch": 1.89622227103009, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.429872989654541, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6942118406295776, + "num_tokens": 446467633.0, + "step": 17267 + }, + { + "epoch": 1.8963320887327038, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.572981119155884, + "learning_rate": 1e-06, + "loss": 1.0523, + "mean_token_accuracy": 0.6988776326179504, + "num_tokens": 446495235.0, + "step": 17268 + }, + { + "epoch": 1.8964419064353173, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4269378185272217, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6943854689598083, + "num_tokens": 446526089.0, + "step": 17269 + }, + { + "epoch": 1.896551724137931, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8130295276641846, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7106900215148926, + "num_tokens": 446546020.0, + "step": 17270 + }, + { + "epoch": 1.8966615418405448, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.477140426635742, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7111574411392212, + "num_tokens": 446573748.0, + "step": 17271 + }, + { + "epoch": 1.8967713595431583, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7522456645965576, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7191193103790283, + "num_tokens": 446600106.0, + "step": 17272 + }, + { + "epoch": 1.8968811772457719, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0418410301208496, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7166827917098999, + "num_tokens": 446621102.0, + "step": 17273 + }, + { + "epoch": 1.8969909949483856, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5728847980499268, + "learning_rate": 1e-06, + "loss": 1.1261, + "mean_token_accuracy": 0.6734662055969238, + "num_tokens": 446647799.0, + "step": 17274 + }, + { + "epoch": 1.8971008126509994, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5849785804748535, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7030967473983765, + "num_tokens": 446673387.0, + "step": 17275 + }, + { + "epoch": 1.8972106303536131, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.622676134109497, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7035354375839233, + "num_tokens": 446698085.0, + "step": 17276 + }, + { + "epoch": 1.8973204480562267, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.648322343826294, + "learning_rate": 1e-06, + "loss": 1.0964, + "mean_token_accuracy": 0.677061915397644, + "num_tokens": 446724016.0, + "step": 17277 + }, + { + "epoch": 1.8974302657588402, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6024861335754395, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7178809642791748, + "num_tokens": 446749350.0, + "step": 17278 + }, + { + "epoch": 1.897540083461454, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.469882011413574, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.713902473449707, + "num_tokens": 446779779.0, + "step": 17279 + }, + { + "epoch": 1.8976499011640677, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5230588912963867, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7187836766242981, + "num_tokens": 446805939.0, + "step": 17280 + }, + { + "epoch": 1.8977597188666813, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4223897457122803, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7025677561759949, + "num_tokens": 446837087.0, + "step": 17281 + }, + { + "epoch": 1.897869536569295, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.661879777908325, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7092487812042236, + "num_tokens": 446862282.0, + "step": 17282 + }, + { + "epoch": 1.8979793542719086, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.354640007019043, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7190977931022644, + "num_tokens": 446892472.0, + "step": 17283 + }, + { + "epoch": 1.8980891719745223, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4459943771362305, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7230824828147888, + "num_tokens": 446917216.0, + "step": 17284 + }, + { + "epoch": 1.898198989677136, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4822561740875244, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7086723446846008, + "num_tokens": 446942707.0, + "step": 17285 + }, + { + "epoch": 1.8983088073797496, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.489640951156616, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7040435671806335, + "num_tokens": 446970127.0, + "step": 17286 + }, + { + "epoch": 1.8984186250823631, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.986222267150879, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.6972920894622803, + "num_tokens": 446990487.0, + "step": 17287 + }, + { + "epoch": 1.898528442784977, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5693306922912598, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7158923149108887, + "num_tokens": 447016971.0, + "step": 17288 + }, + { + "epoch": 1.8986382604875907, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3282666206359863, + "learning_rate": 1e-06, + "loss": 1.144, + "mean_token_accuracy": 0.6761881113052368, + "num_tokens": 447049599.0, + "step": 17289 + }, + { + "epoch": 1.8987480781902044, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4532082080841064, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.705945611000061, + "num_tokens": 447077909.0, + "step": 17290 + }, + { + "epoch": 1.898857895892818, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5649616718292236, + "learning_rate": 1e-06, + "loss": 0.9176, + "mean_token_accuracy": 0.7316728830337524, + "num_tokens": 447102460.0, + "step": 17291 + }, + { + "epoch": 1.8989677135954315, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5849976539611816, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.709500789642334, + "num_tokens": 447127190.0, + "step": 17292 + }, + { + "epoch": 1.8990775312980452, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.775855779647827, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7222670912742615, + "num_tokens": 447152851.0, + "step": 17293 + }, + { + "epoch": 1.899187349000659, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.441620349884033, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.699359655380249, + "num_tokens": 447181968.0, + "step": 17294 + }, + { + "epoch": 1.8992971667032725, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.665623903274536, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7132018804550171, + "num_tokens": 447204277.0, + "step": 17295 + }, + { + "epoch": 1.899406984405886, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6936087608337402, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7199634909629822, + "num_tokens": 447226752.0, + "step": 17296 + }, + { + "epoch": 1.8995168021084998, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.938894748687744, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.732843816280365, + "num_tokens": 447248056.0, + "step": 17297 + }, + { + "epoch": 1.8996266198111136, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.666163206100464, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6978996396064758, + "num_tokens": 447273500.0, + "step": 17298 + }, + { + "epoch": 1.8997364375137273, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.407059669494629, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.705912709236145, + "num_tokens": 447302646.0, + "step": 17299 + }, + { + "epoch": 1.8998462552163409, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5169968605041504, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7060704231262207, + "num_tokens": 447328778.0, + "step": 17300 + }, + { + "epoch": 1.8999560729189544, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.617504119873047, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7082422971725464, + "num_tokens": 447354432.0, + "step": 17301 + }, + { + "epoch": 1.9000658906215682, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3891379833221436, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7162604331970215, + "num_tokens": 447381128.0, + "step": 17302 + }, + { + "epoch": 1.900175708324182, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7405433654785156, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7030078172683716, + "num_tokens": 447406027.0, + "step": 17303 + }, + { + "epoch": 1.9002855260267957, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.533717155456543, + "learning_rate": 1e-06, + "loss": 0.9035, + "mean_token_accuracy": 0.735715925693512, + "num_tokens": 447430713.0, + "step": 17304 + }, + { + "epoch": 1.9003953437294092, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.78678035736084, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7131858468055725, + "num_tokens": 447455069.0, + "step": 17305 + }, + { + "epoch": 1.9005051614320227, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4422481060028076, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.7000463604927063, + "num_tokens": 447483421.0, + "step": 17306 + }, + { + "epoch": 1.9006149791346365, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.371818780899048, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6961544156074524, + "num_tokens": 447511348.0, + "step": 17307 + }, + { + "epoch": 1.9007247968372503, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 7.952418327331543, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7196868062019348, + "num_tokens": 447533441.0, + "step": 17308 + }, + { + "epoch": 1.9008346145398638, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.608440399169922, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7082182168960571, + "num_tokens": 447558387.0, + "step": 17309 + }, + { + "epoch": 1.9009444322424773, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.742677927017212, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6912627816200256, + "num_tokens": 447592242.0, + "step": 17310 + }, + { + "epoch": 1.901054249945091, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.69049072265625, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6879490613937378, + "num_tokens": 447615537.0, + "step": 17311 + }, + { + "epoch": 1.9011640676477048, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3005430698394775, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.712602972984314, + "num_tokens": 447645715.0, + "step": 17312 + }, + { + "epoch": 1.9012738853503186, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4156434535980225, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7176952958106995, + "num_tokens": 447674361.0, + "step": 17313 + }, + { + "epoch": 1.9013837030529321, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.539360761642456, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6972261667251587, + "num_tokens": 447700851.0, + "step": 17314 + }, + { + "epoch": 1.9014935207555457, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4098095893859863, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7062075734138489, + "num_tokens": 447729638.0, + "step": 17315 + }, + { + "epoch": 1.9016033384581594, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2356889247894287, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6905412673950195, + "num_tokens": 447760890.0, + "step": 17316 + }, + { + "epoch": 1.9017131561607732, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.346001625061035, + "learning_rate": 1e-06, + "loss": 0.8955, + "mean_token_accuracy": 0.7319837808609009, + "num_tokens": 447785658.0, + "step": 17317 + }, + { + "epoch": 1.9018229738633867, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5693626403808594, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6949474811553955, + "num_tokens": 447811159.0, + "step": 17318 + }, + { + "epoch": 1.9019327915660005, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5445499420166016, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7167112827301025, + "num_tokens": 447838044.0, + "step": 17319 + }, + { + "epoch": 1.902042609268614, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.589210271835327, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7163288593292236, + "num_tokens": 447865087.0, + "step": 17320 + }, + { + "epoch": 1.9021524269712278, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8673341274261475, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6994121074676514, + "num_tokens": 447890684.0, + "step": 17321 + }, + { + "epoch": 1.9022622446738415, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.561774730682373, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7115986347198486, + "num_tokens": 447915223.0, + "step": 17322 + }, + { + "epoch": 1.902372062376455, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4918646812438965, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6944859027862549, + "num_tokens": 447943562.0, + "step": 17323 + }, + { + "epoch": 1.9024818800790686, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6767501831054688, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7107723951339722, + "num_tokens": 447967291.0, + "step": 17324 + }, + { + "epoch": 1.9025916977816824, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.585192918777466, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6985695362091064, + "num_tokens": 447993169.0, + "step": 17325 + }, + { + "epoch": 1.9027015154842961, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3834774494171143, + "learning_rate": 1e-06, + "loss": 1.0845, + "mean_token_accuracy": 0.6973822116851807, + "num_tokens": 448021760.0, + "step": 17326 + }, + { + "epoch": 1.9028113331869099, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.385850191116333, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.684170126914978, + "num_tokens": 448054045.0, + "step": 17327 + }, + { + "epoch": 1.9029211508895234, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.652744770050049, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7088824510574341, + "num_tokens": 448083280.0, + "step": 17328 + }, + { + "epoch": 1.903030968592137, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.435835838317871, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.7047601938247681, + "num_tokens": 448110152.0, + "step": 17329 + }, + { + "epoch": 1.9031407862947507, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4534943103790283, + "learning_rate": 1e-06, + "loss": 1.1142, + "mean_token_accuracy": 0.6867026686668396, + "num_tokens": 448141465.0, + "step": 17330 + }, + { + "epoch": 1.9032506039973645, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5263893604278564, + "learning_rate": 1e-06, + "loss": 0.9042, + "mean_token_accuracy": 0.7235698699951172, + "num_tokens": 448165574.0, + "step": 17331 + }, + { + "epoch": 1.903360421699978, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.42152738571167, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.6866754293441772, + "num_tokens": 448193919.0, + "step": 17332 + }, + { + "epoch": 1.9034702394025917, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4606714248657227, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7144617438316345, + "num_tokens": 448219763.0, + "step": 17333 + }, + { + "epoch": 1.9035800571052053, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 4.104879856109619, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6824004650115967, + "num_tokens": 448251785.0, + "step": 17334 + }, + { + "epoch": 1.903689874807819, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.836510181427002, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7068266868591309, + "num_tokens": 448278044.0, + "step": 17335 + }, + { + "epoch": 1.9037996925104328, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.61409592628479, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7165600061416626, + "num_tokens": 448303866.0, + "step": 17336 + }, + { + "epoch": 1.9039095102130463, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.466688871383667, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6942951679229736, + "num_tokens": 448334191.0, + "step": 17337 + }, + { + "epoch": 1.9040193279156599, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5492334365844727, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6999450922012329, + "num_tokens": 448361400.0, + "step": 17338 + }, + { + "epoch": 1.9041291456182736, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.859076738357544, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6967495679855347, + "num_tokens": 448381714.0, + "step": 17339 + }, + { + "epoch": 1.9042389633208874, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.584172248840332, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.6997646689414978, + "num_tokens": 448407440.0, + "step": 17340 + }, + { + "epoch": 1.9043487810235011, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7209534645080566, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7110002040863037, + "num_tokens": 448429954.0, + "step": 17341 + }, + { + "epoch": 1.9044585987261147, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7022018432617188, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.7243002653121948, + "num_tokens": 448455485.0, + "step": 17342 + }, + { + "epoch": 1.9045684164287282, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6407363414764404, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7276722192764282, + "num_tokens": 448480074.0, + "step": 17343 + }, + { + "epoch": 1.904678234131342, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6846580505371094, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7214725017547607, + "num_tokens": 448507473.0, + "step": 17344 + }, + { + "epoch": 1.9047880518339557, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8062877655029297, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7026258707046509, + "num_tokens": 448529085.0, + "step": 17345 + }, + { + "epoch": 1.9048978695365693, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.682605743408203, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.7043056488037109, + "num_tokens": 448553224.0, + "step": 17346 + }, + { + "epoch": 1.9050076872391828, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6509933471679688, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6884094476699829, + "num_tokens": 448577920.0, + "step": 17347 + }, + { + "epoch": 1.9051175049417965, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4833109378814697, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.71270352602005, + "num_tokens": 448604254.0, + "step": 17348 + }, + { + "epoch": 1.9052273226444103, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8085994720458984, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7061192989349365, + "num_tokens": 448626666.0, + "step": 17349 + }, + { + "epoch": 1.905337140347024, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3334269523620605, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6966781616210938, + "num_tokens": 448661643.0, + "step": 17350 + }, + { + "epoch": 1.9054469580496376, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6044058799743652, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7013444304466248, + "num_tokens": 448688672.0, + "step": 17351 + }, + { + "epoch": 1.9055567757522511, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1478490829467773, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7100502252578735, + "num_tokens": 448707192.0, + "step": 17352 + }, + { + "epoch": 1.9056665934548649, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6883625984191895, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6876640319824219, + "num_tokens": 448733459.0, + "step": 17353 + }, + { + "epoch": 1.9057764111574786, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.496861219406128, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7197142839431763, + "num_tokens": 448758398.0, + "step": 17354 + }, + { + "epoch": 1.9058862288600924, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 16.851388931274414, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7033545970916748, + "num_tokens": 448779754.0, + "step": 17355 + }, + { + "epoch": 1.905996046562706, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 4.0378499031066895, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6933484077453613, + "num_tokens": 448804077.0, + "step": 17356 + }, + { + "epoch": 1.9061058642653195, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.000793933868408, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7126491069793701, + "num_tokens": 448825247.0, + "step": 17357 + }, + { + "epoch": 1.9062156819679332, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.3027989864349365, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7139842510223389, + "num_tokens": 448850877.0, + "step": 17358 + }, + { + "epoch": 1.906325499670547, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.740521192550659, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.6942141056060791, + "num_tokens": 448873759.0, + "step": 17359 + }, + { + "epoch": 1.9064353173731605, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5467796325683594, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6894592046737671, + "num_tokens": 448902244.0, + "step": 17360 + }, + { + "epoch": 1.906545135075774, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.271908760070801, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.7094751596450806, + "num_tokens": 448937166.0, + "step": 17361 + }, + { + "epoch": 1.9066549527783878, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4652063846588135, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7074791193008423, + "num_tokens": 448961982.0, + "step": 17362 + }, + { + "epoch": 1.9067647704810016, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.509868860244751, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6950307488441467, + "num_tokens": 448993537.0, + "step": 17363 + }, + { + "epoch": 1.9068745881836153, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5703749656677246, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7096180319786072, + "num_tokens": 449020416.0, + "step": 17364 + }, + { + "epoch": 1.9069844058862289, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.563835620880127, + "learning_rate": 1e-06, + "loss": 0.9775, + "mean_token_accuracy": 0.7239198088645935, + "num_tokens": 449046315.0, + "step": 17365 + }, + { + "epoch": 1.9070942235888424, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7736833095550537, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6904650926589966, + "num_tokens": 449072172.0, + "step": 17366 + }, + { + "epoch": 1.9072040412914562, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 5.494665145874023, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7161482572555542, + "num_tokens": 449094238.0, + "step": 17367 + }, + { + "epoch": 1.90731385899407, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.2703402042388916, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.7046764492988586, + "num_tokens": 449118465.0, + "step": 17368 + }, + { + "epoch": 1.9074236766966837, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.950968027114868, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6914399266242981, + "num_tokens": 449140523.0, + "step": 17369 + }, + { + "epoch": 1.9075334943992972, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.717751979827881, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6954478621482849, + "num_tokens": 449167458.0, + "step": 17370 + }, + { + "epoch": 1.9076433121019107, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8467133045196533, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7167450785636902, + "num_tokens": 449187811.0, + "step": 17371 + }, + { + "epoch": 1.9077531298045245, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6048583984375, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7257741093635559, + "num_tokens": 449212999.0, + "step": 17372 + }, + { + "epoch": 1.9078629475071383, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.46785044670105, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6866664886474609, + "num_tokens": 449238024.0, + "step": 17373 + }, + { + "epoch": 1.9079727652097518, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4061226844787598, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7068315744400024, + "num_tokens": 449266862.0, + "step": 17374 + }, + { + "epoch": 1.9080825829123653, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2922441959381104, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7192394733428955, + "num_tokens": 449296136.0, + "step": 17375 + }, + { + "epoch": 1.908192400614979, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5271005630493164, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6902061700820923, + "num_tokens": 449322698.0, + "step": 17376 + }, + { + "epoch": 1.9083022183175928, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.531817674636841, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6869114637374878, + "num_tokens": 449351229.0, + "step": 17377 + }, + { + "epoch": 1.9084120360202066, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7485034465789795, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7182416915893555, + "num_tokens": 449374457.0, + "step": 17378 + }, + { + "epoch": 1.9085218537228201, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.319087028503418, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7007580399513245, + "num_tokens": 449401738.0, + "step": 17379 + }, + { + "epoch": 1.9086316714254337, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5311009883880615, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.6997793912887573, + "num_tokens": 449428375.0, + "step": 17380 + }, + { + "epoch": 1.9087414891280474, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4547176361083984, + "learning_rate": 1e-06, + "loss": 1.1697, + "mean_token_accuracy": 0.670649528503418, + "num_tokens": 449460367.0, + "step": 17381 + }, + { + "epoch": 1.9088513068306612, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.697704315185547, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7238823175430298, + "num_tokens": 449481939.0, + "step": 17382 + }, + { + "epoch": 1.9089611245332747, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.509136199951172, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7034524083137512, + "num_tokens": 449507779.0, + "step": 17383 + }, + { + "epoch": 1.9090709422358885, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.563499689102173, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7134767770767212, + "num_tokens": 449536049.0, + "step": 17384 + }, + { + "epoch": 1.909180759938502, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9043102264404297, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7224422693252563, + "num_tokens": 449556604.0, + "step": 17385 + }, + { + "epoch": 1.9092905776411158, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.558260679244995, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7108554840087891, + "num_tokens": 449583690.0, + "step": 17386 + }, + { + "epoch": 1.9094003953437295, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6249239444732666, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7192895412445068, + "num_tokens": 449609306.0, + "step": 17387 + }, + { + "epoch": 1.909510213046343, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7368342876434326, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7113131284713745, + "num_tokens": 449637102.0, + "step": 17388 + }, + { + "epoch": 1.9096200307489566, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5827887058258057, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7180095911026001, + "num_tokens": 449659902.0, + "step": 17389 + }, + { + "epoch": 1.9097298484515703, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5560107231140137, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.7111793756484985, + "num_tokens": 449685570.0, + "step": 17390 + }, + { + "epoch": 1.909839666154184, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.768986463546753, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7118328809738159, + "num_tokens": 449709286.0, + "step": 17391 + }, + { + "epoch": 1.9099494838567979, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5720510482788086, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7011345028877258, + "num_tokens": 449735615.0, + "step": 17392 + }, + { + "epoch": 1.9100593015594114, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4064834117889404, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6990554332733154, + "num_tokens": 449762331.0, + "step": 17393 + }, + { + "epoch": 1.910169119262025, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.557863712310791, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6980355381965637, + "num_tokens": 449795113.0, + "step": 17394 + }, + { + "epoch": 1.9102789369646387, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 16.735767364501953, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7153458595275879, + "num_tokens": 449821302.0, + "step": 17395 + }, + { + "epoch": 1.9103887546672524, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.674243927001953, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6982647180557251, + "num_tokens": 449847719.0, + "step": 17396 + }, + { + "epoch": 1.910498572369866, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6930830478668213, + "learning_rate": 1e-06, + "loss": 1.0821, + "mean_token_accuracy": 0.6836965680122375, + "num_tokens": 449874132.0, + "step": 17397 + }, + { + "epoch": 1.9106083900724797, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.951850175857544, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7152355909347534, + "num_tokens": 449892049.0, + "step": 17398 + }, + { + "epoch": 1.9107182077750933, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.451648712158203, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.7046926021575928, + "num_tokens": 449918827.0, + "step": 17399 + }, + { + "epoch": 1.910828025477707, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.1525232791900635, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.697669267654419, + "num_tokens": 449940306.0, + "step": 17400 + }, + { + "epoch": 1.9109378431803208, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4717636108398438, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6912921667098999, + "num_tokens": 449968701.0, + "step": 17401 + }, + { + "epoch": 1.9110476608829343, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.568176507949829, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6839593648910522, + "num_tokens": 449994837.0, + "step": 17402 + }, + { + "epoch": 1.9111574785855479, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6428592205047607, + "learning_rate": 1e-06, + "loss": 1.0591, + "mean_token_accuracy": 0.6920042037963867, + "num_tokens": 450021966.0, + "step": 17403 + }, + { + "epoch": 1.9112672962881616, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.423435926437378, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.692084789276123, + "num_tokens": 450050687.0, + "step": 17404 + }, + { + "epoch": 1.9113771139907754, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6680233478546143, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7060255408287048, + "num_tokens": 450076238.0, + "step": 17405 + }, + { + "epoch": 1.9114869316933891, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6717052459716797, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7110050320625305, + "num_tokens": 450096935.0, + "step": 17406 + }, + { + "epoch": 1.9115967493960027, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6451547145843506, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.7035327553749084, + "num_tokens": 450120989.0, + "step": 17407 + }, + { + "epoch": 1.9117065670986162, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.578744649887085, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7131129503250122, + "num_tokens": 450148570.0, + "step": 17408 + }, + { + "epoch": 1.91181638480123, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.0908255577087402, + "learning_rate": 1e-06, + "loss": 0.8616, + "mean_token_accuracy": 0.7467880845069885, + "num_tokens": 450180517.0, + "step": 17409 + }, + { + "epoch": 1.9119262025038437, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6640512943267822, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7196766138076782, + "num_tokens": 450204352.0, + "step": 17410 + }, + { + "epoch": 1.9120360202064572, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.98457670211792, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7052825689315796, + "num_tokens": 450223829.0, + "step": 17411 + }, + { + "epoch": 1.9121458379090708, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5864524841308594, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6979870796203613, + "num_tokens": 450249659.0, + "step": 17412 + }, + { + "epoch": 1.9122556556116845, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.783902883529663, + "learning_rate": 1e-06, + "loss": 0.9231, + "mean_token_accuracy": 0.7271740436553955, + "num_tokens": 450271616.0, + "step": 17413 + }, + { + "epoch": 1.9123654733142983, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.642793893814087, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7272081971168518, + "num_tokens": 450293969.0, + "step": 17414 + }, + { + "epoch": 1.912475291016912, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8002076148986816, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7186129093170166, + "num_tokens": 450317278.0, + "step": 17415 + }, + { + "epoch": 1.9125851087195256, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7001898288726807, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6949472427368164, + "num_tokens": 450343184.0, + "step": 17416 + }, + { + "epoch": 1.9126949264221391, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9032604694366455, + "learning_rate": 1e-06, + "loss": 0.9187, + "mean_token_accuracy": 0.735028862953186, + "num_tokens": 450360937.0, + "step": 17417 + }, + { + "epoch": 1.9128047441247529, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.555206298828125, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7070658206939697, + "num_tokens": 450384911.0, + "step": 17418 + }, + { + "epoch": 1.9129145618273666, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.977020263671875, + "learning_rate": 1e-06, + "loss": 0.9045, + "mean_token_accuracy": 0.7314678430557251, + "num_tokens": 450403271.0, + "step": 17419 + }, + { + "epoch": 1.9130243795299804, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7450459003448486, + "learning_rate": 1e-06, + "loss": 0.9124, + "mean_token_accuracy": 0.728313684463501, + "num_tokens": 450425048.0, + "step": 17420 + }, + { + "epoch": 1.913134197232594, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5986685752868652, + "learning_rate": 1e-06, + "loss": 0.8733, + "mean_token_accuracy": 0.7430688142776489, + "num_tokens": 450449798.0, + "step": 17421 + }, + { + "epoch": 1.9132440149352075, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.899491548538208, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7122362852096558, + "num_tokens": 450471716.0, + "step": 17422 + }, + { + "epoch": 1.9133538326378212, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.735903739929199, + "learning_rate": 1e-06, + "loss": 1.1392, + "mean_token_accuracy": 0.6800966262817383, + "num_tokens": 450499535.0, + "step": 17423 + }, + { + "epoch": 1.913463650340435, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4497385025024414, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7093583345413208, + "num_tokens": 450523739.0, + "step": 17424 + }, + { + "epoch": 1.9135734680430485, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.358715057373047, + "learning_rate": 1e-06, + "loss": 1.1652, + "mean_token_accuracy": 0.6669344902038574, + "num_tokens": 450556393.0, + "step": 17425 + }, + { + "epoch": 1.913683285745662, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7339365482330322, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7121556997299194, + "num_tokens": 450577971.0, + "step": 17426 + }, + { + "epoch": 1.9137931034482758, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.571465253829956, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6953975558280945, + "num_tokens": 450609518.0, + "step": 17427 + }, + { + "epoch": 1.9139029211508896, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9262032508850098, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7246311902999878, + "num_tokens": 450629031.0, + "step": 17428 + }, + { + "epoch": 1.9140127388535033, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5920774936676025, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7127127647399902, + "num_tokens": 450656033.0, + "step": 17429 + }, + { + "epoch": 1.9141225565561168, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7115750312805176, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6960732936859131, + "num_tokens": 450681630.0, + "step": 17430 + }, + { + "epoch": 1.9142323742587304, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.827376365661621, + "learning_rate": 1e-06, + "loss": 0.8965, + "mean_token_accuracy": 0.7376335263252258, + "num_tokens": 450701588.0, + "step": 17431 + }, + { + "epoch": 1.9143421919613441, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6410746574401855, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7165033221244812, + "num_tokens": 450724736.0, + "step": 17432 + }, + { + "epoch": 1.914452009663958, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6972248554229736, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7108157873153687, + "num_tokens": 450748266.0, + "step": 17433 + }, + { + "epoch": 1.9145618273665717, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.480242967605591, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.7112275958061218, + "num_tokens": 450777544.0, + "step": 17434 + }, + { + "epoch": 1.9146716450691852, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 16.709754943847656, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7078644633293152, + "num_tokens": 450802881.0, + "step": 17435 + }, + { + "epoch": 1.9147814627717987, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.6385159492492676, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6939654350280762, + "num_tokens": 450832711.0, + "step": 17436 + }, + { + "epoch": 1.9148912804744125, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4768669605255127, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6965532898902893, + "num_tokens": 450862156.0, + "step": 17437 + }, + { + "epoch": 1.9150010981770262, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.650301933288574, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.687952995300293, + "num_tokens": 450887857.0, + "step": 17438 + }, + { + "epoch": 1.9151109158796398, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.514961004257202, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7038799524307251, + "num_tokens": 450916891.0, + "step": 17439 + }, + { + "epoch": 1.9152207335822533, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.3535547256469727, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.691241979598999, + "num_tokens": 450943925.0, + "step": 17440 + }, + { + "epoch": 1.915330551284867, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8781065940856934, + "learning_rate": 1e-06, + "loss": 0.9011, + "mean_token_accuracy": 0.7373975515365601, + "num_tokens": 450964413.0, + "step": 17441 + }, + { + "epoch": 1.9154403689874808, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5289204120635986, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7072701454162598, + "num_tokens": 450989912.0, + "step": 17442 + }, + { + "epoch": 1.9155501866900946, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 80.54702758789062, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7013860940933228, + "num_tokens": 451014695.0, + "step": 17443 + }, + { + "epoch": 1.9156600043927081, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.119870185852051, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6923164129257202, + "num_tokens": 451035290.0, + "step": 17444 + }, + { + "epoch": 1.9157698220953217, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.330878973007202, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.7144362926483154, + "num_tokens": 451057017.0, + "step": 17445 + }, + { + "epoch": 1.9158796397979354, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5682036876678467, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7191835641860962, + "num_tokens": 451081191.0, + "step": 17446 + }, + { + "epoch": 1.9159894575005492, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5692098140716553, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7143767476081848, + "num_tokens": 451107356.0, + "step": 17447 + }, + { + "epoch": 1.9160992752031627, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.2902989387512207, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6949077844619751, + "num_tokens": 451142762.0, + "step": 17448 + }, + { + "epoch": 1.9162090929057765, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4048869609832764, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7129322290420532, + "num_tokens": 451170603.0, + "step": 17449 + }, + { + "epoch": 1.91631891060839, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.99706768989563, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6926323771476746, + "num_tokens": 451200087.0, + "step": 17450 + }, + { + "epoch": 1.9164287283110037, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.775367021560669, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6953076720237732, + "num_tokens": 451223252.0, + "step": 17451 + }, + { + "epoch": 1.9165385460136175, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.695014476776123, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6993244290351868, + "num_tokens": 451245659.0, + "step": 17452 + }, + { + "epoch": 1.916648363716231, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6377742290496826, + "learning_rate": 1e-06, + "loss": 0.9606, + "mean_token_accuracy": 0.720655620098114, + "num_tokens": 451267037.0, + "step": 17453 + }, + { + "epoch": 1.9167581814188446, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7601070404052734, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7110207080841064, + "num_tokens": 451288055.0, + "step": 17454 + }, + { + "epoch": 1.9168679991214583, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.910381555557251, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6936976909637451, + "num_tokens": 451309635.0, + "step": 17455 + }, + { + "epoch": 1.916977816824072, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.405076265335083, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7071874737739563, + "num_tokens": 451341564.0, + "step": 17456 + }, + { + "epoch": 1.9170876345266858, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.380995512008667, + "learning_rate": 1e-06, + "loss": 1.1292, + "mean_token_accuracy": 0.6813504099845886, + "num_tokens": 451369371.0, + "step": 17457 + }, + { + "epoch": 1.9171974522292994, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5886950492858887, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7148915529251099, + "num_tokens": 451393862.0, + "step": 17458 + }, + { + "epoch": 1.917307269931913, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5194499492645264, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.7049257755279541, + "num_tokens": 451421284.0, + "step": 17459 + }, + { + "epoch": 1.9174170876345267, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6640350818634033, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.6949501633644104, + "num_tokens": 451445744.0, + "step": 17460 + }, + { + "epoch": 1.9175269053371404, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5633561611175537, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.715142011642456, + "num_tokens": 451469195.0, + "step": 17461 + }, + { + "epoch": 1.917636723039754, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6669161319732666, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7065106630325317, + "num_tokens": 451492045.0, + "step": 17462 + }, + { + "epoch": 1.9177465407423677, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9077060222625732, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7044220566749573, + "num_tokens": 451514767.0, + "step": 17463 + }, + { + "epoch": 1.9178563584449813, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.574376106262207, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7100423574447632, + "num_tokens": 451541390.0, + "step": 17464 + }, + { + "epoch": 1.917966176147595, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.639292001724243, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7107197046279907, + "num_tokens": 451565160.0, + "step": 17465 + }, + { + "epoch": 1.9180759938502088, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.696995258331299, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6926628947257996, + "num_tokens": 451587544.0, + "step": 17466 + }, + { + "epoch": 1.9181858115528223, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.2113659381866455, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.7295615077018738, + "num_tokens": 451612067.0, + "step": 17467 + }, + { + "epoch": 1.9182956292554358, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.533327102661133, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7266075611114502, + "num_tokens": 451638093.0, + "step": 17468 + }, + { + "epoch": 1.9184054469580496, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.576545476913452, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7087130546569824, + "num_tokens": 451665022.0, + "step": 17469 + }, + { + "epoch": 1.9185152646606634, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.899648904800415, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.6980928778648376, + "num_tokens": 451692246.0, + "step": 17470 + }, + { + "epoch": 1.9186250823632771, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7181224822998047, + "learning_rate": 1e-06, + "loss": 0.9085, + "mean_token_accuracy": 0.7299113869667053, + "num_tokens": 451715934.0, + "step": 17471 + }, + { + "epoch": 1.9187349000658906, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2661852836608887, + "learning_rate": 1e-06, + "loss": 1.1152, + "mean_token_accuracy": 0.6795744895935059, + "num_tokens": 451751060.0, + "step": 17472 + }, + { + "epoch": 1.9188447177685042, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.393474817276001, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6991559267044067, + "num_tokens": 451782865.0, + "step": 17473 + }, + { + "epoch": 1.918954535471118, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.411681652069092, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7112857699394226, + "num_tokens": 451812176.0, + "step": 17474 + }, + { + "epoch": 1.9190643531737317, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7245166301727295, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.7295775413513184, + "num_tokens": 451835676.0, + "step": 17475 + }, + { + "epoch": 1.9191741708763452, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.415163040161133, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.7009470462799072, + "num_tokens": 451864951.0, + "step": 17476 + }, + { + "epoch": 1.9192839885789588, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5423226356506348, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7058411240577698, + "num_tokens": 451888418.0, + "step": 17477 + }, + { + "epoch": 1.9193938062815725, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5350890159606934, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6884663105010986, + "num_tokens": 451917245.0, + "step": 17478 + }, + { + "epoch": 1.9195036239841863, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6484851837158203, + "learning_rate": 1e-06, + "loss": 0.9031, + "mean_token_accuracy": 0.7346479892730713, + "num_tokens": 451940714.0, + "step": 17479 + }, + { + "epoch": 1.9196134416868, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.485393524169922, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6985271573066711, + "num_tokens": 451971952.0, + "step": 17480 + }, + { + "epoch": 1.9197232593894136, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.514448404312134, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6898072957992554, + "num_tokens": 452002024.0, + "step": 17481 + }, + { + "epoch": 1.919833077092027, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.538811445236206, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7066928148269653, + "num_tokens": 452029407.0, + "step": 17482 + }, + { + "epoch": 1.9199428947946409, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5384559631347656, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7013881206512451, + "num_tokens": 452056850.0, + "step": 17483 + }, + { + "epoch": 1.9200527124972546, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4476511478424072, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7011309862136841, + "num_tokens": 452084598.0, + "step": 17484 + }, + { + "epoch": 1.9201625301998684, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 4.55547571182251, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7152249813079834, + "num_tokens": 452112922.0, + "step": 17485 + }, + { + "epoch": 1.920272347902482, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.353091239929199, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.6977295279502869, + "num_tokens": 452142665.0, + "step": 17486 + }, + { + "epoch": 1.9203821656050954, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6907293796539307, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7090909481048584, + "num_tokens": 452165606.0, + "step": 17487 + }, + { + "epoch": 1.9204919833077092, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.823812246322632, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7330377697944641, + "num_tokens": 452184198.0, + "step": 17488 + }, + { + "epoch": 1.920601801010323, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8157026767730713, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7021353840827942, + "num_tokens": 452206770.0, + "step": 17489 + }, + { + "epoch": 1.9207116187129365, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6276183128356934, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7041671276092529, + "num_tokens": 452233802.0, + "step": 17490 + }, + { + "epoch": 1.92082143641555, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.901854991912842, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.706964373588562, + "num_tokens": 452257788.0, + "step": 17491 + }, + { + "epoch": 1.9209312541181638, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.434190273284912, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7002814412117004, + "num_tokens": 452285869.0, + "step": 17492 + }, + { + "epoch": 1.9210410718207775, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3419888019561768, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7148288488388062, + "num_tokens": 452317624.0, + "step": 17493 + }, + { + "epoch": 1.9211508895233913, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8265883922576904, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7105494737625122, + "num_tokens": 452341393.0, + "step": 17494 + }, + { + "epoch": 1.9212607072260048, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.695237636566162, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7058220505714417, + "num_tokens": 452366505.0, + "step": 17495 + }, + { + "epoch": 1.9213705249286184, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0719282627105713, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7274506092071533, + "num_tokens": 452387653.0, + "step": 17496 + }, + { + "epoch": 1.9214803426312321, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.686997413635254, + "learning_rate": 1e-06, + "loss": 0.92, + "mean_token_accuracy": 0.7223279476165771, + "num_tokens": 452410836.0, + "step": 17497 + }, + { + "epoch": 1.9215901603338459, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.45971417427063, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6996064782142639, + "num_tokens": 452441639.0, + "step": 17498 + }, + { + "epoch": 1.9216999780364594, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.846118688583374, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7026838064193726, + "num_tokens": 452462803.0, + "step": 17499 + }, + { + "epoch": 1.9218097957390732, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.682218551635742, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7196574807167053, + "num_tokens": 452485567.0, + "step": 17500 + }, + { + "epoch": 1.9219196134416867, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5443403720855713, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6919618844985962, + "num_tokens": 452512177.0, + "step": 17501 + }, + { + "epoch": 1.9220294311443005, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.348576784133911, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.712679922580719, + "num_tokens": 452540671.0, + "step": 17502 + }, + { + "epoch": 1.9221392488469142, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2535154819488525, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.7061878442764282, + "num_tokens": 452572954.0, + "step": 17503 + }, + { + "epoch": 1.9222490665495278, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6917734146118164, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.703067421913147, + "num_tokens": 452595664.0, + "step": 17504 + }, + { + "epoch": 1.9223588842521413, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8549771308898926, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7159496545791626, + "num_tokens": 452616085.0, + "step": 17505 + }, + { + "epoch": 1.922468701954755, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7793960571289062, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6878277063369751, + "num_tokens": 452641245.0, + "step": 17506 + }, + { + "epoch": 1.9225785196573688, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.605926513671875, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.6983850002288818, + "num_tokens": 452666302.0, + "step": 17507 + }, + { + "epoch": 1.9226883373599826, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7213308811187744, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7125471830368042, + "num_tokens": 452689351.0, + "step": 17508 + }, + { + "epoch": 1.922798155062596, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.9867565631866455, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7132700085639954, + "num_tokens": 452709981.0, + "step": 17509 + }, + { + "epoch": 1.9229079727652096, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6747446060180664, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7044909596443176, + "num_tokens": 452739967.0, + "step": 17510 + }, + { + "epoch": 1.9230177904678234, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6014230251312256, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.7065591812133789, + "num_tokens": 452766530.0, + "step": 17511 + }, + { + "epoch": 1.9231276081704372, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5312459468841553, + "learning_rate": 1e-06, + "loss": 1.1208, + "mean_token_accuracy": 0.6820018291473389, + "num_tokens": 452795231.0, + "step": 17512 + }, + { + "epoch": 1.9232374258730507, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.536571979522705, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7035970687866211, + "num_tokens": 452823578.0, + "step": 17513 + }, + { + "epoch": 1.9233472435756644, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.337826728820801, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6985647678375244, + "num_tokens": 452855213.0, + "step": 17514 + }, + { + "epoch": 1.923457061278278, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.501056432723999, + "learning_rate": 1e-06, + "loss": 1.163, + "mean_token_accuracy": 0.676645040512085, + "num_tokens": 452885761.0, + "step": 17515 + }, + { + "epoch": 1.9235668789808917, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7148916721343994, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7218948602676392, + "num_tokens": 452909656.0, + "step": 17516 + }, + { + "epoch": 1.9236766966835055, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.287860631942749, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7093194723129272, + "num_tokens": 452945783.0, + "step": 17517 + }, + { + "epoch": 1.923786514386119, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.818197011947632, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7116075754165649, + "num_tokens": 452967176.0, + "step": 17518 + }, + { + "epoch": 1.9238963320887326, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6544928550720215, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6889806389808655, + "num_tokens": 452991130.0, + "step": 17519 + }, + { + "epoch": 1.9240061497913463, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4273312091827393, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7175679206848145, + "num_tokens": 453017318.0, + "step": 17520 + }, + { + "epoch": 1.92411596749396, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.737410306930542, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6900328397750854, + "num_tokens": 453042097.0, + "step": 17521 + }, + { + "epoch": 1.9242257851965738, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5286436080932617, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6971150636672974, + "num_tokens": 453070639.0, + "step": 17522 + }, + { + "epoch": 1.9243356028991874, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7040798664093018, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7300500869750977, + "num_tokens": 453092593.0, + "step": 17523 + }, + { + "epoch": 1.924445420601801, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.30621600151062, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.7304849624633789, + "num_tokens": 453121695.0, + "step": 17524 + }, + { + "epoch": 1.9245552383044147, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3741023540496826, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6948605179786682, + "num_tokens": 453150979.0, + "step": 17525 + }, + { + "epoch": 1.9246650560070284, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2818715572357178, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6928369402885437, + "num_tokens": 453184250.0, + "step": 17526 + }, + { + "epoch": 1.924774873709642, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 20.39499855041504, + "learning_rate": 1e-06, + "loss": 1.1183, + "mean_token_accuracy": 0.6760238409042358, + "num_tokens": 453215976.0, + "step": 17527 + }, + { + "epoch": 1.9248846914122557, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8471007347106934, + "learning_rate": 1e-06, + "loss": 0.8793, + "mean_token_accuracy": 0.7405095100402832, + "num_tokens": 453239722.0, + "step": 17528 + }, + { + "epoch": 1.9249945091148692, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7077696323394775, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6885137557983398, + "num_tokens": 453267887.0, + "step": 17529 + }, + { + "epoch": 1.925104326817483, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.383068323135376, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6958346366882324, + "num_tokens": 453297401.0, + "step": 17530 + }, + { + "epoch": 1.9252141445200968, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2451744079589844, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7091437578201294, + "num_tokens": 453328276.0, + "step": 17531 + }, + { + "epoch": 1.9253239622227103, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2728331089019775, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.6895771026611328, + "num_tokens": 453361354.0, + "step": 17532 + }, + { + "epoch": 1.9254337799253238, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6702206134796143, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7098416090011597, + "num_tokens": 453385604.0, + "step": 17533 + }, + { + "epoch": 1.9255435976279376, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6383185386657715, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7157568335533142, + "num_tokens": 453411241.0, + "step": 17534 + }, + { + "epoch": 1.9256534153305513, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7721643447875977, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.7364157438278198, + "num_tokens": 453432239.0, + "step": 17535 + }, + { + "epoch": 1.925763233033165, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.59590220451355, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7061194181442261, + "num_tokens": 453461818.0, + "step": 17536 + }, + { + "epoch": 1.9258730507357786, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.894923448562622, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7144364714622498, + "num_tokens": 453482038.0, + "step": 17537 + }, + { + "epoch": 1.9259828684383922, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4937734603881836, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7160948514938354, + "num_tokens": 453508784.0, + "step": 17538 + }, + { + "epoch": 1.926092686141006, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.402273178100586, + "learning_rate": 1e-06, + "loss": 1.1358, + "mean_token_accuracy": 0.6822861433029175, + "num_tokens": 453538687.0, + "step": 17539 + }, + { + "epoch": 1.9262025038436197, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6161251068115234, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.6871189475059509, + "num_tokens": 453564911.0, + "step": 17540 + }, + { + "epoch": 1.9263123215462332, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3794727325439453, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6837995648384094, + "num_tokens": 453599779.0, + "step": 17541 + }, + { + "epoch": 1.9264221392488468, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.666095733642578, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6953318119049072, + "num_tokens": 453624549.0, + "step": 17542 + }, + { + "epoch": 1.9265319569514605, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5767178535461426, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.708365797996521, + "num_tokens": 453648337.0, + "step": 17543 + }, + { + "epoch": 1.9266417746540743, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4059112071990967, + "learning_rate": 1e-06, + "loss": 0.8742, + "mean_token_accuracy": 0.7388761639595032, + "num_tokens": 453675954.0, + "step": 17544 + }, + { + "epoch": 1.926751592356688, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5668935775756836, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6866903305053711, + "num_tokens": 453702509.0, + "step": 17545 + }, + { + "epoch": 1.9268614100593016, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4667365550994873, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7243276834487915, + "num_tokens": 453730629.0, + "step": 17546 + }, + { + "epoch": 1.926971227761915, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.932608127593994, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7138679623603821, + "num_tokens": 453752699.0, + "step": 17547 + }, + { + "epoch": 1.9270810454645289, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.571721076965332, + "learning_rate": 1e-06, + "loss": 1.1073, + "mean_token_accuracy": 0.681488573551178, + "num_tokens": 453779843.0, + "step": 17548 + }, + { + "epoch": 1.9271908631671426, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4620625972747803, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7106131911277771, + "num_tokens": 453805125.0, + "step": 17549 + }, + { + "epoch": 1.9273006808697564, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.569197416305542, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6923784017562866, + "num_tokens": 453834983.0, + "step": 17550 + }, + { + "epoch": 1.92741049857237, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5321693420410156, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7169893383979797, + "num_tokens": 453861241.0, + "step": 17551 + }, + { + "epoch": 1.9275203162749834, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5110771656036377, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7001973390579224, + "num_tokens": 453887612.0, + "step": 17552 + }, + { + "epoch": 1.9276301339775972, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6891725063323975, + "learning_rate": 1e-06, + "loss": 0.9203, + "mean_token_accuracy": 0.7288249731063843, + "num_tokens": 453909501.0, + "step": 17553 + }, + { + "epoch": 1.927739951680211, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.93410587310791, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7181516289710999, + "num_tokens": 453930561.0, + "step": 17554 + }, + { + "epoch": 1.9278497693828245, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4627301692962646, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6796217560768127, + "num_tokens": 453960082.0, + "step": 17555 + }, + { + "epoch": 1.927959587085438, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4249181747436523, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7048011422157288, + "num_tokens": 453994622.0, + "step": 17556 + }, + { + "epoch": 1.9280694047880518, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5520598888397217, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7218076586723328, + "num_tokens": 454019594.0, + "step": 17557 + }, + { + "epoch": 1.9281792224906655, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.753286361694336, + "learning_rate": 1e-06, + "loss": 0.9013, + "mean_token_accuracy": 0.737479567527771, + "num_tokens": 454042410.0, + "step": 17558 + }, + { + "epoch": 1.9282890401932793, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4210216999053955, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7153551578521729, + "num_tokens": 454074885.0, + "step": 17559 + }, + { + "epoch": 1.9283988578958928, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.690310478210449, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6868091821670532, + "num_tokens": 454099407.0, + "step": 17560 + }, + { + "epoch": 1.9285086755985064, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5376482009887695, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7059573531150818, + "num_tokens": 454125813.0, + "step": 17561 + }, + { + "epoch": 1.9286184933011201, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6167914867401123, + "learning_rate": 1e-06, + "loss": 0.8965, + "mean_token_accuracy": 0.7320337891578674, + "num_tokens": 454152954.0, + "step": 17562 + }, + { + "epoch": 1.9287283110037339, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4746506214141846, + "learning_rate": 1e-06, + "loss": 0.9761, + "mean_token_accuracy": 0.7138800024986267, + "num_tokens": 454179351.0, + "step": 17563 + }, + { + "epoch": 1.9288381287063474, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3839733600616455, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7012737989425659, + "num_tokens": 454205510.0, + "step": 17564 + }, + { + "epoch": 1.9289479464089612, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6454403400421143, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7201645374298096, + "num_tokens": 454228764.0, + "step": 17565 + }, + { + "epoch": 1.9290577641115747, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.843989849090576, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7157708406448364, + "num_tokens": 454248196.0, + "step": 17566 + }, + { + "epoch": 1.9291675818141885, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6936604976654053, + "learning_rate": 1e-06, + "loss": 0.9526, + "mean_token_accuracy": 0.720498263835907, + "num_tokens": 454269712.0, + "step": 17567 + }, + { + "epoch": 1.9292773995168022, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5118324756622314, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.6858503818511963, + "num_tokens": 454297607.0, + "step": 17568 + }, + { + "epoch": 1.9293872172194158, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.679229497909546, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7079147100448608, + "num_tokens": 454321168.0, + "step": 17569 + }, + { + "epoch": 1.9294970349220293, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2748870849609375, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7046093344688416, + "num_tokens": 454352864.0, + "step": 17570 + }, + { + "epoch": 1.929606852624643, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.0651257038116455, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7184995412826538, + "num_tokens": 454377003.0, + "step": 17571 + }, + { + "epoch": 1.9297166703272568, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.730552911758423, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6938049793243408, + "num_tokens": 454403443.0, + "step": 17572 + }, + { + "epoch": 1.9298264880298706, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.574774742126465, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7110016345977783, + "num_tokens": 454428688.0, + "step": 17573 + }, + { + "epoch": 1.929936305732484, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.916801929473877, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.7330470085144043, + "num_tokens": 454448632.0, + "step": 17574 + }, + { + "epoch": 1.9300461234350976, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4410388469696045, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7177149057388306, + "num_tokens": 454478079.0, + "step": 17575 + }, + { + "epoch": 1.9301559411377114, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5707104206085205, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.7169221043586731, + "num_tokens": 454503560.0, + "step": 17576 + }, + { + "epoch": 1.9302657588403251, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.452545404434204, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6904337406158447, + "num_tokens": 454532357.0, + "step": 17577 + }, + { + "epoch": 1.9303755765429387, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3082103729248047, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7003409266471863, + "num_tokens": 454561608.0, + "step": 17578 + }, + { + "epoch": 1.9304853942455524, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.497840404510498, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6994558572769165, + "num_tokens": 454590535.0, + "step": 17579 + }, + { + "epoch": 1.930595211948166, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7122418880462646, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7254398465156555, + "num_tokens": 454613021.0, + "step": 17580 + }, + { + "epoch": 1.9307050296507797, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7175121307373047, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6892046928405762, + "num_tokens": 454638183.0, + "step": 17581 + }, + { + "epoch": 1.9308148473533935, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4655654430389404, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7111591100692749, + "num_tokens": 454664254.0, + "step": 17582 + }, + { + "epoch": 1.930924665056007, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2113044261932373, + "learning_rate": 1e-06, + "loss": 1.1087, + "mean_token_accuracy": 0.6839405298233032, + "num_tokens": 454702713.0, + "step": 17583 + }, + { + "epoch": 1.9310344827586206, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5178451538085938, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6998717188835144, + "num_tokens": 454729770.0, + "step": 17584 + }, + { + "epoch": 1.9311443004612343, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5170702934265137, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7075554132461548, + "num_tokens": 454754170.0, + "step": 17585 + }, + { + "epoch": 1.931254118163848, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.642948865890503, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7005536556243896, + "num_tokens": 454779132.0, + "step": 17586 + }, + { + "epoch": 1.9313639358664618, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.677400588989258, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6933383941650391, + "num_tokens": 454803807.0, + "step": 17587 + }, + { + "epoch": 1.9314737535690754, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6456377506256104, + "learning_rate": 1e-06, + "loss": 1.0927, + "mean_token_accuracy": 0.6888792514801025, + "num_tokens": 454831878.0, + "step": 17588 + }, + { + "epoch": 1.931583571271689, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5331757068634033, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7010417580604553, + "num_tokens": 454859770.0, + "step": 17589 + }, + { + "epoch": 1.9316933889743026, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.471923351287842, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6944278478622437, + "num_tokens": 454905089.0, + "step": 17590 + }, + { + "epoch": 1.9318032066769164, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.360250234603882, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7056692838668823, + "num_tokens": 454934153.0, + "step": 17591 + }, + { + "epoch": 1.93191302437953, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 20.531482696533203, + "learning_rate": 1e-06, + "loss": 1.1082, + "mean_token_accuracy": 0.6743394136428833, + "num_tokens": 454965708.0, + "step": 17592 + }, + { + "epoch": 1.9320228420821435, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.654379367828369, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6874114274978638, + "num_tokens": 454994316.0, + "step": 17593 + }, + { + "epoch": 1.9321326597847572, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8815300464630127, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7196812033653259, + "num_tokens": 455016511.0, + "step": 17594 + }, + { + "epoch": 1.932242477487371, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.927516460418701, + "learning_rate": 1e-06, + "loss": 0.901, + "mean_token_accuracy": 0.738101065158844, + "num_tokens": 455036822.0, + "step": 17595 + }, + { + "epoch": 1.9323522951899847, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5392699241638184, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7079464793205261, + "num_tokens": 455065624.0, + "step": 17596 + }, + { + "epoch": 1.9324621128925983, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.573871374130249, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7044065594673157, + "num_tokens": 455092259.0, + "step": 17597 + }, + { + "epoch": 1.9325719305952118, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.71628475189209, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7047284245491028, + "num_tokens": 455114492.0, + "step": 17598 + }, + { + "epoch": 1.9326817482978256, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.797605514526367, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7204580307006836, + "num_tokens": 455135634.0, + "step": 17599 + }, + { + "epoch": 1.9327915660004393, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.546645402908325, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7043169736862183, + "num_tokens": 455167182.0, + "step": 17600 + }, + { + "epoch": 1.932901383703053, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.647139549255371, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7034508585929871, + "num_tokens": 455191465.0, + "step": 17601 + }, + { + "epoch": 1.9330112014056666, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.2357606887817383, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7069647312164307, + "num_tokens": 455223966.0, + "step": 17602 + }, + { + "epoch": 1.9331210191082802, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.48714280128479, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7036831378936768, + "num_tokens": 455250006.0, + "step": 17603 + }, + { + "epoch": 1.933230836810894, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.781039237976074, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7005795240402222, + "num_tokens": 455273183.0, + "step": 17604 + }, + { + "epoch": 1.9333406545135077, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.554719924926758, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6906037330627441, + "num_tokens": 455302408.0, + "step": 17605 + }, + { + "epoch": 1.9334504722161212, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.8558764457702637, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6980804204940796, + "num_tokens": 455330213.0, + "step": 17606 + }, + { + "epoch": 1.9335602899187347, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7637250423431396, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6829845309257507, + "num_tokens": 455354541.0, + "step": 17607 + }, + { + "epoch": 1.9336701076213485, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.767774820327759, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6975252628326416, + "num_tokens": 455382122.0, + "step": 17608 + }, + { + "epoch": 1.9337799253239623, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.595151662826538, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.6994737386703491, + "num_tokens": 455408109.0, + "step": 17609 + }, + { + "epoch": 1.933889743026576, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.818082332611084, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7158289551734924, + "num_tokens": 455429011.0, + "step": 17610 + }, + { + "epoch": 1.9339995607291895, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.642465114593506, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7071840167045593, + "num_tokens": 455453465.0, + "step": 17611 + }, + { + "epoch": 1.934109378431803, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7054567337036133, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6992269158363342, + "num_tokens": 455484298.0, + "step": 17612 + }, + { + "epoch": 1.9342191961344168, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5500171184539795, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6935076117515564, + "num_tokens": 455510085.0, + "step": 17613 + }, + { + "epoch": 1.9343290138370306, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5641086101531982, + "learning_rate": 1e-06, + "loss": 1.1007, + "mean_token_accuracy": 0.6889059543609619, + "num_tokens": 455534673.0, + "step": 17614 + }, + { + "epoch": 1.9344388315396444, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5239510536193848, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.7019385099411011, + "num_tokens": 455559873.0, + "step": 17615 + }, + { + "epoch": 1.934548649242258, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4107489585876465, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7040340304374695, + "num_tokens": 455588007.0, + "step": 17616 + }, + { + "epoch": 1.9346584669448714, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.474254846572876, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.723878026008606, + "num_tokens": 455612967.0, + "step": 17617 + }, + { + "epoch": 1.9347682846474852, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4919981956481934, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7012083530426025, + "num_tokens": 455641419.0, + "step": 17618 + }, + { + "epoch": 1.934878102350099, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.847771406173706, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7137618064880371, + "num_tokens": 455663530.0, + "step": 17619 + }, + { + "epoch": 1.9349879200527125, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4696526527404785, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7061209082603455, + "num_tokens": 455690929.0, + "step": 17620 + }, + { + "epoch": 1.935097737755326, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6039633750915527, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.7072576284408569, + "num_tokens": 455716597.0, + "step": 17621 + }, + { + "epoch": 1.9352075554579398, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5425257682800293, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.6912511587142944, + "num_tokens": 455745358.0, + "step": 17622 + }, + { + "epoch": 1.9353173731605535, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5189733505249023, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6911517381668091, + "num_tokens": 455773826.0, + "step": 17623 + }, + { + "epoch": 1.9354271908631673, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5042366981506348, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6889460682868958, + "num_tokens": 455802065.0, + "step": 17624 + }, + { + "epoch": 1.9355370085657808, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4124233722686768, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.705447793006897, + "num_tokens": 455831596.0, + "step": 17625 + }, + { + "epoch": 1.9356468262683943, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.377440929412842, + "learning_rate": 1e-06, + "loss": 1.099, + "mean_token_accuracy": 0.6857219934463501, + "num_tokens": 455860969.0, + "step": 17626 + }, + { + "epoch": 1.935756643971008, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6693503856658936, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7059390544891357, + "num_tokens": 455884957.0, + "step": 17627 + }, + { + "epoch": 1.9358664616736219, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.460232973098755, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6853557825088501, + "num_tokens": 455912229.0, + "step": 17628 + }, + { + "epoch": 1.9359762793762354, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5067927837371826, + "learning_rate": 1e-06, + "loss": 1.106, + "mean_token_accuracy": 0.685492217540741, + "num_tokens": 455939141.0, + "step": 17629 + }, + { + "epoch": 1.9360860970788492, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.326650381088257, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6917693018913269, + "num_tokens": 455970042.0, + "step": 17630 + }, + { + "epoch": 1.9361959147814627, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.655109405517578, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7238448262214661, + "num_tokens": 455991574.0, + "step": 17631 + }, + { + "epoch": 1.9363057324840764, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.758200168609619, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7118119597434998, + "num_tokens": 456014313.0, + "step": 17632 + }, + { + "epoch": 1.9364155501866902, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6115403175354004, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7232616543769836, + "num_tokens": 456038844.0, + "step": 17633 + }, + { + "epoch": 1.9365253678893037, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5535004138946533, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6955825090408325, + "num_tokens": 456065470.0, + "step": 17634 + }, + { + "epoch": 1.9366351855919173, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.408942937850952, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.712404727935791, + "num_tokens": 456094501.0, + "step": 17635 + }, + { + "epoch": 1.936745003294531, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5660312175750732, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7050513625144958, + "num_tokens": 456121203.0, + "step": 17636 + }, + { + "epoch": 1.9368548209971448, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5855445861816406, + "learning_rate": 1e-06, + "loss": 1.1013, + "mean_token_accuracy": 0.6841187477111816, + "num_tokens": 456147979.0, + "step": 17637 + }, + { + "epoch": 1.9369646386997585, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.709726572036743, + "learning_rate": 1e-06, + "loss": 0.8732, + "mean_token_accuracy": 0.7341009974479675, + "num_tokens": 456169219.0, + "step": 17638 + }, + { + "epoch": 1.937074456402372, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.512644052505493, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.6953650116920471, + "num_tokens": 456197959.0, + "step": 17639 + }, + { + "epoch": 1.9371842741049856, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4749882221221924, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7040262818336487, + "num_tokens": 456225101.0, + "step": 17640 + }, + { + "epoch": 1.9372940918075994, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6080596446990967, + "learning_rate": 1e-06, + "loss": 1.0992, + "mean_token_accuracy": 0.6904500722885132, + "num_tokens": 456251271.0, + "step": 17641 + }, + { + "epoch": 1.9374039095102131, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.547787666320801, + "learning_rate": 1e-06, + "loss": 1.0766, + "mean_token_accuracy": 0.6888495087623596, + "num_tokens": 456279573.0, + "step": 17642 + }, + { + "epoch": 1.9375137272128267, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.550337314605713, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.70683354139328, + "num_tokens": 456304106.0, + "step": 17643 + }, + { + "epoch": 1.9376235449154404, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7703521251678467, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.698625922203064, + "num_tokens": 456326000.0, + "step": 17644 + }, + { + "epoch": 1.937733362618054, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4097180366516113, + "learning_rate": 1e-06, + "loss": 1.0745, + "mean_token_accuracy": 0.6895209550857544, + "num_tokens": 456357308.0, + "step": 17645 + }, + { + "epoch": 1.9378431803206677, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.669132709503174, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7004273533821106, + "num_tokens": 456379764.0, + "step": 17646 + }, + { + "epoch": 1.9379529980232815, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6170403957366943, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.718514084815979, + "num_tokens": 456404692.0, + "step": 17647 + }, + { + "epoch": 1.938062815725895, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.704115867614746, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6922774910926819, + "num_tokens": 456429190.0, + "step": 17648 + }, + { + "epoch": 1.9381726334285085, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.594214916229248, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.704723060131073, + "num_tokens": 456459855.0, + "step": 17649 + }, + { + "epoch": 1.9382824511311223, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5640318393707275, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6999709606170654, + "num_tokens": 456489508.0, + "step": 17650 + }, + { + "epoch": 1.938392268833736, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.769339084625244, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6924120187759399, + "num_tokens": 456514346.0, + "step": 17651 + }, + { + "epoch": 1.9385020865363498, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2999064922332764, + "learning_rate": 1e-06, + "loss": 1.0794, + "mean_token_accuracy": 0.6826459169387817, + "num_tokens": 456547450.0, + "step": 17652 + }, + { + "epoch": 1.9386119042389633, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.988203525543213, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7328174710273743, + "num_tokens": 456565621.0, + "step": 17653 + }, + { + "epoch": 1.9387217219415769, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.929222822189331, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6950131058692932, + "num_tokens": 456585911.0, + "step": 17654 + }, + { + "epoch": 1.9388315396441906, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4496958255767822, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.6893158555030823, + "num_tokens": 456614068.0, + "step": 17655 + }, + { + "epoch": 1.9389413573468044, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.364058494567871, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6981126070022583, + "num_tokens": 456646528.0, + "step": 17656 + }, + { + "epoch": 1.939051175049418, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.584671974182129, + "learning_rate": 1e-06, + "loss": 1.0962, + "mean_token_accuracy": 0.6885700225830078, + "num_tokens": 456672194.0, + "step": 17657 + }, + { + "epoch": 1.9391609927520315, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.879297971725464, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7032356262207031, + "num_tokens": 456692698.0, + "step": 17658 + }, + { + "epoch": 1.9392708104546452, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5790209770202637, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7035194635391235, + "num_tokens": 456715664.0, + "step": 17659 + }, + { + "epoch": 1.939380628157259, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.514941453933716, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7081798315048218, + "num_tokens": 456741854.0, + "step": 17660 + }, + { + "epoch": 1.9394904458598727, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.505992889404297, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6974030137062073, + "num_tokens": 456768364.0, + "step": 17661 + }, + { + "epoch": 1.9396002635624863, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.435885190963745, + "learning_rate": 1e-06, + "loss": 1.135, + "mean_token_accuracy": 0.6811519861221313, + "num_tokens": 456799532.0, + "step": 17662 + }, + { + "epoch": 1.9397100812650998, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.623929500579834, + "learning_rate": 1e-06, + "loss": 0.9336, + "mean_token_accuracy": 0.7286906838417053, + "num_tokens": 456823868.0, + "step": 17663 + }, + { + "epoch": 1.9398198989677136, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.0714759826660156, + "learning_rate": 1e-06, + "loss": 0.9273, + "mean_token_accuracy": 0.7353564500808716, + "num_tokens": 456843968.0, + "step": 17664 + }, + { + "epoch": 1.9399297166703273, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9054527282714844, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7190191745758057, + "num_tokens": 456862883.0, + "step": 17665 + }, + { + "epoch": 1.940039534372941, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.757728099822998, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7019658088684082, + "num_tokens": 456886445.0, + "step": 17666 + }, + { + "epoch": 1.9401493520755546, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.001594305038452, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7089350819587708, + "num_tokens": 456908085.0, + "step": 17667 + }, + { + "epoch": 1.9402591697781681, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.526545763015747, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7122334837913513, + "num_tokens": 456937381.0, + "step": 17668 + }, + { + "epoch": 1.940368987480782, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.579681873321533, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7160854339599609, + "num_tokens": 456961903.0, + "step": 17669 + }, + { + "epoch": 1.9404788051833957, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.661442995071411, + "learning_rate": 1e-06, + "loss": 1.1246, + "mean_token_accuracy": 0.6801378130912781, + "num_tokens": 456986959.0, + "step": 17670 + }, + { + "epoch": 1.9405886228860092, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8413519859313965, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7000909447669983, + "num_tokens": 457006516.0, + "step": 17671 + }, + { + "epoch": 1.9406984405886227, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.3371880054473877, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7159112691879272, + "num_tokens": 457038982.0, + "step": 17672 + }, + { + "epoch": 1.9408082582912365, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5832016468048096, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7144492864608765, + "num_tokens": 457070470.0, + "step": 17673 + }, + { + "epoch": 1.9409180759938502, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.8779079914093018, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7245911359786987, + "num_tokens": 457090603.0, + "step": 17674 + }, + { + "epoch": 1.941027893696464, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6729159355163574, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7066106796264648, + "num_tokens": 457116720.0, + "step": 17675 + }, + { + "epoch": 1.9411377113990775, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.680717945098877, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7235709428787231, + "num_tokens": 457138482.0, + "step": 17676 + }, + { + "epoch": 1.941247529101691, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.504833221435547, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7148187756538391, + "num_tokens": 457162724.0, + "step": 17677 + }, + { + "epoch": 1.9413573468043048, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5791890621185303, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7197723388671875, + "num_tokens": 457187631.0, + "step": 17678 + }, + { + "epoch": 1.9414671645069186, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4045252799987793, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7035250663757324, + "num_tokens": 457217336.0, + "step": 17679 + }, + { + "epoch": 1.9415769822095323, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.718862533569336, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7220385074615479, + "num_tokens": 457239708.0, + "step": 17680 + }, + { + "epoch": 1.9416867999121459, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.33888840675354, + "learning_rate": 1e-06, + "loss": 1.1147, + "mean_token_accuracy": 0.682703971862793, + "num_tokens": 457273370.0, + "step": 17681 + }, + { + "epoch": 1.9417966176147594, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4525389671325684, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.7071372270584106, + "num_tokens": 457303652.0, + "step": 17682 + }, + { + "epoch": 1.9419064353173732, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6543197631835938, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.7053172588348389, + "num_tokens": 457329030.0, + "step": 17683 + }, + { + "epoch": 1.942016253019987, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.076748847961426, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.7182127237319946, + "num_tokens": 457346283.0, + "step": 17684 + }, + { + "epoch": 1.9421260707226005, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.139439344406128, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7122165560722351, + "num_tokens": 457366083.0, + "step": 17685 + }, + { + "epoch": 1.942235888425214, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4323055744171143, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6886329650878906, + "num_tokens": 457397897.0, + "step": 17686 + }, + { + "epoch": 1.9423457061278278, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6171839237213135, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7035304307937622, + "num_tokens": 457422345.0, + "step": 17687 + }, + { + "epoch": 1.9424555238304415, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.536322832107544, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7130482792854309, + "num_tokens": 457448229.0, + "step": 17688 + }, + { + "epoch": 1.9425653415330553, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4331624507904053, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.7019094228744507, + "num_tokens": 457477440.0, + "step": 17689 + }, + { + "epoch": 1.9426751592356688, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.493161678314209, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6912314891815186, + "num_tokens": 457505166.0, + "step": 17690 + }, + { + "epoch": 1.9427849769382823, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2810332775115967, + "learning_rate": 1e-06, + "loss": 1.1141, + "mean_token_accuracy": 0.6890949606895447, + "num_tokens": 457536226.0, + "step": 17691 + }, + { + "epoch": 1.942894794640896, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3597729206085205, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.7030118703842163, + "num_tokens": 457565023.0, + "step": 17692 + }, + { + "epoch": 1.9430046123435099, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4225692749023438, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7133113145828247, + "num_tokens": 457592425.0, + "step": 17693 + }, + { + "epoch": 1.9431144300461234, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4661271572113037, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6938539743423462, + "num_tokens": 457621475.0, + "step": 17694 + }, + { + "epoch": 1.9432242477487371, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3562607765197754, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6949107050895691, + "num_tokens": 457652399.0, + "step": 17695 + }, + { + "epoch": 1.9433340654513507, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6924362182617188, + "learning_rate": 1e-06, + "loss": 1.1428, + "mean_token_accuracy": 0.6703262329101562, + "num_tokens": 457680403.0, + "step": 17696 + }, + { + "epoch": 1.9434438831539644, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.536865234375, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.7018365859985352, + "num_tokens": 457706330.0, + "step": 17697 + }, + { + "epoch": 1.9435537008565782, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.714921236038208, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6830121278762817, + "num_tokens": 457728851.0, + "step": 17698 + }, + { + "epoch": 1.9436635185591917, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9623312950134277, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7093071341514587, + "num_tokens": 457749633.0, + "step": 17699 + }, + { + "epoch": 1.9437733362618053, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7118372917175293, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7036237716674805, + "num_tokens": 457774304.0, + "step": 17700 + }, + { + "epoch": 1.943883153964419, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5774643421173096, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.730661153793335, + "num_tokens": 457801445.0, + "step": 17701 + }, + { + "epoch": 1.9439929716670328, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5441439151763916, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7132924199104309, + "num_tokens": 457825155.0, + "step": 17702 + }, + { + "epoch": 1.9441027893696465, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.544635772705078, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.703639566898346, + "num_tokens": 457852221.0, + "step": 17703 + }, + { + "epoch": 1.94421260707226, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5279324054718018, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6992652416229248, + "num_tokens": 457879946.0, + "step": 17704 + }, + { + "epoch": 1.9443224247748736, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5101161003112793, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.7052497267723083, + "num_tokens": 457910546.0, + "step": 17705 + }, + { + "epoch": 1.9444322424774874, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6902055740356445, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7173280119895935, + "num_tokens": 457933478.0, + "step": 17706 + }, + { + "epoch": 1.9445420601801011, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3756160736083984, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6992260217666626, + "num_tokens": 457964229.0, + "step": 17707 + }, + { + "epoch": 1.9446518778827147, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.532247543334961, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.6942797899246216, + "num_tokens": 457988781.0, + "step": 17708 + }, + { + "epoch": 1.9447616955853284, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.633802652359009, + "learning_rate": 1e-06, + "loss": 1.0811, + "mean_token_accuracy": 0.6916851997375488, + "num_tokens": 458016238.0, + "step": 17709 + }, + { + "epoch": 1.944871513287942, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.373504638671875, + "learning_rate": 1e-06, + "loss": 1.1089, + "mean_token_accuracy": 0.680629551410675, + "num_tokens": 458047596.0, + "step": 17710 + }, + { + "epoch": 1.9449813309905557, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.482774019241333, + "learning_rate": 1e-06, + "loss": 1.108, + "mean_token_accuracy": 0.6866865754127502, + "num_tokens": 458077736.0, + "step": 17711 + }, + { + "epoch": 1.9450911486931695, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.832472801208496, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.7023819088935852, + "num_tokens": 458102156.0, + "step": 17712 + }, + { + "epoch": 1.945200966395783, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.436845302581787, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7118016481399536, + "num_tokens": 458132768.0, + "step": 17713 + }, + { + "epoch": 1.9453107840983965, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.896916389465332, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6905349493026733, + "num_tokens": 458159677.0, + "step": 17714 + }, + { + "epoch": 1.9454206018010103, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.535017251968384, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.7044060230255127, + "num_tokens": 458184722.0, + "step": 17715 + }, + { + "epoch": 1.945530419503624, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.262707471847534, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7051305770874023, + "num_tokens": 458217156.0, + "step": 17716 + }, + { + "epoch": 1.9456402372062378, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.747591257095337, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7062557935714722, + "num_tokens": 458238832.0, + "step": 17717 + }, + { + "epoch": 1.9457500549088513, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6704354286193848, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6960539817810059, + "num_tokens": 458262370.0, + "step": 17718 + }, + { + "epoch": 1.9458598726114649, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9435505867004395, + "learning_rate": 1e-06, + "loss": 0.9157, + "mean_token_accuracy": 0.7254140377044678, + "num_tokens": 458281145.0, + "step": 17719 + }, + { + "epoch": 1.9459696903140786, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.655609369277954, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7054352164268494, + "num_tokens": 458308876.0, + "step": 17720 + }, + { + "epoch": 1.9460795080166924, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.355818033218384, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6886810064315796, + "num_tokens": 458338453.0, + "step": 17721 + }, + { + "epoch": 1.946189325719306, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.489724636077881, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6927657127380371, + "num_tokens": 458364645.0, + "step": 17722 + }, + { + "epoch": 1.9462991434219195, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5864686965942383, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6931816339492798, + "num_tokens": 458389238.0, + "step": 17723 + }, + { + "epoch": 1.9464089611245332, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.712035894393921, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7029837369918823, + "num_tokens": 458412718.0, + "step": 17724 + }, + { + "epoch": 1.946518778827147, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4271624088287354, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7218292355537415, + "num_tokens": 458440883.0, + "step": 17725 + }, + { + "epoch": 1.9466285965297607, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3871161937713623, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.6979827880859375, + "num_tokens": 458471234.0, + "step": 17726 + }, + { + "epoch": 1.9467384142323743, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.914952039718628, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.6994489431381226, + "num_tokens": 458491313.0, + "step": 17727 + }, + { + "epoch": 1.9468482319349878, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.3269922733306885, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7135981917381287, + "num_tokens": 458521269.0, + "step": 17728 + }, + { + "epoch": 1.9469580496376016, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4913792610168457, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6923319101333618, + "num_tokens": 458549404.0, + "step": 17729 + }, + { + "epoch": 1.9470678673402153, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.579470634460449, + "learning_rate": 1e-06, + "loss": 1.0727, + "mean_token_accuracy": 0.6904841661453247, + "num_tokens": 458576406.0, + "step": 17730 + }, + { + "epoch": 1.947177685042829, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.0616183280944824, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7082865238189697, + "num_tokens": 458601780.0, + "step": 17731 + }, + { + "epoch": 1.9472875027454426, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7235965728759766, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7210395336151123, + "num_tokens": 458626965.0, + "step": 17732 + }, + { + "epoch": 1.9473973204480561, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6588690280914307, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7217360138893127, + "num_tokens": 458652567.0, + "step": 17733 + }, + { + "epoch": 1.94750713815067, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.86333966255188, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6978355646133423, + "num_tokens": 458678063.0, + "step": 17734 + }, + { + "epoch": 1.9476169558532836, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.684906244277954, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7155123949050903, + "num_tokens": 458703002.0, + "step": 17735 + }, + { + "epoch": 1.9477267735558972, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.65592098236084, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7137562036514282, + "num_tokens": 458726273.0, + "step": 17736 + }, + { + "epoch": 1.9478365912585107, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7079975605010986, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7066459059715271, + "num_tokens": 458748947.0, + "step": 17737 + }, + { + "epoch": 1.9479464089611245, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.945641279220581, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7044533491134644, + "num_tokens": 458773159.0, + "step": 17738 + }, + { + "epoch": 1.9480562266637382, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.649991035461426, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6983405947685242, + "num_tokens": 458802385.0, + "step": 17739 + }, + { + "epoch": 1.948166044366352, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.72453236579895, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7085862159729004, + "num_tokens": 458828374.0, + "step": 17740 + }, + { + "epoch": 1.9482758620689655, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.590930938720703, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6926769018173218, + "num_tokens": 458852194.0, + "step": 17741 + }, + { + "epoch": 1.948385679771579, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.455482006072998, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7242677211761475, + "num_tokens": 458879585.0, + "step": 17742 + }, + { + "epoch": 1.9484954974741928, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3288991451263428, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.7045962810516357, + "num_tokens": 458910244.0, + "step": 17743 + }, + { + "epoch": 1.9486053151768066, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4436142444610596, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7017096281051636, + "num_tokens": 458937185.0, + "step": 17744 + }, + { + "epoch": 1.94871513287942, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3729608058929443, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7259868383407593, + "num_tokens": 458963727.0, + "step": 17745 + }, + { + "epoch": 1.9488249505820339, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3900420665740967, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6890676021575928, + "num_tokens": 458996251.0, + "step": 17746 + }, + { + "epoch": 1.9489347682846474, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.314707040786743, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7014208436012268, + "num_tokens": 459027137.0, + "step": 17747 + }, + { + "epoch": 1.9490445859872612, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6398959159851074, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7207357287406921, + "num_tokens": 459050086.0, + "step": 17748 + }, + { + "epoch": 1.949154403689875, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6886746883392334, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7196431159973145, + "num_tokens": 459073889.0, + "step": 17749 + }, + { + "epoch": 1.9492642213924884, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.610499143600464, + "learning_rate": 1e-06, + "loss": 1.0928, + "mean_token_accuracy": 0.6983228921890259, + "num_tokens": 459100458.0, + "step": 17750 + }, + { + "epoch": 1.949374039095102, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.933516502380371, + "learning_rate": 1e-06, + "loss": 0.9144, + "mean_token_accuracy": 0.7215348482131958, + "num_tokens": 459119890.0, + "step": 17751 + }, + { + "epoch": 1.9494838567977157, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6612937450408936, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6929172873497009, + "num_tokens": 459147450.0, + "step": 17752 + }, + { + "epoch": 1.9495936745003295, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5123867988586426, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.7341275215148926, + "num_tokens": 459172636.0, + "step": 17753 + }, + { + "epoch": 1.9497034922029433, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.504254102706909, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.6934157609939575, + "num_tokens": 459200885.0, + "step": 17754 + }, + { + "epoch": 1.9498133099055568, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3151426315307617, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7168468832969666, + "num_tokens": 459229408.0, + "step": 17755 + }, + { + "epoch": 1.9499231276081703, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.612316370010376, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7252404689788818, + "num_tokens": 459254265.0, + "step": 17756 + }, + { + "epoch": 1.950032945310784, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6594796180725098, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6958403587341309, + "num_tokens": 459280332.0, + "step": 17757 + }, + { + "epoch": 1.9501427630133978, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5807251930236816, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7092535495758057, + "num_tokens": 459305595.0, + "step": 17758 + }, + { + "epoch": 1.9502525807160114, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7592737674713135, + "learning_rate": 1e-06, + "loss": 0.9211, + "mean_token_accuracy": 0.729812502861023, + "num_tokens": 459325299.0, + "step": 17759 + }, + { + "epoch": 1.9503623984186251, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.644841194152832, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.7031141519546509, + "num_tokens": 459352836.0, + "step": 17760 + }, + { + "epoch": 1.9504722161212387, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6263389587402344, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.691046416759491, + "num_tokens": 459379227.0, + "step": 17761 + }, + { + "epoch": 1.9505820338238524, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.520150899887085, + "learning_rate": 1e-06, + "loss": 1.0898, + "mean_token_accuracy": 0.6871645450592041, + "num_tokens": 459411692.0, + "step": 17762 + }, + { + "epoch": 1.9506918515264662, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.295379877090454, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7129451036453247, + "num_tokens": 459445915.0, + "step": 17763 + }, + { + "epoch": 1.9508016692290797, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6663944721221924, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6871554255485535, + "num_tokens": 459471396.0, + "step": 17764 + }, + { + "epoch": 1.9509114869316933, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.8856654167175293, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.7067012190818787, + "num_tokens": 459493047.0, + "step": 17765 + }, + { + "epoch": 1.951021304634307, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8799703121185303, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7101244926452637, + "num_tokens": 459515684.0, + "step": 17766 + }, + { + "epoch": 1.9511311223369208, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6543967723846436, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7134534120559692, + "num_tokens": 459540503.0, + "step": 17767 + }, + { + "epoch": 1.9512409400395345, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5142769813537598, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7020133137702942, + "num_tokens": 459569352.0, + "step": 17768 + }, + { + "epoch": 1.951350757742148, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7861053943634033, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7291458249092102, + "num_tokens": 459591727.0, + "step": 17769 + }, + { + "epoch": 1.9514605754447616, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7991178035736084, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7195385694503784, + "num_tokens": 459613008.0, + "step": 17770 + }, + { + "epoch": 1.9515703931473753, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.646486759185791, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7132648229598999, + "num_tokens": 459640519.0, + "step": 17771 + }, + { + "epoch": 1.951680210849989, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4848384857177734, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.7012941837310791, + "num_tokens": 459668946.0, + "step": 17772 + }, + { + "epoch": 1.9517900285526026, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.222903251647949, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7132903337478638, + "num_tokens": 459701833.0, + "step": 17773 + }, + { + "epoch": 1.9518998462552162, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4328904151916504, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7162765264511108, + "num_tokens": 459729092.0, + "step": 17774 + }, + { + "epoch": 1.95200966395783, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6486661434173584, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.692208468914032, + "num_tokens": 459755210.0, + "step": 17775 + }, + { + "epoch": 1.9521194816604437, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.351417303085327, + "learning_rate": 1e-06, + "loss": 1.1125, + "mean_token_accuracy": 0.6822412014007568, + "num_tokens": 459787978.0, + "step": 17776 + }, + { + "epoch": 1.9522292993630574, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5182993412017822, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7163563966751099, + "num_tokens": 459814198.0, + "step": 17777 + }, + { + "epoch": 1.952339117065671, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4810733795166016, + "learning_rate": 1e-06, + "loss": 0.8916, + "mean_token_accuracy": 0.7342938184738159, + "num_tokens": 459843631.0, + "step": 17778 + }, + { + "epoch": 1.9524489347682845, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7623603343963623, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7141160368919373, + "num_tokens": 459865903.0, + "step": 17779 + }, + { + "epoch": 1.9525587524708983, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7249600887298584, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6872820258140564, + "num_tokens": 459890911.0, + "step": 17780 + }, + { + "epoch": 1.952668570173512, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4820406436920166, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7187837362289429, + "num_tokens": 459916634.0, + "step": 17781 + }, + { + "epoch": 1.9527783878761258, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5880560874938965, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6922311782836914, + "num_tokens": 459945596.0, + "step": 17782 + }, + { + "epoch": 1.9528882055787393, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.628624439239502, + "learning_rate": 1e-06, + "loss": 1.1283, + "mean_token_accuracy": 0.6729483604431152, + "num_tokens": 459971509.0, + "step": 17783 + }, + { + "epoch": 1.9529980232813529, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7824172973632812, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7029749155044556, + "num_tokens": 459995866.0, + "step": 17784 + }, + { + "epoch": 1.9531078409839666, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7391750812530518, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.7044487595558167, + "num_tokens": 460019422.0, + "step": 17785 + }, + { + "epoch": 1.9532176586865804, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.408625602722168, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6958932876586914, + "num_tokens": 460051120.0, + "step": 17786 + }, + { + "epoch": 1.953327476389194, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.490079641342163, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.714866042137146, + "num_tokens": 460075998.0, + "step": 17787 + }, + { + "epoch": 1.9534372940918074, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5001039505004883, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7255364656448364, + "num_tokens": 460102115.0, + "step": 17788 + }, + { + "epoch": 1.9535471117944212, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.806976556777954, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7019677758216858, + "num_tokens": 460124918.0, + "step": 17789 + }, + { + "epoch": 1.953656929497035, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.588282585144043, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6958069801330566, + "num_tokens": 460151773.0, + "step": 17790 + }, + { + "epoch": 1.9537667471996487, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.50400972366333, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6999582052230835, + "num_tokens": 460180962.0, + "step": 17791 + }, + { + "epoch": 1.9538765649022622, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.716820240020752, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.6817334294319153, + "num_tokens": 460204462.0, + "step": 17792 + }, + { + "epoch": 1.9539863826048758, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4889588356018066, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6885616779327393, + "num_tokens": 460231154.0, + "step": 17793 + }, + { + "epoch": 1.9540962003074895, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7912914752960205, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7213298082351685, + "num_tokens": 460251325.0, + "step": 17794 + }, + { + "epoch": 1.9542060180101033, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5765814781188965, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6985254883766174, + "num_tokens": 460275225.0, + "step": 17795 + }, + { + "epoch": 1.954315835712717, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.769209861755371, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7174165844917297, + "num_tokens": 460300141.0, + "step": 17796 + }, + { + "epoch": 1.9544256534153306, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.1409387588500977, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7153816819190979, + "num_tokens": 460321850.0, + "step": 17797 + }, + { + "epoch": 1.9545354711179441, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3558928966522217, + "learning_rate": 1e-06, + "loss": 0.9471, + "mean_token_accuracy": 0.723936140537262, + "num_tokens": 460353093.0, + "step": 17798 + }, + { + "epoch": 1.9546452888205579, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.614112615585327, + "learning_rate": 1e-06, + "loss": 1.1638, + "mean_token_accuracy": 0.673917829990387, + "num_tokens": 460380598.0, + "step": 17799 + }, + { + "epoch": 1.9547551065231716, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.604154348373413, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.6994951367378235, + "num_tokens": 460405826.0, + "step": 17800 + }, + { + "epoch": 1.9548649242257852, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.63857364654541, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7096942663192749, + "num_tokens": 460429248.0, + "step": 17801 + }, + { + "epoch": 1.9549747419283987, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5378081798553467, + "learning_rate": 1e-06, + "loss": 0.9356, + "mean_token_accuracy": 0.7235250473022461, + "num_tokens": 460453280.0, + "step": 17802 + }, + { + "epoch": 1.9550845596310125, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.631962776184082, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7116267681121826, + "num_tokens": 460476265.0, + "step": 17803 + }, + { + "epoch": 1.9551943773336262, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3886194229125977, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.7029871344566345, + "num_tokens": 460506821.0, + "step": 17804 + }, + { + "epoch": 1.95530419503624, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4174680709838867, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.689690351486206, + "num_tokens": 460535198.0, + "step": 17805 + }, + { + "epoch": 1.9554140127388535, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6916158199310303, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6931878328323364, + "num_tokens": 460559697.0, + "step": 17806 + }, + { + "epoch": 1.955523830441467, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.438983917236328, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6954292058944702, + "num_tokens": 460587873.0, + "step": 17807 + }, + { + "epoch": 1.9556336481440808, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4856112003326416, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6842519044876099, + "num_tokens": 460613492.0, + "step": 17808 + }, + { + "epoch": 1.9557434658466946, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.77453875541687, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7269850373268127, + "num_tokens": 460633636.0, + "step": 17809 + }, + { + "epoch": 1.955853283549308, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.472923755645752, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7166121006011963, + "num_tokens": 460660312.0, + "step": 17810 + }, + { + "epoch": 1.9559631012519219, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.7674193382263184, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7074668407440186, + "num_tokens": 460678619.0, + "step": 17811 + }, + { + "epoch": 1.9560729189545354, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.698075532913208, + "learning_rate": 1e-06, + "loss": 0.9429, + "mean_token_accuracy": 0.7191872596740723, + "num_tokens": 460699511.0, + "step": 17812 + }, + { + "epoch": 1.9561827366571491, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5554678440093994, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7110400199890137, + "num_tokens": 460726563.0, + "step": 17813 + }, + { + "epoch": 1.956292554359763, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.522799491882324, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.723594069480896, + "num_tokens": 460755421.0, + "step": 17814 + }, + { + "epoch": 1.9564023720623764, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.465834379196167, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.708114743232727, + "num_tokens": 460781504.0, + "step": 17815 + }, + { + "epoch": 1.95651218976499, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.950852155685425, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7110382318496704, + "num_tokens": 460808805.0, + "step": 17816 + }, + { + "epoch": 1.9566220074676037, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4376699924468994, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6990116238594055, + "num_tokens": 460836943.0, + "step": 17817 + }, + { + "epoch": 1.9567318251702175, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.440253973007202, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6938384175300598, + "num_tokens": 460863149.0, + "step": 17818 + }, + { + "epoch": 1.9568416428728312, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.444854736328125, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7179419994354248, + "num_tokens": 460891102.0, + "step": 17819 + }, + { + "epoch": 1.9569514605754448, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.528059720993042, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6929998397827148, + "num_tokens": 460917830.0, + "step": 17820 + }, + { + "epoch": 1.9570612782780583, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.454317808151245, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6986867189407349, + "num_tokens": 460943706.0, + "step": 17821 + }, + { + "epoch": 1.957171095980672, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.316263198852539, + "learning_rate": 1e-06, + "loss": 0.9421, + "mean_token_accuracy": 0.7205588817596436, + "num_tokens": 460970926.0, + "step": 17822 + }, + { + "epoch": 1.9572809136832858, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5019049644470215, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7155680060386658, + "num_tokens": 460995967.0, + "step": 17823 + }, + { + "epoch": 1.9573907313858994, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.404597043991089, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6997069120407104, + "num_tokens": 461026643.0, + "step": 17824 + }, + { + "epoch": 1.9575005490885131, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5966267585754395, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.694018542766571, + "num_tokens": 461052393.0, + "step": 17825 + }, + { + "epoch": 1.9576103667911267, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.379647970199585, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6978441476821899, + "num_tokens": 461083688.0, + "step": 17826 + }, + { + "epoch": 1.9577201844937404, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.710365056991577, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7043046951293945, + "num_tokens": 461106965.0, + "step": 17827 + }, + { + "epoch": 1.9578300021963542, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.537565231323242, + "learning_rate": 1e-06, + "loss": 1.1076, + "mean_token_accuracy": 0.6763496398925781, + "num_tokens": 461134676.0, + "step": 17828 + }, + { + "epoch": 1.9579398198989677, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5101563930511475, + "learning_rate": 1e-06, + "loss": 1.1085, + "mean_token_accuracy": 0.6849504709243774, + "num_tokens": 461163256.0, + "step": 17829 + }, + { + "epoch": 1.9580496376015812, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.883549213409424, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7240495681762695, + "num_tokens": 461186258.0, + "step": 17830 + }, + { + "epoch": 1.958159455304195, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3170716762542725, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6916059851646423, + "num_tokens": 461218074.0, + "step": 17831 + }, + { + "epoch": 1.9582692730068088, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5075368881225586, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7123321294784546, + "num_tokens": 461243234.0, + "step": 17832 + }, + { + "epoch": 1.9583790907094225, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.429713249206543, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6991865634918213, + "num_tokens": 461272954.0, + "step": 17833 + }, + { + "epoch": 1.958488908412036, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.727665901184082, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7257266044616699, + "num_tokens": 461295956.0, + "step": 17834 + }, + { + "epoch": 1.9585987261146496, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4782586097717285, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7040084600448608, + "num_tokens": 461323378.0, + "step": 17835 + }, + { + "epoch": 1.9587085438172633, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7905380725860596, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6981545090675354, + "num_tokens": 461349861.0, + "step": 17836 + }, + { + "epoch": 1.958818361519877, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.799764633178711, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7157289385795593, + "num_tokens": 461370899.0, + "step": 17837 + }, + { + "epoch": 1.9589281792224906, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.50549578666687, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.721060037612915, + "num_tokens": 461397994.0, + "step": 17838 + }, + { + "epoch": 1.9590379969251042, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8032875061035156, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7079274654388428, + "num_tokens": 461421247.0, + "step": 17839 + }, + { + "epoch": 1.959147814627718, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.772059679031372, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.7014445662498474, + "num_tokens": 461443922.0, + "step": 17840 + }, + { + "epoch": 1.9592576323303317, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7893948554992676, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7099595069885254, + "num_tokens": 461466502.0, + "step": 17841 + }, + { + "epoch": 1.9593674500329454, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.335617780685425, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6866099834442139, + "num_tokens": 461502352.0, + "step": 17842 + }, + { + "epoch": 1.959477267735559, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.682837963104248, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7139806747436523, + "num_tokens": 461526845.0, + "step": 17843 + }, + { + "epoch": 1.9595870854381725, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.457380533218384, + "learning_rate": 1e-06, + "loss": 0.9283, + "mean_token_accuracy": 0.7243907451629639, + "num_tokens": 461552144.0, + "step": 17844 + }, + { + "epoch": 1.9596969031407863, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8728389739990234, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7102600932121277, + "num_tokens": 461581473.0, + "step": 17845 + }, + { + "epoch": 1.9598067208434, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4720733165740967, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7165802717208862, + "num_tokens": 461607343.0, + "step": 17846 + }, + { + "epoch": 1.9599165385460138, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4552130699157715, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7021059393882751, + "num_tokens": 461634676.0, + "step": 17847 + }, + { + "epoch": 1.9600263562486273, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.571371555328369, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6923315525054932, + "num_tokens": 461667066.0, + "step": 17848 + }, + { + "epoch": 1.9601361739512408, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8429508209228516, + "learning_rate": 1e-06, + "loss": 0.8261, + "mean_token_accuracy": 0.7530161142349243, + "num_tokens": 461687263.0, + "step": 17849 + }, + { + "epoch": 1.9602459916538546, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4921798706054688, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7247138023376465, + "num_tokens": 461712768.0, + "step": 17850 + }, + { + "epoch": 1.9603558093564684, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.528120756149292, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.686812698841095, + "num_tokens": 461738675.0, + "step": 17851 + }, + { + "epoch": 1.960465627059082, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.561526298522949, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7051456570625305, + "num_tokens": 461764686.0, + "step": 17852 + }, + { + "epoch": 1.9605754447616954, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.528235912322998, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7043510675430298, + "num_tokens": 461792000.0, + "step": 17853 + }, + { + "epoch": 1.9606852624643092, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.837127685546875, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.701418399810791, + "num_tokens": 461815714.0, + "step": 17854 + }, + { + "epoch": 1.960795080166923, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5291922092437744, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7029225826263428, + "num_tokens": 461840276.0, + "step": 17855 + }, + { + "epoch": 1.9609048978695367, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.304349184036255, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6823112368583679, + "num_tokens": 461871940.0, + "step": 17856 + }, + { + "epoch": 1.9610147155721502, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.786020040512085, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.722053050994873, + "num_tokens": 461900224.0, + "step": 17857 + }, + { + "epoch": 1.9611245332747638, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.608003854751587, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7073098421096802, + "num_tokens": 461925930.0, + "step": 17858 + }, + { + "epoch": 1.9612343509773775, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4909214973449707, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7103444933891296, + "num_tokens": 461951480.0, + "step": 17859 + }, + { + "epoch": 1.9613441686799913, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5877859592437744, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7044941186904907, + "num_tokens": 461975796.0, + "step": 17860 + }, + { + "epoch": 1.961453986382605, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.872203826904297, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7053977847099304, + "num_tokens": 461996344.0, + "step": 17861 + }, + { + "epoch": 1.9615638040852186, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.468897581100464, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.719472348690033, + "num_tokens": 462023553.0, + "step": 17862 + }, + { + "epoch": 1.9616736217878321, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4755871295928955, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6931284070014954, + "num_tokens": 462049915.0, + "step": 17863 + }, + { + "epoch": 1.9617834394904459, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4734690189361572, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.689998984336853, + "num_tokens": 462079966.0, + "step": 17864 + }, + { + "epoch": 1.9618932571930596, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.809515953063965, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.7024123668670654, + "num_tokens": 462103348.0, + "step": 17865 + }, + { + "epoch": 1.9620030748956732, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4937660694122314, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.7046117782592773, + "num_tokens": 462131788.0, + "step": 17866 + }, + { + "epoch": 1.9621128925982867, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.835693597793579, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7046338319778442, + "num_tokens": 462154304.0, + "step": 17867 + }, + { + "epoch": 1.9622227103009005, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.542178153991699, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7208881378173828, + "num_tokens": 462180987.0, + "step": 17868 + }, + { + "epoch": 1.9623325280035142, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.459764003753662, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7068004012107849, + "num_tokens": 462210465.0, + "step": 17869 + }, + { + "epoch": 1.962442345706128, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5784921646118164, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.7021092176437378, + "num_tokens": 462236827.0, + "step": 17870 + }, + { + "epoch": 1.9625521634087415, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.592236280441284, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7091925740242004, + "num_tokens": 462264662.0, + "step": 17871 + }, + { + "epoch": 1.962661981111355, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7474477291107178, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7085440158843994, + "num_tokens": 462288633.0, + "step": 17872 + }, + { + "epoch": 1.9627717988139688, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.526615858078003, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6958152055740356, + "num_tokens": 462315207.0, + "step": 17873 + }, + { + "epoch": 1.9628816165165826, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.598940849304199, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.6973661184310913, + "num_tokens": 462343361.0, + "step": 17874 + }, + { + "epoch": 1.962991434219196, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.737264394760132, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6988447308540344, + "num_tokens": 462365597.0, + "step": 17875 + }, + { + "epoch": 1.9631012519218098, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5777554512023926, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6878172755241394, + "num_tokens": 462391846.0, + "step": 17876 + }, + { + "epoch": 1.9632110696244234, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.597306489944458, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7057768106460571, + "num_tokens": 462418944.0, + "step": 17877 + }, + { + "epoch": 1.9633208873270371, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6572537422180176, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6978373527526855, + "num_tokens": 462446632.0, + "step": 17878 + }, + { + "epoch": 1.963430705029651, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.673349618911743, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7187570333480835, + "num_tokens": 462471488.0, + "step": 17879 + }, + { + "epoch": 1.9635405227322644, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.399127960205078, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6817606687545776, + "num_tokens": 462499777.0, + "step": 17880 + }, + { + "epoch": 1.963650340434878, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.212070941925049, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7030833959579468, + "num_tokens": 462521390.0, + "step": 17881 + }, + { + "epoch": 1.9637601581374917, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7447407245635986, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7119774222373962, + "num_tokens": 462548402.0, + "step": 17882 + }, + { + "epoch": 1.9638699758401055, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7840192317962646, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7070754766464233, + "num_tokens": 462579022.0, + "step": 17883 + }, + { + "epoch": 1.9639797935427192, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7577385902404785, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.720112681388855, + "num_tokens": 462600350.0, + "step": 17884 + }, + { + "epoch": 1.9640896112453328, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.385140895843506, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7018015384674072, + "num_tokens": 462628836.0, + "step": 17885 + }, + { + "epoch": 1.9641994289479463, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8851125240325928, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.715582013130188, + "num_tokens": 462648404.0, + "step": 17886 + }, + { + "epoch": 1.96430924665056, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.374009370803833, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.716783881187439, + "num_tokens": 462675662.0, + "step": 17887 + }, + { + "epoch": 1.9644190643531738, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5547263622283936, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.6985654234886169, + "num_tokens": 462702373.0, + "step": 17888 + }, + { + "epoch": 1.9645288820557874, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8009097576141357, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7117328643798828, + "num_tokens": 462724181.0, + "step": 17889 + }, + { + "epoch": 1.964638699758401, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5438075065612793, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6972914338111877, + "num_tokens": 462751807.0, + "step": 17890 + }, + { + "epoch": 1.9647485174610146, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.622938632965088, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7038732171058655, + "num_tokens": 462783331.0, + "step": 17891 + }, + { + "epoch": 1.9648583351636284, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4002628326416016, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7192236185073853, + "num_tokens": 462812266.0, + "step": 17892 + }, + { + "epoch": 1.9649681528662422, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.422426462173462, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7101036906242371, + "num_tokens": 462841811.0, + "step": 17893 + }, + { + "epoch": 1.9650779705688557, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5694217681884766, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7061976194381714, + "num_tokens": 462870788.0, + "step": 17894 + }, + { + "epoch": 1.9651877882714692, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.030381441116333, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6846446394920349, + "num_tokens": 462899120.0, + "step": 17895 + }, + { + "epoch": 1.965297605974083, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3837695121765137, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.718828558921814, + "num_tokens": 462928270.0, + "step": 17896 + }, + { + "epoch": 1.9654074236766967, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.284888982772827, + "learning_rate": 1e-06, + "loss": 1.1224, + "mean_token_accuracy": 0.6865088939666748, + "num_tokens": 462960055.0, + "step": 17897 + }, + { + "epoch": 1.9655172413793105, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3903908729553223, + "learning_rate": 1e-06, + "loss": 0.9995, + "mean_token_accuracy": 0.7080150842666626, + "num_tokens": 462989831.0, + "step": 17898 + }, + { + "epoch": 1.965627059081924, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8780086040496826, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6870568990707397, + "num_tokens": 463011219.0, + "step": 17899 + }, + { + "epoch": 1.9657368767845376, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.638671875, + "learning_rate": 1e-06, + "loss": 1.1236, + "mean_token_accuracy": 0.6799188256263733, + "num_tokens": 463047470.0, + "step": 17900 + }, + { + "epoch": 1.9658466944871513, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.39886212348938, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.71023029088974, + "num_tokens": 463068746.0, + "step": 17901 + }, + { + "epoch": 1.965956512189765, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6535420417785645, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7128201723098755, + "num_tokens": 463095132.0, + "step": 17902 + }, + { + "epoch": 1.9660663298923786, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.300539255142212, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7200243473052979, + "num_tokens": 463124970.0, + "step": 17903 + }, + { + "epoch": 1.9661761475949922, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6209919452667236, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7027310729026794, + "num_tokens": 463154144.0, + "step": 17904 + }, + { + "epoch": 1.966285965297606, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.817812919616699, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.7024908661842346, + "num_tokens": 463179158.0, + "step": 17905 + }, + { + "epoch": 1.9663957830002197, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.428074359893799, + "learning_rate": 1e-06, + "loss": 0.8857, + "mean_token_accuracy": 0.73567795753479, + "num_tokens": 463206314.0, + "step": 17906 + }, + { + "epoch": 1.9665056007028334, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9023354053497314, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.7048486471176147, + "num_tokens": 463228266.0, + "step": 17907 + }, + { + "epoch": 1.966615418405447, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.564288854598999, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.6910507082939148, + "num_tokens": 463255744.0, + "step": 17908 + }, + { + "epoch": 1.9667252361080605, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4671549797058105, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6972687244415283, + "num_tokens": 463287159.0, + "step": 17909 + }, + { + "epoch": 1.9668350538106742, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5751636028289795, + "learning_rate": 1e-06, + "loss": 0.8986, + "mean_token_accuracy": 0.7331518530845642, + "num_tokens": 463313671.0, + "step": 17910 + }, + { + "epoch": 1.966944871513288, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.831587553024292, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6999640464782715, + "num_tokens": 463338766.0, + "step": 17911 + }, + { + "epoch": 1.9670546892159018, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.928391695022583, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.709852933883667, + "num_tokens": 463365966.0, + "step": 17912 + }, + { + "epoch": 1.9671645069185153, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6518375873565674, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7048194408416748, + "num_tokens": 463389649.0, + "step": 17913 + }, + { + "epoch": 1.9672743246211288, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.387166738510132, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7022121548652649, + "num_tokens": 463418898.0, + "step": 17914 + }, + { + "epoch": 1.9673841423237426, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.67391300201416, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6968786120414734, + "num_tokens": 463442296.0, + "step": 17915 + }, + { + "epoch": 1.9674939600263563, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4554085731506348, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7017397880554199, + "num_tokens": 463470572.0, + "step": 17916 + }, + { + "epoch": 1.9676037777289699, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.519050359725952, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.7005383968353271, + "num_tokens": 463501542.0, + "step": 17917 + }, + { + "epoch": 1.9677135954315834, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5487022399902344, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6879422664642334, + "num_tokens": 463528562.0, + "step": 17918 + }, + { + "epoch": 1.9678234131341972, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.5091946125030518, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7081195116043091, + "num_tokens": 463557046.0, + "step": 17919 + }, + { + "epoch": 1.967933230836811, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.6061291694641113, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7077958583831787, + "num_tokens": 463581800.0, + "step": 17920 + }, + { + "epoch": 1.9680430485394247, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.601048231124878, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7088341116905212, + "num_tokens": 463607472.0, + "step": 17921 + }, + { + "epoch": 1.9681528662420382, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.56785249710083, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6857417225837708, + "num_tokens": 463635685.0, + "step": 17922 + }, + { + "epoch": 1.9682626839446518, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4985640048980713, + "learning_rate": 1e-06, + "loss": 1.1423, + "mean_token_accuracy": 0.6822182536125183, + "num_tokens": 463667783.0, + "step": 17923 + }, + { + "epoch": 1.9683725016472655, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.4102354049682617, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.7036446928977966, + "num_tokens": 463699725.0, + "step": 17924 + }, + { + "epoch": 1.9684823193498793, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.553105592727661, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7229016423225403, + "num_tokens": 463721407.0, + "step": 17925 + }, + { + "epoch": 1.9685921370524928, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.671118974685669, + "learning_rate": 1e-06, + "loss": 1.1099, + "mean_token_accuracy": 0.694238543510437, + "num_tokens": 463749097.0, + "step": 17926 + }, + { + "epoch": 1.9687019547551066, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.3800811767578125, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7105584740638733, + "num_tokens": 463777892.0, + "step": 17927 + }, + { + "epoch": 1.96881177245772, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.689202070236206, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7098289728164673, + "num_tokens": 463802494.0, + "step": 17928 + }, + { + "epoch": 1.9689215901603339, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.782752752304077, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.6975191235542297, + "num_tokens": 463829345.0, + "step": 17929 + }, + { + "epoch": 1.9690314078629476, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.650684118270874, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6999223232269287, + "num_tokens": 463852750.0, + "step": 17930 + }, + { + "epoch": 1.9691412255655611, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7305092811584473, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7211889028549194, + "num_tokens": 463876341.0, + "step": 17931 + }, + { + "epoch": 1.9692510432681747, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 16.692598342895508, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.679548978805542, + "num_tokens": 463903977.0, + "step": 17932 + }, + { + "epoch": 1.9693608609707884, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4849982261657715, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6917499303817749, + "num_tokens": 463935305.0, + "step": 17933 + }, + { + "epoch": 1.9694706786734022, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4473557472229004, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7099161148071289, + "num_tokens": 463964995.0, + "step": 17934 + }, + { + "epoch": 1.969580496376016, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3811893463134766, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7029954791069031, + "num_tokens": 463995158.0, + "step": 17935 + }, + { + "epoch": 1.9696903140786295, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4159014225006104, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7012101411819458, + "num_tokens": 464022617.0, + "step": 17936 + }, + { + "epoch": 1.969800131781243, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6574230194091797, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7153137922286987, + "num_tokens": 464044568.0, + "step": 17937 + }, + { + "epoch": 1.9699099494838568, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7300398349761963, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7164117097854614, + "num_tokens": 464066711.0, + "step": 17938 + }, + { + "epoch": 1.9700197671864705, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5387587547302246, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6993190050125122, + "num_tokens": 464096915.0, + "step": 17939 + }, + { + "epoch": 1.970129584889084, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6918370723724365, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7021875977516174, + "num_tokens": 464123541.0, + "step": 17940 + }, + { + "epoch": 1.9702394025916978, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.413943290710449, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6986626386642456, + "num_tokens": 464151300.0, + "step": 17941 + }, + { + "epoch": 1.9703492202943114, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4153919219970703, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6968031525611877, + "num_tokens": 464178356.0, + "step": 17942 + }, + { + "epoch": 1.9704590379969251, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5691728591918945, + "learning_rate": 1e-06, + "loss": 1.1177, + "mean_token_accuracy": 0.6810572147369385, + "num_tokens": 464206079.0, + "step": 17943 + }, + { + "epoch": 1.9705688556995389, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.459261417388916, + "learning_rate": 1e-06, + "loss": 1.1166, + "mean_token_accuracy": 0.6771228313446045, + "num_tokens": 464238899.0, + "step": 17944 + }, + { + "epoch": 1.9706786734021524, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4346485137939453, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6965119242668152, + "num_tokens": 464269439.0, + "step": 17945 + }, + { + "epoch": 1.970788491104766, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7440836429595947, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7118310928344727, + "num_tokens": 464292768.0, + "step": 17946 + }, + { + "epoch": 1.9708983088073797, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4323067665100098, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7040259838104248, + "num_tokens": 464322796.0, + "step": 17947 + }, + { + "epoch": 1.9710081265099935, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2837507724761963, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7046963572502136, + "num_tokens": 464353679.0, + "step": 17948 + }, + { + "epoch": 1.9711179442126072, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 3.115839958190918, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.713020920753479, + "num_tokens": 464372732.0, + "step": 17949 + }, + { + "epoch": 1.9712277619152208, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.402555227279663, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7194617986679077, + "num_tokens": 464405373.0, + "step": 17950 + }, + { + "epoch": 1.9713375796178343, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7156054973602295, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6968752145767212, + "num_tokens": 464430443.0, + "step": 17951 + }, + { + "epoch": 1.971447397320448, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3950228691101074, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6871237754821777, + "num_tokens": 464457515.0, + "step": 17952 + }, + { + "epoch": 1.9715572150230618, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.926490068435669, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6995725631713867, + "num_tokens": 464477028.0, + "step": 17953 + }, + { + "epoch": 1.9716670327256753, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.602799892425537, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7000031471252441, + "num_tokens": 464502140.0, + "step": 17954 + }, + { + "epoch": 1.9717768504282889, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.393789052963257, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7132266163825989, + "num_tokens": 464530992.0, + "step": 17955 + }, + { + "epoch": 1.9718866681309026, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3496310710906982, + "learning_rate": 1e-06, + "loss": 1.1156, + "mean_token_accuracy": 0.6773433089256287, + "num_tokens": 464565271.0, + "step": 17956 + }, + { + "epoch": 1.9719964858335164, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4780802726745605, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7326341867446899, + "num_tokens": 464592186.0, + "step": 17957 + }, + { + "epoch": 1.9721063035361301, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5205111503601074, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7134780287742615, + "num_tokens": 464616573.0, + "step": 17958 + }, + { + "epoch": 1.9722161212387437, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6679065227508545, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6875897645950317, + "num_tokens": 464641020.0, + "step": 17959 + }, + { + "epoch": 1.9723259389413572, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.37351131439209, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.722533106803894, + "num_tokens": 464671457.0, + "step": 17960 + }, + { + "epoch": 1.972435756643971, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.292894124984741, + "learning_rate": 1e-06, + "loss": 0.9636, + "mean_token_accuracy": 0.7212452292442322, + "num_tokens": 464701479.0, + "step": 17961 + }, + { + "epoch": 1.9725455743465847, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.437398672103882, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6937415599822998, + "num_tokens": 464730943.0, + "step": 17962 + }, + { + "epoch": 1.9726553920491985, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 16.73897361755371, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6957876682281494, + "num_tokens": 464753566.0, + "step": 17963 + }, + { + "epoch": 1.972765209751812, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7369158267974854, + "learning_rate": 1e-06, + "loss": 1.1073, + "mean_token_accuracy": 0.6840450763702393, + "num_tokens": 464781410.0, + "step": 17964 + }, + { + "epoch": 1.9728750274544256, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.896221876144409, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7119919061660767, + "num_tokens": 464806484.0, + "step": 17965 + }, + { + "epoch": 1.9729848451570393, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.635284662246704, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7007225751876831, + "num_tokens": 464833348.0, + "step": 17966 + }, + { + "epoch": 1.973094662859653, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9091148376464844, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6943002343177795, + "num_tokens": 464856986.0, + "step": 17967 + }, + { + "epoch": 1.9732044805622666, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3072433471679688, + "learning_rate": 1e-06, + "loss": 0.919, + "mean_token_accuracy": 0.7324379682540894, + "num_tokens": 464887986.0, + "step": 17968 + }, + { + "epoch": 1.9733142982648801, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.002457618713379, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7167880535125732, + "num_tokens": 464907957.0, + "step": 17969 + }, + { + "epoch": 1.973424115967494, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5878617763519287, + "learning_rate": 1e-06, + "loss": 0.9397, + "mean_token_accuracy": 0.722944438457489, + "num_tokens": 464933222.0, + "step": 17970 + }, + { + "epoch": 1.9735339336701077, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.614753246307373, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7019971609115601, + "num_tokens": 464959829.0, + "step": 17971 + }, + { + "epoch": 1.9736437513727214, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.611776113510132, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7124389410018921, + "num_tokens": 464983039.0, + "step": 17972 + }, + { + "epoch": 1.973753569075335, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.467358350753784, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.6863244771957397, + "num_tokens": 465012388.0, + "step": 17973 + }, + { + "epoch": 1.9738633867779485, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.700636386871338, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6925675272941589, + "num_tokens": 465042863.0, + "step": 17974 + }, + { + "epoch": 1.9739732044805622, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.586287021636963, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7147328853607178, + "num_tokens": 465066140.0, + "step": 17975 + }, + { + "epoch": 1.974083022183176, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5503270626068115, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.715498685836792, + "num_tokens": 465091795.0, + "step": 17976 + }, + { + "epoch": 1.9741928398857898, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6420862674713135, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.720220148563385, + "num_tokens": 465115843.0, + "step": 17977 + }, + { + "epoch": 1.9743026575884033, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3194901943206787, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7050374746322632, + "num_tokens": 465143631.0, + "step": 17978 + }, + { + "epoch": 1.9744124752910168, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.51440691947937, + "learning_rate": 1e-06, + "loss": 1.1093, + "mean_token_accuracy": 0.6821839809417725, + "num_tokens": 465170006.0, + "step": 17979 + }, + { + "epoch": 1.9745222929936306, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.523589849472046, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7065871953964233, + "num_tokens": 465196759.0, + "step": 17980 + }, + { + "epoch": 1.9746321106962443, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4537816047668457, + "learning_rate": 1e-06, + "loss": 0.9039, + "mean_token_accuracy": 0.7355426549911499, + "num_tokens": 465223186.0, + "step": 17981 + }, + { + "epoch": 1.9747419283988579, + "ewc_loss": 9.715557098388672e-06, + "grad_norm": 2.7303593158721924, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7017508745193481, + "num_tokens": 465244970.0, + "step": 17982 + }, + { + "epoch": 1.9748517461014714, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6048367023468018, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7128514051437378, + "num_tokens": 465267666.0, + "step": 17983 + }, + { + "epoch": 1.9749615638040852, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3349435329437256, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6815305948257446, + "num_tokens": 465300502.0, + "step": 17984 + }, + { + "epoch": 1.975071381506699, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4127893447875977, + "learning_rate": 1e-06, + "loss": 1.1282, + "mean_token_accuracy": 0.6746846437454224, + "num_tokens": 465330538.0, + "step": 17985 + }, + { + "epoch": 1.9751811992093127, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5470571517944336, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.707697331905365, + "num_tokens": 465356579.0, + "step": 17986 + }, + { + "epoch": 1.9752910169119262, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9573304653167725, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.717923104763031, + "num_tokens": 465375455.0, + "step": 17987 + }, + { + "epoch": 1.9754008346145397, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.432183027267456, + "learning_rate": 1e-06, + "loss": 1.1152, + "mean_token_accuracy": 0.6738738417625427, + "num_tokens": 465406089.0, + "step": 17988 + }, + { + "epoch": 1.9755106523171535, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6515421867370605, + "learning_rate": 1e-06, + "loss": 0.938, + "mean_token_accuracy": 0.7262434363365173, + "num_tokens": 465428073.0, + "step": 17989 + }, + { + "epoch": 1.9756204700197673, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4717795848846436, + "learning_rate": 1e-06, + "loss": 1.0523, + "mean_token_accuracy": 0.7007315158843994, + "num_tokens": 465454586.0, + "step": 17990 + }, + { + "epoch": 1.9757302877223808, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.515108346939087, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7047051787376404, + "num_tokens": 465479020.0, + "step": 17991 + }, + { + "epoch": 1.9758401054249946, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7756612300872803, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7129090428352356, + "num_tokens": 465499620.0, + "step": 17992 + }, + { + "epoch": 1.975949923127608, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.775143623352051, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7158503532409668, + "num_tokens": 465523489.0, + "step": 17993 + }, + { + "epoch": 1.9760597408302218, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6804349422454834, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7140852808952332, + "num_tokens": 465550394.0, + "step": 17994 + }, + { + "epoch": 1.9761695585328356, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.611492872238159, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7207425832748413, + "num_tokens": 465574074.0, + "step": 17995 + }, + { + "epoch": 1.9762793762354491, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3018951416015625, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7243587970733643, + "num_tokens": 465610552.0, + "step": 17996 + }, + { + "epoch": 1.9763891939380627, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.480687141418457, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6907488703727722, + "num_tokens": 465638687.0, + "step": 17997 + }, + { + "epoch": 1.9764990116406764, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.93778657913208, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.707696795463562, + "num_tokens": 465659469.0, + "step": 17998 + }, + { + "epoch": 1.9766088293432902, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3175623416900635, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7035328149795532, + "num_tokens": 465690690.0, + "step": 17999 + }, + { + "epoch": 1.976718647045904, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6937618255615234, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7189657092094421, + "num_tokens": 465713521.0, + "step": 18000 + }, + { + "epoch": 1.9768284647485175, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3980023860931396, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6951135396957397, + "num_tokens": 465741454.0, + "step": 18001 + }, + { + "epoch": 1.976938282451131, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6429319381713867, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.7038059234619141, + "num_tokens": 465767905.0, + "step": 18002 + }, + { + "epoch": 1.9770481001537448, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.533968448638916, + "learning_rate": 1e-06, + "loss": 1.116, + "mean_token_accuracy": 0.6780576109886169, + "num_tokens": 465795347.0, + "step": 18003 + }, + { + "epoch": 1.9771579178563585, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.1727969646453857, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7157117128372192, + "num_tokens": 465812978.0, + "step": 18004 + }, + { + "epoch": 1.977267735558972, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5807583332061768, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6997404098510742, + "num_tokens": 465838833.0, + "step": 18005 + }, + { + "epoch": 1.9773775532615858, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5771565437316895, + "learning_rate": 1e-06, + "loss": 0.9278, + "mean_token_accuracy": 0.7225284576416016, + "num_tokens": 465865017.0, + "step": 18006 + }, + { + "epoch": 1.9774873709641994, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.688431978225708, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6959635019302368, + "num_tokens": 465887636.0, + "step": 18007 + }, + { + "epoch": 1.977597188666813, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5887837409973145, + "learning_rate": 1e-06, + "loss": 1.0857, + "mean_token_accuracy": 0.6874618530273438, + "num_tokens": 465917893.0, + "step": 18008 + }, + { + "epoch": 1.9777070063694269, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4920222759246826, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.713474690914154, + "num_tokens": 465946956.0, + "step": 18009 + }, + { + "epoch": 1.9778168240720404, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7529008388519287, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7125906944274902, + "num_tokens": 465971921.0, + "step": 18010 + }, + { + "epoch": 1.977926641774654, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5930585861206055, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7037943601608276, + "num_tokens": 465998156.0, + "step": 18011 + }, + { + "epoch": 1.9780364594772677, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3151867389678955, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7016332149505615, + "num_tokens": 466029346.0, + "step": 18012 + }, + { + "epoch": 1.9781462771798815, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4473371505737305, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7147634029388428, + "num_tokens": 466059698.0, + "step": 18013 + }, + { + "epoch": 1.9782560948824952, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5986130237579346, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6876174211502075, + "num_tokens": 466084530.0, + "step": 18014 + }, + { + "epoch": 1.9783659125851087, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.301523447036743, + "learning_rate": 1e-06, + "loss": 1.1108, + "mean_token_accuracy": 0.6800493001937866, + "num_tokens": 466118536.0, + "step": 18015 + }, + { + "epoch": 1.9784757302877223, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6022112369537354, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7137460708618164, + "num_tokens": 466142907.0, + "step": 18016 + }, + { + "epoch": 1.978585547990336, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.451014757156372, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7097633481025696, + "num_tokens": 466175850.0, + "step": 18017 + }, + { + "epoch": 1.9786953656929498, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.598127841949463, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6853780746459961, + "num_tokens": 466201036.0, + "step": 18018 + }, + { + "epoch": 1.9788051833955633, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7219512462615967, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7166340351104736, + "num_tokens": 466221700.0, + "step": 18019 + }, + { + "epoch": 1.9789150010981769, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.810006618499756, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.706695556640625, + "num_tokens": 466241325.0, + "step": 18020 + }, + { + "epoch": 1.9790248188007906, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.237610101699829, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6975257396697998, + "num_tokens": 466272510.0, + "step": 18021 + }, + { + "epoch": 1.9791346365034044, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.561068058013916, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7048538327217102, + "num_tokens": 466297912.0, + "step": 18022 + }, + { + "epoch": 1.9792444542060181, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5295770168304443, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7016761898994446, + "num_tokens": 466324169.0, + "step": 18023 + }, + { + "epoch": 1.9793542719086317, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.544095993041992, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6963990330696106, + "num_tokens": 466351938.0, + "step": 18024 + }, + { + "epoch": 1.9794640896112452, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5106587409973145, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.6958663463592529, + "num_tokens": 466381266.0, + "step": 18025 + }, + { + "epoch": 1.979573907313859, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.435883045196533, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.7174580097198486, + "num_tokens": 466408555.0, + "step": 18026 + }, + { + "epoch": 1.9796837250164727, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.749262809753418, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7029731273651123, + "num_tokens": 466429737.0, + "step": 18027 + }, + { + "epoch": 1.9797935427190865, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.601494312286377, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7083020806312561, + "num_tokens": 466454995.0, + "step": 18028 + }, + { + "epoch": 1.9799033604217, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3297901153564453, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6920741200447083, + "num_tokens": 466485326.0, + "step": 18029 + }, + { + "epoch": 1.9800131781243135, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.36600661277771, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.6856762170791626, + "num_tokens": 466517105.0, + "step": 18030 + }, + { + "epoch": 1.9801229958269273, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5210213661193848, + "learning_rate": 1e-06, + "loss": 0.9403, + "mean_token_accuracy": 0.7194008827209473, + "num_tokens": 466542822.0, + "step": 18031 + }, + { + "epoch": 1.980232813529541, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.760528802871704, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7115240097045898, + "num_tokens": 466566406.0, + "step": 18032 + }, + { + "epoch": 1.9803426312321546, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.495403289794922, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.7064968347549438, + "num_tokens": 466594203.0, + "step": 18033 + }, + { + "epoch": 1.9804524489347681, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.347012996673584, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7021676898002625, + "num_tokens": 466624685.0, + "step": 18034 + }, + { + "epoch": 1.9805622666373819, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3297312259674072, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.705070972442627, + "num_tokens": 466652801.0, + "step": 18035 + }, + { + "epoch": 1.9806720843399956, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6037535667419434, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.7050204277038574, + "num_tokens": 466678910.0, + "step": 18036 + }, + { + "epoch": 1.9807819020426094, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.47623872756958, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.710712730884552, + "num_tokens": 466707742.0, + "step": 18037 + }, + { + "epoch": 1.980891719745223, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6257128715515137, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6962659358978271, + "num_tokens": 466730741.0, + "step": 18038 + }, + { + "epoch": 1.9810015374478365, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5426785945892334, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6988227963447571, + "num_tokens": 466757928.0, + "step": 18039 + }, + { + "epoch": 1.9811113551504502, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.744351863861084, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7010622024536133, + "num_tokens": 466781909.0, + "step": 18040 + }, + { + "epoch": 1.981221172853064, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6800355911254883, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.6969038844108582, + "num_tokens": 466807173.0, + "step": 18041 + }, + { + "epoch": 1.9813309905556777, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.750995635986328, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.716273844242096, + "num_tokens": 466829869.0, + "step": 18042 + }, + { + "epoch": 1.9814408082582913, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4214954376220703, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7129772901535034, + "num_tokens": 466857696.0, + "step": 18043 + }, + { + "epoch": 1.9815506259609048, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7240118980407715, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7201969623565674, + "num_tokens": 466879323.0, + "step": 18044 + }, + { + "epoch": 1.9816604436635186, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.59909987449646, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7043231129646301, + "num_tokens": 466904558.0, + "step": 18045 + }, + { + "epoch": 1.9817702613661323, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9664833545684814, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7071581482887268, + "num_tokens": 466924408.0, + "step": 18046 + }, + { + "epoch": 1.9818800790687459, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7197928428649902, + "learning_rate": 1e-06, + "loss": 0.898, + "mean_token_accuracy": 0.7330957651138306, + "num_tokens": 466945415.0, + "step": 18047 + }, + { + "epoch": 1.9819898967713594, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.506211280822754, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7086670398712158, + "num_tokens": 466971936.0, + "step": 18048 + }, + { + "epoch": 1.9820997144739732, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5455198287963867, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6990664005279541, + "num_tokens": 466999690.0, + "step": 18049 + }, + { + "epoch": 1.982209532176587, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.791187047958374, + "learning_rate": 1e-06, + "loss": 0.9313, + "mean_token_accuracy": 0.7303097248077393, + "num_tokens": 467019527.0, + "step": 18050 + }, + { + "epoch": 1.9823193498792007, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.384786605834961, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7150759100914001, + "num_tokens": 467047899.0, + "step": 18051 + }, + { + "epoch": 1.9824291675818142, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7780895233154297, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7164911031723022, + "num_tokens": 467069050.0, + "step": 18052 + }, + { + "epoch": 1.9825389852844277, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.688188076019287, + "learning_rate": 1e-06, + "loss": 0.9315, + "mean_token_accuracy": 0.7274301052093506, + "num_tokens": 467091624.0, + "step": 18053 + }, + { + "epoch": 1.9826488029870415, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5601906776428223, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6844936609268188, + "num_tokens": 467119004.0, + "step": 18054 + }, + { + "epoch": 1.9827586206896552, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.384859085083008, + "learning_rate": 1e-06, + "loss": 1.1343, + "mean_token_accuracy": 0.6730988025665283, + "num_tokens": 467150163.0, + "step": 18055 + }, + { + "epoch": 1.9828684383922688, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4969303607940674, + "learning_rate": 1e-06, + "loss": 0.9349, + "mean_token_accuracy": 0.7288898229598999, + "num_tokens": 467180353.0, + "step": 18056 + }, + { + "epoch": 1.9829782560948825, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.715664863586426, + "learning_rate": 1e-06, + "loss": 0.8727, + "mean_token_accuracy": 0.740764319896698, + "num_tokens": 467200651.0, + "step": 18057 + }, + { + "epoch": 1.983088073797496, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.835719347000122, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7231107950210571, + "num_tokens": 467225799.0, + "step": 18058 + }, + { + "epoch": 1.9831978915001098, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.097818374633789, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.715083122253418, + "num_tokens": 467243075.0, + "step": 18059 + }, + { + "epoch": 1.9833077092027236, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4889657497406006, + "learning_rate": 1e-06, + "loss": 1.1064, + "mean_token_accuracy": 0.6837705373764038, + "num_tokens": 467272737.0, + "step": 18060 + }, + { + "epoch": 1.9834175269053371, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.257831573486328, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7212197780609131, + "num_tokens": 467301245.0, + "step": 18061 + }, + { + "epoch": 1.9835273446079507, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3507258892059326, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6930093765258789, + "num_tokens": 467331176.0, + "step": 18062 + }, + { + "epoch": 1.9836371623105644, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9861044883728027, + "learning_rate": 1e-06, + "loss": 0.9034, + "mean_token_accuracy": 0.728649377822876, + "num_tokens": 467350422.0, + "step": 18063 + }, + { + "epoch": 1.9837469800131782, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.468261241912842, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.718220591545105, + "num_tokens": 467377058.0, + "step": 18064 + }, + { + "epoch": 1.983856797715792, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.572936534881592, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6863330602645874, + "num_tokens": 467405378.0, + "step": 18065 + }, + { + "epoch": 1.9839666154184055, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4933691024780273, + "learning_rate": 1e-06, + "loss": 1.0466, + "mean_token_accuracy": 0.7022722959518433, + "num_tokens": 467433938.0, + "step": 18066 + }, + { + "epoch": 1.984076433121019, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6436607837677, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6842217445373535, + "num_tokens": 467456400.0, + "step": 18067 + }, + { + "epoch": 1.9841862508236328, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.539808988571167, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6985584497451782, + "num_tokens": 467485415.0, + "step": 18068 + }, + { + "epoch": 1.9842960685262465, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.459055185317993, + "learning_rate": 1e-06, + "loss": 1.1359, + "mean_token_accuracy": 0.6747524738311768, + "num_tokens": 467516613.0, + "step": 18069 + }, + { + "epoch": 1.98440588622886, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.474301815032959, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7083233594894409, + "num_tokens": 467541554.0, + "step": 18070 + }, + { + "epoch": 1.9845157039314738, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7353298664093018, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.7224185466766357, + "num_tokens": 467562551.0, + "step": 18071 + }, + { + "epoch": 1.9846255216340873, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5112242698669434, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7020613551139832, + "num_tokens": 467587875.0, + "step": 18072 + }, + { + "epoch": 1.984735339336701, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.742506742477417, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7215876579284668, + "num_tokens": 467609473.0, + "step": 18073 + }, + { + "epoch": 1.9848451570393149, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.840474843978882, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7182133197784424, + "num_tokens": 467638318.0, + "step": 18074 + }, + { + "epoch": 1.9849549747419284, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4502646923065186, + "learning_rate": 1e-06, + "loss": 1.1145, + "mean_token_accuracy": 0.6783862113952637, + "num_tokens": 467669113.0, + "step": 18075 + }, + { + "epoch": 1.985064792444542, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.788378953933716, + "learning_rate": 1e-06, + "loss": 1.0678, + "mean_token_accuracy": 0.6865909099578857, + "num_tokens": 467695282.0, + "step": 18076 + }, + { + "epoch": 1.9851746101471557, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6817429065704346, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7086577415466309, + "num_tokens": 467718225.0, + "step": 18077 + }, + { + "epoch": 1.9852844278497694, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.787175178527832, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6833056211471558, + "num_tokens": 467741915.0, + "step": 18078 + }, + { + "epoch": 1.9853942455523832, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.792982339859009, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7150759696960449, + "num_tokens": 467762792.0, + "step": 18079 + }, + { + "epoch": 1.9855040632549967, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7415027618408203, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7047113180160522, + "num_tokens": 467784467.0, + "step": 18080 + }, + { + "epoch": 1.9856138809576103, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5616002082824707, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7166549563407898, + "num_tokens": 467809037.0, + "step": 18081 + }, + { + "epoch": 1.985723698660224, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.784406900405884, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7209831476211548, + "num_tokens": 467833099.0, + "step": 18082 + }, + { + "epoch": 1.9858335163628378, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.516268730163574, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6990468502044678, + "num_tokens": 467861147.0, + "step": 18083 + }, + { + "epoch": 1.9859433340654513, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5914785861968994, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6949659585952759, + "num_tokens": 467886383.0, + "step": 18084 + }, + { + "epoch": 1.9860531517680649, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.88023042678833, + "learning_rate": 1e-06, + "loss": 0.9136, + "mean_token_accuracy": 0.7281739711761475, + "num_tokens": 467904883.0, + "step": 18085 + }, + { + "epoch": 1.9861629694706786, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.7635743618011475, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.705231785774231, + "num_tokens": 467927722.0, + "step": 18086 + }, + { + "epoch": 1.9862727871732924, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5775275230407715, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7259520292282104, + "num_tokens": 467954533.0, + "step": 18087 + }, + { + "epoch": 1.9863826048759061, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.561933755874634, + "learning_rate": 1e-06, + "loss": 0.9055, + "mean_token_accuracy": 0.7339881658554077, + "num_tokens": 467980580.0, + "step": 18088 + }, + { + "epoch": 1.9864924225785197, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.808969259262085, + "learning_rate": 1e-06, + "loss": 0.916, + "mean_token_accuracy": 0.7264971137046814, + "num_tokens": 468001046.0, + "step": 18089 + }, + { + "epoch": 1.9866022402811332, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.726839065551758, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7213977575302124, + "num_tokens": 468022546.0, + "step": 18090 + }, + { + "epoch": 1.986712057983747, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.6663100719451904, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7055222988128662, + "num_tokens": 468045996.0, + "step": 18091 + }, + { + "epoch": 1.9868218756863607, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.6118125915527344, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7241045236587524, + "num_tokens": 468073492.0, + "step": 18092 + }, + { + "epoch": 1.9869316933889745, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.799394130706787, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6937566995620728, + "num_tokens": 468097229.0, + "step": 18093 + }, + { + "epoch": 1.987041511091588, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.490363597869873, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6872450113296509, + "num_tokens": 468124449.0, + "step": 18094 + }, + { + "epoch": 1.9871513287942015, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5696072578430176, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6882708668708801, + "num_tokens": 468153002.0, + "step": 18095 + }, + { + "epoch": 1.9872611464968153, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6611688137054443, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7111175060272217, + "num_tokens": 468175728.0, + "step": 18096 + }, + { + "epoch": 1.987370964199429, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2814109325408936, + "learning_rate": 1e-06, + "loss": 1.1155, + "mean_token_accuracy": 0.6771449446678162, + "num_tokens": 468208067.0, + "step": 18097 + }, + { + "epoch": 1.9874807819020426, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6663379669189453, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7205581068992615, + "num_tokens": 468229495.0, + "step": 18098 + }, + { + "epoch": 1.9875905996046561, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5382919311523438, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7161027193069458, + "num_tokens": 468256501.0, + "step": 18099 + }, + { + "epoch": 1.9877004173072699, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.6857359409332275, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.7256609797477722, + "num_tokens": 468280381.0, + "step": 18100 + }, + { + "epoch": 1.9878102350098836, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5029594898223877, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6973223090171814, + "num_tokens": 468311940.0, + "step": 18101 + }, + { + "epoch": 1.9879200527124974, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.972590446472168, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7051964402198792, + "num_tokens": 468335383.0, + "step": 18102 + }, + { + "epoch": 1.988029870415111, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.7236673831939697, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6917626857757568, + "num_tokens": 468358251.0, + "step": 18103 + }, + { + "epoch": 1.9881396881177245, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7524678707122803, + "learning_rate": 1e-06, + "loss": 0.908, + "mean_token_accuracy": 0.7323663234710693, + "num_tokens": 468381560.0, + "step": 18104 + }, + { + "epoch": 1.9882495058203382, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.748054027557373, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6925861835479736, + "num_tokens": 468405235.0, + "step": 18105 + }, + { + "epoch": 1.988359323522952, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4663245677948, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7141268253326416, + "num_tokens": 468433779.0, + "step": 18106 + }, + { + "epoch": 1.9884691412255655, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4103074073791504, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.719723641872406, + "num_tokens": 468465278.0, + "step": 18107 + }, + { + "epoch": 1.9885789589281793, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7287051677703857, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.7212907075881958, + "num_tokens": 468489088.0, + "step": 18108 + }, + { + "epoch": 1.9886887766307928, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4753732681274414, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6968363523483276, + "num_tokens": 468521256.0, + "step": 18109 + }, + { + "epoch": 1.9887985943334066, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.762624979019165, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7226000428199768, + "num_tokens": 468541124.0, + "step": 18110 + }, + { + "epoch": 1.9889084120360203, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.835663318634033, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7087580561637878, + "num_tokens": 468565961.0, + "step": 18111 + }, + { + "epoch": 1.9890182297386338, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.372274875640869, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.6974191665649414, + "num_tokens": 468595579.0, + "step": 18112 + }, + { + "epoch": 1.9891280474412474, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.586897611618042, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7115621566772461, + "num_tokens": 468621118.0, + "step": 18113 + }, + { + "epoch": 1.9892378651438611, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.828007936477661, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7145547270774841, + "num_tokens": 468641545.0, + "step": 18114 + }, + { + "epoch": 1.989347682846475, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4742000102996826, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7031780481338501, + "num_tokens": 468668558.0, + "step": 18115 + }, + { + "epoch": 1.9894575005490887, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.461205244064331, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6868588924407959, + "num_tokens": 468697598.0, + "step": 18116 + }, + { + "epoch": 1.9895673182517022, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.404419183731079, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6955916881561279, + "num_tokens": 468726401.0, + "step": 18117 + }, + { + "epoch": 1.9896771359543157, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.456604242324829, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6929098963737488, + "num_tokens": 468748943.0, + "step": 18118 + }, + { + "epoch": 1.9897869536569295, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.41168475151062, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7032774090766907, + "num_tokens": 468777685.0, + "step": 18119 + }, + { + "epoch": 1.9898967713595432, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.8063318729400635, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7105388641357422, + "num_tokens": 468800301.0, + "step": 18120 + }, + { + "epoch": 1.9900065890621568, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.523063898086548, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7112386226654053, + "num_tokens": 468826461.0, + "step": 18121 + }, + { + "epoch": 1.9901164067647705, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2687931060791016, + "learning_rate": 1e-06, + "loss": 1.1593, + "mean_token_accuracy": 0.6815884113311768, + "num_tokens": 468858553.0, + "step": 18122 + }, + { + "epoch": 1.990226224467384, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3260223865509033, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.683866024017334, + "num_tokens": 468888908.0, + "step": 18123 + }, + { + "epoch": 1.9903360421699978, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.218580484390259, + "learning_rate": 1e-06, + "loss": 1.1104, + "mean_token_accuracy": 0.6766536235809326, + "num_tokens": 468922491.0, + "step": 18124 + }, + { + "epoch": 1.9904458598726116, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.397486448287964, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6964969635009766, + "num_tokens": 468954283.0, + "step": 18125 + }, + { + "epoch": 1.9905556775752251, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7521567344665527, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6990936994552612, + "num_tokens": 468977253.0, + "step": 18126 + }, + { + "epoch": 1.9906654952778386, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.753340244293213, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.715813398361206, + "num_tokens": 468997728.0, + "step": 18127 + }, + { + "epoch": 1.9907753129804524, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.515200614929199, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7096653580665588, + "num_tokens": 469022017.0, + "step": 18128 + }, + { + "epoch": 1.9908851306830662, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.594696521759033, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7079326510429382, + "num_tokens": 469045797.0, + "step": 18129 + }, + { + "epoch": 1.99099494838568, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.492349863052368, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6817429065704346, + "num_tokens": 469077174.0, + "step": 18130 + }, + { + "epoch": 1.9911047660882935, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.552626848220825, + "learning_rate": 1e-06, + "loss": 1.1283, + "mean_token_accuracy": 0.677765965461731, + "num_tokens": 469104610.0, + "step": 18131 + }, + { + "epoch": 1.991214583790907, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.326979160308838, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6976194381713867, + "num_tokens": 469136746.0, + "step": 18132 + }, + { + "epoch": 1.9913244014935207, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3046786785125732, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.694456160068512, + "num_tokens": 469167149.0, + "step": 18133 + }, + { + "epoch": 1.9914342191961345, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.429931879043579, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.6984484791755676, + "num_tokens": 469194948.0, + "step": 18134 + }, + { + "epoch": 1.991544036898748, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.469650983810425, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7027844190597534, + "num_tokens": 469220938.0, + "step": 18135 + }, + { + "epoch": 1.9916538546013618, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.189748525619507, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7110840082168579, + "num_tokens": 469257117.0, + "step": 18136 + }, + { + "epoch": 1.9917636723039753, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5554065704345703, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7066588401794434, + "num_tokens": 469284383.0, + "step": 18137 + }, + { + "epoch": 1.991873490006589, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.593047857284546, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7138453722000122, + "num_tokens": 469307692.0, + "step": 18138 + }, + { + "epoch": 1.9919833077092028, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7490458488464355, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7016254663467407, + "num_tokens": 469330147.0, + "step": 18139 + }, + { + "epoch": 1.9920931254118164, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5677101612091064, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7097582817077637, + "num_tokens": 469355986.0, + "step": 18140 + }, + { + "epoch": 1.99220294311443, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6452560424804688, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7081495523452759, + "num_tokens": 469384324.0, + "step": 18141 + }, + { + "epoch": 1.9923127608170437, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3401291370391846, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7120329141616821, + "num_tokens": 469414795.0, + "step": 18142 + }, + { + "epoch": 1.9924225785196574, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.702916383743286, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.6935703158378601, + "num_tokens": 469437569.0, + "step": 18143 + }, + { + "epoch": 1.9925323962222712, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6016831398010254, + "learning_rate": 1e-06, + "loss": 0.916, + "mean_token_accuracy": 0.7262091636657715, + "num_tokens": 469461788.0, + "step": 18144 + }, + { + "epoch": 1.9926422139248847, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5913772583007812, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6982179284095764, + "num_tokens": 469489058.0, + "step": 18145 + }, + { + "epoch": 1.9927520316274983, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.014270067214966, + "learning_rate": 1e-06, + "loss": 1.1241, + "mean_token_accuracy": 0.680127739906311, + "num_tokens": 469518606.0, + "step": 18146 + }, + { + "epoch": 1.992861849330112, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.727149248123169, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7080522179603577, + "num_tokens": 469542224.0, + "step": 18147 + }, + { + "epoch": 1.9929716670327258, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.410890817642212, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7006517648696899, + "num_tokens": 469571020.0, + "step": 18148 + }, + { + "epoch": 1.9930814847353393, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.425110340118408, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7125259637832642, + "num_tokens": 469598688.0, + "step": 18149 + }, + { + "epoch": 1.9931913024379528, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2807233333587646, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7131413817405701, + "num_tokens": 469628287.0, + "step": 18150 + }, + { + "epoch": 1.9933011201405666, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2960550785064697, + "learning_rate": 1e-06, + "loss": 1.0839, + "mean_token_accuracy": 0.6930412650108337, + "num_tokens": 469661848.0, + "step": 18151 + }, + { + "epoch": 1.9934109378431804, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.084583044052124, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.715535044670105, + "num_tokens": 469691358.0, + "step": 18152 + }, + { + "epoch": 1.993520755545794, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.521347761154175, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.706446647644043, + "num_tokens": 469719859.0, + "step": 18153 + }, + { + "epoch": 1.9936305732484076, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.621860980987549, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6908571720123291, + "num_tokens": 469744208.0, + "step": 18154 + }, + { + "epoch": 1.9937403909510212, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.481351852416992, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.690739631652832, + "num_tokens": 469775148.0, + "step": 18155 + }, + { + "epoch": 1.993850208653635, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.9210007190704346, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7383068203926086, + "num_tokens": 469796709.0, + "step": 18156 + }, + { + "epoch": 1.9939600263562487, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.870973587036133, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7028740048408508, + "num_tokens": 469819795.0, + "step": 18157 + }, + { + "epoch": 1.9940698440588625, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5572497844696045, + "learning_rate": 1e-06, + "loss": 0.9149, + "mean_token_accuracy": 0.727931559085846, + "num_tokens": 469844011.0, + "step": 18158 + }, + { + "epoch": 1.994179661761476, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.666774272918701, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.703565239906311, + "num_tokens": 469867553.0, + "step": 18159 + }, + { + "epoch": 1.9942894794640895, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.5310451984405518, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7118639349937439, + "num_tokens": 469896905.0, + "step": 18160 + }, + { + "epoch": 1.9943992971667033, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.831766128540039, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6876781582832336, + "num_tokens": 469920200.0, + "step": 18161 + }, + { + "epoch": 1.994509114869317, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.720017194747925, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7063779830932617, + "num_tokens": 469944573.0, + "step": 18162 + }, + { + "epoch": 1.9946189325719306, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 4.353960990905762, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7049314975738525, + "num_tokens": 469967220.0, + "step": 18163 + }, + { + "epoch": 1.994728750274544, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.852544069290161, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7148642539978027, + "num_tokens": 469987942.0, + "step": 18164 + }, + { + "epoch": 1.9948385679771579, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.477149486541748, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7059437036514282, + "num_tokens": 470017012.0, + "step": 18165 + }, + { + "epoch": 1.9949483856797716, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5298726558685303, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7105669975280762, + "num_tokens": 470042843.0, + "step": 18166 + }, + { + "epoch": 1.9950582033823854, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5389177799224854, + "learning_rate": 1e-06, + "loss": 0.9307, + "mean_token_accuracy": 0.7304897308349609, + "num_tokens": 470069002.0, + "step": 18167 + }, + { + "epoch": 1.995168021084999, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.703874111175537, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7183971405029297, + "num_tokens": 470089815.0, + "step": 18168 + }, + { + "epoch": 1.9952778387876124, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2472286224365234, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6968566179275513, + "num_tokens": 470121345.0, + "step": 18169 + }, + { + "epoch": 1.9953876564902262, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3708033561706543, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.7187154293060303, + "num_tokens": 470148531.0, + "step": 18170 + }, + { + "epoch": 1.99549747419284, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.613306999206543, + "learning_rate": 1e-06, + "loss": 1.1114, + "mean_token_accuracy": 0.6871893405914307, + "num_tokens": 470174290.0, + "step": 18171 + }, + { + "epoch": 1.9956072918954535, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5306594371795654, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.7008469104766846, + "num_tokens": 470201166.0, + "step": 18172 + }, + { + "epoch": 1.9957171095980673, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.817652463912964, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7058829069137573, + "num_tokens": 470222186.0, + "step": 18173 + }, + { + "epoch": 1.9958269273006808, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3498194217681885, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7212943434715271, + "num_tokens": 470253168.0, + "step": 18174 + }, + { + "epoch": 1.9959367450032945, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.538120985031128, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7042233347892761, + "num_tokens": 470281135.0, + "step": 18175 + }, + { + "epoch": 1.9960465627059083, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7177093029022217, + "learning_rate": 1e-06, + "loss": 1.0929, + "mean_token_accuracy": 0.6983686685562134, + "num_tokens": 470307947.0, + "step": 18176 + }, + { + "epoch": 1.9961563804085218, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.345937967300415, + "learning_rate": 1e-06, + "loss": 0.9119, + "mean_token_accuracy": 0.7386663556098938, + "num_tokens": 470335331.0, + "step": 18177 + }, + { + "epoch": 1.9962661981111354, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.0735888481140137, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6974121928215027, + "num_tokens": 470365413.0, + "step": 18178 + }, + { + "epoch": 1.9963760158137491, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3731956481933594, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.6894793510437012, + "num_tokens": 470396737.0, + "step": 18179 + }, + { + "epoch": 1.9964858335163629, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.553231716156006, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6978927850723267, + "num_tokens": 470422654.0, + "step": 18180 + }, + { + "epoch": 1.9965956512189766, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.5148324966430664, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.711839497089386, + "num_tokens": 470449058.0, + "step": 18181 + }, + { + "epoch": 1.9967054689215902, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.689795732498169, + "learning_rate": 1e-06, + "loss": 0.9265, + "mean_token_accuracy": 0.7271699905395508, + "num_tokens": 470471733.0, + "step": 18182 + }, + { + "epoch": 1.9968152866242037, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.682420492172241, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7028124332427979, + "num_tokens": 470494067.0, + "step": 18183 + }, + { + "epoch": 1.9969251043268175, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.495361566543579, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7026454210281372, + "num_tokens": 470521583.0, + "step": 18184 + }, + { + "epoch": 1.9970349220294312, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6020238399505615, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7031129598617554, + "num_tokens": 470545232.0, + "step": 18185 + }, + { + "epoch": 1.9971447397320448, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.35152530670166, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7096062898635864, + "num_tokens": 470573581.0, + "step": 18186 + }, + { + "epoch": 1.9972545574346585, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7087032794952393, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7115363478660583, + "num_tokens": 470594915.0, + "step": 18187 + }, + { + "epoch": 1.997364375137272, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.701927423477173, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7104600667953491, + "num_tokens": 470619696.0, + "step": 18188 + }, + { + "epoch": 1.9974741928398858, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.512896776199341, + "learning_rate": 1e-06, + "loss": 1.1012, + "mean_token_accuracy": 0.6831088066101074, + "num_tokens": 470648134.0, + "step": 18189 + }, + { + "epoch": 1.9975840105424996, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5652918815612793, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7190098762512207, + "num_tokens": 470672694.0, + "step": 18190 + }, + { + "epoch": 1.997693828245113, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6258039474487305, + "learning_rate": 1e-06, + "loss": 0.9118, + "mean_token_accuracy": 0.7422475218772888, + "num_tokens": 470698698.0, + "step": 18191 + }, + { + "epoch": 1.9978036459477266, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.494736671447754, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7160835266113281, + "num_tokens": 470725993.0, + "step": 18192 + }, + { + "epoch": 1.9979134636503404, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.46746563911438, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.7017541527748108, + "num_tokens": 470751735.0, + "step": 18193 + }, + { + "epoch": 1.9980232813529542, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 3.3185155391693115, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7235062718391418, + "num_tokens": 470768977.0, + "step": 18194 + }, + { + "epoch": 1.998133099055568, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.404613494873047, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.702072262763977, + "num_tokens": 470797383.0, + "step": 18195 + }, + { + "epoch": 1.9982429167581814, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.788475513458252, + "learning_rate": 1e-06, + "loss": 0.8919, + "mean_token_accuracy": 0.7294958233833313, + "num_tokens": 470818262.0, + "step": 18196 + }, + { + "epoch": 1.998352734460795, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.511843204498291, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6924527287483215, + "num_tokens": 470845484.0, + "step": 18197 + }, + { + "epoch": 1.9984625521634087, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.6250507831573486, + "learning_rate": 1e-06, + "loss": 1.1132, + "mean_token_accuracy": 0.6852778196334839, + "num_tokens": 470870283.0, + "step": 18198 + }, + { + "epoch": 1.9985723698660225, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.70622181892395, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7142691612243652, + "num_tokens": 470891092.0, + "step": 18199 + }, + { + "epoch": 1.998682187568636, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.653085947036743, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7126355171203613, + "num_tokens": 470916592.0, + "step": 18200 + }, + { + "epoch": 1.9987920052712496, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5981621742248535, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7229572534561157, + "num_tokens": 470943251.0, + "step": 18201 + }, + { + "epoch": 1.9989018229738633, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.531799077987671, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7228537797927856, + "num_tokens": 470970150.0, + "step": 18202 + }, + { + "epoch": 1.999011640676477, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3814611434936523, + "learning_rate": 1e-06, + "loss": 1.1074, + "mean_token_accuracy": 0.677422285079956, + "num_tokens": 471001184.0, + "step": 18203 + }, + { + "epoch": 1.9991214583790908, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5891571044921875, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.7107831239700317, + "num_tokens": 471025144.0, + "step": 18204 + }, + { + "epoch": 1.9992312760817044, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.2363827228546143, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6922593116760254, + "num_tokens": 471059042.0, + "step": 18205 + }, + { + "epoch": 1.999341093784318, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3697378635406494, + "learning_rate": 1e-06, + "loss": 1.1297, + "mean_token_accuracy": 0.6768703460693359, + "num_tokens": 471090984.0, + "step": 18206 + }, + { + "epoch": 1.9994509114869317, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.707028865814209, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7182003259658813, + "num_tokens": 471111508.0, + "step": 18207 + }, + { + "epoch": 1.9995607291895454, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5855660438537598, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6958682537078857, + "num_tokens": 471136394.0, + "step": 18208 + }, + { + "epoch": 1.9996705468921592, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.528665781021118, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.700326144695282, + "num_tokens": 471164069.0, + "step": 18209 + }, + { + "epoch": 1.9997803645947727, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.860698938369751, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7059531211853027, + "num_tokens": 471189023.0, + "step": 18210 + }, + { + "epoch": 1.9998901822973862, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.784088134765625, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7204357385635376, + "num_tokens": 471209443.0, + "step": 18211 + }, + { + "epoch": 2.0, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.688687801361084, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6897197961807251, + "num_tokens": 471235014.0, + "step": 18212 + }, + { + "epoch": 2.0001098177026138, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.693499803543091, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7064075469970703, + "num_tokens": 471260654.0, + "step": 18213 + }, + { + "epoch": 2.0002196354052275, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.4830093383789062, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.708328366279602, + "num_tokens": 471286421.0, + "step": 18214 + }, + { + "epoch": 2.000329453107841, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.5986521244049072, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7083996534347534, + "num_tokens": 471310258.0, + "step": 18215 + }, + { + "epoch": 2.0004392708104546, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.40944766998291, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7028636336326599, + "num_tokens": 471338723.0, + "step": 18216 + }, + { + "epoch": 2.0005490885130683, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.3592844009399414, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7093387842178345, + "num_tokens": 471369007.0, + "step": 18217 + }, + { + "epoch": 2.000658906215682, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7904090881347656, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7205862402915955, + "num_tokens": 471390448.0, + "step": 18218 + }, + { + "epoch": 2.0007687239182954, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.7380173206329346, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6994705200195312, + "num_tokens": 471420002.0, + "step": 18219 + }, + { + "epoch": 2.000878541620909, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.72951602935791, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7048794031143188, + "num_tokens": 471442498.0, + "step": 18220 + }, + { + "epoch": 2.000988359323523, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.833698272705078, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7146323919296265, + "num_tokens": 471463307.0, + "step": 18221 + }, + { + "epoch": 2.0010981770261367, + "ewc_loss": 9.775161743164062e-06, + "grad_norm": 2.701122999191284, + "learning_rate": 1e-06, + "loss": 0.8833, + "mean_token_accuracy": 0.7324655055999756, + "num_tokens": 471487919.0, + "step": 18222 + }, + { + "epoch": 2.0012079947287504, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5861926078796387, + "learning_rate": 1e-06, + "loss": 0.9155, + "mean_token_accuracy": 0.7376096248626709, + "num_tokens": 471514828.0, + "step": 18223 + }, + { + "epoch": 2.0013178124313638, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.8171324729919434, + "learning_rate": 1e-06, + "loss": 0.8912, + "mean_token_accuracy": 0.7363811731338501, + "num_tokens": 471537025.0, + "step": 18224 + }, + { + "epoch": 2.0014276301339775, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.845109701156616, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7019584774971008, + "num_tokens": 471557930.0, + "step": 18225 + }, + { + "epoch": 2.0015374478365913, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5215415954589844, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.7011542320251465, + "num_tokens": 471582641.0, + "step": 18226 + }, + { + "epoch": 2.001647265539205, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.729802370071411, + "learning_rate": 1e-06, + "loss": 0.8653, + "mean_token_accuracy": 0.7340224981307983, + "num_tokens": 471604738.0, + "step": 18227 + }, + { + "epoch": 2.001757083241819, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.508641481399536, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6809378862380981, + "num_tokens": 471634425.0, + "step": 18228 + }, + { + "epoch": 2.001866900944432, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.564056396484375, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7064837217330933, + "num_tokens": 471660871.0, + "step": 18229 + }, + { + "epoch": 2.001976718647046, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.641662359237671, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.707841694355011, + "num_tokens": 471688335.0, + "step": 18230 + }, + { + "epoch": 2.0020865363496596, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.446608781814575, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.6890878081321716, + "num_tokens": 471718337.0, + "step": 18231 + }, + { + "epoch": 2.0021963540522734, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4201996326446533, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6960936188697815, + "num_tokens": 471751966.0, + "step": 18232 + }, + { + "epoch": 2.0023061717548867, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.7672019004821777, + "learning_rate": 1e-06, + "loss": 0.9088, + "mean_token_accuracy": 0.7292948365211487, + "num_tokens": 471773386.0, + "step": 18233 + }, + { + "epoch": 2.0024159894575004, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.7090930938720703, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.714362621307373, + "num_tokens": 471795477.0, + "step": 18234 + }, + { + "epoch": 2.002525807160114, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.320134401321411, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.7028212547302246, + "num_tokens": 471825957.0, + "step": 18235 + }, + { + "epoch": 2.002635624862728, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.436555862426758, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7020918130874634, + "num_tokens": 471853767.0, + "step": 18236 + }, + { + "epoch": 2.0027454425653417, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4577643871307373, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.696712851524353, + "num_tokens": 471883082.0, + "step": 18237 + }, + { + "epoch": 2.002855260267955, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4804859161376953, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.71519535779953, + "num_tokens": 471910218.0, + "step": 18238 + }, + { + "epoch": 2.0029650779705688, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5470385551452637, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7054171562194824, + "num_tokens": 471936993.0, + "step": 18239 + }, + { + "epoch": 2.0030748956731825, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 16.62060546875, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6842606067657471, + "num_tokens": 471964226.0, + "step": 18240 + }, + { + "epoch": 2.0031847133757963, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.9502577781677246, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7113487124443054, + "num_tokens": 471986803.0, + "step": 18241 + }, + { + "epoch": 2.00329453107841, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.9331133365631104, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7071877121925354, + "num_tokens": 472010234.0, + "step": 18242 + }, + { + "epoch": 2.0034043487810234, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.8760416507720947, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7147666215896606, + "num_tokens": 472032126.0, + "step": 18243 + }, + { + "epoch": 2.003514166483637, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.8128764629364014, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7016329765319824, + "num_tokens": 472057020.0, + "step": 18244 + }, + { + "epoch": 2.003623984186251, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.811518669128418, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6942787170410156, + "num_tokens": 472079754.0, + "step": 18245 + }, + { + "epoch": 2.0037338018888646, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.398564100265503, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7291561365127563, + "num_tokens": 472107357.0, + "step": 18246 + }, + { + "epoch": 2.003843619591478, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.6739847660064697, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7167923450469971, + "num_tokens": 472132314.0, + "step": 18247 + }, + { + "epoch": 2.0039534372940917, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.634031295776367, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.700063169002533, + "num_tokens": 472159056.0, + "step": 18248 + }, + { + "epoch": 2.0040632549967055, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.447953939437866, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7268723249435425, + "num_tokens": 472187610.0, + "step": 18249 + }, + { + "epoch": 2.004173072699319, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.7022242546081543, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.710324764251709, + "num_tokens": 472211027.0, + "step": 18250 + }, + { + "epoch": 2.004282890401933, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.402731418609619, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7073774337768555, + "num_tokens": 472240380.0, + "step": 18251 + }, + { + "epoch": 2.0043927081045463, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.426525115966797, + "learning_rate": 1e-06, + "loss": 0.931, + "mean_token_accuracy": 0.728814423084259, + "num_tokens": 472270893.0, + "step": 18252 + }, + { + "epoch": 2.00450252580716, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5726191997528076, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7293678522109985, + "num_tokens": 472296851.0, + "step": 18253 + }, + { + "epoch": 2.004612343509774, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.561358690261841, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7082395553588867, + "num_tokens": 472324270.0, + "step": 18254 + }, + { + "epoch": 2.0047221612123876, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5286378860473633, + "learning_rate": 1e-06, + "loss": 0.9395, + "mean_token_accuracy": 0.7195831537246704, + "num_tokens": 472350052.0, + "step": 18255 + }, + { + "epoch": 2.0048319789150013, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.7128489017486572, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7106088399887085, + "num_tokens": 472377760.0, + "step": 18256 + }, + { + "epoch": 2.0049417966176146, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 17.015371322631836, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.7007224559783936, + "num_tokens": 472413478.0, + "step": 18257 + }, + { + "epoch": 2.0050516143202284, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.553797960281372, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.683598518371582, + "num_tokens": 472441679.0, + "step": 18258 + }, + { + "epoch": 2.005161432022842, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.601545810699463, + "learning_rate": 1e-06, + "loss": 0.9326, + "mean_token_accuracy": 0.725425124168396, + "num_tokens": 472467174.0, + "step": 18259 + }, + { + "epoch": 2.005271249725456, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.5780887603759766, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7202446460723877, + "num_tokens": 472491450.0, + "step": 18260 + }, + { + "epoch": 2.005381067428069, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 3.047321319580078, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7247304916381836, + "num_tokens": 472513867.0, + "step": 18261 + }, + { + "epoch": 2.005490885130683, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.686208486557007, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7171250581741333, + "num_tokens": 472538909.0, + "step": 18262 + }, + { + "epoch": 2.0056007028332967, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.3228416442871094, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7019845247268677, + "num_tokens": 472573516.0, + "step": 18263 + }, + { + "epoch": 2.0057105205359105, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.2503490447998047, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.6966346502304077, + "num_tokens": 472604040.0, + "step": 18264 + }, + { + "epoch": 2.0058203382385242, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 3.091425895690918, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7169686555862427, + "num_tokens": 472628842.0, + "step": 18265 + }, + { + "epoch": 2.0059301559411375, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.781485080718994, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6861828565597534, + "num_tokens": 472653793.0, + "step": 18266 + }, + { + "epoch": 2.0060399736437513, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5722365379333496, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6823611855506897, + "num_tokens": 472682750.0, + "step": 18267 + }, + { + "epoch": 2.006149791346365, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.587242603302002, + "learning_rate": 1e-06, + "loss": 1.0935, + "mean_token_accuracy": 0.6847563982009888, + "num_tokens": 472708280.0, + "step": 18268 + }, + { + "epoch": 2.006259609048979, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.627206563949585, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6977027654647827, + "num_tokens": 472732535.0, + "step": 18269 + }, + { + "epoch": 2.0063694267515926, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4887332916259766, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6883208155632019, + "num_tokens": 472761119.0, + "step": 18270 + }, + { + "epoch": 2.006479244454206, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.771979808807373, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7224063873291016, + "num_tokens": 472782492.0, + "step": 18271 + }, + { + "epoch": 2.0065890621568196, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.542449951171875, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7009332180023193, + "num_tokens": 472809592.0, + "step": 18272 + }, + { + "epoch": 2.0066988798594334, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.503525733947754, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7040234804153442, + "num_tokens": 472838346.0, + "step": 18273 + }, + { + "epoch": 2.006808697562047, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.4578888416290283, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.689937174320221, + "num_tokens": 472868408.0, + "step": 18274 + }, + { + "epoch": 2.0069185152646605, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.9490585327148438, + "learning_rate": 1e-06, + "loss": 0.9397, + "mean_token_accuracy": 0.7232121229171753, + "num_tokens": 472891999.0, + "step": 18275 + }, + { + "epoch": 2.0070283329672742, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.8892245292663574, + "learning_rate": 1e-06, + "loss": 0.9299, + "mean_token_accuracy": 0.7250620126724243, + "num_tokens": 472912478.0, + "step": 18276 + }, + { + "epoch": 2.007138150669888, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.7862331867218018, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7001574039459229, + "num_tokens": 472937871.0, + "step": 18277 + }, + { + "epoch": 2.0072479683725017, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.3887007236480713, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.693398654460907, + "num_tokens": 472968653.0, + "step": 18278 + }, + { + "epoch": 2.0073577860751155, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.743598222732544, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7111375331878662, + "num_tokens": 472993966.0, + "step": 18279 + }, + { + "epoch": 2.007467603777729, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.6309995651245117, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7061044573783875, + "num_tokens": 473019198.0, + "step": 18280 + }, + { + "epoch": 2.0075774214803426, + "ewc_loss": 9.834766387939453e-06, + "grad_norm": 2.5150012969970703, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.7052508592605591, + "num_tokens": 473047784.0, + "step": 18281 + }, + { + "epoch": 2.0076872391829563, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.6033554077148438, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.697705090045929, + "num_tokens": 473073103.0, + "step": 18282 + }, + { + "epoch": 2.00779705688557, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.6590332984924316, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7217848300933838, + "num_tokens": 473098209.0, + "step": 18283 + }, + { + "epoch": 2.0079068745881834, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.5278608798980713, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6795655488967896, + "num_tokens": 473123539.0, + "step": 18284 + }, + { + "epoch": 2.008016692290797, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.3219516277313232, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.6881309151649475, + "num_tokens": 473158145.0, + "step": 18285 + }, + { + "epoch": 2.008126509993411, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.5740442276000977, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.6976882815361023, + "num_tokens": 473185679.0, + "step": 18286 + }, + { + "epoch": 2.0082363276960247, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.827096939086914, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7173179984092712, + "num_tokens": 473205197.0, + "step": 18287 + }, + { + "epoch": 2.0083461453986384, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.6338448524475098, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7189590930938721, + "num_tokens": 473230948.0, + "step": 18288 + }, + { + "epoch": 2.0084559631012517, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.959838628768921, + "learning_rate": 1e-06, + "loss": 0.8916, + "mean_token_accuracy": 0.7373129725456238, + "num_tokens": 473251055.0, + "step": 18289 + }, + { + "epoch": 2.0085657808038655, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.8597800731658936, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7102775573730469, + "num_tokens": 473274792.0, + "step": 18290 + }, + { + "epoch": 2.0086755985064793, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 5.289602756500244, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.7323136925697327, + "num_tokens": 473304591.0, + "step": 18291 + }, + { + "epoch": 2.008785416209093, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.420081853866577, + "learning_rate": 1e-06, + "loss": 0.9316, + "mean_token_accuracy": 0.7161567807197571, + "num_tokens": 473333610.0, + "step": 18292 + }, + { + "epoch": 2.0088952339117068, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 3.0031344890594482, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7124133706092834, + "num_tokens": 473352052.0, + "step": 18293 + }, + { + "epoch": 2.00900505161432, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.3239598274230957, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.691544234752655, + "num_tokens": 473384444.0, + "step": 18294 + }, + { + "epoch": 2.009114869316934, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.303318738937378, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7017688155174255, + "num_tokens": 473415544.0, + "step": 18295 + }, + { + "epoch": 2.0092246870195476, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.4742591381073, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.7036906480789185, + "num_tokens": 473443069.0, + "step": 18296 + }, + { + "epoch": 2.0093345047221614, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.554877758026123, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7197328805923462, + "num_tokens": 473467650.0, + "step": 18297 + }, + { + "epoch": 2.0094443224247747, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.3691861629486084, + "learning_rate": 1e-06, + "loss": 0.891, + "mean_token_accuracy": 0.7210041880607605, + "num_tokens": 473495979.0, + "step": 18298 + }, + { + "epoch": 2.0095541401273884, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.5572757720947266, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.721778392791748, + "num_tokens": 473521867.0, + "step": 18299 + }, + { + "epoch": 2.009663957830002, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.418515682220459, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7130432724952698, + "num_tokens": 473551775.0, + "step": 18300 + }, + { + "epoch": 2.009773775532616, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 3.029045581817627, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7284024953842163, + "num_tokens": 473570580.0, + "step": 18301 + }, + { + "epoch": 2.0098835932352297, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 3.033625841140747, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7159848213195801, + "num_tokens": 473589849.0, + "step": 18302 + }, + { + "epoch": 2.009993410937843, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.3998210430145264, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7089710831642151, + "num_tokens": 473621943.0, + "step": 18303 + }, + { + "epoch": 2.0101032286404568, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.747551679611206, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7152067422866821, + "num_tokens": 473644412.0, + "step": 18304 + }, + { + "epoch": 2.0102130463430705, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.2125706672668457, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7203769087791443, + "num_tokens": 473681233.0, + "step": 18305 + }, + { + "epoch": 2.0103228640456843, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.701059341430664, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7188469767570496, + "num_tokens": 473703760.0, + "step": 18306 + }, + { + "epoch": 2.010432681748298, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.643763303756714, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7059773206710815, + "num_tokens": 473728373.0, + "step": 18307 + }, + { + "epoch": 2.0105424994509113, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.6019349098205566, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7058542966842651, + "num_tokens": 473754675.0, + "step": 18308 + }, + { + "epoch": 2.010652317153525, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.8742592334747314, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7137537598609924, + "num_tokens": 473776730.0, + "step": 18309 + }, + { + "epoch": 2.010762134856139, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.459228992462158, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.7043566107749939, + "num_tokens": 473806682.0, + "step": 18310 + }, + { + "epoch": 2.0108719525587526, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.493145704269409, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.686482310295105, + "num_tokens": 473837602.0, + "step": 18311 + }, + { + "epoch": 2.010981770261366, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.7106544971466064, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7172356247901917, + "num_tokens": 473861516.0, + "step": 18312 + }, + { + "epoch": 2.0110915879639797, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 3.054260015487671, + "learning_rate": 1e-06, + "loss": 0.9423, + "mean_token_accuracy": 0.7266539931297302, + "num_tokens": 473887367.0, + "step": 18313 + }, + { + "epoch": 2.0112014056665934, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.554513454437256, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7026274800300598, + "num_tokens": 473914911.0, + "step": 18314 + }, + { + "epoch": 2.011311223369207, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7649807929992676, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.700437605381012, + "num_tokens": 473941329.0, + "step": 18315 + }, + { + "epoch": 2.011421041071821, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.9569530487060547, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7059156894683838, + "num_tokens": 473962147.0, + "step": 18316 + }, + { + "epoch": 2.0115308587744343, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.49887752532959, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6980708837509155, + "num_tokens": 473989895.0, + "step": 18317 + }, + { + "epoch": 2.011640676477048, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.70192813873291, + "learning_rate": 1e-06, + "loss": 0.8677, + "mean_token_accuracy": 0.742425799369812, + "num_tokens": 474010572.0, + "step": 18318 + }, + { + "epoch": 2.011750494179662, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.714653491973877, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7093684077262878, + "num_tokens": 474032697.0, + "step": 18319 + }, + { + "epoch": 2.0118603118822755, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.4112868309020996, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7324308753013611, + "num_tokens": 474063383.0, + "step": 18320 + }, + { + "epoch": 2.0119701295848893, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.631047487258911, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7109490633010864, + "num_tokens": 474091568.0, + "step": 18321 + }, + { + "epoch": 2.0120799472875026, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7278809547424316, + "learning_rate": 1e-06, + "loss": 0.8978, + "mean_token_accuracy": 0.7339351177215576, + "num_tokens": 474113882.0, + "step": 18322 + }, + { + "epoch": 2.0121897649901164, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.4617199897766113, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7123229503631592, + "num_tokens": 474143063.0, + "step": 18323 + }, + { + "epoch": 2.01229958269273, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.650278329849243, + "learning_rate": 1e-06, + "loss": 0.8561, + "mean_token_accuracy": 0.7500616908073425, + "num_tokens": 474167338.0, + "step": 18324 + }, + { + "epoch": 2.012409400395344, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.875779151916504, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7237765192985535, + "num_tokens": 474191108.0, + "step": 18325 + }, + { + "epoch": 2.012519218097957, + "ewc_loss": 9.894371032714844e-06, + "grad_norm": 2.591089963912964, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7069939374923706, + "num_tokens": 474216320.0, + "step": 18326 + }, + { + "epoch": 2.012629035800571, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.4741389751434326, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7091538310050964, + "num_tokens": 474248732.0, + "step": 18327 + }, + { + "epoch": 2.0127388535031847, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7808837890625, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.7309689521789551, + "num_tokens": 474270619.0, + "step": 18328 + }, + { + "epoch": 2.0128486712057985, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.810131549835205, + "learning_rate": 1e-06, + "loss": 0.8604, + "mean_token_accuracy": 0.7459219694137573, + "num_tokens": 474293168.0, + "step": 18329 + }, + { + "epoch": 2.0129584889084122, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.6170144081115723, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7113686800003052, + "num_tokens": 474318595.0, + "step": 18330 + }, + { + "epoch": 2.0130683066110255, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.445744276046753, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7191533446311951, + "num_tokens": 474346507.0, + "step": 18331 + }, + { + "epoch": 2.0131781243136393, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.510340452194214, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6944875121116638, + "num_tokens": 474374731.0, + "step": 18332 + }, + { + "epoch": 2.013287942016253, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.763741970062256, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7245193123817444, + "num_tokens": 474396677.0, + "step": 18333 + }, + { + "epoch": 2.013397759718867, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.455897092819214, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7022672891616821, + "num_tokens": 474424538.0, + "step": 18334 + }, + { + "epoch": 2.01350757742148, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.870999813079834, + "learning_rate": 1e-06, + "loss": 0.9063, + "mean_token_accuracy": 0.7327434420585632, + "num_tokens": 474444738.0, + "step": 18335 + }, + { + "epoch": 2.013617395124094, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.933584213256836, + "learning_rate": 1e-06, + "loss": 0.882, + "mean_token_accuracy": 0.734912633895874, + "num_tokens": 474464517.0, + "step": 18336 + }, + { + "epoch": 2.0137272128267076, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.383650064468384, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7067292332649231, + "num_tokens": 474495613.0, + "step": 18337 + }, + { + "epoch": 2.0138370305293214, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.633824586868286, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6889063715934753, + "num_tokens": 474521657.0, + "step": 18338 + }, + { + "epoch": 2.013946848231935, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.4508752822875977, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7062917351722717, + "num_tokens": 474549732.0, + "step": 18339 + }, + { + "epoch": 2.0140566659345485, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.6144440174102783, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7093744874000549, + "num_tokens": 474578853.0, + "step": 18340 + }, + { + "epoch": 2.014166483637162, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.3745412826538086, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7045097351074219, + "num_tokens": 474609898.0, + "step": 18341 + }, + { + "epoch": 2.014276301339776, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 3.0686259269714355, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7268175482749939, + "num_tokens": 474630096.0, + "step": 18342 + }, + { + "epoch": 2.0143861190423897, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.622068405151367, + "learning_rate": 1e-06, + "loss": 0.9042, + "mean_token_accuracy": 0.7322026491165161, + "num_tokens": 474655330.0, + "step": 18343 + }, + { + "epoch": 2.0144959367450035, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.692535400390625, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7038698792457581, + "num_tokens": 474679402.0, + "step": 18344 + }, + { + "epoch": 2.014605754447617, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.4578142166137695, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7191994190216064, + "num_tokens": 474709822.0, + "step": 18345 + }, + { + "epoch": 2.0147155721502306, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.640455484390259, + "learning_rate": 1e-06, + "loss": 0.894, + "mean_token_accuracy": 0.7374804019927979, + "num_tokens": 474734308.0, + "step": 18346 + }, + { + "epoch": 2.0148253898528443, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.723964214324951, + "learning_rate": 1e-06, + "loss": 0.9191, + "mean_token_accuracy": 0.7284277677536011, + "num_tokens": 474757181.0, + "step": 18347 + }, + { + "epoch": 2.014935207555458, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.8198206424713135, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7033989429473877, + "num_tokens": 474779652.0, + "step": 18348 + }, + { + "epoch": 2.0150450252580714, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7132577896118164, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7077155113220215, + "num_tokens": 474802679.0, + "step": 18349 + }, + { + "epoch": 2.015154842960685, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7791199684143066, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7078887820243835, + "num_tokens": 474826802.0, + "step": 18350 + }, + { + "epoch": 2.015264660663299, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.5961923599243164, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.708784818649292, + "num_tokens": 474852591.0, + "step": 18351 + }, + { + "epoch": 2.0153744783659127, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.5606184005737305, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7228678464889526, + "num_tokens": 474881174.0, + "step": 18352 + }, + { + "epoch": 2.0154842960685264, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.4155197143554688, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7136269807815552, + "num_tokens": 474912081.0, + "step": 18353 + }, + { + "epoch": 2.0155941137711397, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.686783790588379, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.6950740814208984, + "num_tokens": 474935198.0, + "step": 18354 + }, + { + "epoch": 2.0157039314737535, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.865346670150757, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.7006121873855591, + "num_tokens": 474958871.0, + "step": 18355 + }, + { + "epoch": 2.0158137491763672, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.8440146446228027, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7097601890563965, + "num_tokens": 474982605.0, + "step": 18356 + }, + { + "epoch": 2.015923566878981, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7279200553894043, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7037312984466553, + "num_tokens": 475006223.0, + "step": 18357 + }, + { + "epoch": 2.0160333845815948, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.5708653926849365, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7058889269828796, + "num_tokens": 475034253.0, + "step": 18358 + }, + { + "epoch": 2.016143202284208, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7117960453033447, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.703307032585144, + "num_tokens": 475064327.0, + "step": 18359 + }, + { + "epoch": 2.016253019986822, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.658031940460205, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7133334279060364, + "num_tokens": 475088498.0, + "step": 18360 + }, + { + "epoch": 2.0163628376894356, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.5334596633911133, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7144098281860352, + "num_tokens": 475113631.0, + "step": 18361 + }, + { + "epoch": 2.0164726553920493, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.361406087875366, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7043075561523438, + "num_tokens": 475144330.0, + "step": 18362 + }, + { + "epoch": 2.0165824730946627, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.6040232181549072, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7017518281936646, + "num_tokens": 475173602.0, + "step": 18363 + }, + { + "epoch": 2.0166922907972764, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.657534122467041, + "learning_rate": 1e-06, + "loss": 0.9566, + "mean_token_accuracy": 0.7148788571357727, + "num_tokens": 475197199.0, + "step": 18364 + }, + { + "epoch": 2.01680210849989, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.309095621109009, + "learning_rate": 1e-06, + "loss": 0.8994, + "mean_token_accuracy": 0.7340322732925415, + "num_tokens": 475228261.0, + "step": 18365 + }, + { + "epoch": 2.016911926202504, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.811927318572998, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7005569934844971, + "num_tokens": 475255246.0, + "step": 18366 + }, + { + "epoch": 2.0170217439051177, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7587890625, + "learning_rate": 1e-06, + "loss": 0.9212, + "mean_token_accuracy": 0.7304598689079285, + "num_tokens": 475278080.0, + "step": 18367 + }, + { + "epoch": 2.017131561607731, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.9137370586395264, + "learning_rate": 1e-06, + "loss": 0.9259, + "mean_token_accuracy": 0.7285886406898499, + "num_tokens": 475297692.0, + "step": 18368 + }, + { + "epoch": 2.0172413793103448, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.8244235515594482, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6845564842224121, + "num_tokens": 475323342.0, + "step": 18369 + }, + { + "epoch": 2.0173511970129585, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.5931036472320557, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.715587317943573, + "num_tokens": 475346994.0, + "step": 18370 + }, + { + "epoch": 2.0174610147155723, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.846039295196533, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.710706353187561, + "num_tokens": 475372315.0, + "step": 18371 + }, + { + "epoch": 2.017570832418186, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8584883213043213, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.72549968957901, + "num_tokens": 475396438.0, + "step": 18372 + }, + { + "epoch": 2.0176806501207993, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.6589198112487793, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6974160075187683, + "num_tokens": 475422077.0, + "step": 18373 + }, + { + "epoch": 2.017790467823413, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.563019037246704, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7203197479248047, + "num_tokens": 475447617.0, + "step": 18374 + }, + { + "epoch": 2.017900285526027, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8851258754730225, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7182386517524719, + "num_tokens": 475471261.0, + "step": 18375 + }, + { + "epoch": 2.0180101032286406, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.850968360900879, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7156151533126831, + "num_tokens": 475492817.0, + "step": 18376 + }, + { + "epoch": 2.018119920931254, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.6522598266601562, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7029305696487427, + "num_tokens": 475519515.0, + "step": 18377 + }, + { + "epoch": 2.0182297386338677, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.9077682495117188, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7111669778823853, + "num_tokens": 475546017.0, + "step": 18378 + }, + { + "epoch": 2.0183395563364814, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.608269214630127, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6886507272720337, + "num_tokens": 475576814.0, + "step": 18379 + }, + { + "epoch": 2.018449374039095, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.6005969047546387, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6965241432189941, + "num_tokens": 475604395.0, + "step": 18380 + }, + { + "epoch": 2.018559191741709, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.48222279548645, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.693762481212616, + "num_tokens": 475632406.0, + "step": 18381 + }, + { + "epoch": 2.0186690094443223, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5553839206695557, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7026234865188599, + "num_tokens": 475663063.0, + "step": 18382 + }, + { + "epoch": 2.018778827146936, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.9348227977752686, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.714667558670044, + "num_tokens": 475683086.0, + "step": 18383 + }, + { + "epoch": 2.0188886448495498, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.992096424102783, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7219151854515076, + "num_tokens": 475701641.0, + "step": 18384 + }, + { + "epoch": 2.0189984625521635, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6933436393737793, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7008085250854492, + "num_tokens": 475725651.0, + "step": 18385 + }, + { + "epoch": 2.0191082802547773, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 4.1272759437561035, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7131143808364868, + "num_tokens": 475755963.0, + "step": 18386 + }, + { + "epoch": 2.0192180979573906, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7147974967956543, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.6917526721954346, + "num_tokens": 475782500.0, + "step": 18387 + }, + { + "epoch": 2.0193279156600044, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.8496158123016357, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7013974189758301, + "num_tokens": 475810085.0, + "step": 18388 + }, + { + "epoch": 2.019437733362618, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.869229555130005, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7158231735229492, + "num_tokens": 475831769.0, + "step": 18389 + }, + { + "epoch": 2.019547551065232, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6118788719177246, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6976146697998047, + "num_tokens": 475860288.0, + "step": 18390 + }, + { + "epoch": 2.019657368767845, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7088944911956787, + "learning_rate": 1e-06, + "loss": 0.9007, + "mean_token_accuracy": 0.7329742908477783, + "num_tokens": 475881887.0, + "step": 18391 + }, + { + "epoch": 2.019767186470459, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.7792093753814697, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.6963920593261719, + "num_tokens": 475907087.0, + "step": 18392 + }, + { + "epoch": 2.0198770041730727, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.679232597351074, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7072834968566895, + "num_tokens": 475930532.0, + "step": 18393 + }, + { + "epoch": 2.0199868218756865, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.551400899887085, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.725088357925415, + "num_tokens": 475956755.0, + "step": 18394 + }, + { + "epoch": 2.0200966395783, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.427089214324951, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7116055488586426, + "num_tokens": 475988953.0, + "step": 18395 + }, + { + "epoch": 2.0202064572809135, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.9183614253997803, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7243905067443848, + "num_tokens": 476010678.0, + "step": 18396 + }, + { + "epoch": 2.0203162749835273, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.561037540435791, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7168890237808228, + "num_tokens": 476038599.0, + "step": 18397 + }, + { + "epoch": 2.020426092686141, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7783689498901367, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6975085139274597, + "num_tokens": 476067170.0, + "step": 18398 + }, + { + "epoch": 2.020535910388755, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6126348972320557, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7157989144325256, + "num_tokens": 476093353.0, + "step": 18399 + }, + { + "epoch": 2.020645728091368, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7075955867767334, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7071544528007507, + "num_tokens": 476120892.0, + "step": 18400 + }, + { + "epoch": 2.020755545793982, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4768848419189453, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7089253067970276, + "num_tokens": 476148830.0, + "step": 18401 + }, + { + "epoch": 2.0208653634965956, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5006654262542725, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.685170590877533, + "num_tokens": 476176269.0, + "step": 18402 + }, + { + "epoch": 2.0209751811992094, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.535645008087158, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6844627857208252, + "num_tokens": 476203739.0, + "step": 18403 + }, + { + "epoch": 2.021084998901823, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.810335159301758, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7162328362464905, + "num_tokens": 476227125.0, + "step": 18404 + }, + { + "epoch": 2.0211948166044365, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.56951904296875, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7030000686645508, + "num_tokens": 476253997.0, + "step": 18405 + }, + { + "epoch": 2.02130463430705, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.557441234588623, + "learning_rate": 1e-06, + "loss": 0.8216, + "mean_token_accuracy": 0.7560518980026245, + "num_tokens": 476276686.0, + "step": 18406 + }, + { + "epoch": 2.021414452009664, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3858895301818848, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7020877599716187, + "num_tokens": 476307945.0, + "step": 18407 + }, + { + "epoch": 2.0215242697122777, + "ewc_loss": 9.953975677490234e-06, + "grad_norm": 2.4284043312072754, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6994340419769287, + "num_tokens": 476344153.0, + "step": 18408 + }, + { + "epoch": 2.0216340874148915, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4514596462249756, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.724079430103302, + "num_tokens": 476372534.0, + "step": 18409 + }, + { + "epoch": 2.021743905117505, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5271127223968506, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7008472681045532, + "num_tokens": 476400143.0, + "step": 18410 + }, + { + "epoch": 2.0218537228201185, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8561110496520996, + "learning_rate": 1e-06, + "loss": 0.9313, + "mean_token_accuracy": 0.7319629788398743, + "num_tokens": 476421887.0, + "step": 18411 + }, + { + "epoch": 2.0219635405227323, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5087437629699707, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7106847167015076, + "num_tokens": 476448408.0, + "step": 18412 + }, + { + "epoch": 2.022073358225346, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 3.0007474422454834, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7251436710357666, + "num_tokens": 476467945.0, + "step": 18413 + }, + { + "epoch": 2.0221831759279594, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.308579683303833, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6940744519233704, + "num_tokens": 476504064.0, + "step": 18414 + }, + { + "epoch": 2.022292993630573, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7479259967803955, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.7197374701499939, + "num_tokens": 476528563.0, + "step": 18415 + }, + { + "epoch": 2.022402811333187, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.819549798965454, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7142808437347412, + "num_tokens": 476551278.0, + "step": 18416 + }, + { + "epoch": 2.0225126290358006, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.9346327781677246, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.7337406277656555, + "num_tokens": 476571059.0, + "step": 18417 + }, + { + "epoch": 2.0226224467384144, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.554809093475342, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7157462239265442, + "num_tokens": 476601332.0, + "step": 18418 + }, + { + "epoch": 2.0227322644410277, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 3.1642701625823975, + "learning_rate": 1e-06, + "loss": 0.903, + "mean_token_accuracy": 0.7328683137893677, + "num_tokens": 476620721.0, + "step": 18419 + }, + { + "epoch": 2.0228420821436415, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6615331172943115, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.711112380027771, + "num_tokens": 476647984.0, + "step": 18420 + }, + { + "epoch": 2.0229518998462552, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.541557550430298, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7164542078971863, + "num_tokens": 476675219.0, + "step": 18421 + }, + { + "epoch": 2.023061717548869, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3351352214813232, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6927317976951599, + "num_tokens": 476707059.0, + "step": 18422 + }, + { + "epoch": 2.0231715352514827, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.600374221801758, + "learning_rate": 1e-06, + "loss": 0.9364, + "mean_token_accuracy": 0.7243234515190125, + "num_tokens": 476734933.0, + "step": 18423 + }, + { + "epoch": 2.023281352954096, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.677244186401367, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.727942943572998, + "num_tokens": 476758364.0, + "step": 18424 + }, + { + "epoch": 2.02339117065671, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.351223945617676, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6977189183235168, + "num_tokens": 476788896.0, + "step": 18425 + }, + { + "epoch": 2.0235009883593236, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5068111419677734, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6908254623413086, + "num_tokens": 476814130.0, + "step": 18426 + }, + { + "epoch": 2.0236108060619373, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5934717655181885, + "learning_rate": 1e-06, + "loss": 0.9308, + "mean_token_accuracy": 0.7214035987854004, + "num_tokens": 476839446.0, + "step": 18427 + }, + { + "epoch": 2.0237206237645506, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3768298625946045, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6928750276565552, + "num_tokens": 476871147.0, + "step": 18428 + }, + { + "epoch": 2.0238304414671644, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.491203784942627, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7009674310684204, + "num_tokens": 476897450.0, + "step": 18429 + }, + { + "epoch": 2.023940259169778, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.823694944381714, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6994975805282593, + "num_tokens": 476921078.0, + "step": 18430 + }, + { + "epoch": 2.024050076872392, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5835671424865723, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7091824412345886, + "num_tokens": 476944763.0, + "step": 18431 + }, + { + "epoch": 2.0241598945750057, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.983393907546997, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7237111330032349, + "num_tokens": 476964532.0, + "step": 18432 + }, + { + "epoch": 2.024269712277619, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8783979415893555, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7205666303634644, + "num_tokens": 476984769.0, + "step": 18433 + }, + { + "epoch": 2.0243795299802327, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 3.0986735820770264, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.701910138130188, + "num_tokens": 477010814.0, + "step": 18434 + }, + { + "epoch": 2.0244893476828465, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.685561180114746, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7117600440979004, + "num_tokens": 477037902.0, + "step": 18435 + }, + { + "epoch": 2.0245991653854603, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8140573501586914, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7146947383880615, + "num_tokens": 477062536.0, + "step": 18436 + }, + { + "epoch": 2.024708983088074, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 3.005275011062622, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7167112827301025, + "num_tokens": 477088327.0, + "step": 18437 + }, + { + "epoch": 2.0248188007906873, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6888439655303955, + "learning_rate": 1e-06, + "loss": 0.9262, + "mean_token_accuracy": 0.737968921661377, + "num_tokens": 477111255.0, + "step": 18438 + }, + { + "epoch": 2.024928618493301, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3316915035247803, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7156400680541992, + "num_tokens": 477140742.0, + "step": 18439 + }, + { + "epoch": 2.025038436195915, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.63986873626709, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.7024407386779785, + "num_tokens": 477167026.0, + "step": 18440 + }, + { + "epoch": 2.0251482538985286, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.788052558898926, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7075197696685791, + "num_tokens": 477189984.0, + "step": 18441 + }, + { + "epoch": 2.025258071601142, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5884206295013428, + "learning_rate": 1e-06, + "loss": 1.1319, + "mean_token_accuracy": 0.6774446964263916, + "num_tokens": 477216833.0, + "step": 18442 + }, + { + "epoch": 2.0253678893037557, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7008914947509766, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.704925537109375, + "num_tokens": 477239306.0, + "step": 18443 + }, + { + "epoch": 2.0254777070063694, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8139963150024414, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7075990438461304, + "num_tokens": 477261024.0, + "step": 18444 + }, + { + "epoch": 2.025587524708983, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.933305501937866, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7141444087028503, + "num_tokens": 477287270.0, + "step": 18445 + }, + { + "epoch": 2.025697342411597, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.463190793991089, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7118215560913086, + "num_tokens": 477315531.0, + "step": 18446 + }, + { + "epoch": 2.0258071601142102, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.483391523361206, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7105942964553833, + "num_tokens": 477340295.0, + "step": 18447 + }, + { + "epoch": 2.025916977816824, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.69319748878479, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7237875461578369, + "num_tokens": 477362513.0, + "step": 18448 + }, + { + "epoch": 2.0260267955194378, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4913928508758545, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6853271126747131, + "num_tokens": 477390622.0, + "step": 18449 + }, + { + "epoch": 2.0261366132220515, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6222593784332275, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7155630588531494, + "num_tokens": 477415906.0, + "step": 18450 + }, + { + "epoch": 2.0262464309246653, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5018205642700195, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7107053399085999, + "num_tokens": 477442984.0, + "step": 18451 + }, + { + "epoch": 2.0263562486272786, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.635514974594116, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7238649129867554, + "num_tokens": 477467160.0, + "step": 18452 + }, + { + "epoch": 2.0264660663298923, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.421614408493042, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7170330286026001, + "num_tokens": 477494952.0, + "step": 18453 + }, + { + "epoch": 2.026575884032506, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4808313846588135, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7008496522903442, + "num_tokens": 477523578.0, + "step": 18454 + }, + { + "epoch": 2.02668570173512, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.657578945159912, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.716336727142334, + "num_tokens": 477547502.0, + "step": 18455 + }, + { + "epoch": 2.026795519437733, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.840789318084717, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.6950996518135071, + "num_tokens": 477568762.0, + "step": 18456 + }, + { + "epoch": 2.026905337140347, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 3.0098111629486084, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7215796113014221, + "num_tokens": 477591095.0, + "step": 18457 + }, + { + "epoch": 2.0270151548429607, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.598764657974243, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7160796523094177, + "num_tokens": 477615159.0, + "step": 18458 + }, + { + "epoch": 2.0271249725455744, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5731201171875, + "learning_rate": 1e-06, + "loss": 1.0982, + "mean_token_accuracy": 0.6792304515838623, + "num_tokens": 477646241.0, + "step": 18459 + }, + { + "epoch": 2.027234790248188, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4174935817718506, + "learning_rate": 1e-06, + "loss": 1.1091, + "mean_token_accuracy": 0.6805372834205627, + "num_tokens": 477678701.0, + "step": 18460 + }, + { + "epoch": 2.0273446079508015, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4149739742279053, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6897528171539307, + "num_tokens": 477708276.0, + "step": 18461 + }, + { + "epoch": 2.0274544256534153, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6826977729797363, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7067371010780334, + "num_tokens": 477734701.0, + "step": 18462 + }, + { + "epoch": 2.027564243356029, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5476033687591553, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.704256534576416, + "num_tokens": 477761108.0, + "step": 18463 + }, + { + "epoch": 2.027674061058643, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4990077018737793, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7266761064529419, + "num_tokens": 477787164.0, + "step": 18464 + }, + { + "epoch": 2.027783878761256, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7297701835632324, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7063746452331543, + "num_tokens": 477811238.0, + "step": 18465 + }, + { + "epoch": 2.02789369646387, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.43533992767334, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7048019766807556, + "num_tokens": 477840378.0, + "step": 18466 + }, + { + "epoch": 2.0280035141664836, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4486019611358643, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6953779458999634, + "num_tokens": 477866312.0, + "step": 18467 + }, + { + "epoch": 2.0281133318690974, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7099673748016357, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7282219529151917, + "num_tokens": 477889462.0, + "step": 18468 + }, + { + "epoch": 2.028223149571711, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7280263900756836, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7087336778640747, + "num_tokens": 477912523.0, + "step": 18469 + }, + { + "epoch": 2.0283329672743244, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.877321481704712, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.723457932472229, + "num_tokens": 477938584.0, + "step": 18470 + }, + { + "epoch": 2.028442784976938, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5645527839660645, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7031019926071167, + "num_tokens": 477969511.0, + "step": 18471 + }, + { + "epoch": 2.028552602679552, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6369991302490234, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.692795991897583, + "num_tokens": 477996399.0, + "step": 18472 + }, + { + "epoch": 2.0286624203821657, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.817107677459717, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7170006036758423, + "num_tokens": 478019229.0, + "step": 18473 + }, + { + "epoch": 2.0287722380847795, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.035567045211792, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7226712703704834, + "num_tokens": 478040209.0, + "step": 18474 + }, + { + "epoch": 2.028882055787393, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 80.54898834228516, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7271007895469666, + "num_tokens": 478062881.0, + "step": 18475 + }, + { + "epoch": 2.0289918734900065, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.9611029624938965, + "learning_rate": 1e-06, + "loss": 0.9288, + "mean_token_accuracy": 0.7277148962020874, + "num_tokens": 478085248.0, + "step": 18476 + }, + { + "epoch": 2.0291016911926203, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8089346885681152, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.7077603936195374, + "num_tokens": 478111876.0, + "step": 18477 + }, + { + "epoch": 2.029211508895234, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.504581928253174, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7099536657333374, + "num_tokens": 478140060.0, + "step": 18478 + }, + { + "epoch": 2.0293213265978474, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4519262313842773, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.690762996673584, + "num_tokens": 478173027.0, + "step": 18479 + }, + { + "epoch": 2.029431144300461, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.667310953140259, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7127296924591064, + "num_tokens": 478197596.0, + "step": 18480 + }, + { + "epoch": 2.029540962003075, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.484109401702881, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7218331098556519, + "num_tokens": 478227789.0, + "step": 18481 + }, + { + "epoch": 2.0296507797056886, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4997267723083496, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6998013854026794, + "num_tokens": 478259865.0, + "step": 18482 + }, + { + "epoch": 2.0297605974083024, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.9988787174224854, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7210491895675659, + "num_tokens": 478280288.0, + "step": 18483 + }, + { + "epoch": 2.0298704151109157, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.917795419692993, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7090457677841187, + "num_tokens": 478301159.0, + "step": 18484 + }, + { + "epoch": 2.0299802328135295, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6104373931884766, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7169178128242493, + "num_tokens": 478326072.0, + "step": 18485 + }, + { + "epoch": 2.030090050516143, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.355273723602295, + "learning_rate": 1e-06, + "loss": 1.125, + "mean_token_accuracy": 0.6706990003585815, + "num_tokens": 478362978.0, + "step": 18486 + }, + { + "epoch": 2.030199868218757, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5095348358154297, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7109372615814209, + "num_tokens": 478395259.0, + "step": 18487 + }, + { + "epoch": 2.0303096859213707, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5146687030792236, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.7013294696807861, + "num_tokens": 478424449.0, + "step": 18488 + }, + { + "epoch": 2.030419503623984, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.419198989868164, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6919323205947876, + "num_tokens": 478454349.0, + "step": 18489 + }, + { + "epoch": 2.030529321326598, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.872997760772705, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.6985411643981934, + "num_tokens": 478476519.0, + "step": 18490 + }, + { + "epoch": 2.0306391390292116, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6729471683502197, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7069092392921448, + "num_tokens": 478501584.0, + "step": 18491 + }, + { + "epoch": 2.0307489567318253, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.583749532699585, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.7040327191352844, + "num_tokens": 478527028.0, + "step": 18492 + }, + { + "epoch": 2.0308587744344386, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3758327960968018, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7084686160087585, + "num_tokens": 478556497.0, + "step": 18493 + }, + { + "epoch": 2.0309685921370524, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5930521488189697, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7116199135780334, + "num_tokens": 478581600.0, + "step": 18494 + }, + { + "epoch": 2.031078409839666, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6087710857391357, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6967421174049377, + "num_tokens": 478608224.0, + "step": 18495 + }, + { + "epoch": 2.03118822754228, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8462541103363037, + "learning_rate": 1e-06, + "loss": 0.9504, + "mean_token_accuracy": 0.7222941517829895, + "num_tokens": 478630025.0, + "step": 18496 + }, + { + "epoch": 2.0312980452448937, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.542696237564087, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7057567238807678, + "num_tokens": 478655944.0, + "step": 18497 + }, + { + "epoch": 2.031407862947507, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.742948293685913, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.697266697883606, + "num_tokens": 478682722.0, + "step": 18498 + }, + { + "epoch": 2.0315176806501207, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.700552463531494, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7027826309204102, + "num_tokens": 478712650.0, + "step": 18499 + }, + { + "epoch": 2.0316274983527345, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.167114496231079, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6965129375457764, + "num_tokens": 478750751.0, + "step": 18500 + }, + { + "epoch": 2.0317373160553482, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4274089336395264, + "learning_rate": 1e-06, + "loss": 1.1068, + "mean_token_accuracy": 0.68328458070755, + "num_tokens": 478782213.0, + "step": 18501 + }, + { + "epoch": 2.031847133757962, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.6480865478515625, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.7029216289520264, + "num_tokens": 478806595.0, + "step": 18502 + }, + { + "epoch": 2.0319569514605753, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.2813172340393066, + "learning_rate": 1e-06, + "loss": 1.0832, + "mean_token_accuracy": 0.6841959953308105, + "num_tokens": 478842045.0, + "step": 18503 + }, + { + "epoch": 2.032066769163189, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.53362774848938, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7061434984207153, + "num_tokens": 478870674.0, + "step": 18504 + }, + { + "epoch": 2.032176586865803, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 3.1158416271209717, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7105260491371155, + "num_tokens": 478891731.0, + "step": 18505 + }, + { + "epoch": 2.0322864045684166, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.610114097595215, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7056542634963989, + "num_tokens": 478920241.0, + "step": 18506 + }, + { + "epoch": 2.03239622227103, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.428356885910034, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6887353658676147, + "num_tokens": 478951766.0, + "step": 18507 + }, + { + "epoch": 2.0325060399736437, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4052059650421143, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6985416412353516, + "num_tokens": 478984028.0, + "step": 18508 + }, + { + "epoch": 2.0326158576762574, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3955917358398438, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6949171423912048, + "num_tokens": 479013463.0, + "step": 18509 + }, + { + "epoch": 2.032725675378871, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.658704996109009, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6872282028198242, + "num_tokens": 479038882.0, + "step": 18510 + }, + { + "epoch": 2.032835493081485, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7240495681762695, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6979276537895203, + "num_tokens": 479063048.0, + "step": 18511 + }, + { + "epoch": 2.0329453107840982, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5260536670684814, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6951233148574829, + "num_tokens": 479092793.0, + "step": 18512 + }, + { + "epoch": 2.033055128486712, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7750115394592285, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7169705629348755, + "num_tokens": 479114094.0, + "step": 18513 + }, + { + "epoch": 2.0331649461893258, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4473321437835693, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7244616746902466, + "num_tokens": 479139288.0, + "step": 18514 + }, + { + "epoch": 2.0332747638919395, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.88618803024292, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7140693664550781, + "num_tokens": 479161789.0, + "step": 18515 + }, + { + "epoch": 2.033384581594553, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.645310640335083, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.701349675655365, + "num_tokens": 479187593.0, + "step": 18516 + }, + { + "epoch": 2.0334943992971666, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5156495571136475, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7134572267532349, + "num_tokens": 479215416.0, + "step": 18517 + }, + { + "epoch": 2.0336042169997803, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4741151332855225, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7077200412750244, + "num_tokens": 479242278.0, + "step": 18518 + }, + { + "epoch": 2.033714034702394, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5548479557037354, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6978816986083984, + "num_tokens": 479270308.0, + "step": 18519 + }, + { + "epoch": 2.033823852405008, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.774961233139038, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7227955460548401, + "num_tokens": 479297337.0, + "step": 18520 + }, + { + "epoch": 2.033933670107621, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4740538597106934, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.6978585124015808, + "num_tokens": 479326319.0, + "step": 18521 + }, + { + "epoch": 2.034043487810235, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.4774746894836426, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7098001837730408, + "num_tokens": 479353862.0, + "step": 18522 + }, + { + "epoch": 2.0341533055128487, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5041215419769287, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7085379958152771, + "num_tokens": 479380234.0, + "step": 18523 + }, + { + "epoch": 2.0342631232154624, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7933237552642822, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7195749878883362, + "num_tokens": 479400222.0, + "step": 18524 + }, + { + "epoch": 2.034372940918076, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.7570362091064453, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.7128680944442749, + "num_tokens": 479425746.0, + "step": 18525 + }, + { + "epoch": 2.0344827586206895, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.8951363563537598, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7170854806900024, + "num_tokens": 479449819.0, + "step": 18526 + }, + { + "epoch": 2.0345925763233033, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5674779415130615, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7008823156356812, + "num_tokens": 479475323.0, + "step": 18527 + }, + { + "epoch": 2.034702394025917, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.450296640396118, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6913031339645386, + "num_tokens": 479505738.0, + "step": 18528 + }, + { + "epoch": 2.0348122117285308, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.811265468597412, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7154397368431091, + "num_tokens": 479527435.0, + "step": 18529 + }, + { + "epoch": 2.034922029431144, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.0148277282714844, + "learning_rate": 1e-06, + "loss": 0.866, + "mean_token_accuracy": 0.7409757375717163, + "num_tokens": 479545792.0, + "step": 18530 + }, + { + "epoch": 2.035031847133758, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5361409187316895, + "learning_rate": 1e-06, + "loss": 0.9294, + "mean_token_accuracy": 0.7195329666137695, + "num_tokens": 479570974.0, + "step": 18531 + }, + { + "epoch": 2.0351416648363716, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.200725793838501, + "learning_rate": 1e-06, + "loss": 0.8824, + "mean_token_accuracy": 0.7383430004119873, + "num_tokens": 479588853.0, + "step": 18532 + }, + { + "epoch": 2.0352514825389854, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.9182872772216797, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7167125344276428, + "num_tokens": 479609539.0, + "step": 18533 + }, + { + "epoch": 2.035361300241599, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.6076552867889404, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7175850868225098, + "num_tokens": 479635985.0, + "step": 18534 + }, + { + "epoch": 2.0354711179442124, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.6153693199157715, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7055345773696899, + "num_tokens": 479662486.0, + "step": 18535 + }, + { + "epoch": 2.035580935646826, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.8799736499786377, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7050492167472839, + "num_tokens": 479686596.0, + "step": 18536 + }, + { + "epoch": 2.03569075334944, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.2395074367523193, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7188723683357239, + "num_tokens": 479718032.0, + "step": 18537 + }, + { + "epoch": 2.0358005710520537, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.671196937561035, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7084274291992188, + "num_tokens": 479743117.0, + "step": 18538 + }, + { + "epoch": 2.0359103887546675, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.505465030670166, + "learning_rate": 1e-06, + "loss": 0.9312, + "mean_token_accuracy": 0.7308988571166992, + "num_tokens": 479772035.0, + "step": 18539 + }, + { + "epoch": 2.0360202064572808, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.658756732940674, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7109404802322388, + "num_tokens": 479797667.0, + "step": 18540 + }, + { + "epoch": 2.0361300241598945, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4450902938842773, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6811375617980957, + "num_tokens": 479825969.0, + "step": 18541 + }, + { + "epoch": 2.0362398418625083, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.6546568870544434, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7098875045776367, + "num_tokens": 479851139.0, + "step": 18542 + }, + { + "epoch": 2.036349659565122, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.605269193649292, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.7011826634407043, + "num_tokens": 479876346.0, + "step": 18543 + }, + { + "epoch": 2.0364594772677354, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.1299493312835693, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7094271183013916, + "num_tokens": 479897002.0, + "step": 18544 + }, + { + "epoch": 2.036569294970349, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8015220165252686, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7111208438873291, + "num_tokens": 479919893.0, + "step": 18545 + }, + { + "epoch": 2.036679112672963, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5743887424468994, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7072354555130005, + "num_tokens": 479949082.0, + "step": 18546 + }, + { + "epoch": 2.0367889303755766, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.805881977081299, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.70456463098526, + "num_tokens": 479972061.0, + "step": 18547 + }, + { + "epoch": 2.0368987480781904, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4182469844818115, + "learning_rate": 1e-06, + "loss": 0.9078, + "mean_token_accuracy": 0.7367947101593018, + "num_tokens": 480000751.0, + "step": 18548 + }, + { + "epoch": 2.0370085657808037, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.016369581222534, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7072358727455139, + "num_tokens": 480029222.0, + "step": 18549 + }, + { + "epoch": 2.0371183834834174, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.840945243835449, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7273829579353333, + "num_tokens": 480053003.0, + "step": 18550 + }, + { + "epoch": 2.037228201186031, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4795846939086914, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7147208452224731, + "num_tokens": 480080069.0, + "step": 18551 + }, + { + "epoch": 2.037338018888645, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7913200855255127, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7108814716339111, + "num_tokens": 480103242.0, + "step": 18552 + }, + { + "epoch": 2.0374478365912587, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6442501544952393, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.7085781693458557, + "num_tokens": 480128365.0, + "step": 18553 + }, + { + "epoch": 2.037557654293872, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.77654767036438, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.7298794984817505, + "num_tokens": 480150859.0, + "step": 18554 + }, + { + "epoch": 2.037667471996486, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.595630407333374, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7140641808509827, + "num_tokens": 480176966.0, + "step": 18555 + }, + { + "epoch": 2.0377772896990995, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.0986266136169434, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7218860387802124, + "num_tokens": 480196250.0, + "step": 18556 + }, + { + "epoch": 2.0378871074017133, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.78114652633667, + "learning_rate": 1e-06, + "loss": 0.9096, + "mean_token_accuracy": 0.7276760339736938, + "num_tokens": 480218428.0, + "step": 18557 + }, + { + "epoch": 2.0379969251043266, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4219284057617188, + "learning_rate": 1e-06, + "loss": 0.9332, + "mean_token_accuracy": 0.7259107828140259, + "num_tokens": 480247440.0, + "step": 18558 + }, + { + "epoch": 2.0381067428069404, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.7585666179656982, + "learning_rate": 1e-06, + "loss": 0.8986, + "mean_token_accuracy": 0.7324944734573364, + "num_tokens": 480270897.0, + "step": 18559 + }, + { + "epoch": 2.038216560509554, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.427300453186035, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.7255758047103882, + "num_tokens": 480299102.0, + "step": 18560 + }, + { + "epoch": 2.038326378212168, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.865013599395752, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6979417204856873, + "num_tokens": 480322035.0, + "step": 18561 + }, + { + "epoch": 2.0384361959147816, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.6610801219940186, + "learning_rate": 1e-06, + "loss": 0.938, + "mean_token_accuracy": 0.7234339714050293, + "num_tokens": 480346328.0, + "step": 18562 + }, + { + "epoch": 2.038546013617395, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.7790780067443848, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7178447246551514, + "num_tokens": 480370514.0, + "step": 18563 + }, + { + "epoch": 2.0386558313200087, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.719991683959961, + "learning_rate": 1e-06, + "loss": 0.9129, + "mean_token_accuracy": 0.7244478464126587, + "num_tokens": 480391909.0, + "step": 18564 + }, + { + "epoch": 2.0387656490226225, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.679882049560547, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7199628353118896, + "num_tokens": 480415314.0, + "step": 18565 + }, + { + "epoch": 2.0388754667252362, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4208016395568848, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.6986609697341919, + "num_tokens": 480444693.0, + "step": 18566 + }, + { + "epoch": 2.03898528442785, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5488076210021973, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7054339647293091, + "num_tokens": 480472164.0, + "step": 18567 + }, + { + "epoch": 2.0390951021304633, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.8365743160247803, + "learning_rate": 1e-06, + "loss": 0.9373, + "mean_token_accuracy": 0.7374042868614197, + "num_tokens": 480504905.0, + "step": 18568 + }, + { + "epoch": 2.039204919833077, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.752462387084961, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7052096128463745, + "num_tokens": 480530076.0, + "step": 18569 + }, + { + "epoch": 2.039314737535691, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.502535581588745, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.7008233070373535, + "num_tokens": 480560966.0, + "step": 18570 + }, + { + "epoch": 2.0394245552383046, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4657013416290283, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6913766860961914, + "num_tokens": 480586988.0, + "step": 18571 + }, + { + "epoch": 2.039534372940918, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.441246747970581, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6858090758323669, + "num_tokens": 480615259.0, + "step": 18572 + }, + { + "epoch": 2.0396441906435316, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.3607375621795654, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.697924017906189, + "num_tokens": 480646526.0, + "step": 18573 + }, + { + "epoch": 2.0397540083461454, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.949472665786743, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7045776844024658, + "num_tokens": 480667027.0, + "step": 18574 + }, + { + "epoch": 2.039863826048759, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.7151167392730713, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7125047445297241, + "num_tokens": 480690476.0, + "step": 18575 + }, + { + "epoch": 2.039973643751373, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.593426465988159, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6880294680595398, + "num_tokens": 480719978.0, + "step": 18576 + }, + { + "epoch": 2.0400834614539862, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.3623130321502686, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7199761271476746, + "num_tokens": 480749163.0, + "step": 18577 + }, + { + "epoch": 2.0401932791566, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.885727643966675, + "learning_rate": 1e-06, + "loss": 0.931, + "mean_token_accuracy": 0.7278834581375122, + "num_tokens": 480769114.0, + "step": 18578 + }, + { + "epoch": 2.0403030968592137, + "ewc_loss": 1.0013580322265625e-05, + "grad_norm": 2.5731639862060547, + "learning_rate": 1e-06, + "loss": 0.8666, + "mean_token_accuracy": 0.7436211705207825, + "num_tokens": 480792241.0, + "step": 18579 + }, + { + "epoch": 2.0404129145618275, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.6485373973846436, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6933497190475464, + "num_tokens": 480817932.0, + "step": 18580 + }, + { + "epoch": 2.0405227322644413, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9602766036987305, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7187355756759644, + "num_tokens": 480839867.0, + "step": 18581 + }, + { + "epoch": 2.0406325499670546, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7402966022491455, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7261202335357666, + "num_tokens": 480863797.0, + "step": 18582 + }, + { + "epoch": 2.0407423676696683, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.625669002532959, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.716184139251709, + "num_tokens": 480894772.0, + "step": 18583 + }, + { + "epoch": 2.040852185372282, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0692436695098877, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7151051163673401, + "num_tokens": 480917888.0, + "step": 18584 + }, + { + "epoch": 2.040962003074896, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8086729049682617, + "learning_rate": 1e-06, + "loss": 0.8795, + "mean_token_accuracy": 0.7329891920089722, + "num_tokens": 480937887.0, + "step": 18585 + }, + { + "epoch": 2.041071820777509, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8102164268493652, + "learning_rate": 1e-06, + "loss": 0.9128, + "mean_token_accuracy": 0.7258495092391968, + "num_tokens": 480959462.0, + "step": 18586 + }, + { + "epoch": 2.041181638480123, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.431060552597046, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6952096819877625, + "num_tokens": 480987958.0, + "step": 18587 + }, + { + "epoch": 2.0412914561827367, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9026379585266113, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6978296041488647, + "num_tokens": 481011290.0, + "step": 18588 + }, + { + "epoch": 2.0414012738853504, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.670186996459961, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7198930978775024, + "num_tokens": 481036578.0, + "step": 18589 + }, + { + "epoch": 2.041511091587964, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.1980807781219482, + "learning_rate": 1e-06, + "loss": 1.0523, + "mean_token_accuracy": 0.6969141960144043, + "num_tokens": 481065866.0, + "step": 18590 + }, + { + "epoch": 2.0416209092905775, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.615767240524292, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6800018548965454, + "num_tokens": 481092331.0, + "step": 18591 + }, + { + "epoch": 2.0417307269931912, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.70998477935791, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7052911520004272, + "num_tokens": 481116641.0, + "step": 18592 + }, + { + "epoch": 2.041840544695805, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.337250232696533, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6966891884803772, + "num_tokens": 481147880.0, + "step": 18593 + }, + { + "epoch": 2.0419503623984188, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.2804512977600098, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7069301605224609, + "num_tokens": 481177967.0, + "step": 18594 + }, + { + "epoch": 2.042060180101032, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7795002460479736, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7143194675445557, + "num_tokens": 481199113.0, + "step": 18595 + }, + { + "epoch": 2.042169997803646, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4777865409851074, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7109922170639038, + "num_tokens": 481224513.0, + "step": 18596 + }, + { + "epoch": 2.0422798155062596, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8064587116241455, + "learning_rate": 1e-06, + "loss": 0.9761, + "mean_token_accuracy": 0.7154329419136047, + "num_tokens": 481247538.0, + "step": 18597 + }, + { + "epoch": 2.0423896332088733, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0664565563201904, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7173285484313965, + "num_tokens": 481268354.0, + "step": 18598 + }, + { + "epoch": 2.042499450911487, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.736119031906128, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6920520067214966, + "num_tokens": 481292886.0, + "step": 18599 + }, + { + "epoch": 2.0426092686141004, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5559725761413574, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.708463191986084, + "num_tokens": 481318875.0, + "step": 18600 + }, + { + "epoch": 2.042719086316714, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.3776602745056152, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6988832950592041, + "num_tokens": 481350672.0, + "step": 18601 + }, + { + "epoch": 2.042828904019328, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.5637094974517822, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6897298097610474, + "num_tokens": 481373624.0, + "step": 18602 + }, + { + "epoch": 2.0429387217219417, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0093071460723877, + "learning_rate": 1e-06, + "loss": 0.8524, + "mean_token_accuracy": 0.7456330060958862, + "num_tokens": 481392991.0, + "step": 18603 + }, + { + "epoch": 2.0430485394245554, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.817636728286743, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.710319995880127, + "num_tokens": 481413218.0, + "step": 18604 + }, + { + "epoch": 2.0431583571271688, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.536330461502075, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6936949491500854, + "num_tokens": 481441656.0, + "step": 18605 + }, + { + "epoch": 2.0432681748297825, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.138974666595459, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.7176071405410767, + "num_tokens": 481459766.0, + "step": 18606 + }, + { + "epoch": 2.0433779925323963, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.480374574661255, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7141680121421814, + "num_tokens": 481486822.0, + "step": 18607 + }, + { + "epoch": 2.04348781023501, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8460304737091064, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7225654125213623, + "num_tokens": 481508810.0, + "step": 18608 + }, + { + "epoch": 2.0435976279376233, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.524427890777588, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.6990548372268677, + "num_tokens": 481535761.0, + "step": 18609 + }, + { + "epoch": 2.043707445640237, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.2049777507781982, + "learning_rate": 1e-06, + "loss": 0.8649, + "mean_token_accuracy": 0.7376067638397217, + "num_tokens": 481567869.0, + "step": 18610 + }, + { + "epoch": 2.043817263342851, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.50510311126709, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6940008401870728, + "num_tokens": 481596756.0, + "step": 18611 + }, + { + "epoch": 2.0439270810454646, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.68969988822937, + "learning_rate": 1e-06, + "loss": 1.1282, + "mean_token_accuracy": 0.6770795583724976, + "num_tokens": 481625764.0, + "step": 18612 + }, + { + "epoch": 2.0440368987480784, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 16.6883544921875, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6992039084434509, + "num_tokens": 481654419.0, + "step": 18613 + }, + { + "epoch": 2.0441467164506917, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.9333488941192627, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.7196593284606934, + "num_tokens": 481673690.0, + "step": 18614 + }, + { + "epoch": 2.0442565341533054, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6437149047851562, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7158827781677246, + "num_tokens": 481700537.0, + "step": 18615 + }, + { + "epoch": 2.044366351855919, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7691433429718018, + "learning_rate": 1e-06, + "loss": 0.9329, + "mean_token_accuracy": 0.7253308892250061, + "num_tokens": 481721716.0, + "step": 18616 + }, + { + "epoch": 2.044476169558533, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7048797607421875, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7114280462265015, + "num_tokens": 481744694.0, + "step": 18617 + }, + { + "epoch": 2.0445859872611467, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4653890132904053, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7114283442497253, + "num_tokens": 481773619.0, + "step": 18618 + }, + { + "epoch": 2.04469580496376, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7692127227783203, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7208110094070435, + "num_tokens": 481796860.0, + "step": 18619 + }, + { + "epoch": 2.044805622666374, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4818899631500244, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7079728841781616, + "num_tokens": 481824971.0, + "step": 18620 + }, + { + "epoch": 2.0449154403689875, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.592163562774658, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7142820358276367, + "num_tokens": 481851380.0, + "step": 18621 + }, + { + "epoch": 2.0450252580716013, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6161162853240967, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7217357158660889, + "num_tokens": 481875181.0, + "step": 18622 + }, + { + "epoch": 2.0451350757742146, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.50655460357666, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6870980262756348, + "num_tokens": 481901916.0, + "step": 18623 + }, + { + "epoch": 2.0452448934768284, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5231211185455322, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7004001140594482, + "num_tokens": 481929392.0, + "step": 18624 + }, + { + "epoch": 2.045354711179442, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.711423397064209, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7086428999900818, + "num_tokens": 481955024.0, + "step": 18625 + }, + { + "epoch": 2.045464528882056, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.9604930877685547, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7136380672454834, + "num_tokens": 481982868.0, + "step": 18626 + }, + { + "epoch": 2.0455743465846696, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.642852783203125, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.696386456489563, + "num_tokens": 482010241.0, + "step": 18627 + }, + { + "epoch": 2.045684164287283, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4357893466949463, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6916848421096802, + "num_tokens": 482038554.0, + "step": 18628 + }, + { + "epoch": 2.0457939819898967, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.344719409942627, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6961238980293274, + "num_tokens": 482071571.0, + "step": 18629 + }, + { + "epoch": 2.0459037996925105, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.856498956680298, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6891628503799438, + "num_tokens": 482094918.0, + "step": 18630 + }, + { + "epoch": 2.046013617395124, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.0846292972564697, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.6965851187705994, + "num_tokens": 482113080.0, + "step": 18631 + }, + { + "epoch": 2.046123435097738, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.553469657897949, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.723181962966919, + "num_tokens": 482139261.0, + "step": 18632 + }, + { + "epoch": 2.0462332528003513, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.1033542156219482, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7164045572280884, + "num_tokens": 482159933.0, + "step": 18633 + }, + { + "epoch": 2.046343070502965, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8494873046875, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7066732048988342, + "num_tokens": 482184511.0, + "step": 18634 + }, + { + "epoch": 2.046452888205579, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.4275407791137695, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7037086486816406, + "num_tokens": 482214353.0, + "step": 18635 + }, + { + "epoch": 2.0465627059081926, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7200143337249756, + "learning_rate": 1e-06, + "loss": 0.9192, + "mean_token_accuracy": 0.7341436743736267, + "num_tokens": 482238019.0, + "step": 18636 + }, + { + "epoch": 2.046672523610806, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7415919303894043, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7087288498878479, + "num_tokens": 482263259.0, + "step": 18637 + }, + { + "epoch": 2.0467823413134196, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.29724383354187, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7091356515884399, + "num_tokens": 482297651.0, + "step": 18638 + }, + { + "epoch": 2.0468921590160334, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4929206371307373, + "learning_rate": 1e-06, + "loss": 1.1193, + "mean_token_accuracy": 0.6770718097686768, + "num_tokens": 482326127.0, + "step": 18639 + }, + { + "epoch": 2.047001976718647, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6478724479675293, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.7072486281394958, + "num_tokens": 482355648.0, + "step": 18640 + }, + { + "epoch": 2.047111794421261, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.3150322437286377, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7158112525939941, + "num_tokens": 482393914.0, + "step": 18641 + }, + { + "epoch": 2.047221612123874, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5848960876464844, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.699435830116272, + "num_tokens": 482419669.0, + "step": 18642 + }, + { + "epoch": 2.047331429826488, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.7271888256073, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7075352668762207, + "num_tokens": 482441513.0, + "step": 18643 + }, + { + "epoch": 2.0474412475291017, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5371766090393066, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6951006650924683, + "num_tokens": 482467869.0, + "step": 18644 + }, + { + "epoch": 2.0475510652317155, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8308770656585693, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7178175449371338, + "num_tokens": 482496206.0, + "step": 18645 + }, + { + "epoch": 2.047660882934329, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8134372234344482, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6951358318328857, + "num_tokens": 482521093.0, + "step": 18646 + }, + { + "epoch": 2.0477707006369426, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.714409112930298, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7109782695770264, + "num_tokens": 482545337.0, + "step": 18647 + }, + { + "epoch": 2.0478805183395563, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.411694049835205, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7375555038452148, + "num_tokens": 482572245.0, + "step": 18648 + }, + { + "epoch": 2.04799033604217, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.533374071121216, + "learning_rate": 1e-06, + "loss": 0.9604, + "mean_token_accuracy": 0.7198333740234375, + "num_tokens": 482600882.0, + "step": 18649 + }, + { + "epoch": 2.048100153744784, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.63889217376709, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7220958471298218, + "num_tokens": 482626059.0, + "step": 18650 + }, + { + "epoch": 2.048209971447397, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6615912914276123, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6887134313583374, + "num_tokens": 482654852.0, + "step": 18651 + }, + { + "epoch": 2.048319789150011, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5729799270629883, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7274361848831177, + "num_tokens": 482681155.0, + "step": 18652 + }, + { + "epoch": 2.0484296068526247, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.991492986679077, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7284025549888611, + "num_tokens": 482699738.0, + "step": 18653 + }, + { + "epoch": 2.0485394245552384, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.558650016784668, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.70052170753479, + "num_tokens": 482729568.0, + "step": 18654 + }, + { + "epoch": 2.048649242257852, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.711501359939575, + "learning_rate": 1e-06, + "loss": 0.9416, + "mean_token_accuracy": 0.7244563102722168, + "num_tokens": 482752127.0, + "step": 18655 + }, + { + "epoch": 2.0487590599604655, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8971259593963623, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7127231359481812, + "num_tokens": 482775429.0, + "step": 18656 + }, + { + "epoch": 2.0488688776630792, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4875524044036865, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.6954770684242249, + "num_tokens": 482803518.0, + "step": 18657 + }, + { + "epoch": 2.048978695365693, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.97594952583313, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.704982340335846, + "num_tokens": 482829798.0, + "step": 18658 + }, + { + "epoch": 2.0490885130683067, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4946343898773193, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7141100168228149, + "num_tokens": 482858683.0, + "step": 18659 + }, + { + "epoch": 2.04919833077092, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5107479095458984, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7040612101554871, + "num_tokens": 482884621.0, + "step": 18660 + }, + { + "epoch": 2.049308148473534, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7257747650146484, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7110780477523804, + "num_tokens": 482910212.0, + "step": 18661 + }, + { + "epoch": 2.0494179661761476, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.600602149963379, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7249391078948975, + "num_tokens": 482938617.0, + "step": 18662 + }, + { + "epoch": 2.0495277838787613, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4387495517730713, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7129272222518921, + "num_tokens": 482968113.0, + "step": 18663 + }, + { + "epoch": 2.049637601581375, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5681350231170654, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6895074248313904, + "num_tokens": 482998324.0, + "step": 18664 + }, + { + "epoch": 2.0497474192839884, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.513367176055908, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6934163570404053, + "num_tokens": 483025136.0, + "step": 18665 + }, + { + "epoch": 2.049857236986602, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9045262336730957, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7185068130493164, + "num_tokens": 483048026.0, + "step": 18666 + }, + { + "epoch": 2.049967054689216, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.88657546043396, + "learning_rate": 1e-06, + "loss": 0.9053, + "mean_token_accuracy": 0.7315558195114136, + "num_tokens": 483069609.0, + "step": 18667 + }, + { + "epoch": 2.0500768723918297, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6668050289154053, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7170780897140503, + "num_tokens": 483092995.0, + "step": 18668 + }, + { + "epoch": 2.0501866900944434, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.496105194091797, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.7049140930175781, + "num_tokens": 483119666.0, + "step": 18669 + }, + { + "epoch": 2.0502965077970567, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.305427312850952, + "learning_rate": 1e-06, + "loss": 0.9253, + "mean_token_accuracy": 0.741545557975769, + "num_tokens": 483150595.0, + "step": 18670 + }, + { + "epoch": 2.0504063254996705, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.768733024597168, + "learning_rate": 1e-06, + "loss": 0.9311, + "mean_token_accuracy": 0.7203732132911682, + "num_tokens": 483173866.0, + "step": 18671 + }, + { + "epoch": 2.0505161432022843, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.051704168319702, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7017353773117065, + "num_tokens": 483195512.0, + "step": 18672 + }, + { + "epoch": 2.050625960904898, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4169514179229736, + "learning_rate": 1e-06, + "loss": 0.931, + "mean_token_accuracy": 0.7292135953903198, + "num_tokens": 483224278.0, + "step": 18673 + }, + { + "epoch": 2.0507357786075113, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4944705963134766, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7153266668319702, + "num_tokens": 483251606.0, + "step": 18674 + }, + { + "epoch": 2.050845596310125, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.877875566482544, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7049741148948669, + "num_tokens": 483276901.0, + "step": 18675 + }, + { + "epoch": 2.050955414012739, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.563199758529663, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7204519510269165, + "num_tokens": 483304570.0, + "step": 18676 + }, + { + "epoch": 2.0510652317153526, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.52888560295105, + "learning_rate": 1e-06, + "loss": 0.9034, + "mean_token_accuracy": 0.7329893112182617, + "num_tokens": 483330328.0, + "step": 18677 + }, + { + "epoch": 2.0511750494179664, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.707387924194336, + "learning_rate": 1e-06, + "loss": 1.1052, + "mean_token_accuracy": 0.6852728128433228, + "num_tokens": 483356872.0, + "step": 18678 + }, + { + "epoch": 2.0512848671205797, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.8037021160125732, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.6980835199356079, + "num_tokens": 483382817.0, + "step": 18679 + }, + { + "epoch": 2.0513946848231934, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.4633283615112305, + "learning_rate": 1e-06, + "loss": 0.904, + "mean_token_accuracy": 0.7317902445793152, + "num_tokens": 483411726.0, + "step": 18680 + }, + { + "epoch": 2.051504502525807, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.492973804473877, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.710358738899231, + "num_tokens": 483440186.0, + "step": 18681 + }, + { + "epoch": 2.051614320228421, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.686598539352417, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7183947563171387, + "num_tokens": 483463864.0, + "step": 18682 + }, + { + "epoch": 2.0517241379310347, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7136387825012207, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7091503143310547, + "num_tokens": 483488716.0, + "step": 18683 + }, + { + "epoch": 2.051833955633648, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.5197081565856934, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6948837637901306, + "num_tokens": 483516897.0, + "step": 18684 + }, + { + "epoch": 2.0519437733362618, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.7874624729156494, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7180285453796387, + "num_tokens": 483539481.0, + "step": 18685 + }, + { + "epoch": 2.0520535910388755, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 3.086416721343994, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7292469143867493, + "num_tokens": 483562160.0, + "step": 18686 + }, + { + "epoch": 2.0521634087414893, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.8455145359039307, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.701179563999176, + "num_tokens": 483587201.0, + "step": 18687 + }, + { + "epoch": 2.0522732264441026, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.762770414352417, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7127670049667358, + "num_tokens": 483610822.0, + "step": 18688 + }, + { + "epoch": 2.0523830441467164, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.820666790008545, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7034870982170105, + "num_tokens": 483632577.0, + "step": 18689 + }, + { + "epoch": 2.05249286184933, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.534162759780884, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.700226902961731, + "num_tokens": 483659643.0, + "step": 18690 + }, + { + "epoch": 2.052602679551944, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.61965274810791, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7095649838447571, + "num_tokens": 483685053.0, + "step": 18691 + }, + { + "epoch": 2.0527124972545576, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8719711303710938, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7055835723876953, + "num_tokens": 483708837.0, + "step": 18692 + }, + { + "epoch": 2.052822314957171, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.739326000213623, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7109745144844055, + "num_tokens": 483732545.0, + "step": 18693 + }, + { + "epoch": 2.0529321326597847, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.83296799659729, + "learning_rate": 1e-06, + "loss": 0.923, + "mean_token_accuracy": 0.7258394956588745, + "num_tokens": 483755112.0, + "step": 18694 + }, + { + "epoch": 2.0530419503623984, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.00520920753479, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7154406309127808, + "num_tokens": 483777305.0, + "step": 18695 + }, + { + "epoch": 2.053151768065012, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7105555534362793, + "learning_rate": 1e-06, + "loss": 0.881, + "mean_token_accuracy": 0.7357212901115417, + "num_tokens": 483800500.0, + "step": 18696 + }, + { + "epoch": 2.0532615857676255, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.641108512878418, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7122982144355774, + "num_tokens": 483826162.0, + "step": 18697 + }, + { + "epoch": 2.0533714034702393, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5458710193634033, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7142760753631592, + "num_tokens": 483855322.0, + "step": 18698 + }, + { + "epoch": 2.053481221172853, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0120339393615723, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7167537808418274, + "num_tokens": 483877944.0, + "step": 18699 + }, + { + "epoch": 2.053591038875467, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.551302194595337, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7120419144630432, + "num_tokens": 483904372.0, + "step": 18700 + }, + { + "epoch": 2.0537008565780805, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.8765640258789062, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.6862255334854126, + "num_tokens": 483933983.0, + "step": 18701 + }, + { + "epoch": 2.053810674280694, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6832151412963867, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7010059356689453, + "num_tokens": 483960201.0, + "step": 18702 + }, + { + "epoch": 2.0539204919833076, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.450731039047241, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.696130633354187, + "num_tokens": 483990681.0, + "step": 18703 + }, + { + "epoch": 2.0540303096859214, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5135765075683594, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7125825881958008, + "num_tokens": 484019494.0, + "step": 18704 + }, + { + "epoch": 2.054140127388535, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.709824323654175, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7061983346939087, + "num_tokens": 484044720.0, + "step": 18705 + }, + { + "epoch": 2.054249945091149, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.696950674057007, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7098368406295776, + "num_tokens": 484071022.0, + "step": 18706 + }, + { + "epoch": 2.054359762793762, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.921686887741089, + "learning_rate": 1e-06, + "loss": 0.8815, + "mean_token_accuracy": 0.7374048829078674, + "num_tokens": 484089798.0, + "step": 18707 + }, + { + "epoch": 2.054469580496376, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6660943031311035, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7172232270240784, + "num_tokens": 484115280.0, + "step": 18708 + }, + { + "epoch": 2.0545793981989897, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.824070692062378, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7013221383094788, + "num_tokens": 484139932.0, + "step": 18709 + }, + { + "epoch": 2.0546892159016035, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.3474278450012207, + "learning_rate": 1e-06, + "loss": 0.91, + "mean_token_accuracy": 0.7319502830505371, + "num_tokens": 484170720.0, + "step": 18710 + }, + { + "epoch": 2.054799033604217, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6774797439575195, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7024338841438293, + "num_tokens": 484199463.0, + "step": 18711 + }, + { + "epoch": 2.0549088513068305, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.900634765625, + "learning_rate": 1e-06, + "loss": 0.9321, + "mean_token_accuracy": 0.7208530306816101, + "num_tokens": 484223309.0, + "step": 18712 + }, + { + "epoch": 2.0550186690094443, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.606154203414917, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6924849152565002, + "num_tokens": 484252023.0, + "step": 18713 + }, + { + "epoch": 2.055128486712058, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.835176467895508, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7113770246505737, + "num_tokens": 484276029.0, + "step": 18714 + }, + { + "epoch": 2.055238304414672, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5698139667510986, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6904159784317017, + "num_tokens": 484305587.0, + "step": 18715 + }, + { + "epoch": 2.055348122117285, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.11857533454895, + "learning_rate": 1e-06, + "loss": 0.9061, + "mean_token_accuracy": 0.7331440448760986, + "num_tokens": 484324145.0, + "step": 18716 + }, + { + "epoch": 2.055457939819899, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4583189487457275, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.7038176655769348, + "num_tokens": 484356125.0, + "step": 18717 + }, + { + "epoch": 2.0555677575225126, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.499995231628418, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7063885927200317, + "num_tokens": 484385600.0, + "step": 18718 + }, + { + "epoch": 2.0556775752251264, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.649318218231201, + "learning_rate": 1e-06, + "loss": 0.9206, + "mean_token_accuracy": 0.7332234382629395, + "num_tokens": 484410154.0, + "step": 18719 + }, + { + "epoch": 2.05578739292774, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.734279155731201, + "learning_rate": 1e-06, + "loss": 0.9457, + "mean_token_accuracy": 0.7261013984680176, + "num_tokens": 484433103.0, + "step": 18720 + }, + { + "epoch": 2.0558972106303535, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.677694082260132, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7089271545410156, + "num_tokens": 484461954.0, + "step": 18721 + }, + { + "epoch": 2.0560070283329672, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.760319709777832, + "learning_rate": 1e-06, + "loss": 0.8781, + "mean_token_accuracy": 0.7382656335830688, + "num_tokens": 484484988.0, + "step": 18722 + }, + { + "epoch": 2.056116846035581, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.421492338180542, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.7062915563583374, + "num_tokens": 484516484.0, + "step": 18723 + }, + { + "epoch": 2.0562266637381947, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.624574661254883, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7040859460830688, + "num_tokens": 484539532.0, + "step": 18724 + }, + { + "epoch": 2.056336481440808, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.704167604446411, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7069315910339355, + "num_tokens": 484564198.0, + "step": 18725 + }, + { + "epoch": 2.056446299143422, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.644984483718872, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.7012820243835449, + "num_tokens": 484590622.0, + "step": 18726 + }, + { + "epoch": 2.0565561168460356, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5684921741485596, + "learning_rate": 1e-06, + "loss": 1.0946, + "mean_token_accuracy": 0.6803895831108093, + "num_tokens": 484619910.0, + "step": 18727 + }, + { + "epoch": 2.0566659345486493, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0202102661132812, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.714537501335144, + "num_tokens": 484640093.0, + "step": 18728 + }, + { + "epoch": 2.056775752251263, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5382633209228516, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.7001075744628906, + "num_tokens": 484669591.0, + "step": 18729 + }, + { + "epoch": 2.0568855699538764, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7481744289398193, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7082208395004272, + "num_tokens": 484693323.0, + "step": 18730 + }, + { + "epoch": 2.05699538765649, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9808268547058105, + "learning_rate": 1e-06, + "loss": 0.8937, + "mean_token_accuracy": 0.7392064929008484, + "num_tokens": 484713895.0, + "step": 18731 + }, + { + "epoch": 2.057105205359104, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.971964120864868, + "learning_rate": 1e-06, + "loss": 0.8474, + "mean_token_accuracy": 0.7491135597229004, + "num_tokens": 484734883.0, + "step": 18732 + }, + { + "epoch": 2.0572150230617177, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7933402061462402, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7182559967041016, + "num_tokens": 484757425.0, + "step": 18733 + }, + { + "epoch": 2.0573248407643314, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6293458938598633, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7112459540367126, + "num_tokens": 484782516.0, + "step": 18734 + }, + { + "epoch": 2.0574346584669447, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8940157890319824, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7199524641036987, + "num_tokens": 484805520.0, + "step": 18735 + }, + { + "epoch": 2.0575444761695585, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6088359355926514, + "learning_rate": 1e-06, + "loss": 1.1137, + "mean_token_accuracy": 0.6848529577255249, + "num_tokens": 484833321.0, + "step": 18736 + }, + { + "epoch": 2.0576542938721722, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.302539348602295, + "learning_rate": 1e-06, + "loss": 1.1246, + "mean_token_accuracy": 0.6754276752471924, + "num_tokens": 484868316.0, + "step": 18737 + }, + { + "epoch": 2.057764111574786, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.613450765609741, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7153843641281128, + "num_tokens": 484892628.0, + "step": 18738 + }, + { + "epoch": 2.0578739292773993, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5041890144348145, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7066607475280762, + "num_tokens": 484919840.0, + "step": 18739 + }, + { + "epoch": 2.057983746980013, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5947391986846924, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7191981673240662, + "num_tokens": 484949046.0, + "step": 18740 + }, + { + "epoch": 2.058093564682627, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6615986824035645, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6991307139396667, + "num_tokens": 484973166.0, + "step": 18741 + }, + { + "epoch": 2.0582033823852406, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6969051361083984, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.6983895301818848, + "num_tokens": 484997745.0, + "step": 18742 + }, + { + "epoch": 2.0583132000878543, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.424579620361328, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.712063729763031, + "num_tokens": 485030912.0, + "step": 18743 + }, + { + "epoch": 2.0584230177904677, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9491591453552246, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7124597430229187, + "num_tokens": 485059141.0, + "step": 18744 + }, + { + "epoch": 2.0585328354930814, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7358973026275635, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7235319018363953, + "num_tokens": 485083742.0, + "step": 18745 + }, + { + "epoch": 2.058642653195695, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9449374675750732, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7177451848983765, + "num_tokens": 485105674.0, + "step": 18746 + }, + { + "epoch": 2.058752470898309, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.47393798828125, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7103500366210938, + "num_tokens": 485132428.0, + "step": 18747 + }, + { + "epoch": 2.0588622886009227, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0228271484375, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7201251983642578, + "num_tokens": 485155653.0, + "step": 18748 + }, + { + "epoch": 2.058972106303536, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.545341730117798, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7097015380859375, + "num_tokens": 485183472.0, + "step": 18749 + }, + { + "epoch": 2.0590819240061498, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.627070188522339, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7041758894920349, + "num_tokens": 485208921.0, + "step": 18750 + }, + { + "epoch": 2.0591917417087635, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.2995240688323975, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7005503177642822, + "num_tokens": 485243201.0, + "step": 18751 + }, + { + "epoch": 2.0593015594113773, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.733548879623413, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7100048065185547, + "num_tokens": 485274135.0, + "step": 18752 + }, + { + "epoch": 2.0594113771139906, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6233785152435303, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7032565474510193, + "num_tokens": 485303106.0, + "step": 18753 + }, + { + "epoch": 2.0595211948166043, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6989307403564453, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7115439176559448, + "num_tokens": 485326573.0, + "step": 18754 + }, + { + "epoch": 2.059631012519218, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5594990253448486, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7122992277145386, + "num_tokens": 485354403.0, + "step": 18755 + }, + { + "epoch": 2.059740830221832, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.6337246894836426, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7021434903144836, + "num_tokens": 485381116.0, + "step": 18756 + }, + { + "epoch": 2.0598506479244456, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.523552894592285, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7252990007400513, + "num_tokens": 485404852.0, + "step": 18757 + }, + { + "epoch": 2.059960465627059, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4843037128448486, + "learning_rate": 1e-06, + "loss": 0.9275, + "mean_token_accuracy": 0.7245641946792603, + "num_tokens": 485429368.0, + "step": 18758 + }, + { + "epoch": 2.0600702833296727, + "ewc_loss": 1.0073184967041016e-05, + "grad_norm": 2.3125195503234863, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.7021788358688354, + "num_tokens": 485462570.0, + "step": 18759 + }, + { + "epoch": 2.0601801010322864, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6549036502838135, + "learning_rate": 1e-06, + "loss": 0.9218, + "mean_token_accuracy": 0.729157567024231, + "num_tokens": 485486324.0, + "step": 18760 + }, + { + "epoch": 2.0602899187349, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.823523998260498, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7127964496612549, + "num_tokens": 485514079.0, + "step": 18761 + }, + { + "epoch": 2.060399736437514, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.0919137001037598, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6925531625747681, + "num_tokens": 485539712.0, + "step": 18762 + }, + { + "epoch": 2.0605095541401273, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.584775447845459, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6965929269790649, + "num_tokens": 485565998.0, + "step": 18763 + }, + { + "epoch": 2.060619371842741, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7515807151794434, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6991070508956909, + "num_tokens": 485590371.0, + "step": 18764 + }, + { + "epoch": 2.060729189545355, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7151825428009033, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7155247330665588, + "num_tokens": 485613661.0, + "step": 18765 + }, + { + "epoch": 2.0608390072479685, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.574885368347168, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.7138479948043823, + "num_tokens": 485640582.0, + "step": 18766 + }, + { + "epoch": 2.060948824950582, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5037121772766113, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6942138671875, + "num_tokens": 485667177.0, + "step": 18767 + }, + { + "epoch": 2.0610586426531956, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5408263206481934, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7215175628662109, + "num_tokens": 485692337.0, + "step": 18768 + }, + { + "epoch": 2.0611684603558094, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.650947332382202, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7029247879981995, + "num_tokens": 485715895.0, + "step": 18769 + }, + { + "epoch": 2.061278278058423, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5889835357666016, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7038543224334717, + "num_tokens": 485740056.0, + "step": 18770 + }, + { + "epoch": 2.061388095761037, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.562502861022949, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7176464200019836, + "num_tokens": 485765531.0, + "step": 18771 + }, + { + "epoch": 2.06149791346365, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.489182710647583, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7139263153076172, + "num_tokens": 485790828.0, + "step": 18772 + }, + { + "epoch": 2.061607731166264, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.506472110748291, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6923065185546875, + "num_tokens": 485821307.0, + "step": 18773 + }, + { + "epoch": 2.0617175488688777, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6543333530426025, + "learning_rate": 1e-06, + "loss": 0.9261, + "mean_token_accuracy": 0.7263111472129822, + "num_tokens": 485842368.0, + "step": 18774 + }, + { + "epoch": 2.0618273665714915, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5100724697113037, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7092198133468628, + "num_tokens": 485869451.0, + "step": 18775 + }, + { + "epoch": 2.0619371842741048, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6619534492492676, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7023271918296814, + "num_tokens": 485894433.0, + "step": 18776 + }, + { + "epoch": 2.0620470019767185, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.9053759574890137, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7232829332351685, + "num_tokens": 485915047.0, + "step": 18777 + }, + { + "epoch": 2.0621568196793323, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6163692474365234, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7039840221405029, + "num_tokens": 485942877.0, + "step": 18778 + }, + { + "epoch": 2.062266637381946, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.750208854675293, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7100785374641418, + "num_tokens": 485966534.0, + "step": 18779 + }, + { + "epoch": 2.06237645508456, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 3.028878688812256, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7228295803070068, + "num_tokens": 485985634.0, + "step": 18780 + }, + { + "epoch": 2.062486272787173, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7258243560791016, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.708304762840271, + "num_tokens": 486012183.0, + "step": 18781 + }, + { + "epoch": 2.062596090489787, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5699453353881836, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.6980103254318237, + "num_tokens": 486041381.0, + "step": 18782 + }, + { + "epoch": 2.0627059081924006, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5784854888916016, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7180178165435791, + "num_tokens": 486067899.0, + "step": 18783 + }, + { + "epoch": 2.0628157258950144, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.665473699569702, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7116062045097351, + "num_tokens": 486090785.0, + "step": 18784 + }, + { + "epoch": 2.062925543597628, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6791200637817383, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7082105875015259, + "num_tokens": 486117247.0, + "step": 18785 + }, + { + "epoch": 2.0630353613002415, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5944530963897705, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7101631164550781, + "num_tokens": 486142138.0, + "step": 18786 + }, + { + "epoch": 2.063145179002855, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.9929208755493164, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.7220320701599121, + "num_tokens": 486162436.0, + "step": 18787 + }, + { + "epoch": 2.063254996705469, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.965555429458618, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7304621934890747, + "num_tokens": 486183170.0, + "step": 18788 + }, + { + "epoch": 2.0633648144080827, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.232745409011841, + "learning_rate": 1e-06, + "loss": 1.0637, + "mean_token_accuracy": 0.6917728185653687, + "num_tokens": 486219250.0, + "step": 18789 + }, + { + "epoch": 2.063474632110696, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 3.5343072414398193, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7005339860916138, + "num_tokens": 486248913.0, + "step": 18790 + }, + { + "epoch": 2.06358444981331, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.822627305984497, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7216994762420654, + "num_tokens": 486270415.0, + "step": 18791 + }, + { + "epoch": 2.0636942675159236, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.68782114982605, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6889461278915405, + "num_tokens": 486298050.0, + "step": 18792 + }, + { + "epoch": 2.0638040852185373, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.757014036178589, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7018424272537231, + "num_tokens": 486321495.0, + "step": 18793 + }, + { + "epoch": 2.063913902921151, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.723926544189453, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7168107032775879, + "num_tokens": 486346931.0, + "step": 18794 + }, + { + "epoch": 2.0640237206237644, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.976327896118164, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6961370706558228, + "num_tokens": 486374258.0, + "step": 18795 + }, + { + "epoch": 2.064133538326378, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5722947120666504, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.6997290253639221, + "num_tokens": 486401074.0, + "step": 18796 + }, + { + "epoch": 2.064243356028992, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.707012176513672, + "learning_rate": 1e-06, + "loss": 0.9471, + "mean_token_accuracy": 0.7191380262374878, + "num_tokens": 486425411.0, + "step": 18797 + }, + { + "epoch": 2.0643531737316057, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.598045825958252, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7074069976806641, + "num_tokens": 486450354.0, + "step": 18798 + }, + { + "epoch": 2.0644629914342194, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.398146390914917, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7213650941848755, + "num_tokens": 486479627.0, + "step": 18799 + }, + { + "epoch": 2.0645728091368327, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.700385093688965, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7117756605148315, + "num_tokens": 486502869.0, + "step": 18800 + }, + { + "epoch": 2.0646826268394465, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.468461751937866, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.696096658706665, + "num_tokens": 486534327.0, + "step": 18801 + }, + { + "epoch": 2.0647924445420602, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6853761672973633, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6946001648902893, + "num_tokens": 486559490.0, + "step": 18802 + }, + { + "epoch": 2.064902262244674, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.899763345718384, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7111759781837463, + "num_tokens": 486579601.0, + "step": 18803 + }, + { + "epoch": 2.0650120799472873, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.799103021621704, + "learning_rate": 1e-06, + "loss": 0.914, + "mean_token_accuracy": 0.7336369752883911, + "num_tokens": 486600957.0, + "step": 18804 + }, + { + "epoch": 2.065121897649901, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.656891345977783, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7142558097839355, + "num_tokens": 486627356.0, + "step": 18805 + }, + { + "epoch": 2.065231715352515, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7513229846954346, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7136122584342957, + "num_tokens": 486653002.0, + "step": 18806 + }, + { + "epoch": 2.0653415330551286, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.628237247467041, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.7229273915290833, + "num_tokens": 486677465.0, + "step": 18807 + }, + { + "epoch": 2.0654513507577423, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.50150465965271, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7024750709533691, + "num_tokens": 486704398.0, + "step": 18808 + }, + { + "epoch": 2.0655611684603556, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.3710036277770996, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.7064776420593262, + "num_tokens": 486735319.0, + "step": 18809 + }, + { + "epoch": 2.0656709861629694, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 3.0100719928741455, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7009608745574951, + "num_tokens": 486756914.0, + "step": 18810 + }, + { + "epoch": 2.065780803865583, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6031181812286377, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7237720489501953, + "num_tokens": 486781933.0, + "step": 18811 + }, + { + "epoch": 2.065890621568197, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.463486909866333, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7067025899887085, + "num_tokens": 486808660.0, + "step": 18812 + }, + { + "epoch": 2.0660004392708107, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5065159797668457, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6963823437690735, + "num_tokens": 486844039.0, + "step": 18813 + }, + { + "epoch": 2.066110256973424, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.413313865661621, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7064919471740723, + "num_tokens": 486878633.0, + "step": 18814 + }, + { + "epoch": 2.0662200746760377, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4449198246002197, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7206889986991882, + "num_tokens": 486907779.0, + "step": 18815 + }, + { + "epoch": 2.0663298923786515, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5526981353759766, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6954935789108276, + "num_tokens": 486938039.0, + "step": 18816 + }, + { + "epoch": 2.0664397100812653, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.455235719680786, + "learning_rate": 1e-06, + "loss": 1.0846, + "mean_token_accuracy": 0.6902773976325989, + "num_tokens": 486967462.0, + "step": 18817 + }, + { + "epoch": 2.0665495277838786, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5516653060913086, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7250386476516724, + "num_tokens": 486994806.0, + "step": 18818 + }, + { + "epoch": 2.0666593454864923, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.5825095176696777, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6914954781532288, + "num_tokens": 487021312.0, + "step": 18819 + }, + { + "epoch": 2.066769163189106, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 3.5918002128601074, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7151042222976685, + "num_tokens": 487048531.0, + "step": 18820 + }, + { + "epoch": 2.06687898089172, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6394550800323486, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7109169960021973, + "num_tokens": 487074614.0, + "step": 18821 + }, + { + "epoch": 2.0669887985943336, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7492926120758057, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7130905389785767, + "num_tokens": 487097613.0, + "step": 18822 + }, + { + "epoch": 2.067098616296947, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6871161460876465, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7019321322441101, + "num_tokens": 487124192.0, + "step": 18823 + }, + { + "epoch": 2.0672084339995607, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.3818628787994385, + "learning_rate": 1e-06, + "loss": 1.1035, + "mean_token_accuracy": 0.6792957782745361, + "num_tokens": 487157659.0, + "step": 18824 + }, + { + "epoch": 2.0673182517021744, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.356544017791748, + "learning_rate": 1e-06, + "loss": 0.8916, + "mean_token_accuracy": 0.7357386946678162, + "num_tokens": 487186665.0, + "step": 18825 + }, + { + "epoch": 2.067428069404788, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.598062753677368, + "learning_rate": 1e-06, + "loss": 0.9035, + "mean_token_accuracy": 0.732140064239502, + "num_tokens": 487213021.0, + "step": 18826 + }, + { + "epoch": 2.0675378871074015, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7471890449523926, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7035621404647827, + "num_tokens": 487236826.0, + "step": 18827 + }, + { + "epoch": 2.0676477048100153, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.798046112060547, + "learning_rate": 1e-06, + "loss": 0.8434, + "mean_token_accuracy": 0.7420099377632141, + "num_tokens": 487257249.0, + "step": 18828 + }, + { + "epoch": 2.067757522512629, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6390345096588135, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7178646326065063, + "num_tokens": 487279719.0, + "step": 18829 + }, + { + "epoch": 2.0678673402152428, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.4981932640075684, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7094422578811646, + "num_tokens": 487305998.0, + "step": 18830 + }, + { + "epoch": 2.0679771579178565, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.436354637145996, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7237539887428284, + "num_tokens": 487335168.0, + "step": 18831 + }, + { + "epoch": 2.06808697562047, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.6224074363708496, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7145061492919922, + "num_tokens": 487359653.0, + "step": 18832 + }, + { + "epoch": 2.0681967933230836, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.899658679962158, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7245112061500549, + "num_tokens": 487379403.0, + "step": 18833 + }, + { + "epoch": 2.0683066110256974, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.716243267059326, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.711225152015686, + "num_tokens": 487403805.0, + "step": 18834 + }, + { + "epoch": 2.068416428728311, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.720353126525879, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7139090299606323, + "num_tokens": 487429302.0, + "step": 18835 + }, + { + "epoch": 2.068526246430925, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.526865243911743, + "learning_rate": 1e-06, + "loss": 0.8963, + "mean_token_accuracy": 0.7340402007102966, + "num_tokens": 487454107.0, + "step": 18836 + }, + { + "epoch": 2.068636064133538, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6573235988616943, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.717118501663208, + "num_tokens": 487480299.0, + "step": 18837 + }, + { + "epoch": 2.068745881836152, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.869286060333252, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.698693037033081, + "num_tokens": 487507207.0, + "step": 18838 + }, + { + "epoch": 2.0688556995387657, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5912020206451416, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7263311147689819, + "num_tokens": 487532758.0, + "step": 18839 + }, + { + "epoch": 2.0689655172413794, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8125014305114746, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7245879173278809, + "num_tokens": 487555940.0, + "step": 18840 + }, + { + "epoch": 2.0690753349439928, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6791696548461914, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.7003511190414429, + "num_tokens": 487581254.0, + "step": 18841 + }, + { + "epoch": 2.0691851526466065, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.927426815032959, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7274644374847412, + "num_tokens": 487600889.0, + "step": 18842 + }, + { + "epoch": 2.0692949703492203, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 3.0888705253601074, + "learning_rate": 1e-06, + "loss": 0.9065, + "mean_token_accuracy": 0.7336221933364868, + "num_tokens": 487620106.0, + "step": 18843 + }, + { + "epoch": 2.069404788051834, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9999516010284424, + "learning_rate": 1e-06, + "loss": 0.9117, + "mean_token_accuracy": 0.7326778769493103, + "num_tokens": 487652019.0, + "step": 18844 + }, + { + "epoch": 2.069514605754448, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4986491203308105, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7119165658950806, + "num_tokens": 487682072.0, + "step": 18845 + }, + { + "epoch": 2.069624423457061, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.425136089324951, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7041226029396057, + "num_tokens": 487712303.0, + "step": 18846 + }, + { + "epoch": 2.069734241159675, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7735652923583984, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7191374897956848, + "num_tokens": 487733598.0, + "step": 18847 + }, + { + "epoch": 2.0698440588622886, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5479156970977783, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7193487882614136, + "num_tokens": 487758116.0, + "step": 18848 + }, + { + "epoch": 2.0699538765649024, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5628905296325684, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6957432627677917, + "num_tokens": 487784928.0, + "step": 18849 + }, + { + "epoch": 2.070063694267516, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6963281631469727, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7116485834121704, + "num_tokens": 487806999.0, + "step": 18850 + }, + { + "epoch": 2.0701735119701294, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6690499782562256, + "learning_rate": 1e-06, + "loss": 0.8911, + "mean_token_accuracy": 0.7315724492073059, + "num_tokens": 487829487.0, + "step": 18851 + }, + { + "epoch": 2.070283329672743, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8102877140045166, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.7030133605003357, + "num_tokens": 487857701.0, + "step": 18852 + }, + { + "epoch": 2.070393147375357, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.417287588119507, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7049661874771118, + "num_tokens": 487886019.0, + "step": 18853 + }, + { + "epoch": 2.0705029650779707, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.7733819484710693, + "learning_rate": 1e-06, + "loss": 0.9135, + "mean_token_accuracy": 0.7328646779060364, + "num_tokens": 487906054.0, + "step": 18854 + }, + { + "epoch": 2.070612782780584, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4575204849243164, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7218503952026367, + "num_tokens": 487932732.0, + "step": 18855 + }, + { + "epoch": 2.070722600483198, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8208131790161133, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.6989176273345947, + "num_tokens": 487954950.0, + "step": 18856 + }, + { + "epoch": 2.0708324181858115, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.2326176166534424, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7077929973602295, + "num_tokens": 487989789.0, + "step": 18857 + }, + { + "epoch": 2.0709422358884253, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.635913610458374, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7098312377929688, + "num_tokens": 488013438.0, + "step": 18858 + }, + { + "epoch": 2.071052053591039, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8907530307769775, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7133017778396606, + "num_tokens": 488033967.0, + "step": 18859 + }, + { + "epoch": 2.0711618712936524, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8914990425109863, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7134532332420349, + "num_tokens": 488055618.0, + "step": 18860 + }, + { + "epoch": 2.071271688996266, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 3.2633056640625, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7137411832809448, + "num_tokens": 488077824.0, + "step": 18861 + }, + { + "epoch": 2.07138150669888, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6103408336639404, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6981300115585327, + "num_tokens": 488103374.0, + "step": 18862 + }, + { + "epoch": 2.0714913244014936, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5610272884368896, + "learning_rate": 1e-06, + "loss": 0.921, + "mean_token_accuracy": 0.7262622117996216, + "num_tokens": 488129776.0, + "step": 18863 + }, + { + "epoch": 2.0716011421041074, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.655714273452759, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7010836005210876, + "num_tokens": 488154567.0, + "step": 18864 + }, + { + "epoch": 2.0717109598067207, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8143396377563477, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7149088382720947, + "num_tokens": 488175092.0, + "step": 18865 + }, + { + "epoch": 2.0718207775093345, + "ewc_loss": 1.0132789611816406e-05, + "grad_norm": 2.577101945877075, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6860570311546326, + "num_tokens": 488202329.0, + "step": 18866 + }, + { + "epoch": 2.0719305952119482, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5270493030548096, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7036103010177612, + "num_tokens": 488229448.0, + "step": 18867 + }, + { + "epoch": 2.072040412914562, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.002471923828125, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.71451336145401, + "num_tokens": 488248341.0, + "step": 18868 + }, + { + "epoch": 2.0721502306171753, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.502633810043335, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.690256655216217, + "num_tokens": 488276403.0, + "step": 18869 + }, + { + "epoch": 2.072260048319789, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8249588012695312, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.697210431098938, + "num_tokens": 488300829.0, + "step": 18870 + }, + { + "epoch": 2.072369866022403, + "ewc_loss": 1.0728836059570312e-05, + "grad_norm": 80.53890991210938, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.7011973261833191, + "num_tokens": 488330090.0, + "step": 18871 + }, + { + "epoch": 2.0724796837250166, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7028870582580566, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7144365310668945, + "num_tokens": 488357430.0, + "step": 18872 + }, + { + "epoch": 2.0725895014276303, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.736528158187866, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7090044021606445, + "num_tokens": 488382389.0, + "step": 18873 + }, + { + "epoch": 2.0726993191302436, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.766758441925049, + "learning_rate": 1e-06, + "loss": 0.982, + "mean_token_accuracy": 0.7236614227294922, + "num_tokens": 488405819.0, + "step": 18874 + }, + { + "epoch": 2.0728091368328574, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6922764778137207, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7055689096450806, + "num_tokens": 488428902.0, + "step": 18875 + }, + { + "epoch": 2.072918954535471, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.660592794418335, + "learning_rate": 1e-06, + "loss": 0.9149, + "mean_token_accuracy": 0.7299160957336426, + "num_tokens": 488451267.0, + "step": 18876 + }, + { + "epoch": 2.073028772238085, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5070455074310303, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7150033712387085, + "num_tokens": 488477786.0, + "step": 18877 + }, + { + "epoch": 2.073138589940698, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6330199241638184, + "learning_rate": 1e-06, + "loss": 0.935, + "mean_token_accuracy": 0.7247593998908997, + "num_tokens": 488501779.0, + "step": 18878 + }, + { + "epoch": 2.073248407643312, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6184206008911133, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7079163193702698, + "num_tokens": 488525927.0, + "step": 18879 + }, + { + "epoch": 2.0733582253459257, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.808995008468628, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7117328643798828, + "num_tokens": 488549374.0, + "step": 18880 + }, + { + "epoch": 2.0734680430485395, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.41882586479187, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6978633403778076, + "num_tokens": 488581234.0, + "step": 18881 + }, + { + "epoch": 2.0735778607511532, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.79838228225708, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7160818576812744, + "num_tokens": 488604300.0, + "step": 18882 + }, + { + "epoch": 2.0736876784537666, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.58773136138916, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7045631408691406, + "num_tokens": 488628493.0, + "step": 18883 + }, + { + "epoch": 2.0737974961563803, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8295977115631104, + "learning_rate": 1e-06, + "loss": 0.9124, + "mean_token_accuracy": 0.7307091951370239, + "num_tokens": 488649579.0, + "step": 18884 + }, + { + "epoch": 2.073907313858994, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.529245615005493, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6978271007537842, + "num_tokens": 488679453.0, + "step": 18885 + }, + { + "epoch": 2.074017131561608, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4346916675567627, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7167028188705444, + "num_tokens": 488706913.0, + "step": 18886 + }, + { + "epoch": 2.0741269492642216, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.522294282913208, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7172395586967468, + "num_tokens": 488733613.0, + "step": 18887 + }, + { + "epoch": 2.074236766966835, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.569528579711914, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.7003829479217529, + "num_tokens": 488760801.0, + "step": 18888 + }, + { + "epoch": 2.0743465846694487, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.564222812652588, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7008645534515381, + "num_tokens": 488788687.0, + "step": 18889 + }, + { + "epoch": 2.0744564023720624, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5892107486724854, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7118926048278809, + "num_tokens": 488814072.0, + "step": 18890 + }, + { + "epoch": 2.074566220074676, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.574744462966919, + "learning_rate": 1e-06, + "loss": 0.9171, + "mean_token_accuracy": 0.7369585037231445, + "num_tokens": 488841485.0, + "step": 18891 + }, + { + "epoch": 2.07467603777729, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.609071731567383, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7163265943527222, + "num_tokens": 488867353.0, + "step": 18892 + }, + { + "epoch": 2.0747858554799032, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8691864013671875, + "learning_rate": 1e-06, + "loss": 1.0974, + "mean_token_accuracy": 0.681613028049469, + "num_tokens": 488896078.0, + "step": 18893 + }, + { + "epoch": 2.074895673182517, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.881133794784546, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7081345915794373, + "num_tokens": 488917064.0, + "step": 18894 + }, + { + "epoch": 2.0750054908851308, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.545156240463257, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7050237655639648, + "num_tokens": 488945488.0, + "step": 18895 + }, + { + "epoch": 2.0751153085877445, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.652088165283203, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7165853977203369, + "num_tokens": 488968188.0, + "step": 18896 + }, + { + "epoch": 2.075225126290358, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9942612648010254, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7195136547088623, + "num_tokens": 488989368.0, + "step": 18897 + }, + { + "epoch": 2.0753349439929716, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4730031490325928, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7031989097595215, + "num_tokens": 489019061.0, + "step": 18898 + }, + { + "epoch": 2.0754447616955853, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4613163471221924, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6973927021026611, + "num_tokens": 489049187.0, + "step": 18899 + }, + { + "epoch": 2.075554579398199, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.786062240600586, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7183228731155396, + "num_tokens": 489072262.0, + "step": 18900 + }, + { + "epoch": 2.075664397100813, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6356201171875, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7141270041465759, + "num_tokens": 489098065.0, + "step": 18901 + }, + { + "epoch": 2.075774214803426, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6149184703826904, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7122993469238281, + "num_tokens": 489122738.0, + "step": 18902 + }, + { + "epoch": 2.07588403250604, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.661245584487915, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6972047090530396, + "num_tokens": 489149658.0, + "step": 18903 + }, + { + "epoch": 2.0759938502086537, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7732439041137695, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7233526706695557, + "num_tokens": 489170115.0, + "step": 18904 + }, + { + "epoch": 2.0761036679112674, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4933269023895264, + "learning_rate": 1e-06, + "loss": 1.0833, + "mean_token_accuracy": 0.6869477033615112, + "num_tokens": 489198899.0, + "step": 18905 + }, + { + "epoch": 2.0762134856138807, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7714433670043945, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7074720859527588, + "num_tokens": 489220994.0, + "step": 18906 + }, + { + "epoch": 2.0763233033164945, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5455284118652344, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.704124927520752, + "num_tokens": 489247684.0, + "step": 18907 + }, + { + "epoch": 2.0764331210191083, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6611294746398926, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7186364531517029, + "num_tokens": 489274000.0, + "step": 18908 + }, + { + "epoch": 2.076542938721722, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5801782608032227, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.715985119342804, + "num_tokens": 489301138.0, + "step": 18909 + }, + { + "epoch": 2.0766527564243358, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.646286725997925, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7144149541854858, + "num_tokens": 489323949.0, + "step": 18910 + }, + { + "epoch": 2.076762574126949, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.638767719268799, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7065883278846741, + "num_tokens": 489348177.0, + "step": 18911 + }, + { + "epoch": 2.076872391829563, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.765766143798828, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7009344100952148, + "num_tokens": 489370653.0, + "step": 18912 + }, + { + "epoch": 2.0769822095321766, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.660134792327881, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6987802982330322, + "num_tokens": 489396076.0, + "step": 18913 + }, + { + "epoch": 2.0770920272347904, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.2464540004730225, + "learning_rate": 1e-06, + "loss": 0.8882, + "mean_token_accuracy": 0.7442946434020996, + "num_tokens": 489422322.0, + "step": 18914 + }, + { + "epoch": 2.077201844937404, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5756232738494873, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6966814994812012, + "num_tokens": 489452241.0, + "step": 18915 + }, + { + "epoch": 2.0773116626400174, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.955744981765747, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7174662351608276, + "num_tokens": 489471717.0, + "step": 18916 + }, + { + "epoch": 2.077421480342631, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.677460193634033, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.7337023019790649, + "num_tokens": 489493599.0, + "step": 18917 + }, + { + "epoch": 2.077531298045245, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.6934564113616943, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6918835043907166, + "num_tokens": 489517991.0, + "step": 18918 + }, + { + "epoch": 2.0776411157478587, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3970844745635986, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6873970627784729, + "num_tokens": 489550331.0, + "step": 18919 + }, + { + "epoch": 2.077750933450472, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4914629459381104, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7009029984474182, + "num_tokens": 489576755.0, + "step": 18920 + }, + { + "epoch": 2.0778607511530858, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6647536754608154, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.7186647653579712, + "num_tokens": 489603232.0, + "step": 18921 + }, + { + "epoch": 2.0779705688556995, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.405107259750366, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.688679575920105, + "num_tokens": 489633861.0, + "step": 18922 + }, + { + "epoch": 2.0780803865583133, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.9188551902770996, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7057668566703796, + "num_tokens": 489656715.0, + "step": 18923 + }, + { + "epoch": 2.078190204260927, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8838348388671875, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7156139612197876, + "num_tokens": 489678745.0, + "step": 18924 + }, + { + "epoch": 2.0783000219635404, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.480469226837158, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7056695222854614, + "num_tokens": 489705042.0, + "step": 18925 + }, + { + "epoch": 2.078409839666154, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.646214246749878, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7111684083938599, + "num_tokens": 489732389.0, + "step": 18926 + }, + { + "epoch": 2.078519657368768, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7413289546966553, + "learning_rate": 1e-06, + "loss": 1.0954, + "mean_token_accuracy": 0.6812376976013184, + "num_tokens": 489756253.0, + "step": 18927 + }, + { + "epoch": 2.0786294750713816, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7005183696746826, + "learning_rate": 1e-06, + "loss": 0.908, + "mean_token_accuracy": 0.7292916178703308, + "num_tokens": 489779374.0, + "step": 18928 + }, + { + "epoch": 2.078739292773995, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.756462812423706, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7032109498977661, + "num_tokens": 489802145.0, + "step": 18929 + }, + { + "epoch": 2.0788491104766087, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4877922534942627, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6981229186058044, + "num_tokens": 489832481.0, + "step": 18930 + }, + { + "epoch": 2.0789589281792225, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6591315269470215, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7006591558456421, + "num_tokens": 489855325.0, + "step": 18931 + }, + { + "epoch": 2.079068745881836, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.716649055480957, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7065886855125427, + "num_tokens": 489880338.0, + "step": 18932 + }, + { + "epoch": 2.07917856358445, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7058005332946777, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.717552661895752, + "num_tokens": 489904091.0, + "step": 18933 + }, + { + "epoch": 2.0792883812870633, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8235678672790527, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.704806387424469, + "num_tokens": 489928676.0, + "step": 18934 + }, + { + "epoch": 2.079398198989677, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3780548572540283, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7084385752677917, + "num_tokens": 489956846.0, + "step": 18935 + }, + { + "epoch": 2.079508016692291, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6314430236816406, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.7050997614860535, + "num_tokens": 489981568.0, + "step": 18936 + }, + { + "epoch": 2.0796178343949046, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.543952703475952, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7218879461288452, + "num_tokens": 490007835.0, + "step": 18937 + }, + { + "epoch": 2.0797276520975183, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8433361053466797, + "learning_rate": 1e-06, + "loss": 0.7953, + "mean_token_accuracy": 0.7645035982131958, + "num_tokens": 490027937.0, + "step": 18938 + }, + { + "epoch": 2.0798374698001316, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3180577754974365, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6983168125152588, + "num_tokens": 490063001.0, + "step": 18939 + }, + { + "epoch": 2.0799472875027454, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5964391231536865, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7219159603118896, + "num_tokens": 490088863.0, + "step": 18940 + }, + { + "epoch": 2.080057105205359, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8189780712127686, + "learning_rate": 1e-06, + "loss": 0.8724, + "mean_token_accuracy": 0.7399966716766357, + "num_tokens": 490110820.0, + "step": 18941 + }, + { + "epoch": 2.080166922907973, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4355504512786865, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6993204355239868, + "num_tokens": 490140847.0, + "step": 18942 + }, + { + "epoch": 2.0802767406105866, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5428786277770996, + "learning_rate": 1e-06, + "loss": 1.1346, + "mean_token_accuracy": 0.6844685077667236, + "num_tokens": 490166679.0, + "step": 18943 + }, + { + "epoch": 2.0803865583132, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.874070167541504, + "learning_rate": 1e-06, + "loss": 0.9404, + "mean_token_accuracy": 0.7175896167755127, + "num_tokens": 490187894.0, + "step": 18944 + }, + { + "epoch": 2.0804963760158137, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.212444543838501, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7069872617721558, + "num_tokens": 490221796.0, + "step": 18945 + }, + { + "epoch": 2.0806061937184275, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7310469150543213, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.707048237323761, + "num_tokens": 490243502.0, + "step": 18946 + }, + { + "epoch": 2.0807160114210412, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4899919033050537, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7225744724273682, + "num_tokens": 490271915.0, + "step": 18947 + }, + { + "epoch": 2.0808258291236545, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.434692144393921, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.711785078048706, + "num_tokens": 490300156.0, + "step": 18948 + }, + { + "epoch": 2.0809356468262683, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4044137001037598, + "learning_rate": 1e-06, + "loss": 1.0986, + "mean_token_accuracy": 0.6806545257568359, + "num_tokens": 490330308.0, + "step": 18949 + }, + { + "epoch": 2.081045464528882, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4010350704193115, + "learning_rate": 1e-06, + "loss": 1.1228, + "mean_token_accuracy": 0.6933448910713196, + "num_tokens": 490358907.0, + "step": 18950 + }, + { + "epoch": 2.081155282231496, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.642836809158325, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7002561688423157, + "num_tokens": 490383541.0, + "step": 18951 + }, + { + "epoch": 2.0812650999341096, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.507335901260376, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7156574726104736, + "num_tokens": 490409125.0, + "step": 18952 + }, + { + "epoch": 2.081374917636723, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6694724559783936, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7171198129653931, + "num_tokens": 490432306.0, + "step": 18953 + }, + { + "epoch": 2.0814847353393366, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.900543689727783, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7016140818595886, + "num_tokens": 490454009.0, + "step": 18954 + }, + { + "epoch": 2.0815945530419504, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9469172954559326, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.698005735874176, + "num_tokens": 490482565.0, + "step": 18955 + }, + { + "epoch": 2.081704370744564, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4309637546539307, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6866270303726196, + "num_tokens": 490519834.0, + "step": 18956 + }, + { + "epoch": 2.0818141884471775, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5443549156188965, + "learning_rate": 1e-06, + "loss": 0.8889, + "mean_token_accuracy": 0.7346216440200806, + "num_tokens": 490546074.0, + "step": 18957 + }, + { + "epoch": 2.0819240061497912, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.59309458732605, + "learning_rate": 1e-06, + "loss": 0.8608, + "mean_token_accuracy": 0.7365238070487976, + "num_tokens": 490568911.0, + "step": 18958 + }, + { + "epoch": 2.082033823852405, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.932529926300049, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6932616233825684, + "num_tokens": 490594591.0, + "step": 18959 + }, + { + "epoch": 2.0821436415550187, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.686856746673584, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7115527987480164, + "num_tokens": 490619546.0, + "step": 18960 + }, + { + "epoch": 2.0822534592576325, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5391829013824463, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7125033736228943, + "num_tokens": 490648240.0, + "step": 18961 + }, + { + "epoch": 2.082363276960246, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.784797191619873, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7072534561157227, + "num_tokens": 490672451.0, + "step": 18962 + }, + { + "epoch": 2.0824730946628596, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6173007488250732, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7230291366577148, + "num_tokens": 490698015.0, + "step": 18963 + }, + { + "epoch": 2.0825829123654733, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.616556167602539, + "learning_rate": 1e-06, + "loss": 0.9269, + "mean_token_accuracy": 0.7255810499191284, + "num_tokens": 490721407.0, + "step": 18964 + }, + { + "epoch": 2.082692730068087, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7469112873077393, + "learning_rate": 1e-06, + "loss": 0.8988, + "mean_token_accuracy": 0.732830286026001, + "num_tokens": 490743600.0, + "step": 18965 + }, + { + "epoch": 2.082802547770701, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.704373359680176, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.709321141242981, + "num_tokens": 490766922.0, + "step": 18966 + }, + { + "epoch": 2.082912365473314, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8987083435058594, + "learning_rate": 1e-06, + "loss": 0.9183, + "mean_token_accuracy": 0.724724292755127, + "num_tokens": 490787071.0, + "step": 18967 + }, + { + "epoch": 2.083022183175928, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.403841257095337, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7130873799324036, + "num_tokens": 490818928.0, + "step": 18968 + }, + { + "epoch": 2.0831320008785417, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.478076457977295, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.6987279057502747, + "num_tokens": 490845146.0, + "step": 18969 + }, + { + "epoch": 2.0832418185811554, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.678422689437866, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.706947922706604, + "num_tokens": 490870347.0, + "step": 18970 + }, + { + "epoch": 2.0833516362837687, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.837883234024048, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.7039774656295776, + "num_tokens": 490897708.0, + "step": 18971 + }, + { + "epoch": 2.0834614539863825, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.487154006958008, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6972754001617432, + "num_tokens": 490926444.0, + "step": 18972 + }, + { + "epoch": 2.0835712716889963, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.634610652923584, + "learning_rate": 1e-06, + "loss": 0.9268, + "mean_token_accuracy": 0.7235714793205261, + "num_tokens": 490951897.0, + "step": 18973 + }, + { + "epoch": 2.08368108939161, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.2227365970611572, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6876958012580872, + "num_tokens": 490987745.0, + "step": 18974 + }, + { + "epoch": 2.0837909070942238, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3755345344543457, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6939451098442078, + "num_tokens": 491018623.0, + "step": 18975 + }, + { + "epoch": 2.083900724796837, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5229296684265137, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7132041454315186, + "num_tokens": 491042741.0, + "step": 18976 + }, + { + "epoch": 2.084010542499451, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5640344619750977, + "learning_rate": 1e-06, + "loss": 1.0006, + "mean_token_accuracy": 0.7099236249923706, + "num_tokens": 491068750.0, + "step": 18977 + }, + { + "epoch": 2.0841203602020646, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7430787086486816, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.6913489103317261, + "num_tokens": 491092239.0, + "step": 18978 + }, + { + "epoch": 2.0842301779046783, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.958183765411377, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.725228488445282, + "num_tokens": 491114475.0, + "step": 18979 + }, + { + "epoch": 2.084339995607292, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.890207290649414, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.7241989970207214, + "num_tokens": 491134611.0, + "step": 18980 + }, + { + "epoch": 2.0844498133099054, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8797810077667236, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7117431163787842, + "num_tokens": 491155132.0, + "step": 18981 + }, + { + "epoch": 2.084559631012519, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.797699213027954, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7209469079971313, + "num_tokens": 491178243.0, + "step": 18982 + }, + { + "epoch": 2.084669448715133, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.510408639907837, + "learning_rate": 1e-06, + "loss": 1.1321, + "mean_token_accuracy": 0.6861540079116821, + "num_tokens": 491209504.0, + "step": 18983 + }, + { + "epoch": 2.0847792664177467, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9089102745056152, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7086893916130066, + "num_tokens": 491234330.0, + "step": 18984 + }, + { + "epoch": 2.08488908412036, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5504608154296875, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7099457383155823, + "num_tokens": 491262831.0, + "step": 18985 + }, + { + "epoch": 2.0849989018229738, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.665055274963379, + "learning_rate": 1e-06, + "loss": 1.0905, + "mean_token_accuracy": 0.6862456798553467, + "num_tokens": 491290913.0, + "step": 18986 + }, + { + "epoch": 2.0851087195255875, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.23120379447937, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6906238794326782, + "num_tokens": 491314614.0, + "step": 18987 + }, + { + "epoch": 2.0852185372282013, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6879608631134033, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6920599341392517, + "num_tokens": 491344133.0, + "step": 18988 + }, + { + "epoch": 2.085328354930815, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.569406032562256, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7084077596664429, + "num_tokens": 491370174.0, + "step": 18989 + }, + { + "epoch": 2.0854381726334283, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5235519409179688, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7156322598457336, + "num_tokens": 491396116.0, + "step": 18990 + }, + { + "epoch": 2.085547990336042, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.3245794773101807, + "learning_rate": 1e-06, + "loss": 1.1204, + "mean_token_accuracy": 0.6825066804885864, + "num_tokens": 491430346.0, + "step": 18991 + }, + { + "epoch": 2.085657808038656, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.486294746398926, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7098212838172913, + "num_tokens": 491458979.0, + "step": 18992 + }, + { + "epoch": 2.0857676257412696, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6870052814483643, + "learning_rate": 1e-06, + "loss": 0.942, + "mean_token_accuracy": 0.7245096564292908, + "num_tokens": 491484176.0, + "step": 18993 + }, + { + "epoch": 2.0858774434438834, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.621706485748291, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7048680186271667, + "num_tokens": 491509964.0, + "step": 18994 + }, + { + "epoch": 2.0859872611464967, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.97597074508667, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7093602418899536, + "num_tokens": 491531663.0, + "step": 18995 + }, + { + "epoch": 2.0860970788491104, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9157283306121826, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6997553110122681, + "num_tokens": 491554738.0, + "step": 18996 + }, + { + "epoch": 2.086206896551724, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7308919429779053, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.719517707824707, + "num_tokens": 491578738.0, + "step": 18997 + }, + { + "epoch": 2.086316714254338, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6084752082824707, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6919032335281372, + "num_tokens": 491606169.0, + "step": 18998 + }, + { + "epoch": 2.0864265319569513, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 3.051114082336426, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7206810116767883, + "num_tokens": 491634519.0, + "step": 18999 + }, + { + "epoch": 2.086536349659565, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.636333465576172, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6864327192306519, + "num_tokens": 491664802.0, + "step": 19000 + }, + { + "epoch": 2.086646167362179, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.236156463623047, + "learning_rate": 1e-06, + "loss": 0.979, + "mean_token_accuracy": 0.7226859331130981, + "num_tokens": 491699769.0, + "step": 19001 + }, + { + "epoch": 2.0867559850647925, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5418143272399902, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7085860967636108, + "num_tokens": 491726122.0, + "step": 19002 + }, + { + "epoch": 2.0868658027674063, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.609282970428467, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7165615558624268, + "num_tokens": 491751062.0, + "step": 19003 + }, + { + "epoch": 2.0869756204700196, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6044325828552246, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7151103615760803, + "num_tokens": 491777880.0, + "step": 19004 + }, + { + "epoch": 2.0870854381726334, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.481700897216797, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7100245952606201, + "num_tokens": 491803602.0, + "step": 19005 + }, + { + "epoch": 2.087195255875247, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4958348274230957, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7096314430236816, + "num_tokens": 491833549.0, + "step": 19006 + }, + { + "epoch": 2.087305073577861, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.879262685775757, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6876590847969055, + "num_tokens": 491855771.0, + "step": 19007 + }, + { + "epoch": 2.087414891280474, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4651544094085693, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7116214632987976, + "num_tokens": 491883597.0, + "step": 19008 + }, + { + "epoch": 2.087524708983088, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.5364904403686523, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.6981831789016724, + "num_tokens": 491912746.0, + "step": 19009 + }, + { + "epoch": 2.0876345266857017, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.666813373565674, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7155271768569946, + "num_tokens": 491937029.0, + "step": 19010 + }, + { + "epoch": 2.0877443443883155, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.9081268310546875, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7216812372207642, + "num_tokens": 491959920.0, + "step": 19011 + }, + { + "epoch": 2.087854162090929, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.4160022735595703, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7220036387443542, + "num_tokens": 491989044.0, + "step": 19012 + }, + { + "epoch": 2.0879639797935425, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.879053831100464, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7126408219337463, + "num_tokens": 492013024.0, + "step": 19013 + }, + { + "epoch": 2.0880737974961563, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.462775707244873, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7056946158409119, + "num_tokens": 492044906.0, + "step": 19014 + }, + { + "epoch": 2.08818361519877, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6223599910736084, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.708694577217102, + "num_tokens": 492070927.0, + "step": 19015 + }, + { + "epoch": 2.088293432901384, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.7142834663391113, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7017303705215454, + "num_tokens": 492095050.0, + "step": 19016 + }, + { + "epoch": 2.0884032506039976, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.874666929244995, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7133963108062744, + "num_tokens": 492115775.0, + "step": 19017 + }, + { + "epoch": 2.088513068306611, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7244889736175537, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7061207294464111, + "num_tokens": 492139040.0, + "step": 19018 + }, + { + "epoch": 2.0886228860092246, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.70449161529541, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.7018303871154785, + "num_tokens": 492167280.0, + "step": 19019 + }, + { + "epoch": 2.0887327037118384, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7475638389587402, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.700147807598114, + "num_tokens": 492193315.0, + "step": 19020 + }, + { + "epoch": 2.088842521414452, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8693718910217285, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7022366523742676, + "num_tokens": 492217710.0, + "step": 19021 + }, + { + "epoch": 2.0889523391170655, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.150362968444824, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.7002352476119995, + "num_tokens": 492238561.0, + "step": 19022 + }, + { + "epoch": 2.089062156819679, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7918081283569336, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7071153521537781, + "num_tokens": 492261352.0, + "step": 19023 + }, + { + "epoch": 2.089171974522293, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6701605319976807, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7055696249008179, + "num_tokens": 492286699.0, + "step": 19024 + }, + { + "epoch": 2.0892817922249067, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.686603546142578, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7196149826049805, + "num_tokens": 492309374.0, + "step": 19025 + }, + { + "epoch": 2.0893916099275205, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.243358850479126, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.7006618976593018, + "num_tokens": 492342546.0, + "step": 19026 + }, + { + "epoch": 2.089501427630134, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.458638906478882, + "learning_rate": 1e-06, + "loss": 0.9189, + "mean_token_accuracy": 0.7304254770278931, + "num_tokens": 492369335.0, + "step": 19027 + }, + { + "epoch": 2.0896112453327476, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7486941814422607, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7025066614151001, + "num_tokens": 492393485.0, + "step": 19028 + }, + { + "epoch": 2.0897210630353613, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7438082695007324, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7078359127044678, + "num_tokens": 492417059.0, + "step": 19029 + }, + { + "epoch": 2.089830880737975, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.061450719833374, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7076923847198486, + "num_tokens": 492438264.0, + "step": 19030 + }, + { + "epoch": 2.089940698440589, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7121243476867676, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6929193735122681, + "num_tokens": 492463019.0, + "step": 19031 + }, + { + "epoch": 2.090050516143202, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.57132887840271, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7068779468536377, + "num_tokens": 492489882.0, + "step": 19032 + }, + { + "epoch": 2.090160333845816, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.938633918762207, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7207009792327881, + "num_tokens": 492514387.0, + "step": 19033 + }, + { + "epoch": 2.0902701515484297, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.751537799835205, + "learning_rate": 1e-06, + "loss": 0.9195, + "mean_token_accuracy": 0.727053165435791, + "num_tokens": 492535709.0, + "step": 19034 + }, + { + "epoch": 2.0903799692510434, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.705099582672119, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7134301662445068, + "num_tokens": 492558357.0, + "step": 19035 + }, + { + "epoch": 2.0904897869536567, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4787709712982178, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.7004320025444031, + "num_tokens": 492585376.0, + "step": 19036 + }, + { + "epoch": 2.0905996046562705, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 6.59023904800415, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6903704404830933, + "num_tokens": 492614595.0, + "step": 19037 + }, + { + "epoch": 2.0907094223588842, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.097949981689453, + "learning_rate": 1e-06, + "loss": 0.9369, + "mean_token_accuracy": 0.7235755920410156, + "num_tokens": 492632727.0, + "step": 19038 + }, + { + "epoch": 2.090819240061498, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.66744327545166, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.697737455368042, + "num_tokens": 492662994.0, + "step": 19039 + }, + { + "epoch": 2.0909290577641118, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.5550410747528076, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7094235420227051, + "num_tokens": 492690017.0, + "step": 19040 + }, + { + "epoch": 2.091038875466725, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5287373065948486, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7121333479881287, + "num_tokens": 492718107.0, + "step": 19041 + }, + { + "epoch": 2.091148693169339, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8886420726776123, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7122633457183838, + "num_tokens": 492741305.0, + "step": 19042 + }, + { + "epoch": 2.0912585108719526, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.663667917251587, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7063726186752319, + "num_tokens": 492766787.0, + "step": 19043 + }, + { + "epoch": 2.0913683285745663, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.666276216506958, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.710976243019104, + "num_tokens": 492793789.0, + "step": 19044 + }, + { + "epoch": 2.09147814627718, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.560490369796753, + "learning_rate": 1e-06, + "loss": 0.9395, + "mean_token_accuracy": 0.7233811020851135, + "num_tokens": 492820917.0, + "step": 19045 + }, + { + "epoch": 2.0915879639797934, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5023839473724365, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6897794604301453, + "num_tokens": 492849776.0, + "step": 19046 + }, + { + "epoch": 2.091697781682407, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.647247791290283, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7227246761322021, + "num_tokens": 492877582.0, + "step": 19047 + }, + { + "epoch": 2.091807599385021, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.91857647895813, + "learning_rate": 1e-06, + "loss": 1.1269, + "mean_token_accuracy": 0.6850353479385376, + "num_tokens": 492906048.0, + "step": 19048 + }, + { + "epoch": 2.0919174170876347, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.783095598220825, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.7051366567611694, + "num_tokens": 492928817.0, + "step": 19049 + }, + { + "epoch": 2.092027234790248, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.74727463722229, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.716983437538147, + "num_tokens": 492951138.0, + "step": 19050 + }, + { + "epoch": 2.0921370524928617, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.731515645980835, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7191010117530823, + "num_tokens": 492985746.0, + "step": 19051 + }, + { + "epoch": 2.0922468701954755, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.732429027557373, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7153841853141785, + "num_tokens": 493008945.0, + "step": 19052 + }, + { + "epoch": 2.0923566878980893, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.0282251834869385, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7166595458984375, + "num_tokens": 493032172.0, + "step": 19053 + }, + { + "epoch": 2.092466505600703, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.337470769882202, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.6975820064544678, + "num_tokens": 493064895.0, + "step": 19054 + }, + { + "epoch": 2.0925763233033163, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4729366302490234, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6791779398918152, + "num_tokens": 493094111.0, + "step": 19055 + }, + { + "epoch": 2.09268614100593, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5401203632354736, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7173638343811035, + "num_tokens": 493120831.0, + "step": 19056 + }, + { + "epoch": 2.092795958708544, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4833755493164062, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7126524448394775, + "num_tokens": 493148749.0, + "step": 19057 + }, + { + "epoch": 2.0929057764111576, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4442336559295654, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7024039626121521, + "num_tokens": 493187015.0, + "step": 19058 + }, + { + "epoch": 2.093015594113771, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.8270552158355713, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7013558149337769, + "num_tokens": 493211688.0, + "step": 19059 + }, + { + "epoch": 2.0931254118163847, + "ewc_loss": 1.0192394256591797e-05, + "grad_norm": 2.6730692386627197, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.714227557182312, + "num_tokens": 493238035.0, + "step": 19060 + }, + { + "epoch": 2.0932352295189984, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 16.71100616455078, + "learning_rate": 1e-06, + "loss": 0.9309, + "mean_token_accuracy": 0.7262318730354309, + "num_tokens": 493265714.0, + "step": 19061 + }, + { + "epoch": 2.093345047221612, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.59753680229187, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.730379045009613, + "num_tokens": 493294297.0, + "step": 19062 + }, + { + "epoch": 2.093454864924226, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8267791271209717, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7297709584236145, + "num_tokens": 493315677.0, + "step": 19063 + }, + { + "epoch": 2.0935646826268393, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6277918815612793, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6913443803787231, + "num_tokens": 493342213.0, + "step": 19064 + }, + { + "epoch": 2.093674500329453, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6818227767944336, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7068414688110352, + "num_tokens": 493365307.0, + "step": 19065 + }, + { + "epoch": 2.0937843180320668, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3715932369232178, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6977947950363159, + "num_tokens": 493396239.0, + "step": 19066 + }, + { + "epoch": 2.0938941357346805, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.157784938812256, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7077344655990601, + "num_tokens": 493416150.0, + "step": 19067 + }, + { + "epoch": 2.0940039534372943, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7395453453063965, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6873164772987366, + "num_tokens": 493440906.0, + "step": 19068 + }, + { + "epoch": 2.0941137711399076, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6640689373016357, + "learning_rate": 1e-06, + "loss": 0.9281, + "mean_token_accuracy": 0.7235940098762512, + "num_tokens": 493467869.0, + "step": 19069 + }, + { + "epoch": 2.0942235888425214, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5691072940826416, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7028574347496033, + "num_tokens": 493496899.0, + "step": 19070 + }, + { + "epoch": 2.094333406545135, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7436017990112305, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7210485935211182, + "num_tokens": 493518423.0, + "step": 19071 + }, + { + "epoch": 2.094443224247749, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.596991777420044, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7095401287078857, + "num_tokens": 493543772.0, + "step": 19072 + }, + { + "epoch": 2.0945530419503626, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6965994834899902, + "learning_rate": 1e-06, + "loss": 0.9416, + "mean_token_accuracy": 0.7237226963043213, + "num_tokens": 493569302.0, + "step": 19073 + }, + { + "epoch": 2.094662859652976, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 8.079215049743652, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7131962776184082, + "num_tokens": 493590965.0, + "step": 19074 + }, + { + "epoch": 2.0947726773555897, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 4.480747222900391, + "learning_rate": 1e-06, + "loss": 1.0876, + "mean_token_accuracy": 0.689691424369812, + "num_tokens": 493618059.0, + "step": 19075 + }, + { + "epoch": 2.0948824950582035, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6046600341796875, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.716477632522583, + "num_tokens": 493644737.0, + "step": 19076 + }, + { + "epoch": 2.094992312760817, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.711665630340576, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.696806013584137, + "num_tokens": 493670418.0, + "step": 19077 + }, + { + "epoch": 2.0951021304634305, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.703754425048828, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7216609120368958, + "num_tokens": 493700392.0, + "step": 19078 + }, + { + "epoch": 2.0952119481660443, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.443838119506836, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.7245601415634155, + "num_tokens": 493727488.0, + "step": 19079 + }, + { + "epoch": 2.095321765868658, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.2576606273651123, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7295169234275818, + "num_tokens": 493750634.0, + "step": 19080 + }, + { + "epoch": 2.095431583571272, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.1739487648010254, + "learning_rate": 1e-06, + "loss": 0.8913, + "mean_token_accuracy": 0.7300471067428589, + "num_tokens": 493766122.0, + "step": 19081 + }, + { + "epoch": 2.0955414012738856, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.765930652618408, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6859880685806274, + "num_tokens": 493792108.0, + "step": 19082 + }, + { + "epoch": 2.095651218976499, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5641818046569824, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7100349068641663, + "num_tokens": 493818761.0, + "step": 19083 + }, + { + "epoch": 2.0957610366791126, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7531955242156982, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7185677289962769, + "num_tokens": 493842579.0, + "step": 19084 + }, + { + "epoch": 2.0958708543817264, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 7.947229862213135, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7018884420394897, + "num_tokens": 493866585.0, + "step": 19085 + }, + { + "epoch": 2.09598067208434, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7099313735961914, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.70564204454422, + "num_tokens": 493891549.0, + "step": 19086 + }, + { + "epoch": 2.0960904897869534, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8043792247772217, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6893944144248962, + "num_tokens": 493917048.0, + "step": 19087 + }, + { + "epoch": 2.096200307489567, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9290080070495605, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7092488408088684, + "num_tokens": 493936071.0, + "step": 19088 + }, + { + "epoch": 2.096310125192181, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.833125352859497, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7106322050094604, + "num_tokens": 493957816.0, + "step": 19089 + }, + { + "epoch": 2.0964199428947947, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4897170066833496, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.695885956287384, + "num_tokens": 493983395.0, + "step": 19090 + }, + { + "epoch": 2.0965297605974085, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5296173095703125, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6900938749313354, + "num_tokens": 494012211.0, + "step": 19091 + }, + { + "epoch": 2.096639578300022, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.584447145462036, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.696686863899231, + "num_tokens": 494038658.0, + "step": 19092 + }, + { + "epoch": 2.0967493960026355, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6086018085479736, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7073248624801636, + "num_tokens": 494064052.0, + "step": 19093 + }, + { + "epoch": 2.0968592137052493, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.648554563522339, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7138251662254333, + "num_tokens": 494089375.0, + "step": 19094 + }, + { + "epoch": 2.096969031407863, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.8781230449676514, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7074695825576782, + "num_tokens": 494118483.0, + "step": 19095 + }, + { + "epoch": 2.097078849110477, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.8213250637054443, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.690691351890564, + "num_tokens": 494141958.0, + "step": 19096 + }, + { + "epoch": 2.09718866681309, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4274983406066895, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7054082155227661, + "num_tokens": 494170703.0, + "step": 19097 + }, + { + "epoch": 2.097298484515704, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 16.602832794189453, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.7008893489837646, + "num_tokens": 494194926.0, + "step": 19098 + }, + { + "epoch": 2.0974083022183176, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6743321418762207, + "learning_rate": 1e-06, + "loss": 0.9194, + "mean_token_accuracy": 0.7317857146263123, + "num_tokens": 494220067.0, + "step": 19099 + }, + { + "epoch": 2.0975181199209314, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.475886344909668, + "learning_rate": 1e-06, + "loss": 0.9283, + "mean_token_accuracy": 0.7228895425796509, + "num_tokens": 494246743.0, + "step": 19100 + }, + { + "epoch": 2.0976279376235447, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.85595440864563, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6972647905349731, + "num_tokens": 494267627.0, + "step": 19101 + }, + { + "epoch": 2.0977377553261585, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.2302699089050293, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7015800476074219, + "num_tokens": 494298287.0, + "step": 19102 + }, + { + "epoch": 2.0978475730287722, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.530843734741211, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7072802782058716, + "num_tokens": 494323848.0, + "step": 19103 + }, + { + "epoch": 2.097957390731386, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7388641834259033, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7155821919441223, + "num_tokens": 494348007.0, + "step": 19104 + }, + { + "epoch": 2.0980672084339997, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6295273303985596, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.708839476108551, + "num_tokens": 494372682.0, + "step": 19105 + }, + { + "epoch": 2.098177026136613, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.570126533508301, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7236340045928955, + "num_tokens": 494398639.0, + "step": 19106 + }, + { + "epoch": 2.098286843839227, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8677175045013428, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7029395699501038, + "num_tokens": 494420579.0, + "step": 19107 + }, + { + "epoch": 2.0983966615418406, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8465688228607178, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7023484706878662, + "num_tokens": 494444564.0, + "step": 19108 + }, + { + "epoch": 2.0985064792444543, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6604177951812744, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7059508562088013, + "num_tokens": 494468730.0, + "step": 19109 + }, + { + "epoch": 2.0986162969470676, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.718768358230591, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7025012969970703, + "num_tokens": 494494472.0, + "step": 19110 + }, + { + "epoch": 2.0987261146496814, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.9044747352600098, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.6934031248092651, + "num_tokens": 494523174.0, + "step": 19111 + }, + { + "epoch": 2.098835932352295, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8756120204925537, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7123128175735474, + "num_tokens": 494545274.0, + "step": 19112 + }, + { + "epoch": 2.098945750054909, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7682743072509766, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7121458649635315, + "num_tokens": 494566831.0, + "step": 19113 + }, + { + "epoch": 2.0990555677575227, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4195406436920166, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7047302722930908, + "num_tokens": 494598634.0, + "step": 19114 + }, + { + "epoch": 2.099165385460136, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5548014640808105, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7013338804244995, + "num_tokens": 494624358.0, + "step": 19115 + }, + { + "epoch": 2.0992752031627497, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5239956378936768, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7127648591995239, + "num_tokens": 494652414.0, + "step": 19116 + }, + { + "epoch": 2.0993850208653635, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5495331287384033, + "learning_rate": 1e-06, + "loss": 0.9205, + "mean_token_accuracy": 0.7295700907707214, + "num_tokens": 494677253.0, + "step": 19117 + }, + { + "epoch": 2.0994948385679773, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.899078369140625, + "learning_rate": 1e-06, + "loss": 0.916, + "mean_token_accuracy": 0.7259206771850586, + "num_tokens": 494697568.0, + "step": 19118 + }, + { + "epoch": 2.099604656270591, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.99216365814209, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7067824602127075, + "num_tokens": 494723570.0, + "step": 19119 + }, + { + "epoch": 2.0997144739732043, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8690059185028076, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7189740538597107, + "num_tokens": 494744675.0, + "step": 19120 + }, + { + "epoch": 2.099824291675818, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.435285806655884, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7106839418411255, + "num_tokens": 494771771.0, + "step": 19121 + }, + { + "epoch": 2.099934109378432, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6850597858428955, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7025455832481384, + "num_tokens": 494796506.0, + "step": 19122 + }, + { + "epoch": 2.1000439270810456, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.2645881175994873, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.719730019569397, + "num_tokens": 494829970.0, + "step": 19123 + }, + { + "epoch": 2.1001537447836593, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.6003496646881104, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6988072395324707, + "num_tokens": 494856312.0, + "step": 19124 + }, + { + "epoch": 2.1002635624862727, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5638198852539062, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7193727493286133, + "num_tokens": 494887483.0, + "step": 19125 + }, + { + "epoch": 2.1003733801888864, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7624807357788086, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7149405479431152, + "num_tokens": 494910484.0, + "step": 19126 + }, + { + "epoch": 2.1004831978915, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4902727603912354, + "learning_rate": 1e-06, + "loss": 0.9238, + "mean_token_accuracy": 0.7288792133331299, + "num_tokens": 494942311.0, + "step": 19127 + }, + { + "epoch": 2.100593015594114, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9579033851623535, + "learning_rate": 1e-06, + "loss": 0.888, + "mean_token_accuracy": 0.7348076105117798, + "num_tokens": 494963926.0, + "step": 19128 + }, + { + "epoch": 2.1007028332967272, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.383409023284912, + "learning_rate": 1e-06, + "loss": 0.8804, + "mean_token_accuracy": 0.7451044321060181, + "num_tokens": 494992019.0, + "step": 19129 + }, + { + "epoch": 2.100812650999341, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4754245281219482, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6887043714523315, + "num_tokens": 495028204.0, + "step": 19130 + }, + { + "epoch": 2.1009224687019548, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8493845462799072, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7189433574676514, + "num_tokens": 495048069.0, + "step": 19131 + }, + { + "epoch": 2.1010322864045685, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4698877334594727, + "learning_rate": 1e-06, + "loss": 0.9223, + "mean_token_accuracy": 0.7293491363525391, + "num_tokens": 495075347.0, + "step": 19132 + }, + { + "epoch": 2.1011421041071823, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7441751956939697, + "learning_rate": 1e-06, + "loss": 0.9106, + "mean_token_accuracy": 0.7343818545341492, + "num_tokens": 495099759.0, + "step": 19133 + }, + { + "epoch": 2.1012519218097956, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4925193786621094, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7094433307647705, + "num_tokens": 495128353.0, + "step": 19134 + }, + { + "epoch": 2.1013617395124093, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.023310899734497, + "learning_rate": 1e-06, + "loss": 0.9271, + "mean_token_accuracy": 0.7276715636253357, + "num_tokens": 495148321.0, + "step": 19135 + }, + { + "epoch": 2.101471557215023, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.577463150024414, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7060434818267822, + "num_tokens": 495173397.0, + "step": 19136 + }, + { + "epoch": 2.101581374917637, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.5817787647247314, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6967308521270752, + "num_tokens": 495197877.0, + "step": 19137 + }, + { + "epoch": 2.10169119262025, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3834638595581055, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7070910334587097, + "num_tokens": 495227587.0, + "step": 19138 + }, + { + "epoch": 2.101801010322864, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 16.71004295349121, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7013180255889893, + "num_tokens": 495253735.0, + "step": 19139 + }, + { + "epoch": 2.1019108280254777, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.8458821773529053, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.712548017501831, + "num_tokens": 495275860.0, + "step": 19140 + }, + { + "epoch": 2.1020206457280914, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.9136860370635986, + "learning_rate": 1e-06, + "loss": 0.9097, + "mean_token_accuracy": 0.734135627746582, + "num_tokens": 495295552.0, + "step": 19141 + }, + { + "epoch": 2.102130463430705, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.7851381301879883, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7129664421081543, + "num_tokens": 495318799.0, + "step": 19142 + }, + { + "epoch": 2.1022402811333185, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.882037401199341, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7210547924041748, + "num_tokens": 495340919.0, + "step": 19143 + }, + { + "epoch": 2.1023500988359323, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.518528699874878, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7069653272628784, + "num_tokens": 495366914.0, + "step": 19144 + }, + { + "epoch": 2.102459916538546, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 3.0055363178253174, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7180953621864319, + "num_tokens": 495384879.0, + "step": 19145 + }, + { + "epoch": 2.10256973424116, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.770280122756958, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7025409936904907, + "num_tokens": 495407837.0, + "step": 19146 + }, + { + "epoch": 2.1026795519437735, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.64033579826355, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7200324535369873, + "num_tokens": 495434520.0, + "step": 19147 + }, + { + "epoch": 2.102789369646387, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 3.196998119354248, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7218101024627686, + "num_tokens": 495452239.0, + "step": 19148 + }, + { + "epoch": 2.1028991873490006, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.572134017944336, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.7042989730834961, + "num_tokens": 495478247.0, + "step": 19149 + }, + { + "epoch": 2.1030090050516144, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.506652355194092, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6955206394195557, + "num_tokens": 495507686.0, + "step": 19150 + }, + { + "epoch": 2.103118822754228, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.4263296127319336, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.7032663226127625, + "num_tokens": 495542033.0, + "step": 19151 + }, + { + "epoch": 2.1032286404568414, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.5109777450561523, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7325394153594971, + "num_tokens": 495568759.0, + "step": 19152 + }, + { + "epoch": 2.103338458159455, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.604926586151123, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.7075526714324951, + "num_tokens": 495591660.0, + "step": 19153 + }, + { + "epoch": 2.103448275862069, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.4619836807250977, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7205322980880737, + "num_tokens": 495618271.0, + "step": 19154 + }, + { + "epoch": 2.1035580935646827, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.8204658031463623, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6947652101516724, + "num_tokens": 495642256.0, + "step": 19155 + }, + { + "epoch": 2.1036679112672965, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.9415931701660156, + "learning_rate": 1e-06, + "loss": 0.8997, + "mean_token_accuracy": 0.7332866787910461, + "num_tokens": 495662454.0, + "step": 19156 + }, + { + "epoch": 2.1037777289699098, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.926434278488159, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6997250318527222, + "num_tokens": 495690054.0, + "step": 19157 + }, + { + "epoch": 2.1038875466725235, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.737962007522583, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7211692929267883, + "num_tokens": 495713834.0, + "step": 19158 + }, + { + "epoch": 2.1039973643751373, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5585591793060303, + "learning_rate": 1e-06, + "loss": 0.8927, + "mean_token_accuracy": 0.7371946573257446, + "num_tokens": 495740116.0, + "step": 19159 + }, + { + "epoch": 2.104107182077751, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.8924548625946045, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7146043181419373, + "num_tokens": 495762199.0, + "step": 19160 + }, + { + "epoch": 2.104216999780365, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.5979502201080322, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7199798226356506, + "num_tokens": 495787374.0, + "step": 19161 + }, + { + "epoch": 2.104326817482978, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.3328802585601807, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.7033818364143372, + "num_tokens": 495820269.0, + "step": 19162 + }, + { + "epoch": 2.104436635185592, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.444962978363037, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7015253305435181, + "num_tokens": 495848752.0, + "step": 19163 + }, + { + "epoch": 2.1045464528882056, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8115627765655518, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7086687088012695, + "num_tokens": 495871746.0, + "step": 19164 + }, + { + "epoch": 2.1046562705908194, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.7377660274505615, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.6982108950614929, + "num_tokens": 495894868.0, + "step": 19165 + }, + { + "epoch": 2.1047660882934327, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3843135833740234, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6954474449157715, + "num_tokens": 495923860.0, + "step": 19166 + }, + { + "epoch": 2.1048759059960465, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.220055103302002, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7139836549758911, + "num_tokens": 495957197.0, + "step": 19167 + }, + { + "epoch": 2.10498572369866, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.4356913566589355, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6971855163574219, + "num_tokens": 495985904.0, + "step": 19168 + }, + { + "epoch": 2.105095541401274, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.3863728046417236, + "learning_rate": 1e-06, + "loss": 1.1217, + "mean_token_accuracy": 0.66959547996521, + "num_tokens": 496015293.0, + "step": 19169 + }, + { + "epoch": 2.1052053591038877, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.616032123565674, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.706455409526825, + "num_tokens": 496042205.0, + "step": 19170 + }, + { + "epoch": 2.105315176806501, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.595294237136841, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7053661942481995, + "num_tokens": 496067972.0, + "step": 19171 + }, + { + "epoch": 2.105424994509115, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.5825130939483643, + "learning_rate": 1e-06, + "loss": 0.9403, + "mean_token_accuracy": 0.727429986000061, + "num_tokens": 496092607.0, + "step": 19172 + }, + { + "epoch": 2.1055348122117286, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.4484658241271973, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7095991969108582, + "num_tokens": 496122886.0, + "step": 19173 + }, + { + "epoch": 2.1056446299143423, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.851909875869751, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.6940969228744507, + "num_tokens": 496146924.0, + "step": 19174 + }, + { + "epoch": 2.105754447616956, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.8246071338653564, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.714701771736145, + "num_tokens": 496166899.0, + "step": 19175 + }, + { + "epoch": 2.1058642653195694, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.777698278427124, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6958256959915161, + "num_tokens": 496196377.0, + "step": 19176 + }, + { + "epoch": 2.105974083022183, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8274545669555664, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7061386108398438, + "num_tokens": 496223225.0, + "step": 19177 + }, + { + "epoch": 2.106083900724797, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.59082293510437, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.6931052803993225, + "num_tokens": 496250112.0, + "step": 19178 + }, + { + "epoch": 2.1061937184274107, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.917405366897583, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7180904746055603, + "num_tokens": 496279922.0, + "step": 19179 + }, + { + "epoch": 2.106303536130024, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.722935914993286, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6912345886230469, + "num_tokens": 496304436.0, + "step": 19180 + }, + { + "epoch": 2.1064133538326377, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.4243059158325195, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.697301983833313, + "num_tokens": 496334696.0, + "step": 19181 + }, + { + "epoch": 2.1065231715352515, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.6730103492736816, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7151328921318054, + "num_tokens": 496358059.0, + "step": 19182 + }, + { + "epoch": 2.1066329892378652, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.494072675704956, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7059019804000854, + "num_tokens": 496387740.0, + "step": 19183 + }, + { + "epoch": 2.106742806940479, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.617398500442505, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6981905698776245, + "num_tokens": 496415509.0, + "step": 19184 + }, + { + "epoch": 2.1068526246430923, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8730673789978027, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.705411970615387, + "num_tokens": 496437011.0, + "step": 19185 + }, + { + "epoch": 2.106962442345706, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4256272315979004, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.704112708568573, + "num_tokens": 496469315.0, + "step": 19186 + }, + { + "epoch": 2.10707226004832, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.6156959533691406, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.7024219036102295, + "num_tokens": 496496872.0, + "step": 19187 + }, + { + "epoch": 2.1071820777509336, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.412332057952881, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7196270227432251, + "num_tokens": 496526371.0, + "step": 19188 + }, + { + "epoch": 2.107291895453547, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.820547342300415, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7064584493637085, + "num_tokens": 496552103.0, + "step": 19189 + }, + { + "epoch": 2.1074017131561606, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.5129361152648926, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6966370344161987, + "num_tokens": 496579357.0, + "step": 19190 + }, + { + "epoch": 2.1075115308587744, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.6813831329345703, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6910269856452942, + "num_tokens": 496605039.0, + "step": 19191 + }, + { + "epoch": 2.107621348561388, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 3.0390102863311768, + "learning_rate": 1e-06, + "loss": 0.8801, + "mean_token_accuracy": 0.7354111075401306, + "num_tokens": 496621766.0, + "step": 19192 + }, + { + "epoch": 2.107731166264002, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.599750518798828, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.7182739973068237, + "num_tokens": 496649137.0, + "step": 19193 + }, + { + "epoch": 2.1078409839666152, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.620220422744751, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7123196125030518, + "num_tokens": 496675600.0, + "step": 19194 + }, + { + "epoch": 2.107950801669229, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.582504987716675, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7084081768989563, + "num_tokens": 496702492.0, + "step": 19195 + }, + { + "epoch": 2.1080606193718427, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.624185800552368, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7243395447731018, + "num_tokens": 496729085.0, + "step": 19196 + }, + { + "epoch": 2.1081704370744565, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6703412532806396, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7073694467544556, + "num_tokens": 496752094.0, + "step": 19197 + }, + { + "epoch": 2.1082802547770703, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.307487964630127, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6824772953987122, + "num_tokens": 496785434.0, + "step": 19198 + }, + { + "epoch": 2.1083900724796836, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.43430233001709, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6974756717681885, + "num_tokens": 496814915.0, + "step": 19199 + }, + { + "epoch": 2.1084998901822973, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.4591917991638184, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7149553298950195, + "num_tokens": 496842686.0, + "step": 19200 + }, + { + "epoch": 2.108609707884911, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.0128848552703857, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7088003754615784, + "num_tokens": 496863892.0, + "step": 19201 + }, + { + "epoch": 2.108719525587525, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.663239002227783, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.7118902206420898, + "num_tokens": 496886522.0, + "step": 19202 + }, + { + "epoch": 2.108829343290138, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.6531965732574463, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7173874378204346, + "num_tokens": 496909244.0, + "step": 19203 + }, + { + "epoch": 2.108939160992752, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7830076217651367, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7008448839187622, + "num_tokens": 496935758.0, + "step": 19204 + }, + { + "epoch": 2.1090489786953657, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.80845308303833, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7137300372123718, + "num_tokens": 496960666.0, + "step": 19205 + }, + { + "epoch": 2.1091587963979794, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.561452627182007, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.7057021856307983, + "num_tokens": 496987818.0, + "step": 19206 + }, + { + "epoch": 2.109268614100593, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.7460174560546875, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7096503973007202, + "num_tokens": 497013415.0, + "step": 19207 + }, + { + "epoch": 2.1093784318032065, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 3.027277708053589, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7132951021194458, + "num_tokens": 497034957.0, + "step": 19208 + }, + { + "epoch": 2.1094882495058203, + "ewc_loss": 1.0251998901367188e-05, + "grad_norm": 2.7934021949768066, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7186710834503174, + "num_tokens": 497057299.0, + "step": 19209 + }, + { + "epoch": 2.109598067208434, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.9991402626037598, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6997678279876709, + "num_tokens": 497083160.0, + "step": 19210 + }, + { + "epoch": 2.1097078849110478, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6244328022003174, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.7046630382537842, + "num_tokens": 497110314.0, + "step": 19211 + }, + { + "epoch": 2.1098177026136615, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.753404140472412, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7300516366958618, + "num_tokens": 497133958.0, + "step": 19212 + }, + { + "epoch": 2.109927520316275, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.8079206943511963, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7145652770996094, + "num_tokens": 497156841.0, + "step": 19213 + }, + { + "epoch": 2.1100373380188886, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.616642951965332, + "learning_rate": 1e-06, + "loss": 0.9109, + "mean_token_accuracy": 0.7293950319290161, + "num_tokens": 497181270.0, + "step": 19214 + }, + { + "epoch": 2.1101471557215024, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.7245712280273438, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7117272615432739, + "num_tokens": 497206080.0, + "step": 19215 + }, + { + "epoch": 2.110256973424116, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.643216609954834, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6994214057922363, + "num_tokens": 497236514.0, + "step": 19216 + }, + { + "epoch": 2.1103667911267294, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4588124752044678, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6969054341316223, + "num_tokens": 497265054.0, + "step": 19217 + }, + { + "epoch": 2.110476608829343, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.924002170562744, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7246667742729187, + "num_tokens": 497284782.0, + "step": 19218 + }, + { + "epoch": 2.110586426531957, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.366835832595825, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.707517147064209, + "num_tokens": 497315170.0, + "step": 19219 + }, + { + "epoch": 2.1106962442345707, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7717089653015137, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6929112076759338, + "num_tokens": 497342680.0, + "step": 19220 + }, + { + "epoch": 2.1108060619371845, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.82250714302063, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.6985548734664917, + "num_tokens": 497365246.0, + "step": 19221 + }, + { + "epoch": 2.1109158796397978, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7767112255096436, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.705970287322998, + "num_tokens": 497386614.0, + "step": 19222 + }, + { + "epoch": 2.1110256973424115, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.032419204711914, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7080286741256714, + "num_tokens": 497408631.0, + "step": 19223 + }, + { + "epoch": 2.1111355150450253, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.428529739379883, + "learning_rate": 1e-06, + "loss": 0.901, + "mean_token_accuracy": 0.7341173887252808, + "num_tokens": 497435262.0, + "step": 19224 + }, + { + "epoch": 2.111245332747639, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4899537563323975, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7174257040023804, + "num_tokens": 497462587.0, + "step": 19225 + }, + { + "epoch": 2.111355150450253, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6587467193603516, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7163656949996948, + "num_tokens": 497486300.0, + "step": 19226 + }, + { + "epoch": 2.111464968152866, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8168296813964844, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7182369828224182, + "num_tokens": 497507945.0, + "step": 19227 + }, + { + "epoch": 2.11157478585548, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 9.6950101852417, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6875524520874023, + "num_tokens": 497540400.0, + "step": 19228 + }, + { + "epoch": 2.1116846035580936, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8454971313476562, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7156922221183777, + "num_tokens": 497562841.0, + "step": 19229 + }, + { + "epoch": 2.1117944212607074, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6951911449432373, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6974356770515442, + "num_tokens": 497586063.0, + "step": 19230 + }, + { + "epoch": 2.1119042389633207, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8638718128204346, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7166382670402527, + "num_tokens": 497611524.0, + "step": 19231 + }, + { + "epoch": 2.1120140566659344, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6312851905822754, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7136796712875366, + "num_tokens": 497636834.0, + "step": 19232 + }, + { + "epoch": 2.112123874368548, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6810786724090576, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7120281457901001, + "num_tokens": 497662050.0, + "step": 19233 + }, + { + "epoch": 2.112233692071162, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.586552619934082, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7120522260665894, + "num_tokens": 497691458.0, + "step": 19234 + }, + { + "epoch": 2.1123435097737757, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.799724817276001, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7131425142288208, + "num_tokens": 497713683.0, + "step": 19235 + }, + { + "epoch": 2.112453327476389, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6540908813476562, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.710734486579895, + "num_tokens": 497740788.0, + "step": 19236 + }, + { + "epoch": 2.112563145179003, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.583498477935791, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7003530263900757, + "num_tokens": 497767673.0, + "step": 19237 + }, + { + "epoch": 2.1126729628816165, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.760518789291382, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7295688390731812, + "num_tokens": 497787761.0, + "step": 19238 + }, + { + "epoch": 2.1127827805842303, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4728260040283203, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7014147043228149, + "num_tokens": 497818542.0, + "step": 19239 + }, + { + "epoch": 2.1128925982868436, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6781294345855713, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6984374523162842, + "num_tokens": 497845783.0, + "step": 19240 + }, + { + "epoch": 2.1130024159894574, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.408731460571289, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.7036141157150269, + "num_tokens": 497876838.0, + "step": 19241 + }, + { + "epoch": 2.113112233692071, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.5992400646209717, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.698161780834198, + "num_tokens": 497906561.0, + "step": 19242 + }, + { + "epoch": 2.113222051394685, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.4577131271362305, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6851634979248047, + "num_tokens": 497936046.0, + "step": 19243 + }, + { + "epoch": 2.1133318690972986, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.79032564163208, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7065238952636719, + "num_tokens": 497959831.0, + "step": 19244 + }, + { + "epoch": 2.113441686799912, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.75207781791687, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7136503458023071, + "num_tokens": 497986435.0, + "step": 19245 + }, + { + "epoch": 2.1135515045025257, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.440612554550171, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.702454686164856, + "num_tokens": 498016562.0, + "step": 19246 + }, + { + "epoch": 2.1136613222051395, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.5267624855041504, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7207373380661011, + "num_tokens": 498042669.0, + "step": 19247 + }, + { + "epoch": 2.1137711399077532, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.9786388874053955, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7010812759399414, + "num_tokens": 498062798.0, + "step": 19248 + }, + { + "epoch": 2.113880957610367, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.6243066787719727, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6999561190605164, + "num_tokens": 498090075.0, + "step": 19249 + }, + { + "epoch": 2.1139907753129803, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.4883410930633545, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.701260507106781, + "num_tokens": 498119591.0, + "step": 19250 + }, + { + "epoch": 2.114100593015594, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.33135986328125, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7051588296890259, + "num_tokens": 498154592.0, + "step": 19251 + }, + { + "epoch": 2.114210410718208, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.340555191040039, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.6961474418640137, + "num_tokens": 498185943.0, + "step": 19252 + }, + { + "epoch": 2.1143202284208216, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.6906096935272217, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6919000744819641, + "num_tokens": 498212773.0, + "step": 19253 + }, + { + "epoch": 2.1144300461234353, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.356499671936035, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7161036729812622, + "num_tokens": 498240260.0, + "step": 19254 + }, + { + "epoch": 2.1145398638260486, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.743642568588257, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7039744853973389, + "num_tokens": 498264336.0, + "step": 19255 + }, + { + "epoch": 2.1146496815286624, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.62037992477417, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.6997854709625244, + "num_tokens": 498293781.0, + "step": 19256 + }, + { + "epoch": 2.114759499231276, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.5049889087677, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7017456293106079, + "num_tokens": 498322343.0, + "step": 19257 + }, + { + "epoch": 2.11486931693389, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6447336673736572, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7218605279922485, + "num_tokens": 498352780.0, + "step": 19258 + }, + { + "epoch": 2.114979134636503, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.526672840118408, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.717552125453949, + "num_tokens": 498377441.0, + "step": 19259 + }, + { + "epoch": 2.115088952339117, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.860614061355591, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7028955221176147, + "num_tokens": 498397880.0, + "step": 19260 + }, + { + "epoch": 2.1151987700417307, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.565333604812622, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6975213885307312, + "num_tokens": 498426227.0, + "step": 19261 + }, + { + "epoch": 2.1153085877443445, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.522080898284912, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.700334906578064, + "num_tokens": 498453888.0, + "step": 19262 + }, + { + "epoch": 2.1154184054469582, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.578228712081909, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7060943841934204, + "num_tokens": 498482431.0, + "step": 19263 + }, + { + "epoch": 2.1155282231495716, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.362882614135742, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6945641040802002, + "num_tokens": 498517225.0, + "step": 19264 + }, + { + "epoch": 2.1156380408521853, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.6356403827667236, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7123047709465027, + "num_tokens": 498542827.0, + "step": 19265 + }, + { + "epoch": 2.115747858554799, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.3508481979370117, + "learning_rate": 1e-06, + "loss": 1.1194, + "mean_token_accuracy": 0.6802529096603394, + "num_tokens": 498574854.0, + "step": 19266 + }, + { + "epoch": 2.115857676257413, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.740774631500244, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.70328289270401, + "num_tokens": 498600075.0, + "step": 19267 + }, + { + "epoch": 2.115967493960026, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.4390738010406494, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.7000199556350708, + "num_tokens": 498627775.0, + "step": 19268 + }, + { + "epoch": 2.11607731166264, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.5094680786132812, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7010670304298401, + "num_tokens": 498652204.0, + "step": 19269 + }, + { + "epoch": 2.1161871293652537, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 3.2461671829223633, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7194100618362427, + "num_tokens": 498672745.0, + "step": 19270 + }, + { + "epoch": 2.1162969470678674, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 8.050369262695312, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7123523950576782, + "num_tokens": 498695867.0, + "step": 19271 + }, + { + "epoch": 2.116406764770481, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0549824237823486, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7096428275108337, + "num_tokens": 498715297.0, + "step": 19272 + }, + { + "epoch": 2.1165165824730945, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.712482452392578, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7106231451034546, + "num_tokens": 498739354.0, + "step": 19273 + }, + { + "epoch": 2.1166264001757082, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.8025944232940674, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7205737233161926, + "num_tokens": 498762108.0, + "step": 19274 + }, + { + "epoch": 2.116736217878322, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.54671573638916, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7127987146377563, + "num_tokens": 498789026.0, + "step": 19275 + }, + { + "epoch": 2.1168460355809358, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.546360731124878, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7056844830513, + "num_tokens": 498814879.0, + "step": 19276 + }, + { + "epoch": 2.1169558532835495, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.683088779449463, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.717555820941925, + "num_tokens": 498836829.0, + "step": 19277 + }, + { + "epoch": 2.117065670986163, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.641063928604126, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7013641595840454, + "num_tokens": 498861658.0, + "step": 19278 + }, + { + "epoch": 2.1171754886887766, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 3.0096681118011475, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7090319991111755, + "num_tokens": 498880637.0, + "step": 19279 + }, + { + "epoch": 2.1172853063913903, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.8594412803649902, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6951741576194763, + "num_tokens": 498907284.0, + "step": 19280 + }, + { + "epoch": 2.117395124094004, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.5092999935150146, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.709212064743042, + "num_tokens": 498936942.0, + "step": 19281 + }, + { + "epoch": 2.1175049417966174, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.388486385345459, + "learning_rate": 1e-06, + "loss": 0.9376, + "mean_token_accuracy": 0.7314419746398926, + "num_tokens": 498966621.0, + "step": 19282 + }, + { + "epoch": 2.117614759499231, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.79042649269104, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7165753841400146, + "num_tokens": 498990140.0, + "step": 19283 + }, + { + "epoch": 2.117724577201845, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7697322368621826, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7090625166893005, + "num_tokens": 499011949.0, + "step": 19284 + }, + { + "epoch": 2.1178343949044587, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.507084846496582, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7147848606109619, + "num_tokens": 499038072.0, + "step": 19285 + }, + { + "epoch": 2.1179442126070724, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.772505283355713, + "learning_rate": 1e-06, + "loss": 0.8686, + "mean_token_accuracy": 0.7413439750671387, + "num_tokens": 499058058.0, + "step": 19286 + }, + { + "epoch": 2.1180540303096858, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3674509525299072, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6960199475288391, + "num_tokens": 499090487.0, + "step": 19287 + }, + { + "epoch": 2.1181638480122995, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6661133766174316, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6969200372695923, + "num_tokens": 499117531.0, + "step": 19288 + }, + { + "epoch": 2.1182736657149133, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.496403455734253, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7212349772453308, + "num_tokens": 499146062.0, + "step": 19289 + }, + { + "epoch": 2.118383483417527, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5207180976867676, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6870813369750977, + "num_tokens": 499178273.0, + "step": 19290 + }, + { + "epoch": 2.118493301120141, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3465840816497803, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7087578773498535, + "num_tokens": 499209831.0, + "step": 19291 + }, + { + "epoch": 2.118603118822754, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.581540822982788, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7034215331077576, + "num_tokens": 499236931.0, + "step": 19292 + }, + { + "epoch": 2.118712936525368, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.261307954788208, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6984984874725342, + "num_tokens": 499268058.0, + "step": 19293 + }, + { + "epoch": 2.1188227542279816, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.601447343826294, + "learning_rate": 1e-06, + "loss": 0.9332, + "mean_token_accuracy": 0.7252930402755737, + "num_tokens": 499294168.0, + "step": 19294 + }, + { + "epoch": 2.1189325719305954, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.635174036026001, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7026903033256531, + "num_tokens": 499321024.0, + "step": 19295 + }, + { + "epoch": 2.1190423896332087, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 7.856555938720703, + "learning_rate": 1e-06, + "loss": 0.9254, + "mean_token_accuracy": 0.7285026907920837, + "num_tokens": 499352640.0, + "step": 19296 + }, + { + "epoch": 2.1191522073358224, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.647799491882324, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.7069410085678101, + "num_tokens": 499379641.0, + "step": 19297 + }, + { + "epoch": 2.119262025038436, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5462634563446045, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.7003019452095032, + "num_tokens": 499411356.0, + "step": 19298 + }, + { + "epoch": 2.11937184274105, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8506152629852295, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7242655158042908, + "num_tokens": 499431949.0, + "step": 19299 + }, + { + "epoch": 2.1194816604436637, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 3.1520299911499023, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7094000577926636, + "num_tokens": 499448969.0, + "step": 19300 + }, + { + "epoch": 2.119591478146277, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.763396978378296, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.708530843257904, + "num_tokens": 499474514.0, + "step": 19301 + }, + { + "epoch": 2.1197012958488908, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6236658096313477, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7150925397872925, + "num_tokens": 499497833.0, + "step": 19302 + }, + { + "epoch": 2.1198111135515045, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.439518451690674, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.717936635017395, + "num_tokens": 499527147.0, + "step": 19303 + }, + { + "epoch": 2.1199209312541183, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.6502864360809326, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6991088390350342, + "num_tokens": 499553289.0, + "step": 19304 + }, + { + "epoch": 2.120030748956732, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 3.2626237869262695, + "learning_rate": 1e-06, + "loss": 0.887, + "mean_token_accuracy": 0.7343809008598328, + "num_tokens": 499575098.0, + "step": 19305 + }, + { + "epoch": 2.1201405666593454, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5109479427337646, + "learning_rate": 1e-06, + "loss": 1.068, + "mean_token_accuracy": 0.6917496919631958, + "num_tokens": 499605573.0, + "step": 19306 + }, + { + "epoch": 2.120250384361959, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7838242053985596, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.712497353553772, + "num_tokens": 499630419.0, + "step": 19307 + }, + { + "epoch": 2.120360202064573, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.585747718811035, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7150107622146606, + "num_tokens": 499655863.0, + "step": 19308 + }, + { + "epoch": 2.1204700197671866, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.342315912246704, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6926841735839844, + "num_tokens": 499689025.0, + "step": 19309 + }, + { + "epoch": 2.1205798374698, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.423248767852783, + "learning_rate": 1e-06, + "loss": 0.9573, + "mean_token_accuracy": 0.7157080769538879, + "num_tokens": 499717713.0, + "step": 19310 + }, + { + "epoch": 2.1206896551724137, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8230302333831787, + "learning_rate": 1e-06, + "loss": 0.9251, + "mean_token_accuracy": 0.7287055850028992, + "num_tokens": 499739851.0, + "step": 19311 + }, + { + "epoch": 2.1207994728750275, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.4553885459899902, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7140878438949585, + "num_tokens": 499770583.0, + "step": 19312 + }, + { + "epoch": 2.120909290577641, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.54923152923584, + "learning_rate": 1e-06, + "loss": 0.9262, + "mean_token_accuracy": 0.7281773090362549, + "num_tokens": 499798119.0, + "step": 19313 + }, + { + "epoch": 2.121019108280255, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.3810231685638428, + "learning_rate": 1e-06, + "loss": 1.0977, + "mean_token_accuracy": 0.6825064420700073, + "num_tokens": 499828541.0, + "step": 19314 + }, + { + "epoch": 2.1211289259828683, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.871999740600586, + "learning_rate": 1e-06, + "loss": 0.9099, + "mean_token_accuracy": 0.7331022024154663, + "num_tokens": 499849585.0, + "step": 19315 + }, + { + "epoch": 2.121238743685482, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.7212374210357666, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7182613611221313, + "num_tokens": 499871411.0, + "step": 19316 + }, + { + "epoch": 2.121348561388096, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.7440266609191895, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7089800834655762, + "num_tokens": 499895292.0, + "step": 19317 + }, + { + "epoch": 2.1214583790907096, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.6171395778656006, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6983696222305298, + "num_tokens": 499925480.0, + "step": 19318 + }, + { + "epoch": 2.121568196793323, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.9924468994140625, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6997768878936768, + "num_tokens": 499948452.0, + "step": 19319 + }, + { + "epoch": 2.1216780144959366, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.7076048851013184, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7130934596061707, + "num_tokens": 499972771.0, + "step": 19320 + }, + { + "epoch": 2.1217878321985504, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0320217609405518, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.711018443107605, + "num_tokens": 499993166.0, + "step": 19321 + }, + { + "epoch": 2.121897649901164, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.429450511932373, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7217576503753662, + "num_tokens": 500024026.0, + "step": 19322 + }, + { + "epoch": 2.122007467603778, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.474003314971924, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7113242149353027, + "num_tokens": 500052901.0, + "step": 19323 + }, + { + "epoch": 2.122117285306391, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.582284450531006, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7047239542007446, + "num_tokens": 500076201.0, + "step": 19324 + }, + { + "epoch": 2.122227103009005, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.5649197101593018, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.693021833896637, + "num_tokens": 500102261.0, + "step": 19325 + }, + { + "epoch": 2.1223369207116187, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.2212047576904297, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7296600937843323, + "num_tokens": 500121653.0, + "step": 19326 + }, + { + "epoch": 2.1224467384142325, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5845227241516113, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7154746055603027, + "num_tokens": 500150513.0, + "step": 19327 + }, + { + "epoch": 2.1225565561168462, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6082346439361572, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.7000164985656738, + "num_tokens": 500176494.0, + "step": 19328 + }, + { + "epoch": 2.1226663738194596, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8928608894348145, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7243412733078003, + "num_tokens": 500197664.0, + "step": 19329 + }, + { + "epoch": 2.1227761915220733, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6126320362091064, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6969437599182129, + "num_tokens": 500225618.0, + "step": 19330 + }, + { + "epoch": 2.122886009224687, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7175779342651367, + "learning_rate": 1e-06, + "loss": 0.9604, + "mean_token_accuracy": 0.7116475701332092, + "num_tokens": 500250309.0, + "step": 19331 + }, + { + "epoch": 2.122995826927301, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9247894287109375, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7173537611961365, + "num_tokens": 500274450.0, + "step": 19332 + }, + { + "epoch": 2.123105644629914, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5083670616149902, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.7026424407958984, + "num_tokens": 500305517.0, + "step": 19333 + }, + { + "epoch": 2.123215462332528, + "ewc_loss": 1.0311603546142578e-05, + "grad_norm": 2.711982488632202, + "learning_rate": 1e-06, + "loss": 0.8779, + "mean_token_accuracy": 0.7437187433242798, + "num_tokens": 500329715.0, + "step": 19334 + }, + { + "epoch": 2.1233252800351416, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.806692123413086, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7239729166030884, + "num_tokens": 500352933.0, + "step": 19335 + }, + { + "epoch": 2.1234350977377554, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.884243965148926, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7069426774978638, + "num_tokens": 500374961.0, + "step": 19336 + }, + { + "epoch": 2.123544915440369, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.475972890853882, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7148369550704956, + "num_tokens": 500404202.0, + "step": 19337 + }, + { + "epoch": 2.1236547331429825, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4635112285614014, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7110031843185425, + "num_tokens": 500431137.0, + "step": 19338 + }, + { + "epoch": 2.1237645508455962, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.526078939437866, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7236425876617432, + "num_tokens": 500456704.0, + "step": 19339 + }, + { + "epoch": 2.12387436854821, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.842750072479248, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.714004397392273, + "num_tokens": 500479314.0, + "step": 19340 + }, + { + "epoch": 2.1239841862508237, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.668405771255493, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.7041693925857544, + "num_tokens": 500502601.0, + "step": 19341 + }, + { + "epoch": 2.1240940039534375, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.775895595550537, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6999590992927551, + "num_tokens": 500525637.0, + "step": 19342 + }, + { + "epoch": 2.124203821656051, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.72281551361084, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7050792574882507, + "num_tokens": 500551665.0, + "step": 19343 + }, + { + "epoch": 2.1243136393586646, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5741147994995117, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.698145866394043, + "num_tokens": 500580210.0, + "step": 19344 + }, + { + "epoch": 2.1244234570612783, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6561286449432373, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6983798742294312, + "num_tokens": 500605288.0, + "step": 19345 + }, + { + "epoch": 2.124533274763892, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3482964038848877, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.6990112066268921, + "num_tokens": 500634790.0, + "step": 19346 + }, + { + "epoch": 2.1246430924665054, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6127796173095703, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6850998401641846, + "num_tokens": 500661686.0, + "step": 19347 + }, + { + "epoch": 2.124752910169119, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.822533130645752, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.7248404622077942, + "num_tokens": 500682856.0, + "step": 19348 + }, + { + "epoch": 2.124862727871733, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.588263511657715, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.701866865158081, + "num_tokens": 500713087.0, + "step": 19349 + }, + { + "epoch": 2.1249725455743467, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.895094871520996, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7106938362121582, + "num_tokens": 500736800.0, + "step": 19350 + }, + { + "epoch": 2.1250823632769604, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8220512866973877, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.721796989440918, + "num_tokens": 500760036.0, + "step": 19351 + }, + { + "epoch": 2.1251921809795737, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7074549198150635, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7082789540290833, + "num_tokens": 500786040.0, + "step": 19352 + }, + { + "epoch": 2.1253019986821875, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4186692237854004, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7113312482833862, + "num_tokens": 500813124.0, + "step": 19353 + }, + { + "epoch": 2.1254118163848013, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5893566608428955, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7180017232894897, + "num_tokens": 500838213.0, + "step": 19354 + }, + { + "epoch": 2.125521634087415, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7758448123931885, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7143799066543579, + "num_tokens": 500863071.0, + "step": 19355 + }, + { + "epoch": 2.1256314517900288, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.80499267578125, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7055343389511108, + "num_tokens": 500889884.0, + "step": 19356 + }, + { + "epoch": 2.125741269492642, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.432328224182129, + "learning_rate": 1e-06, + "loss": 0.9212, + "mean_token_accuracy": 0.728537917137146, + "num_tokens": 500917979.0, + "step": 19357 + }, + { + "epoch": 2.125851087195256, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6032629013061523, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7046602368354797, + "num_tokens": 500944795.0, + "step": 19358 + }, + { + "epoch": 2.1259609048978696, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.511457681655884, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7055363059043884, + "num_tokens": 500973310.0, + "step": 19359 + }, + { + "epoch": 2.1260707226004834, + "ewc_loss": 1.1026859283447266e-05, + "grad_norm": 12.801403045654297, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.708652913570404, + "num_tokens": 501001309.0, + "step": 19360 + }, + { + "epoch": 2.1261805403030967, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8782124519348145, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.6893413662910461, + "num_tokens": 501031692.0, + "step": 19361 + }, + { + "epoch": 2.1262903580057104, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9491653442382812, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7067223787307739, + "num_tokens": 501055722.0, + "step": 19362 + }, + { + "epoch": 2.126400175708324, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.981388092041016, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7123172283172607, + "num_tokens": 501079954.0, + "step": 19363 + }, + { + "epoch": 2.126509993410938, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.840177059173584, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7210185527801514, + "num_tokens": 501102979.0, + "step": 19364 + }, + { + "epoch": 2.1266198111135517, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.605285406112671, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7131356000900269, + "num_tokens": 501128962.0, + "step": 19365 + }, + { + "epoch": 2.126729628816165, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.606572151184082, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7148273587226868, + "num_tokens": 501153358.0, + "step": 19366 + }, + { + "epoch": 2.1268394465187788, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.380614757537842, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7106130123138428, + "num_tokens": 501183138.0, + "step": 19367 + }, + { + "epoch": 2.1269492642213925, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9703171253204346, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7127161026000977, + "num_tokens": 501202590.0, + "step": 19368 + }, + { + "epoch": 2.1270590819240063, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.940627098083496, + "learning_rate": 1e-06, + "loss": 0.9462, + "mean_token_accuracy": 0.7210314273834229, + "num_tokens": 501224546.0, + "step": 19369 + }, + { + "epoch": 2.1271688996266196, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7110049724578857, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7219959497451782, + "num_tokens": 501246890.0, + "step": 19370 + }, + { + "epoch": 2.1272787173292333, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8376786708831787, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7178009748458862, + "num_tokens": 501271218.0, + "step": 19371 + }, + { + "epoch": 2.127388535031847, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.677104949951172, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6978362202644348, + "num_tokens": 501298334.0, + "step": 19372 + }, + { + "epoch": 2.127498352734461, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9660086631774902, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7160822153091431, + "num_tokens": 501318225.0, + "step": 19373 + }, + { + "epoch": 2.1276081704370746, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8520798683166504, + "learning_rate": 1e-06, + "loss": 0.8971, + "mean_token_accuracy": 0.7284121513366699, + "num_tokens": 501337679.0, + "step": 19374 + }, + { + "epoch": 2.127717988139688, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.705839157104492, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7309085726737976, + "num_tokens": 501359800.0, + "step": 19375 + }, + { + "epoch": 2.1278278058423017, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6093108654022217, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.69864821434021, + "num_tokens": 501384640.0, + "step": 19376 + }, + { + "epoch": 2.1279376235449154, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.662781000137329, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7174569368362427, + "num_tokens": 501408346.0, + "step": 19377 + }, + { + "epoch": 2.128047441247529, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6659154891967773, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.715372622013092, + "num_tokens": 501434397.0, + "step": 19378 + }, + { + "epoch": 2.128157258950143, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8209376335144043, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7093102931976318, + "num_tokens": 501458928.0, + "step": 19379 + }, + { + "epoch": 2.1282670766527563, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.79957914352417, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7082186937332153, + "num_tokens": 501482911.0, + "step": 19380 + }, + { + "epoch": 2.12837689435537, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5254385471343994, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7044757604598999, + "num_tokens": 501511662.0, + "step": 19381 + }, + { + "epoch": 2.128486712057984, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4483237266540527, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7095366716384888, + "num_tokens": 501540540.0, + "step": 19382 + }, + { + "epoch": 2.1285965297605975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.393817901611328, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6857786178588867, + "num_tokens": 501570794.0, + "step": 19383 + }, + { + "epoch": 2.1287063474632113, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6038455963134766, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6928653717041016, + "num_tokens": 501598342.0, + "step": 19384 + }, + { + "epoch": 2.1288161651658246, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9200668334960938, + "learning_rate": 1e-06, + "loss": 0.9327, + "mean_token_accuracy": 0.7294163703918457, + "num_tokens": 501617992.0, + "step": 19385 + }, + { + "epoch": 2.1289259828684384, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7591168880462646, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7038167715072632, + "num_tokens": 501640359.0, + "step": 19386 + }, + { + "epoch": 2.129035800571052, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6269235610961914, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7137207984924316, + "num_tokens": 501666252.0, + "step": 19387 + }, + { + "epoch": 2.129145618273666, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.806752920150757, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6834284067153931, + "num_tokens": 501696625.0, + "step": 19388 + }, + { + "epoch": 2.129255435976279, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.775824546813965, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6853355169296265, + "num_tokens": 501721953.0, + "step": 19389 + }, + { + "epoch": 2.129365253678893, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4130067825317383, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7255323529243469, + "num_tokens": 501751758.0, + "step": 19390 + }, + { + "epoch": 2.1294750713815067, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.480776786804199, + "learning_rate": 1e-06, + "loss": 0.8972, + "mean_token_accuracy": 0.7386515736579895, + "num_tokens": 501777933.0, + "step": 19391 + }, + { + "epoch": 2.1295848890841205, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.103076219558716, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6799103021621704, + "num_tokens": 501805384.0, + "step": 19392 + }, + { + "epoch": 2.1296947067867342, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.87412428855896, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.7302035093307495, + "num_tokens": 501825156.0, + "step": 19393 + }, + { + "epoch": 2.1298045244893475, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7438533306121826, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7331026196479797, + "num_tokens": 501847690.0, + "step": 19394 + }, + { + "epoch": 2.1299143421919613, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.789788246154785, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.716689944267273, + "num_tokens": 501871064.0, + "step": 19395 + }, + { + "epoch": 2.130024159894575, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5242490768432617, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6951101422309875, + "num_tokens": 501898573.0, + "step": 19396 + }, + { + "epoch": 2.130133977597189, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.52489972114563, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7058389186859131, + "num_tokens": 501926272.0, + "step": 19397 + }, + { + "epoch": 2.130243795299802, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4659006595611572, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6821867823600769, + "num_tokens": 501954207.0, + "step": 19398 + }, + { + "epoch": 2.130353613002416, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6632041931152344, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7087427377700806, + "num_tokens": 501977226.0, + "step": 19399 + }, + { + "epoch": 2.1304634307050296, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.607696771621704, + "learning_rate": 1e-06, + "loss": 0.9201, + "mean_token_accuracy": 0.7270513772964478, + "num_tokens": 502001304.0, + "step": 19400 + }, + { + "epoch": 2.1305732484076434, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7511744499206543, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7154141664505005, + "num_tokens": 502023302.0, + "step": 19401 + }, + { + "epoch": 2.130683066110257, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5577588081359863, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7099730372428894, + "num_tokens": 502048565.0, + "step": 19402 + }, + { + "epoch": 2.1307928838128705, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7817542552948, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6974297761917114, + "num_tokens": 502071693.0, + "step": 19403 + }, + { + "epoch": 2.130902701515484, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4235126972198486, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6810173988342285, + "num_tokens": 502102292.0, + "step": 19404 + }, + { + "epoch": 2.131012519218098, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.854909658432007, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7249208688735962, + "num_tokens": 502124153.0, + "step": 19405 + }, + { + "epoch": 2.1311223369207117, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6042234897613525, + "learning_rate": 1e-06, + "loss": 0.9426, + "mean_token_accuracy": 0.7250014543533325, + "num_tokens": 502149127.0, + "step": 19406 + }, + { + "epoch": 2.1312321546233255, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.596543312072754, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7167896032333374, + "num_tokens": 502175688.0, + "step": 19407 + }, + { + "epoch": 2.131341972325939, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0259313583374023, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7109937071800232, + "num_tokens": 502194729.0, + "step": 19408 + }, + { + "epoch": 2.1314517900285526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5218658447265625, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7110282778739929, + "num_tokens": 502223616.0, + "step": 19409 + }, + { + "epoch": 2.1315616077311663, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.839094400405884, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7100400328636169, + "num_tokens": 502246377.0, + "step": 19410 + }, + { + "epoch": 2.13167142543378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.668728828430176, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6934998631477356, + "num_tokens": 502274348.0, + "step": 19411 + }, + { + "epoch": 2.1317812431363934, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3040361404418945, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7195950746536255, + "num_tokens": 502304932.0, + "step": 19412 + }, + { + "epoch": 2.131891060839007, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.384796142578125, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7061153054237366, + "num_tokens": 502334295.0, + "step": 19413 + }, + { + "epoch": 2.132000878541621, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.401171922683716, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7119919657707214, + "num_tokens": 502361123.0, + "step": 19414 + }, + { + "epoch": 2.1321106962442347, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5955090522766113, + "learning_rate": 1e-06, + "loss": 0.9698, + "mean_token_accuracy": 0.7198265790939331, + "num_tokens": 502385868.0, + "step": 19415 + }, + { + "epoch": 2.1322205139468484, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5454771518707275, + "learning_rate": 1e-06, + "loss": 0.8702, + "mean_token_accuracy": 0.7448928356170654, + "num_tokens": 502411193.0, + "step": 19416 + }, + { + "epoch": 2.1323303316494617, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5908868312835693, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7106974124908447, + "num_tokens": 502436624.0, + "step": 19417 + }, + { + "epoch": 2.1324401493520755, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.642164707183838, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6912930011749268, + "num_tokens": 502463703.0, + "step": 19418 + }, + { + "epoch": 2.1325499670546892, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6553139686584473, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.706814706325531, + "num_tokens": 502490220.0, + "step": 19419 + }, + { + "epoch": 2.132659784757303, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3300602436065674, + "learning_rate": 1e-06, + "loss": 0.9034, + "mean_token_accuracy": 0.7406872510910034, + "num_tokens": 502520829.0, + "step": 19420 + }, + { + "epoch": 2.1327696024599163, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5014898777008057, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.698485255241394, + "num_tokens": 502548809.0, + "step": 19421 + }, + { + "epoch": 2.13287942016253, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5501129627227783, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.715686559677124, + "num_tokens": 502572556.0, + "step": 19422 + }, + { + "epoch": 2.132989237865144, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4358763694763184, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7006535530090332, + "num_tokens": 502601085.0, + "step": 19423 + }, + { + "epoch": 2.1330990555677576, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.558281183242798, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.704056441783905, + "num_tokens": 502626949.0, + "step": 19424 + }, + { + "epoch": 2.1332088732703713, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7284600734710693, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7322587370872498, + "num_tokens": 502648130.0, + "step": 19425 + }, + { + "epoch": 2.1333186909729847, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.661684989929199, + "learning_rate": 1e-06, + "loss": 1.1122, + "mean_token_accuracy": 0.6866455078125, + "num_tokens": 502672859.0, + "step": 19426 + }, + { + "epoch": 2.1334285086755984, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.494236469268799, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.686854898929596, + "num_tokens": 502704382.0, + "step": 19427 + }, + { + "epoch": 2.133538326378212, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.506394624710083, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7187668085098267, + "num_tokens": 502734070.0, + "step": 19428 + }, + { + "epoch": 2.133648144080826, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.180727243423462, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7233017683029175, + "num_tokens": 502752470.0, + "step": 19429 + }, + { + "epoch": 2.1337579617834397, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.773324966430664, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6998584270477295, + "num_tokens": 502777691.0, + "step": 19430 + }, + { + "epoch": 2.133867779486053, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7430741786956787, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.7061383128166199, + "num_tokens": 502802280.0, + "step": 19431 + }, + { + "epoch": 2.1339775971886668, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3972160816192627, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.7017605304718018, + "num_tokens": 502836765.0, + "step": 19432 + }, + { + "epoch": 2.1340874148912805, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4880118370056152, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6932408809661865, + "num_tokens": 502865359.0, + "step": 19433 + }, + { + "epoch": 2.1341972325938943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5969815254211426, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7065819501876831, + "num_tokens": 502890557.0, + "step": 19434 + }, + { + "epoch": 2.134307050296508, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.389683485031128, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7085006237030029, + "num_tokens": 502919872.0, + "step": 19435 + }, + { + "epoch": 2.1344168679991213, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.803941488265991, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6859180927276611, + "num_tokens": 502945931.0, + "step": 19436 + }, + { + "epoch": 2.134526685701735, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6157681941986084, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7073142528533936, + "num_tokens": 502973651.0, + "step": 19437 + }, + { + "epoch": 2.134636503404349, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5867042541503906, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7009626626968384, + "num_tokens": 502999893.0, + "step": 19438 + }, + { + "epoch": 2.1347463211069626, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9355499744415283, + "learning_rate": 1e-06, + "loss": 0.9328, + "mean_token_accuracy": 0.7218493223190308, + "num_tokens": 503019470.0, + "step": 19439 + }, + { + "epoch": 2.134856138809576, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7570700645446777, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7066901922225952, + "num_tokens": 503043003.0, + "step": 19440 + }, + { + "epoch": 2.1349659565121897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.563913345336914, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6977342963218689, + "num_tokens": 503072935.0, + "step": 19441 + }, + { + "epoch": 2.1350757742148034, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.693087339401245, + "learning_rate": 1e-06, + "loss": 0.8821, + "mean_token_accuracy": 0.7357511520385742, + "num_tokens": 503097451.0, + "step": 19442 + }, + { + "epoch": 2.135185591917417, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.676771402359009, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.710811972618103, + "num_tokens": 503122261.0, + "step": 19443 + }, + { + "epoch": 2.135295409620031, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.492938280105591, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6932883262634277, + "num_tokens": 503152427.0, + "step": 19444 + }, + { + "epoch": 2.1354052273226443, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7609314918518066, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6915127038955688, + "num_tokens": 503177395.0, + "step": 19445 + }, + { + "epoch": 2.135515045025258, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.908888101577759, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7115598320960999, + "num_tokens": 503200906.0, + "step": 19446 + }, + { + "epoch": 2.1356248627278718, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6849887371063232, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7051994204521179, + "num_tokens": 503224808.0, + "step": 19447 + }, + { + "epoch": 2.1357346804304855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5724222660064697, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6953188180923462, + "num_tokens": 503252145.0, + "step": 19448 + }, + { + "epoch": 2.135844498133099, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4130845069885254, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7263600826263428, + "num_tokens": 503280156.0, + "step": 19449 + }, + { + "epoch": 2.1359543158357126, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6950483322143555, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.7281100749969482, + "num_tokens": 503305478.0, + "step": 19450 + }, + { + "epoch": 2.1360641335383264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.895362615585327, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7029244899749756, + "num_tokens": 503328600.0, + "step": 19451 + }, + { + "epoch": 2.13617395124094, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0935776233673096, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7013592720031738, + "num_tokens": 503351610.0, + "step": 19452 + }, + { + "epoch": 2.136283768943554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.912637710571289, + "learning_rate": 1e-06, + "loss": 0.9006, + "mean_token_accuracy": 0.7362922430038452, + "num_tokens": 503370742.0, + "step": 19453 + }, + { + "epoch": 2.136393586646167, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5571701526641846, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7169455289840698, + "num_tokens": 503397469.0, + "step": 19454 + }, + { + "epoch": 2.136503404348781, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.33689022064209, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7152407169342041, + "num_tokens": 503431983.0, + "step": 19455 + }, + { + "epoch": 2.1366132220513947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.654559850692749, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7153937816619873, + "num_tokens": 503460751.0, + "step": 19456 + }, + { + "epoch": 2.1367230397540085, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6749379634857178, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7038753032684326, + "num_tokens": 503485362.0, + "step": 19457 + }, + { + "epoch": 2.136832857456622, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.647793769836426, + "learning_rate": 1e-06, + "loss": 0.9016, + "mean_token_accuracy": 0.7312015295028687, + "num_tokens": 503508116.0, + "step": 19458 + }, + { + "epoch": 2.1369426751592355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4625356197357178, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7067369222640991, + "num_tokens": 503535310.0, + "step": 19459 + }, + { + "epoch": 2.1370524928618493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5257434844970703, + "learning_rate": 1e-06, + "loss": 1.1258, + "mean_token_accuracy": 0.6770712733268738, + "num_tokens": 503565033.0, + "step": 19460 + }, + { + "epoch": 2.137162310564463, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.827040672302246, + "learning_rate": 1e-06, + "loss": 0.9169, + "mean_token_accuracy": 0.7321760058403015, + "num_tokens": 503585580.0, + "step": 19461 + }, + { + "epoch": 2.137272128267077, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7178313732147217, + "learning_rate": 1e-06, + "loss": 0.9726, + "mean_token_accuracy": 0.7183374166488647, + "num_tokens": 503610320.0, + "step": 19462 + }, + { + "epoch": 2.13738194596969, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5993154048919678, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7173137664794922, + "num_tokens": 503636171.0, + "step": 19463 + }, + { + "epoch": 2.137491763672304, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4974429607391357, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.7037916779518127, + "num_tokens": 503667294.0, + "step": 19464 + }, + { + "epoch": 2.1376015813749176, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5290656089782715, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7211289405822754, + "num_tokens": 503693320.0, + "step": 19465 + }, + { + "epoch": 2.1377113990775314, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5798377990722656, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7214324474334717, + "num_tokens": 503718333.0, + "step": 19466 + }, + { + "epoch": 2.137821216780145, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2332308292388916, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7142462730407715, + "num_tokens": 503750800.0, + "step": 19467 + }, + { + "epoch": 2.1379310344827585, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.563645362854004, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7319560050964355, + "num_tokens": 503776981.0, + "step": 19468 + }, + { + "epoch": 2.138040852185372, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.318664073944092, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7193532586097717, + "num_tokens": 503805743.0, + "step": 19469 + }, + { + "epoch": 2.138150669887986, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.687201499938965, + "learning_rate": 1e-06, + "loss": 0.9164, + "mean_token_accuracy": 0.7341387271881104, + "num_tokens": 503828450.0, + "step": 19470 + }, + { + "epoch": 2.1382604875905997, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.746609687805176, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7194255590438843, + "num_tokens": 503851585.0, + "step": 19471 + }, + { + "epoch": 2.138370305293213, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5659401416778564, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7180159687995911, + "num_tokens": 503877200.0, + "step": 19472 + }, + { + "epoch": 2.138480122995827, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7576913833618164, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7060589790344238, + "num_tokens": 503901070.0, + "step": 19473 + }, + { + "epoch": 2.1385899406984406, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8962059020996094, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6992166042327881, + "num_tokens": 503924246.0, + "step": 19474 + }, + { + "epoch": 2.1386997584010543, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4724197387695312, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6983250975608826, + "num_tokens": 503954041.0, + "step": 19475 + }, + { + "epoch": 2.138809576103668, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8436875343322754, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7071365118026733, + "num_tokens": 503981062.0, + "step": 19476 + }, + { + "epoch": 2.1389193938062814, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0180721282958984, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7111086249351501, + "num_tokens": 504002281.0, + "step": 19477 + }, + { + "epoch": 2.139029211508895, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4935994148254395, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7195479869842529, + "num_tokens": 504032604.0, + "step": 19478 + }, + { + "epoch": 2.139139029211509, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6711931228637695, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6975528597831726, + "num_tokens": 504061406.0, + "step": 19479 + }, + { + "epoch": 2.1392488469141226, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8757269382476807, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7174621820449829, + "num_tokens": 504082777.0, + "step": 19480 + }, + { + "epoch": 2.1393586646167364, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6429615020751953, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.6998339891433716, + "num_tokens": 504107983.0, + "step": 19481 + }, + { + "epoch": 2.1394684823193497, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4514219760894775, + "learning_rate": 1e-06, + "loss": 1.1301, + "mean_token_accuracy": 0.6835300922393799, + "num_tokens": 504139751.0, + "step": 19482 + }, + { + "epoch": 2.1395783000219635, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7758498191833496, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7064566612243652, + "num_tokens": 504163631.0, + "step": 19483 + }, + { + "epoch": 2.1396881177245772, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.688680648803711, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6987114548683167, + "num_tokens": 504188728.0, + "step": 19484 + }, + { + "epoch": 2.139797935427191, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6938812732696533, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7036241292953491, + "num_tokens": 504214656.0, + "step": 19485 + }, + { + "epoch": 2.1399077531298047, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6378889083862305, + "learning_rate": 1e-06, + "loss": 0.8498, + "mean_token_accuracy": 0.741249680519104, + "num_tokens": 504236576.0, + "step": 19486 + }, + { + "epoch": 2.140017570832418, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 5.1009016036987305, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7126486897468567, + "num_tokens": 504263587.0, + "step": 19487 + }, + { + "epoch": 2.140127388535032, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.878875494003296, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7201910018920898, + "num_tokens": 504286875.0, + "step": 19488 + }, + { + "epoch": 2.1402372062376456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.743191957473755, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7233684659004211, + "num_tokens": 504310774.0, + "step": 19489 + }, + { + "epoch": 2.1403470239402593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.783317804336548, + "learning_rate": 1e-06, + "loss": 0.9089, + "mean_token_accuracy": 0.7302727699279785, + "num_tokens": 504332683.0, + "step": 19490 + }, + { + "epoch": 2.1404568416428726, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.71712327003479, + "learning_rate": 1e-06, + "loss": 1.1117, + "mean_token_accuracy": 0.6778925061225891, + "num_tokens": 504359380.0, + "step": 19491 + }, + { + "epoch": 2.1405666593454864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.538531541824341, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7087240219116211, + "num_tokens": 504389134.0, + "step": 19492 + }, + { + "epoch": 2.1406764770481, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6480062007904053, + "learning_rate": 1e-06, + "loss": 1.0471, + "mean_token_accuracy": 0.6968032717704773, + "num_tokens": 504414439.0, + "step": 19493 + }, + { + "epoch": 2.140786294750714, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7798655033111572, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6984063386917114, + "num_tokens": 504437166.0, + "step": 19494 + }, + { + "epoch": 2.1408961124533277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8420770168304443, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.7099018096923828, + "num_tokens": 504460330.0, + "step": 19495 + }, + { + "epoch": 2.141005930155941, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5937917232513428, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6905150413513184, + "num_tokens": 504487957.0, + "step": 19496 + }, + { + "epoch": 2.1411157478585547, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.510145664215088, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.7112432718276978, + "num_tokens": 504516567.0, + "step": 19497 + }, + { + "epoch": 2.1412255655611685, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.536935329437256, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7074528932571411, + "num_tokens": 504545127.0, + "step": 19498 + }, + { + "epoch": 2.1413353832637823, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.231311321258545, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.72138512134552, + "num_tokens": 504563497.0, + "step": 19499 + }, + { + "epoch": 2.1414452009663956, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6332879066467285, + "learning_rate": 1e-06, + "loss": 0.9025, + "mean_token_accuracy": 0.7263641357421875, + "num_tokens": 504587343.0, + "step": 19500 + }, + { + "epoch": 2.1415550186690093, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4028334617614746, + "learning_rate": 1e-06, + "loss": 1.0078, + "mean_token_accuracy": 0.7054048776626587, + "num_tokens": 504620265.0, + "step": 19501 + }, + { + "epoch": 2.141664836371623, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.891120195388794, + "learning_rate": 1e-06, + "loss": 0.8518, + "mean_token_accuracy": 0.7456327676773071, + "num_tokens": 504638830.0, + "step": 19502 + }, + { + "epoch": 2.141774654074237, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7441277503967285, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7093614339828491, + "num_tokens": 504661914.0, + "step": 19503 + }, + { + "epoch": 2.1418844717768506, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.593625068664551, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.70135098695755, + "num_tokens": 504689335.0, + "step": 19504 + }, + { + "epoch": 2.141994289479464, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9439120292663574, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.70698481798172, + "num_tokens": 504718675.0, + "step": 19505 + }, + { + "epoch": 2.1421041071820777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7354156970977783, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7202022075653076, + "num_tokens": 504741613.0, + "step": 19506 + }, + { + "epoch": 2.1422139248846914, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5191867351531982, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.709280252456665, + "num_tokens": 504769859.0, + "step": 19507 + }, + { + "epoch": 2.142323742587305, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5079567432403564, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7253469228744507, + "num_tokens": 504797627.0, + "step": 19508 + }, + { + "epoch": 2.142433560289919, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7628014087677, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7199596166610718, + "num_tokens": 504820414.0, + "step": 19509 + }, + { + "epoch": 2.1425433779925322, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5921573638916016, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7033272385597229, + "num_tokens": 504844428.0, + "step": 19510 + }, + { + "epoch": 2.142653195695146, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6026742458343506, + "learning_rate": 1e-06, + "loss": 1.1067, + "mean_token_accuracy": 0.6804635524749756, + "num_tokens": 504871808.0, + "step": 19511 + }, + { + "epoch": 2.1427630133977598, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.886559009552002, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7077230215072632, + "num_tokens": 504897352.0, + "step": 19512 + }, + { + "epoch": 2.1428728311003735, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.682422637939453, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7201714515686035, + "num_tokens": 504920529.0, + "step": 19513 + }, + { + "epoch": 2.142982648802987, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8878912925720215, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7138831615447998, + "num_tokens": 504943383.0, + "step": 19514 + }, + { + "epoch": 2.1430924665056006, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.615036964416504, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7095012068748474, + "num_tokens": 504972413.0, + "step": 19515 + }, + { + "epoch": 2.1432022842082143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.915696382522583, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7135482430458069, + "num_tokens": 504994458.0, + "step": 19516 + }, + { + "epoch": 2.143312101910828, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6335177421569824, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.718561589717865, + "num_tokens": 505022774.0, + "step": 19517 + }, + { + "epoch": 2.143421919613442, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7406411170959473, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7089271545410156, + "num_tokens": 505046288.0, + "step": 19518 + }, + { + "epoch": 2.143531737316055, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.603055000305176, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7123160362243652, + "num_tokens": 505073053.0, + "step": 19519 + }, + { + "epoch": 2.143641555018669, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.706695318222046, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7043498754501343, + "num_tokens": 505101192.0, + "step": 19520 + }, + { + "epoch": 2.1437513727212827, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.639645576477051, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7064394354820251, + "num_tokens": 505127222.0, + "step": 19521 + }, + { + "epoch": 2.1438611904238964, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.563398838043213, + "learning_rate": 1e-06, + "loss": 1.0966, + "mean_token_accuracy": 0.6829661726951599, + "num_tokens": 505155887.0, + "step": 19522 + }, + { + "epoch": 2.1439710081265098, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.434635639190674, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7035527229309082, + "num_tokens": 505184600.0, + "step": 19523 + }, + { + "epoch": 2.1440808258291235, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9204230308532715, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6993190050125122, + "num_tokens": 505208945.0, + "step": 19524 + }, + { + "epoch": 2.1441906435317373, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9174275398254395, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7314233779907227, + "num_tokens": 505232164.0, + "step": 19525 + }, + { + "epoch": 2.144300461234351, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4750702381134033, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6955564022064209, + "num_tokens": 505260252.0, + "step": 19526 + }, + { + "epoch": 2.144410278936965, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.747260093688965, + "learning_rate": 1e-06, + "loss": 0.9045, + "mean_token_accuracy": 0.7258576154708862, + "num_tokens": 505282177.0, + "step": 19527 + }, + { + "epoch": 2.144520096639578, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7549264430999756, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7094917297363281, + "num_tokens": 505305190.0, + "step": 19528 + }, + { + "epoch": 2.144629914342192, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.806891441345215, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7200374603271484, + "num_tokens": 505329180.0, + "step": 19529 + }, + { + "epoch": 2.1447397320448056, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.538356065750122, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.720360279083252, + "num_tokens": 505358431.0, + "step": 19530 + }, + { + "epoch": 2.1448495497474194, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.305039167404175, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6967353224754333, + "num_tokens": 505388397.0, + "step": 19531 + }, + { + "epoch": 2.144959367450033, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.463182210922241, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7119754552841187, + "num_tokens": 505420726.0, + "step": 19532 + }, + { + "epoch": 2.1450691851526464, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.551464557647705, + "learning_rate": 1e-06, + "loss": 0.9359, + "mean_token_accuracy": 0.7253287434577942, + "num_tokens": 505447288.0, + "step": 19533 + }, + { + "epoch": 2.14517900285526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5069546699523926, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7045527100563049, + "num_tokens": 505481821.0, + "step": 19534 + }, + { + "epoch": 2.145288820557874, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.491027355194092, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.7056065797805786, + "num_tokens": 505510728.0, + "step": 19535 + }, + { + "epoch": 2.1453986382604877, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7910542488098145, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6935165524482727, + "num_tokens": 505533522.0, + "step": 19536 + }, + { + "epoch": 2.1455084559631015, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6268470287323, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6944940090179443, + "num_tokens": 505561078.0, + "step": 19537 + }, + { + "epoch": 2.145618273665715, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.439929723739624, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6888713836669922, + "num_tokens": 505593858.0, + "step": 19538 + }, + { + "epoch": 2.1457280913683285, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.856431007385254, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7024803161621094, + "num_tokens": 505619141.0, + "step": 19539 + }, + { + "epoch": 2.1458379090709423, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.567739486694336, + "learning_rate": 1e-06, + "loss": 0.9284, + "mean_token_accuracy": 0.7269360423088074, + "num_tokens": 505645105.0, + "step": 19540 + }, + { + "epoch": 2.145947726773556, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4083540439605713, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6922734975814819, + "num_tokens": 505675618.0, + "step": 19541 + }, + { + "epoch": 2.1460575444761694, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.955547332763672, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6998675465583801, + "num_tokens": 505701316.0, + "step": 19542 + }, + { + "epoch": 2.146167362178783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.84846568107605, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6965914964675903, + "num_tokens": 505728786.0, + "step": 19543 + }, + { + "epoch": 2.146277179881397, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.680148124694824, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7249730825424194, + "num_tokens": 505754729.0, + "step": 19544 + }, + { + "epoch": 2.1463869975840106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6099753379821777, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6891980767250061, + "num_tokens": 505786922.0, + "step": 19545 + }, + { + "epoch": 2.1464968152866244, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5810446739196777, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.7071515321731567, + "num_tokens": 505816400.0, + "step": 19546 + }, + { + "epoch": 2.1466066329892377, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.535400390625, + "learning_rate": 1e-06, + "loss": 0.897, + "mean_token_accuracy": 0.7424838542938232, + "num_tokens": 505839740.0, + "step": 19547 + }, + { + "epoch": 2.1467164506918515, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6739325523376465, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.717443585395813, + "num_tokens": 505868143.0, + "step": 19548 + }, + { + "epoch": 2.146826268394465, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.763320207595825, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.6993948221206665, + "num_tokens": 505894277.0, + "step": 19549 + }, + { + "epoch": 2.146936086097079, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.6889846324920654, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7088046669960022, + "num_tokens": 505919504.0, + "step": 19550 + }, + { + "epoch": 2.1470459037996923, + "ewc_loss": 1.0371208190917969e-05, + "grad_norm": 2.736844539642334, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7076064348220825, + "num_tokens": 505942117.0, + "step": 19551 + }, + { + "epoch": 2.147155721502306, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.733165979385376, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7021326422691345, + "num_tokens": 505966337.0, + "step": 19552 + }, + { + "epoch": 2.14726553920492, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7881202697753906, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6968739032745361, + "num_tokens": 505991024.0, + "step": 19553 + }, + { + "epoch": 2.1473753569075336, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.516338348388672, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7068555355072021, + "num_tokens": 506018116.0, + "step": 19554 + }, + { + "epoch": 2.1474851746101473, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6615042686462402, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7018092274665833, + "num_tokens": 506042599.0, + "step": 19555 + }, + { + "epoch": 2.1475949923127606, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.568570613861084, + "learning_rate": 1e-06, + "loss": 1.119, + "mean_token_accuracy": 0.6761064529418945, + "num_tokens": 506071190.0, + "step": 19556 + }, + { + "epoch": 2.1477048100153744, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4975526332855225, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7127517461776733, + "num_tokens": 506101070.0, + "step": 19557 + }, + { + "epoch": 2.147814627717988, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.694136381149292, + "learning_rate": 1e-06, + "loss": 0.898, + "mean_token_accuracy": 0.7323001027107239, + "num_tokens": 506123347.0, + "step": 19558 + }, + { + "epoch": 2.147924445420602, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7355384826660156, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7062673568725586, + "num_tokens": 506146118.0, + "step": 19559 + }, + { + "epoch": 2.1480342631232157, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.686002016067505, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6929680705070496, + "num_tokens": 506174054.0, + "step": 19560 + }, + { + "epoch": 2.148144080825829, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6941640377044678, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.723488450050354, + "num_tokens": 506197243.0, + "step": 19561 + }, + { + "epoch": 2.1482538985284427, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.281106948852539, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6924461126327515, + "num_tokens": 506229973.0, + "step": 19562 + }, + { + "epoch": 2.1483637162310565, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8289566040039062, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7112744450569153, + "num_tokens": 506250745.0, + "step": 19563 + }, + { + "epoch": 2.1484735339336702, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.55503511428833, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6999813914299011, + "num_tokens": 506282315.0, + "step": 19564 + }, + { + "epoch": 2.148583351636284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.269864797592163, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.6985419988632202, + "num_tokens": 506316308.0, + "step": 19565 + }, + { + "epoch": 2.1486931693388973, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.574939727783203, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.7010324001312256, + "num_tokens": 506345180.0, + "step": 19566 + }, + { + "epoch": 2.148802987041511, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.651366710662842, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7179884314537048, + "num_tokens": 506369076.0, + "step": 19567 + }, + { + "epoch": 2.148912804744125, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7862625122070312, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6912268400192261, + "num_tokens": 506393479.0, + "step": 19568 + }, + { + "epoch": 2.1490226224467386, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.57322359085083, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7078419923782349, + "num_tokens": 506419932.0, + "step": 19569 + }, + { + "epoch": 2.149132440149352, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4567081928253174, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7014493942260742, + "num_tokens": 506448953.0, + "step": 19570 + }, + { + "epoch": 2.1492422578519657, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4308676719665527, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7097501158714294, + "num_tokens": 506478305.0, + "step": 19571 + }, + { + "epoch": 2.1493520755545794, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6185693740844727, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.7114284634590149, + "num_tokens": 506503441.0, + "step": 19572 + }, + { + "epoch": 2.149461893257193, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.676091432571411, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6960006356239319, + "num_tokens": 506527930.0, + "step": 19573 + }, + { + "epoch": 2.149571710959807, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.501389503479004, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6973443627357483, + "num_tokens": 506557797.0, + "step": 19574 + }, + { + "epoch": 2.1496815286624202, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.555814266204834, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7229629158973694, + "num_tokens": 506584377.0, + "step": 19575 + }, + { + "epoch": 2.149791346365034, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.1957013607025146, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7117730379104614, + "num_tokens": 506603088.0, + "step": 19576 + }, + { + "epoch": 2.1499011640676478, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.742156982421875, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6935714483261108, + "num_tokens": 506630461.0, + "step": 19577 + }, + { + "epoch": 2.1500109817702615, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5261268615722656, + "learning_rate": 1e-06, + "loss": 0.8673, + "mean_token_accuracy": 0.7430182695388794, + "num_tokens": 506654651.0, + "step": 19578 + }, + { + "epoch": 2.150120799472875, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.527775287628174, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6923410892486572, + "num_tokens": 506685387.0, + "step": 19579 + }, + { + "epoch": 2.1502306171754886, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.485034227371216, + "learning_rate": 1e-06, + "loss": 0.8743, + "mean_token_accuracy": 0.733210563659668, + "num_tokens": 506710343.0, + "step": 19580 + }, + { + "epoch": 2.1503404348781023, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.396153211593628, + "learning_rate": 1e-06, + "loss": 0.8921, + "mean_token_accuracy": 0.734693706035614, + "num_tokens": 506736132.0, + "step": 19581 + }, + { + "epoch": 2.150450252580716, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8496198654174805, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7103970050811768, + "num_tokens": 506757653.0, + "step": 19582 + }, + { + "epoch": 2.15056007028333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.927243232727051, + "learning_rate": 1e-06, + "loss": 0.9576, + "mean_token_accuracy": 0.7160931825637817, + "num_tokens": 506777449.0, + "step": 19583 + }, + { + "epoch": 2.150669887985943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5126216411590576, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.689521074295044, + "num_tokens": 506804917.0, + "step": 19584 + }, + { + "epoch": 2.150779705688557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.685159921646118, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7204789519309998, + "num_tokens": 506831845.0, + "step": 19585 + }, + { + "epoch": 2.1508895233911707, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.805582046508789, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7042397260665894, + "num_tokens": 506858898.0, + "step": 19586 + }, + { + "epoch": 2.1509993410937844, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.06754732131958, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7120732069015503, + "num_tokens": 506881919.0, + "step": 19587 + }, + { + "epoch": 2.151109158796398, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8398749828338623, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7221924662590027, + "num_tokens": 506907384.0, + "step": 19588 + }, + { + "epoch": 2.1512189764990115, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.784362316131592, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.6985774040222168, + "num_tokens": 506931758.0, + "step": 19589 + }, + { + "epoch": 2.1513287942016253, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.124527931213379, + "learning_rate": 1e-06, + "loss": 0.9094, + "mean_token_accuracy": 0.733586311340332, + "num_tokens": 506950804.0, + "step": 19590 + }, + { + "epoch": 2.151438611904239, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.780020236968994, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7135366201400757, + "num_tokens": 506973831.0, + "step": 19591 + }, + { + "epoch": 2.1515484296068528, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6247799396514893, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6869986057281494, + "num_tokens": 507002361.0, + "step": 19592 + }, + { + "epoch": 2.151658247309466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5775647163391113, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7116047739982605, + "num_tokens": 507031774.0, + "step": 19593 + }, + { + "epoch": 2.15176806501208, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5786495208740234, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6875314712524414, + "num_tokens": 507056790.0, + "step": 19594 + }, + { + "epoch": 2.1518778827146936, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6993064880371094, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6951709389686584, + "num_tokens": 507082694.0, + "step": 19595 + }, + { + "epoch": 2.1519877004173074, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9361605644226074, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7227951288223267, + "num_tokens": 507102802.0, + "step": 19596 + }, + { + "epoch": 2.152097518119921, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.615776777267456, + "learning_rate": 1e-06, + "loss": 0.9158, + "mean_token_accuracy": 0.7335878014564514, + "num_tokens": 507126400.0, + "step": 19597 + }, + { + "epoch": 2.1522073358225344, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4940199851989746, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7191879749298096, + "num_tokens": 507153282.0, + "step": 19598 + }, + { + "epoch": 2.152317153525148, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8790042400360107, + "learning_rate": 1e-06, + "loss": 0.9253, + "mean_token_accuracy": 0.7242962718009949, + "num_tokens": 507174702.0, + "step": 19599 + }, + { + "epoch": 2.152426971227762, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5598270893096924, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7019857168197632, + "num_tokens": 507202298.0, + "step": 19600 + }, + { + "epoch": 2.1525367889303757, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.808936834335327, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.6937879323959351, + "num_tokens": 507227837.0, + "step": 19601 + }, + { + "epoch": 2.152646606632989, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.534158229827881, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7008781433105469, + "num_tokens": 507255427.0, + "step": 19602 + }, + { + "epoch": 2.1527564243356028, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6214046478271484, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7087329626083374, + "num_tokens": 507282890.0, + "step": 19603 + }, + { + "epoch": 2.1528662420382165, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.733041524887085, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7249990105628967, + "num_tokens": 507306909.0, + "step": 19604 + }, + { + "epoch": 2.1529760597408303, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3548672199249268, + "learning_rate": 1e-06, + "loss": 0.9315, + "mean_token_accuracy": 0.7251639366149902, + "num_tokens": 507338753.0, + "step": 19605 + }, + { + "epoch": 2.153085877443444, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.78943133354187, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6960741281509399, + "num_tokens": 507360979.0, + "step": 19606 + }, + { + "epoch": 2.1531956951460574, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8324899673461914, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7144644856452942, + "num_tokens": 507390687.0, + "step": 19607 + }, + { + "epoch": 2.153305512848671, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6056532859802246, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6839739680290222, + "num_tokens": 507415851.0, + "step": 19608 + }, + { + "epoch": 2.153415330551285, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9849681854248047, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7125201225280762, + "num_tokens": 507436547.0, + "step": 19609 + }, + { + "epoch": 2.1535251482538986, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.850261688232422, + "learning_rate": 1e-06, + "loss": 0.8713, + "mean_token_accuracy": 0.7456300258636475, + "num_tokens": 507456399.0, + "step": 19610 + }, + { + "epoch": 2.1536349659565124, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.622560739517212, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6919406652450562, + "num_tokens": 507483644.0, + "step": 19611 + }, + { + "epoch": 2.1537447836591257, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.676722288131714, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7157580852508545, + "num_tokens": 507507416.0, + "step": 19612 + }, + { + "epoch": 2.1538546013617395, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.60406756401062, + "learning_rate": 1e-06, + "loss": 1.1134, + "mean_token_accuracy": 0.6763567924499512, + "num_tokens": 507534168.0, + "step": 19613 + }, + { + "epoch": 2.153964419064353, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5791444778442383, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7141196727752686, + "num_tokens": 507560904.0, + "step": 19614 + }, + { + "epoch": 2.154074236766967, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8397276401519775, + "learning_rate": 1e-06, + "loss": 0.9223, + "mean_token_accuracy": 0.7330942749977112, + "num_tokens": 507583246.0, + "step": 19615 + }, + { + "epoch": 2.1541840544695807, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.684436559677124, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7214241027832031, + "num_tokens": 507608366.0, + "step": 19616 + }, + { + "epoch": 2.154293872172194, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5734784603118896, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.700933575630188, + "num_tokens": 507637581.0, + "step": 19617 + }, + { + "epoch": 2.154403689874808, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9931447505950928, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7136378884315491, + "num_tokens": 507657453.0, + "step": 19618 + }, + { + "epoch": 2.1545135075774215, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0412049293518066, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6905828714370728, + "num_tokens": 507685092.0, + "step": 19619 + }, + { + "epoch": 2.1546233252800353, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6497488021850586, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7142009735107422, + "num_tokens": 507710248.0, + "step": 19620 + }, + { + "epoch": 2.1547331429826486, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.729645013809204, + "learning_rate": 1e-06, + "loss": 0.9058, + "mean_token_accuracy": 0.7315160632133484, + "num_tokens": 507735881.0, + "step": 19621 + }, + { + "epoch": 2.1548429606852624, + "ewc_loss": 1.0967254638671875e-05, + "grad_norm": 80.54639434814453, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6996420621871948, + "num_tokens": 507762092.0, + "step": 19622 + }, + { + "epoch": 2.154952778387876, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.733288526535034, + "learning_rate": 1e-06, + "loss": 1.0641, + "mean_token_accuracy": 0.6932817101478577, + "num_tokens": 507789685.0, + "step": 19623 + }, + { + "epoch": 2.15506259609049, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6728460788726807, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6914146542549133, + "num_tokens": 507815097.0, + "step": 19624 + }, + { + "epoch": 2.1551724137931036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.637812614440918, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7163493633270264, + "num_tokens": 507839109.0, + "step": 19625 + }, + { + "epoch": 2.155282231495717, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7135651111602783, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6853382587432861, + "num_tokens": 507864532.0, + "step": 19626 + }, + { + "epoch": 2.1553920491983307, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8239355087280273, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7093485593795776, + "num_tokens": 507886816.0, + "step": 19627 + }, + { + "epoch": 2.1555018669009445, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5573177337646484, + "learning_rate": 1e-06, + "loss": 0.934, + "mean_token_accuracy": 0.7262154817581177, + "num_tokens": 507910570.0, + "step": 19628 + }, + { + "epoch": 2.1556116846035582, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8695907592773438, + "learning_rate": 1e-06, + "loss": 0.9282, + "mean_token_accuracy": 0.732985258102417, + "num_tokens": 507930748.0, + "step": 19629 + }, + { + "epoch": 2.1557215023061715, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0799291133880615, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.6978853344917297, + "num_tokens": 507954800.0, + "step": 19630 + }, + { + "epoch": 2.1558313200087853, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.006521701812744, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.715430498123169, + "num_tokens": 507974285.0, + "step": 19631 + }, + { + "epoch": 2.155941137711399, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5057733058929443, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7206198573112488, + "num_tokens": 508001138.0, + "step": 19632 + }, + { + "epoch": 2.156050955414013, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.842539072036743, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7117665410041809, + "num_tokens": 508027055.0, + "step": 19633 + }, + { + "epoch": 2.1561607731166266, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4981374740600586, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6964617967605591, + "num_tokens": 508055242.0, + "step": 19634 + }, + { + "epoch": 2.15627059081924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.532644271850586, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.689373254776001, + "num_tokens": 508084001.0, + "step": 19635 + }, + { + "epoch": 2.1563804085218536, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.426985025405884, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.71184903383255, + "num_tokens": 508111763.0, + "step": 19636 + }, + { + "epoch": 2.1564902262244674, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.524794816970825, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7193280458450317, + "num_tokens": 508137475.0, + "step": 19637 + }, + { + "epoch": 2.156600043927081, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3777785301208496, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7066123485565186, + "num_tokens": 508165164.0, + "step": 19638 + }, + { + "epoch": 2.156709861629695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0019314289093018, + "learning_rate": 1e-06, + "loss": 0.9016, + "mean_token_accuracy": 0.7333044409751892, + "num_tokens": 508183366.0, + "step": 19639 + }, + { + "epoch": 2.1568196793323082, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6046030521392822, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6985181570053101, + "num_tokens": 508210922.0, + "step": 19640 + }, + { + "epoch": 2.156929497034922, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.919792652130127, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7034727334976196, + "num_tokens": 508233612.0, + "step": 19641 + }, + { + "epoch": 2.1570393147375357, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4892475605010986, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6939700841903687, + "num_tokens": 508262678.0, + "step": 19642 + }, + { + "epoch": 2.1571491324401495, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.577953815460205, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6977343559265137, + "num_tokens": 508290696.0, + "step": 19643 + }, + { + "epoch": 2.157258950142763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4296352863311768, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7197540402412415, + "num_tokens": 508321748.0, + "step": 19644 + }, + { + "epoch": 2.1573687678453766, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.627711772918701, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7143814563751221, + "num_tokens": 508345438.0, + "step": 19645 + }, + { + "epoch": 2.1574785855479903, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.525752544403076, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6938610076904297, + "num_tokens": 508372099.0, + "step": 19646 + }, + { + "epoch": 2.157588403250604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6737751960754395, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7138766050338745, + "num_tokens": 508398974.0, + "step": 19647 + }, + { + "epoch": 2.157698220953218, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5018136501312256, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.702125072479248, + "num_tokens": 508428562.0, + "step": 19648 + }, + { + "epoch": 2.157808038655831, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.502246141433716, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6915293335914612, + "num_tokens": 508456487.0, + "step": 19649 + }, + { + "epoch": 2.157917856358445, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7391910552978516, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7192419767379761, + "num_tokens": 508481014.0, + "step": 19650 + }, + { + "epoch": 2.1580276740610587, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6702065467834473, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7117776274681091, + "num_tokens": 508506398.0, + "step": 19651 + }, + { + "epoch": 2.1581374917636724, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7220213413238525, + "learning_rate": 1e-06, + "loss": 0.9688, + "mean_token_accuracy": 0.719260573387146, + "num_tokens": 508530440.0, + "step": 19652 + }, + { + "epoch": 2.1582473094662857, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.852240562438965, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.700511634349823, + "num_tokens": 508553552.0, + "step": 19653 + }, + { + "epoch": 2.1583571271688995, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.468669891357422, + "learning_rate": 1e-06, + "loss": 0.9296, + "mean_token_accuracy": 0.7274922132492065, + "num_tokens": 508579793.0, + "step": 19654 + }, + { + "epoch": 2.1584669448715132, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.371687412261963, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6938148140907288, + "num_tokens": 508614825.0, + "step": 19655 + }, + { + "epoch": 2.158576762574127, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.980581521987915, + "learning_rate": 1e-06, + "loss": 0.9402, + "mean_token_accuracy": 0.7228259444236755, + "num_tokens": 508636198.0, + "step": 19656 + }, + { + "epoch": 2.1586865802767408, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.441251277923584, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.7031095623970032, + "num_tokens": 508665107.0, + "step": 19657 + }, + { + "epoch": 2.158796397979354, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.880647897720337, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7088095545768738, + "num_tokens": 508692624.0, + "step": 19658 + }, + { + "epoch": 2.158906215681968, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5941553115844727, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7120296955108643, + "num_tokens": 508716756.0, + "step": 19659 + }, + { + "epoch": 2.1590160333845816, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4694859981536865, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7036336660385132, + "num_tokens": 508745627.0, + "step": 19660 + }, + { + "epoch": 2.1591258510871953, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.723557233810425, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6938436031341553, + "num_tokens": 508770712.0, + "step": 19661 + }, + { + "epoch": 2.159235668789809, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.473971366882324, + "learning_rate": 1e-06, + "loss": 1.0956, + "mean_token_accuracy": 0.6802756786346436, + "num_tokens": 508800724.0, + "step": 19662 + }, + { + "epoch": 2.1593454864924224, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6377451419830322, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7094081044197083, + "num_tokens": 508826714.0, + "step": 19663 + }, + { + "epoch": 2.159455304195036, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6656322479248047, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.7046518325805664, + "num_tokens": 508852337.0, + "step": 19664 + }, + { + "epoch": 2.15956512189765, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0817956924438477, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6990625262260437, + "num_tokens": 508872900.0, + "step": 19665 + }, + { + "epoch": 2.1596749396002637, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5837643146514893, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7167965769767761, + "num_tokens": 508898461.0, + "step": 19666 + }, + { + "epoch": 2.1597847573028774, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.573685884475708, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.712687611579895, + "num_tokens": 508925970.0, + "step": 19667 + }, + { + "epoch": 2.1598945750054908, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.677948236465454, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6997146606445312, + "num_tokens": 508954639.0, + "step": 19668 + }, + { + "epoch": 2.1600043927081045, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.458763599395752, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.693635106086731, + "num_tokens": 508987944.0, + "step": 19669 + }, + { + "epoch": 2.1601142104107183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.526921033859253, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7134820818901062, + "num_tokens": 509006625.0, + "step": 19670 + }, + { + "epoch": 2.160224028113332, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.486271858215332, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6997570991516113, + "num_tokens": 509036342.0, + "step": 19671 + }, + { + "epoch": 2.1603338458159453, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.507935047149658, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6943103075027466, + "num_tokens": 509064156.0, + "step": 19672 + }, + { + "epoch": 2.160443663518559, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5775489807128906, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.710264265537262, + "num_tokens": 509088825.0, + "step": 19673 + }, + { + "epoch": 2.160553481221173, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.394747495651245, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.7021825313568115, + "num_tokens": 509120868.0, + "step": 19674 + }, + { + "epoch": 2.1606632989237866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.43892765045166, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.687675952911377, + "num_tokens": 509150995.0, + "step": 19675 + }, + { + "epoch": 2.1607731166264004, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.678709030151367, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7243225574493408, + "num_tokens": 509177385.0, + "step": 19676 + }, + { + "epoch": 2.1608829343290137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8063981533050537, + "learning_rate": 1e-06, + "loss": 1.1461, + "mean_token_accuracy": 0.6718899011611938, + "num_tokens": 509203424.0, + "step": 19677 + }, + { + "epoch": 2.1609927520316274, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8186442852020264, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6926937699317932, + "num_tokens": 509228899.0, + "step": 19678 + }, + { + "epoch": 2.161102569734241, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6320199966430664, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7179011106491089, + "num_tokens": 509256290.0, + "step": 19679 + }, + { + "epoch": 2.161212387436855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4580461978912354, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7092894315719604, + "num_tokens": 509283170.0, + "step": 19680 + }, + { + "epoch": 2.1613222051394683, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4745311737060547, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6951280832290649, + "num_tokens": 509312279.0, + "step": 19681 + }, + { + "epoch": 2.161432022842082, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5817694664001465, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.721320629119873, + "num_tokens": 509338350.0, + "step": 19682 + }, + { + "epoch": 2.161541840544696, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.550536632537842, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7111073732376099, + "num_tokens": 509366366.0, + "step": 19683 + }, + { + "epoch": 2.1616516582473095, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9720351696014404, + "learning_rate": 1e-06, + "loss": 0.8197, + "mean_token_accuracy": 0.7536026239395142, + "num_tokens": 509383927.0, + "step": 19684 + }, + { + "epoch": 2.1617614759499233, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6482996940612793, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6925740838050842, + "num_tokens": 509411164.0, + "step": 19685 + }, + { + "epoch": 2.1618712936525366, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.795445680618286, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.696395754814148, + "num_tokens": 509436736.0, + "step": 19686 + }, + { + "epoch": 2.1619811113551504, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7674520015716553, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6866594552993774, + "num_tokens": 509462765.0, + "step": 19687 + }, + { + "epoch": 2.162090929057764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1962192058563232, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7251293063163757, + "num_tokens": 509485486.0, + "step": 19688 + }, + { + "epoch": 2.162200746760378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6086180210113525, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7280071377754211, + "num_tokens": 509511127.0, + "step": 19689 + }, + { + "epoch": 2.1623105644629916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.497410774230957, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7222185134887695, + "num_tokens": 509536626.0, + "step": 19690 + }, + { + "epoch": 2.162420382165605, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7045176029205322, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7274450063705444, + "num_tokens": 509560467.0, + "step": 19691 + }, + { + "epoch": 2.1625301998682187, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7310266494750977, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6929612755775452, + "num_tokens": 509587425.0, + "step": 19692 + }, + { + "epoch": 2.1626400175708325, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7335779666900635, + "learning_rate": 1e-06, + "loss": 0.9314, + "mean_token_accuracy": 0.7209123373031616, + "num_tokens": 509610781.0, + "step": 19693 + }, + { + "epoch": 2.162749835273446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.706514835357666, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7100538015365601, + "num_tokens": 509635032.0, + "step": 19694 + }, + { + "epoch": 2.16285965297606, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520275115966797, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7055371999740601, + "num_tokens": 509663336.0, + "step": 19695 + }, + { + "epoch": 2.1629694706786733, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.950421094894409, + "learning_rate": 1e-06, + "loss": 1.108, + "mean_token_accuracy": 0.6877559423446655, + "num_tokens": 509687814.0, + "step": 19696 + }, + { + "epoch": 2.163079288381287, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5900449752807617, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6980238556861877, + "num_tokens": 509713456.0, + "step": 19697 + }, + { + "epoch": 2.163189106083901, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0608232021331787, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.7343606352806091, + "num_tokens": 509735769.0, + "step": 19698 + }, + { + "epoch": 2.1632989237865146, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6455628871917725, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.704492449760437, + "num_tokens": 509761707.0, + "step": 19699 + }, + { + "epoch": 2.163408741489128, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3820643424987793, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7123814821243286, + "num_tokens": 509793233.0, + "step": 19700 + }, + { + "epoch": 2.1635185591917416, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7473502159118652, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7042142152786255, + "num_tokens": 509820851.0, + "step": 19701 + }, + { + "epoch": 2.1636283768943554, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.490799903869629, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7074934840202332, + "num_tokens": 509848138.0, + "step": 19702 + }, + { + "epoch": 2.163738194596969, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5797600746154785, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7029812932014465, + "num_tokens": 509875119.0, + "step": 19703 + }, + { + "epoch": 2.1638480122995825, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.630539894104004, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7112546563148499, + "num_tokens": 509902039.0, + "step": 19704 + }, + { + "epoch": 2.163957830002196, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6125476360321045, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7154406309127808, + "num_tokens": 509926147.0, + "step": 19705 + }, + { + "epoch": 2.16406764770481, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8111631870269775, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7005653381347656, + "num_tokens": 509949592.0, + "step": 19706 + }, + { + "epoch": 2.1641774654074237, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5597734451293945, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7145940661430359, + "num_tokens": 509975177.0, + "step": 19707 + }, + { + "epoch": 2.1642872831100375, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.723304033279419, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7172448039054871, + "num_tokens": 509998908.0, + "step": 19708 + }, + { + "epoch": 2.164397100812651, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.606696128845215, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7163586020469666, + "num_tokens": 510024977.0, + "step": 19709 + }, + { + "epoch": 2.1645069185152646, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4452340602874756, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7158066630363464, + "num_tokens": 510054957.0, + "step": 19710 + }, + { + "epoch": 2.1646167362178783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.527604341506958, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7175583839416504, + "num_tokens": 510081631.0, + "step": 19711 + }, + { + "epoch": 2.164726553920492, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.673772096633911, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7227848768234253, + "num_tokens": 510107994.0, + "step": 19712 + }, + { + "epoch": 2.164836371623106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4886059761047363, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6833139657974243, + "num_tokens": 510136020.0, + "step": 19713 + }, + { + "epoch": 2.164946189325719, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.600358724594116, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7165296673774719, + "num_tokens": 510161525.0, + "step": 19714 + }, + { + "epoch": 2.165056007028333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.535391092300415, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6910735368728638, + "num_tokens": 510188589.0, + "step": 19715 + }, + { + "epoch": 2.1651658247309467, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3825571537017822, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7130920886993408, + "num_tokens": 510216497.0, + "step": 19716 + }, + { + "epoch": 2.1652756424335604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5175511837005615, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7166934013366699, + "num_tokens": 510240815.0, + "step": 19717 + }, + { + "epoch": 2.165385460136174, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7027947902679443, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7106865644454956, + "num_tokens": 510263707.0, + "step": 19718 + }, + { + "epoch": 2.1654952778387875, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.61269474029541, + "learning_rate": 1e-06, + "loss": 0.8745, + "mean_token_accuracy": 0.739582896232605, + "num_tokens": 510287514.0, + "step": 19719 + }, + { + "epoch": 2.1656050955414012, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5614492893218994, + "learning_rate": 1e-06, + "loss": 0.8961, + "mean_token_accuracy": 0.728824257850647, + "num_tokens": 510315369.0, + "step": 19720 + }, + { + "epoch": 2.165714913244015, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7170228958129883, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.719863772392273, + "num_tokens": 510339585.0, + "step": 19721 + }, + { + "epoch": 2.1658247309466288, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.513993501663208, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6949344873428345, + "num_tokens": 510367976.0, + "step": 19722 + }, + { + "epoch": 2.165934548649242, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6818861961364746, + "learning_rate": 1e-06, + "loss": 0.9167, + "mean_token_accuracy": 0.7296237945556641, + "num_tokens": 510389909.0, + "step": 19723 + }, + { + "epoch": 2.166044366351856, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7027688026428223, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6996737122535706, + "num_tokens": 510412666.0, + "step": 19724 + }, + { + "epoch": 2.1661541840544696, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9882941246032715, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.731547474861145, + "num_tokens": 510431793.0, + "step": 19725 + }, + { + "epoch": 2.1662640017570833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7493326663970947, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.717582643032074, + "num_tokens": 510457084.0, + "step": 19726 + }, + { + "epoch": 2.166373819459697, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5671820640563965, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7112784385681152, + "num_tokens": 510483651.0, + "step": 19727 + }, + { + "epoch": 2.1664836371623104, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7382895946502686, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7206093072891235, + "num_tokens": 510508228.0, + "step": 19728 + }, + { + "epoch": 2.166593454864924, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.64284086227417, + "learning_rate": 1e-06, + "loss": 1.064, + "mean_token_accuracy": 0.6953151226043701, + "num_tokens": 510535082.0, + "step": 19729 + }, + { + "epoch": 2.166703272567538, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.818690538406372, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7268422842025757, + "num_tokens": 510561521.0, + "step": 19730 + }, + { + "epoch": 2.1668130902701517, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.656766891479492, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7166039943695068, + "num_tokens": 510586160.0, + "step": 19731 + }, + { + "epoch": 2.166922907972765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5955872535705566, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7157291769981384, + "num_tokens": 510611053.0, + "step": 19732 + }, + { + "epoch": 2.1670327256753787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536858320236206, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6990274786949158, + "num_tokens": 510639136.0, + "step": 19733 + }, + { + "epoch": 2.1671425433779925, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6844968795776367, + "learning_rate": 1e-06, + "loss": 1.1312, + "mean_token_accuracy": 0.677364706993103, + "num_tokens": 510667618.0, + "step": 19734 + }, + { + "epoch": 2.1672523610806063, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7211761474609375, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7304915189743042, + "num_tokens": 510690341.0, + "step": 19735 + }, + { + "epoch": 2.16736217878322, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5129001140594482, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7109257578849792, + "num_tokens": 510717531.0, + "step": 19736 + }, + { + "epoch": 2.1674719964858333, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.459028959274292, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.6792917251586914, + "num_tokens": 510750234.0, + "step": 19737 + }, + { + "epoch": 2.167581814188447, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5656275749206543, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.7091572284698486, + "num_tokens": 510779786.0, + "step": 19738 + }, + { + "epoch": 2.167691631891061, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7986435890197754, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7168024182319641, + "num_tokens": 510802666.0, + "step": 19739 + }, + { + "epoch": 2.1678014495936746, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.732604742050171, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7104752659797668, + "num_tokens": 510825954.0, + "step": 19740 + }, + { + "epoch": 2.1679112672962884, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8830227851867676, + "learning_rate": 1e-06, + "loss": 0.9077, + "mean_token_accuracy": 0.7354414463043213, + "num_tokens": 510847338.0, + "step": 19741 + }, + { + "epoch": 2.1680210849989017, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6429197788238525, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6962034702301025, + "num_tokens": 510873730.0, + "step": 19742 + }, + { + "epoch": 2.1681309027015154, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.584744691848755, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6936261653900146, + "num_tokens": 510903789.0, + "step": 19743 + }, + { + "epoch": 2.168240720404129, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5877480506896973, + "learning_rate": 1e-06, + "loss": 0.8379, + "mean_token_accuracy": 0.7570334076881409, + "num_tokens": 510929730.0, + "step": 19744 + }, + { + "epoch": 2.168350538106743, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5799968242645264, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7221077084541321, + "num_tokens": 510955077.0, + "step": 19745 + }, + { + "epoch": 2.1684603558093567, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7715773582458496, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7252192497253418, + "num_tokens": 510979271.0, + "step": 19746 + }, + { + "epoch": 2.16857017351197, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.352932929992676, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6928046941757202, + "num_tokens": 511012254.0, + "step": 19747 + }, + { + "epoch": 2.1686799912145838, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1242759227752686, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7213736176490784, + "num_tokens": 511033036.0, + "step": 19748 + }, + { + "epoch": 2.1687898089171975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6035521030426025, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7231690287590027, + "num_tokens": 511058387.0, + "step": 19749 + }, + { + "epoch": 2.1688996266198113, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5503406524658203, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7108365297317505, + "num_tokens": 511084828.0, + "step": 19750 + }, + { + "epoch": 2.1690094443224246, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5534980297088623, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7085233926773071, + "num_tokens": 511112041.0, + "step": 19751 + }, + { + "epoch": 2.1691192620250384, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4862356185913086, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6979382634162903, + "num_tokens": 511143885.0, + "step": 19752 + }, + { + "epoch": 2.169229079727652, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.666576623916626, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7053296566009521, + "num_tokens": 511168625.0, + "step": 19753 + }, + { + "epoch": 2.169338897430266, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4748435020446777, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7079349160194397, + "num_tokens": 511194843.0, + "step": 19754 + }, + { + "epoch": 2.1694487151328796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.850618362426758, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.701547384262085, + "num_tokens": 511216030.0, + "step": 19755 + }, + { + "epoch": 2.169558532835493, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7295942306518555, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.6923962235450745, + "num_tokens": 511241673.0, + "step": 19756 + }, + { + "epoch": 2.1696683505381067, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6601858139038086, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7063517570495605, + "num_tokens": 511265301.0, + "step": 19757 + }, + { + "epoch": 2.1697781682407205, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.947150230407715, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7106980681419373, + "num_tokens": 511284891.0, + "step": 19758 + }, + { + "epoch": 2.169887985943334, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5960545539855957, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6969746351242065, + "num_tokens": 511310615.0, + "step": 19759 + }, + { + "epoch": 2.1699978036459475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.50185489654541, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7177395820617676, + "num_tokens": 511339053.0, + "step": 19760 + }, + { + "epoch": 2.1701076213485613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.868398666381836, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6916584372520447, + "num_tokens": 511371657.0, + "step": 19761 + }, + { + "epoch": 2.170217439051175, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4897618293762207, + "learning_rate": 1e-06, + "loss": 0.942, + "mean_token_accuracy": 0.7219663262367249, + "num_tokens": 511400108.0, + "step": 19762 + }, + { + "epoch": 2.170327256753789, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.646678924560547, + "learning_rate": 1e-06, + "loss": 0.8763, + "mean_token_accuracy": 0.7369855046272278, + "num_tokens": 511422134.0, + "step": 19763 + }, + { + "epoch": 2.1704370744564025, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6109282970428467, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7122566103935242, + "num_tokens": 511452620.0, + "step": 19764 + }, + { + "epoch": 2.170546892159016, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7347075939178467, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7121618986129761, + "num_tokens": 511477119.0, + "step": 19765 + }, + { + "epoch": 2.1706567098616296, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.50675630569458, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7033390998840332, + "num_tokens": 511506474.0, + "step": 19766 + }, + { + "epoch": 2.1707665275642434, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.681331157684326, + "learning_rate": 1e-06, + "loss": 0.9197, + "mean_token_accuracy": 0.7308600544929504, + "num_tokens": 511529501.0, + "step": 19767 + }, + { + "epoch": 2.170876345266857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.489192008972168, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7158653736114502, + "num_tokens": 511558257.0, + "step": 19768 + }, + { + "epoch": 2.170986162969471, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7584080696105957, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7057416439056396, + "num_tokens": 511583088.0, + "step": 19769 + }, + { + "epoch": 2.171095980672084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5277562141418457, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7036813497543335, + "num_tokens": 511612614.0, + "step": 19770 + }, + { + "epoch": 2.171205798374698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8852579593658447, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7205075025558472, + "num_tokens": 511633909.0, + "step": 19771 + }, + { + "epoch": 2.1713156160773117, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.587083339691162, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7024425268173218, + "num_tokens": 511659968.0, + "step": 19772 + }, + { + "epoch": 2.1714254337799255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7809152603149414, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7169928550720215, + "num_tokens": 511682095.0, + "step": 19773 + }, + { + "epoch": 2.171535251482539, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6612229347229004, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6991493105888367, + "num_tokens": 511710990.0, + "step": 19774 + }, + { + "epoch": 2.1716450691851525, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4556868076324463, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7253349423408508, + "num_tokens": 511739624.0, + "step": 19775 + }, + { + "epoch": 2.1717548868877663, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.735010862350464, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.689583957195282, + "num_tokens": 511764140.0, + "step": 19776 + }, + { + "epoch": 2.17186470459038, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5775601863861084, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7170625925064087, + "num_tokens": 511791977.0, + "step": 19777 + }, + { + "epoch": 2.171974522292994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6051220893859863, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7139717936515808, + "num_tokens": 511815727.0, + "step": 19778 + }, + { + "epoch": 2.172084339995607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.489539384841919, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6921286582946777, + "num_tokens": 511844926.0, + "step": 19779 + }, + { + "epoch": 2.172194157698221, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.350958824157715, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7220191359519958, + "num_tokens": 511877010.0, + "step": 19780 + }, + { + "epoch": 2.1723039754008346, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7204768657684326, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7139514684677124, + "num_tokens": 511897821.0, + "step": 19781 + }, + { + "epoch": 2.1724137931034484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7272541522979736, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.703213095664978, + "num_tokens": 511925977.0, + "step": 19782 + }, + { + "epoch": 2.1725236108060617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5708212852478027, + "learning_rate": 1e-06, + "loss": 0.9491, + "mean_token_accuracy": 0.7121815085411072, + "num_tokens": 511949715.0, + "step": 19783 + }, + { + "epoch": 2.1726334285086755, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.575831174850464, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7331061959266663, + "num_tokens": 511977205.0, + "step": 19784 + }, + { + "epoch": 2.1727432462112892, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6506924629211426, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7125211954116821, + "num_tokens": 512002980.0, + "step": 19785 + }, + { + "epoch": 2.172853063913903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9669840335845947, + "learning_rate": 1e-06, + "loss": 0.9378, + "mean_token_accuracy": 0.7249640822410583, + "num_tokens": 512023265.0, + "step": 19786 + }, + { + "epoch": 2.1729628816165167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5147924423217773, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7117043733596802, + "num_tokens": 512049915.0, + "step": 19787 + }, + { + "epoch": 2.17307269931913, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5578413009643555, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7109758853912354, + "num_tokens": 512078755.0, + "step": 19788 + }, + { + "epoch": 2.173182517021744, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.41915225982666, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7230231761932373, + "num_tokens": 512109347.0, + "step": 19789 + }, + { + "epoch": 2.1732923347243576, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.415849208831787, + "learning_rate": 1e-06, + "loss": 0.9399, + "mean_token_accuracy": 0.7177280187606812, + "num_tokens": 512137702.0, + "step": 19790 + }, + { + "epoch": 2.1734021524269713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5563485622406006, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.717487633228302, + "num_tokens": 512163873.0, + "step": 19791 + }, + { + "epoch": 2.173511970129585, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8030104637145996, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7107024192810059, + "num_tokens": 512187004.0, + "step": 19792 + }, + { + "epoch": 2.1736217878321984, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.01483154296875, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7187161445617676, + "num_tokens": 512206001.0, + "step": 19793 + }, + { + "epoch": 2.173731605534812, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4832801818847656, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6930166482925415, + "num_tokens": 512239300.0, + "step": 19794 + }, + { + "epoch": 2.173841423237426, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.558215379714966, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7169535160064697, + "num_tokens": 512267649.0, + "step": 19795 + }, + { + "epoch": 2.1739512409400397, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.882845401763916, + "learning_rate": 1e-06, + "loss": 0.8668, + "mean_token_accuracy": 0.7482835054397583, + "num_tokens": 512287428.0, + "step": 19796 + }, + { + "epoch": 2.1740610586426534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6312429904937744, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7139987349510193, + "num_tokens": 512312435.0, + "step": 19797 + }, + { + "epoch": 2.1741708763452667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6990089416503906, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7209416031837463, + "num_tokens": 512336657.0, + "step": 19798 + }, + { + "epoch": 2.1742806940478805, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5344576835632324, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6916347146034241, + "num_tokens": 512363372.0, + "step": 19799 + }, + { + "epoch": 2.1743905117504942, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.57090425491333, + "learning_rate": 1e-06, + "loss": 1.0466, + "mean_token_accuracy": 0.7049599885940552, + "num_tokens": 512391054.0, + "step": 19800 + }, + { + "epoch": 2.174500329453108, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.874014377593994, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7198344469070435, + "num_tokens": 512413378.0, + "step": 19801 + }, + { + "epoch": 2.1746101471557213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 7.936452388763428, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7132096290588379, + "num_tokens": 512436309.0, + "step": 19802 + }, + { + "epoch": 2.174719964858335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.821580171585083, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7138326168060303, + "num_tokens": 512458612.0, + "step": 19803 + }, + { + "epoch": 2.174829782560949, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.738389015197754, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.691719651222229, + "num_tokens": 512483455.0, + "step": 19804 + }, + { + "epoch": 2.1749396002635626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5739283561706543, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7114453911781311, + "num_tokens": 512509908.0, + "step": 19805 + }, + { + "epoch": 2.1750494179661763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7123920917510986, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.7010653018951416, + "num_tokens": 512534717.0, + "step": 19806 + }, + { + "epoch": 2.1751592356687897, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8421638011932373, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.7168066501617432, + "num_tokens": 512557093.0, + "step": 19807 + }, + { + "epoch": 2.1752690533714034, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6559503078460693, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.692660927772522, + "num_tokens": 512584806.0, + "step": 19808 + }, + { + "epoch": 2.175378871074017, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.615482807159424, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6883452534675598, + "num_tokens": 512610521.0, + "step": 19809 + }, + { + "epoch": 2.175488688776631, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3646464347839355, + "learning_rate": 1e-06, + "loss": 0.8965, + "mean_token_accuracy": 0.733009934425354, + "num_tokens": 512638067.0, + "step": 19810 + }, + { + "epoch": 2.1755985064792442, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.802950143814087, + "learning_rate": 1e-06, + "loss": 0.854, + "mean_token_accuracy": 0.7429080009460449, + "num_tokens": 512657400.0, + "step": 19811 + }, + { + "epoch": 2.175708324181858, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7398412227630615, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7171919345855713, + "num_tokens": 512683831.0, + "step": 19812 + }, + { + "epoch": 2.1758181418844718, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5907678604125977, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7099899649620056, + "num_tokens": 512711437.0, + "step": 19813 + }, + { + "epoch": 2.1759279595870855, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 9.686203002929688, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6984212398529053, + "num_tokens": 512739807.0, + "step": 19814 + }, + { + "epoch": 2.1760377772896993, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3513882160186768, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7226169109344482, + "num_tokens": 512772090.0, + "step": 19815 + }, + { + "epoch": 2.1761475949923126, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9584953784942627, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6998767852783203, + "num_tokens": 512796315.0, + "step": 19816 + }, + { + "epoch": 2.1762574126949263, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.526967763900757, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6926069259643555, + "num_tokens": 512823980.0, + "step": 19817 + }, + { + "epoch": 2.17636723039754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.398987293243408, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7254788279533386, + "num_tokens": 512853809.0, + "step": 19818 + }, + { + "epoch": 2.176477048100154, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.686692476272583, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7219860553741455, + "num_tokens": 512877992.0, + "step": 19819 + }, + { + "epoch": 2.1765868658027676, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8247995376586914, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.7236731648445129, + "num_tokens": 512900783.0, + "step": 19820 + }, + { + "epoch": 2.176696683505381, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5438528060913086, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7073934674263, + "num_tokens": 512926616.0, + "step": 19821 + }, + { + "epoch": 2.1768065012079947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7760472297668457, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7310979962348938, + "num_tokens": 512950669.0, + "step": 19822 + }, + { + "epoch": 2.1769163189106084, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.568932294845581, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6993510723114014, + "num_tokens": 512975107.0, + "step": 19823 + }, + { + "epoch": 2.177026136613222, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.467935800552368, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7282987236976624, + "num_tokens": 513001713.0, + "step": 19824 + }, + { + "epoch": 2.1771359543158355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7653648853302, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6876401901245117, + "num_tokens": 513026129.0, + "step": 19825 + }, + { + "epoch": 2.1772457720184493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6893765926361084, + "learning_rate": 1e-06, + "loss": 0.8943, + "mean_token_accuracy": 0.7314853668212891, + "num_tokens": 513049616.0, + "step": 19826 + }, + { + "epoch": 2.177355589721063, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4888038635253906, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.701324462890625, + "num_tokens": 513076361.0, + "step": 19827 + }, + { + "epoch": 2.177465407423677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.555536985397339, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.704936146736145, + "num_tokens": 513103893.0, + "step": 19828 + }, + { + "epoch": 2.1775752251262905, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8024749755859375, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7063387632369995, + "num_tokens": 513125256.0, + "step": 19829 + }, + { + "epoch": 2.177685042828904, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5186235904693604, + "learning_rate": 1e-06, + "loss": 1.1044, + "mean_token_accuracy": 0.6817115545272827, + "num_tokens": 513152505.0, + "step": 19830 + }, + { + "epoch": 2.1777948605315176, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.346583366394043, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7212902307510376, + "num_tokens": 513182326.0, + "step": 19831 + }, + { + "epoch": 2.1779046782341314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.619746685028076, + "learning_rate": 1e-06, + "loss": 0.9252, + "mean_token_accuracy": 0.7370595335960388, + "num_tokens": 513204562.0, + "step": 19832 + }, + { + "epoch": 2.178014495936745, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.380923271179199, + "learning_rate": 1e-06, + "loss": 0.8873, + "mean_token_accuracy": 0.7394446134567261, + "num_tokens": 513231469.0, + "step": 19833 + }, + { + "epoch": 2.1781243136393584, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.436650037765503, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7180827856063843, + "num_tokens": 513260543.0, + "step": 19834 + }, + { + "epoch": 2.178234131341972, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.373265266418457, + "learning_rate": 1e-06, + "loss": 0.9229, + "mean_token_accuracy": 0.7313358783721924, + "num_tokens": 513283046.0, + "step": 19835 + }, + { + "epoch": 2.178343949044586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5497007369995117, + "learning_rate": 1e-06, + "loss": 1.1402, + "mean_token_accuracy": 0.6728430986404419, + "num_tokens": 513309369.0, + "step": 19836 + }, + { + "epoch": 2.1784537667471997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.650454521179199, + "learning_rate": 1e-06, + "loss": 1.1126, + "mean_token_accuracy": 0.6770306825637817, + "num_tokens": 513337836.0, + "step": 19837 + }, + { + "epoch": 2.1785635844498135, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.304816484451294, + "learning_rate": 1e-06, + "loss": 1.1396, + "mean_token_accuracy": 0.6720850467681885, + "num_tokens": 513370115.0, + "step": 19838 + }, + { + "epoch": 2.1786734021524268, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.444415330886841, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.698607861995697, + "num_tokens": 513398324.0, + "step": 19839 + }, + { + "epoch": 2.1787832198550405, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.50886607170105, + "learning_rate": 1e-06, + "loss": 0.9907, + "mean_token_accuracy": 0.7090809345245361, + "num_tokens": 513427101.0, + "step": 19840 + }, + { + "epoch": 2.1788930375576543, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7334816455841064, + "learning_rate": 1e-06, + "loss": 0.9336, + "mean_token_accuracy": 0.7276982069015503, + "num_tokens": 513450935.0, + "step": 19841 + }, + { + "epoch": 2.179002855260268, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0744593143463135, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.7063617706298828, + "num_tokens": 513476918.0, + "step": 19842 + }, + { + "epoch": 2.179112672962882, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3729708194732666, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7079391479492188, + "num_tokens": 513508803.0, + "step": 19843 + }, + { + "epoch": 2.179222490665495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7475883960723877, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7073369026184082, + "num_tokens": 513533374.0, + "step": 19844 + }, + { + "epoch": 2.179332308368109, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8624022006988525, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7210139632225037, + "num_tokens": 513557068.0, + "step": 19845 + }, + { + "epoch": 2.1794421260707226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8664002418518066, + "learning_rate": 1e-06, + "loss": 0.9508, + "mean_token_accuracy": 0.7176283001899719, + "num_tokens": 513579110.0, + "step": 19846 + }, + { + "epoch": 2.1795519437733364, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.556690216064453, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7043850421905518, + "num_tokens": 513604626.0, + "step": 19847 + }, + { + "epoch": 2.17966176147595, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.567394495010376, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6944843530654907, + "num_tokens": 513630768.0, + "step": 19848 + }, + { + "epoch": 2.1797715791785635, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.206066370010376, + "learning_rate": 1e-06, + "loss": 0.9361, + "mean_token_accuracy": 0.7233351469039917, + "num_tokens": 513656694.0, + "step": 19849 + }, + { + "epoch": 2.179881396881177, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5581040382385254, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6975380182266235, + "num_tokens": 513685769.0, + "step": 19850 + }, + { + "epoch": 2.179991214583791, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3506784439086914, + "learning_rate": 1e-06, + "loss": 1.0925, + "mean_token_accuracy": 0.6858950853347778, + "num_tokens": 513719014.0, + "step": 19851 + }, + { + "epoch": 2.1801010322864047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5139434337615967, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6966307163238525, + "num_tokens": 513745490.0, + "step": 19852 + }, + { + "epoch": 2.180210849989018, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3917651176452637, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6989144086837769, + "num_tokens": 513774431.0, + "step": 19853 + }, + { + "epoch": 2.180320667691632, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.453629732131958, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7101313471794128, + "num_tokens": 513799223.0, + "step": 19854 + }, + { + "epoch": 2.1804304853942456, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5301499366760254, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6991221904754639, + "num_tokens": 513826121.0, + "step": 19855 + }, + { + "epoch": 2.1805403030968593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.394944906234741, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6988511085510254, + "num_tokens": 513857479.0, + "step": 19856 + }, + { + "epoch": 2.180650120799473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.834723711013794, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6956970691680908, + "num_tokens": 513878498.0, + "step": 19857 + }, + { + "epoch": 2.1807599385020864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.751025438308716, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7140433192253113, + "num_tokens": 513899898.0, + "step": 19858 + }, + { + "epoch": 2.1808697562047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.52579927444458, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6989220380783081, + "num_tokens": 513929478.0, + "step": 19859 + }, + { + "epoch": 2.180979573907314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6457982063293457, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.705199122428894, + "num_tokens": 513956905.0, + "step": 19860 + }, + { + "epoch": 2.1810893916099277, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.604572296142578, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7218181490898132, + "num_tokens": 513984325.0, + "step": 19861 + }, + { + "epoch": 2.181199209312541, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561661720275879, + "learning_rate": 1e-06, + "loss": 1.0992, + "mean_token_accuracy": 0.6867942214012146, + "num_tokens": 514013885.0, + "step": 19862 + }, + { + "epoch": 2.1813090270151547, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3100838661193848, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7033714056015015, + "num_tokens": 514045712.0, + "step": 19863 + }, + { + "epoch": 2.1814188447177685, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9364049434661865, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7119097709655762, + "num_tokens": 514067777.0, + "step": 19864 + }, + { + "epoch": 2.1815286624203822, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.064798355102539, + "learning_rate": 1e-06, + "loss": 0.8845, + "mean_token_accuracy": 0.7329709529876709, + "num_tokens": 514084239.0, + "step": 19865 + }, + { + "epoch": 2.181638480122996, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.660527467727661, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6962460279464722, + "num_tokens": 514111071.0, + "step": 19866 + }, + { + "epoch": 2.1817482978256093, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5720136165618896, + "learning_rate": 1e-06, + "loss": 0.9604, + "mean_token_accuracy": 0.7217434644699097, + "num_tokens": 514139515.0, + "step": 19867 + }, + { + "epoch": 2.181858115528223, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5262982845306396, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6940983533859253, + "num_tokens": 514166181.0, + "step": 19868 + }, + { + "epoch": 2.181967933230837, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2887113094329834, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7225779891014099, + "num_tokens": 514194972.0, + "step": 19869 + }, + { + "epoch": 2.1820777509334506, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4418435096740723, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6978379487991333, + "num_tokens": 514224534.0, + "step": 19870 + }, + { + "epoch": 2.1821875686360643, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4104866981506348, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7205506563186646, + "num_tokens": 514256824.0, + "step": 19871 + }, + { + "epoch": 2.1822973863386776, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.389784336090088, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7169355154037476, + "num_tokens": 514284891.0, + "step": 19872 + }, + { + "epoch": 2.1824072040412914, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.894097328186035, + "learning_rate": 1e-06, + "loss": 0.9318, + "mean_token_accuracy": 0.7231783866882324, + "num_tokens": 514306297.0, + "step": 19873 + }, + { + "epoch": 2.182517021743905, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5389976501464844, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7181966304779053, + "num_tokens": 514331575.0, + "step": 19874 + }, + { + "epoch": 2.182626839446519, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.601780414581299, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7103031873703003, + "num_tokens": 514355046.0, + "step": 19875 + }, + { + "epoch": 2.1827366571491327, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5337796211242676, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.712009608745575, + "num_tokens": 514381567.0, + "step": 19876 + }, + { + "epoch": 2.182846474851746, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.738426685333252, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7210476398468018, + "num_tokens": 514404830.0, + "step": 19877 + }, + { + "epoch": 2.1829562925543597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3235154151916504, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7153565883636475, + "num_tokens": 514434722.0, + "step": 19878 + }, + { + "epoch": 2.1830661102569735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.838947296142578, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.696829080581665, + "num_tokens": 514457611.0, + "step": 19879 + }, + { + "epoch": 2.1831759279595873, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.82509708404541, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7206489443778992, + "num_tokens": 514477830.0, + "step": 19880 + }, + { + "epoch": 2.1832857456622006, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.416151285171509, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7197586297988892, + "num_tokens": 514503871.0, + "step": 19881 + }, + { + "epoch": 2.1833955633648143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8346378803253174, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7102583050727844, + "num_tokens": 514525892.0, + "step": 19882 + }, + { + "epoch": 2.183505381067428, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3784751892089844, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6957859992980957, + "num_tokens": 514557113.0, + "step": 19883 + }, + { + "epoch": 2.183615198770042, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1778275966644287, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7151713967323303, + "num_tokens": 514588219.0, + "step": 19884 + }, + { + "epoch": 2.1837250164726556, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1959075927734375, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7177206873893738, + "num_tokens": 514615084.0, + "step": 19885 + }, + { + "epoch": 2.183834834175269, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4652657508850098, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7149824500083923, + "num_tokens": 514639856.0, + "step": 19886 + }, + { + "epoch": 2.1839446518778827, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.030594825744629, + "learning_rate": 1e-06, + "loss": 0.9428, + "mean_token_accuracy": 0.7257370948791504, + "num_tokens": 514659269.0, + "step": 19887 + }, + { + "epoch": 2.1840544695804964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7024829387664795, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7069814205169678, + "num_tokens": 514682620.0, + "step": 19888 + }, + { + "epoch": 2.18416428728311, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.028668165206909, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6958234310150146, + "num_tokens": 514705976.0, + "step": 19889 + }, + { + "epoch": 2.1842741049857235, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6988203525543213, + "learning_rate": 1e-06, + "loss": 0.8971, + "mean_token_accuracy": 0.7371349334716797, + "num_tokens": 514727483.0, + "step": 19890 + }, + { + "epoch": 2.1843839226883373, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7406327724456787, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7063605189323425, + "num_tokens": 514750353.0, + "step": 19891 + }, + { + "epoch": 2.184493740390951, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.602017641067505, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.6992703080177307, + "num_tokens": 514776854.0, + "step": 19892 + }, + { + "epoch": 2.1846035580935648, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.38006591796875, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7097790241241455, + "num_tokens": 514809145.0, + "step": 19893 + }, + { + "epoch": 2.1847133757961785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.148303985595703, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7095315456390381, + "num_tokens": 514837614.0, + "step": 19894 + }, + { + "epoch": 2.184823193498792, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.683417320251465, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6780905723571777, + "num_tokens": 514863122.0, + "step": 19895 + }, + { + "epoch": 2.1849330112014056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8047595024108887, + "learning_rate": 1e-06, + "loss": 0.9369, + "mean_token_accuracy": 0.7287613153457642, + "num_tokens": 514884835.0, + "step": 19896 + }, + { + "epoch": 2.1850428289040194, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6831209659576416, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7158870100975037, + "num_tokens": 514910027.0, + "step": 19897 + }, + { + "epoch": 2.185152646606633, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7544922828674316, + "learning_rate": 1e-06, + "loss": 0.9402, + "mean_token_accuracy": 0.7374780178070068, + "num_tokens": 514934175.0, + "step": 19898 + }, + { + "epoch": 2.185262464309247, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.951767683029175, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.704576849937439, + "num_tokens": 514956012.0, + "step": 19899 + }, + { + "epoch": 2.18537228201186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5633909702301025, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7126561999320984, + "num_tokens": 514982460.0, + "step": 19900 + }, + { + "epoch": 2.185482099714474, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7737910747528076, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.72670578956604, + "num_tokens": 515005156.0, + "step": 19901 + }, + { + "epoch": 2.1855919174170877, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7560312747955322, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.6995229125022888, + "num_tokens": 515030699.0, + "step": 19902 + }, + { + "epoch": 2.1857017351197014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5435240268707275, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7079133987426758, + "num_tokens": 515055876.0, + "step": 19903 + }, + { + "epoch": 2.1858115528223148, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.458022356033325, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6947727203369141, + "num_tokens": 515086888.0, + "step": 19904 + }, + { + "epoch": 2.1859213705249285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4812793731689453, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.715879499912262, + "num_tokens": 515115200.0, + "step": 19905 + }, + { + "epoch": 2.1860311882275423, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5279335975646973, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7101588249206543, + "num_tokens": 515141586.0, + "step": 19906 + }, + { + "epoch": 2.186141005930156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5454649925231934, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.711450457572937, + "num_tokens": 515167298.0, + "step": 19907 + }, + { + "epoch": 2.18625082363277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.666468858718872, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7128417491912842, + "num_tokens": 515190325.0, + "step": 19908 + }, + { + "epoch": 2.186360641335383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6449341773986816, + "learning_rate": 1e-06, + "loss": 0.9313, + "mean_token_accuracy": 0.7235808372497559, + "num_tokens": 515215158.0, + "step": 19909 + }, + { + "epoch": 2.186470459037997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3920953273773193, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7195121049880981, + "num_tokens": 515244882.0, + "step": 19910 + }, + { + "epoch": 2.1865802767406106, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8891801834106445, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7168856859207153, + "num_tokens": 515265951.0, + "step": 19911 + }, + { + "epoch": 2.1866900944432244, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.821082353591919, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7042515277862549, + "num_tokens": 515287808.0, + "step": 19912 + }, + { + "epoch": 2.1867999121458377, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7182698249816895, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7124291658401489, + "num_tokens": 515312532.0, + "step": 19913 + }, + { + "epoch": 2.1869097298484514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.162297487258911, + "learning_rate": 1e-06, + "loss": 0.9593, + "mean_token_accuracy": 0.7268692255020142, + "num_tokens": 515330522.0, + "step": 19914 + }, + { + "epoch": 2.187019547551065, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8150670528411865, + "learning_rate": 1e-06, + "loss": 1.1373, + "mean_token_accuracy": 0.6853125095367432, + "num_tokens": 515353629.0, + "step": 19915 + }, + { + "epoch": 2.187129365253679, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.76545786857605, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7153732776641846, + "num_tokens": 515377622.0, + "step": 19916 + }, + { + "epoch": 2.1872391829562927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.694288492202759, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7005564570426941, + "num_tokens": 515402491.0, + "step": 19917 + }, + { + "epoch": 2.187349000658906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.58278751373291, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7321916818618774, + "num_tokens": 515427669.0, + "step": 19918 + }, + { + "epoch": 2.18745881836152, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.463168144226074, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6810135245323181, + "num_tokens": 515455641.0, + "step": 19919 + }, + { + "epoch": 2.1875686360641335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.557551860809326, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7205580472946167, + "num_tokens": 515480433.0, + "step": 19920 + }, + { + "epoch": 2.1876784537667473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8541784286499023, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7101372480392456, + "num_tokens": 515506270.0, + "step": 19921 + }, + { + "epoch": 2.187788271469361, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6334939002990723, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7087987661361694, + "num_tokens": 515532306.0, + "step": 19922 + }, + { + "epoch": 2.1878980891719744, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.615339756011963, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.6972718834877014, + "num_tokens": 515556984.0, + "step": 19923 + }, + { + "epoch": 2.188007906874588, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.681103467941284, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7064390182495117, + "num_tokens": 515581961.0, + "step": 19924 + }, + { + "epoch": 2.188117724577202, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1344165802001953, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.719272255897522, + "num_tokens": 515599473.0, + "step": 19925 + }, + { + "epoch": 2.1882275422798156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.664591073989868, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6920803189277649, + "num_tokens": 515623550.0, + "step": 19926 + }, + { + "epoch": 2.1883373599824294, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 2.836577892303467, + "learning_rate": 1e-06, + "loss": 0.8677, + "mean_token_accuracy": 0.7381736040115356, + "num_tokens": 515647512.0, + "step": 19927 + }, + { + "epoch": 2.1884471776850427, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0339341163635254, + "learning_rate": 1e-06, + "loss": 0.9022, + "mean_token_accuracy": 0.7371259331703186, + "num_tokens": 515666805.0, + "step": 19928 + }, + { + "epoch": 2.1885569953876565, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.862555980682373, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7172600030899048, + "num_tokens": 515691097.0, + "step": 19929 + }, + { + "epoch": 2.1886668130902702, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4491400718688965, + "learning_rate": 1e-06, + "loss": 1.1352, + "mean_token_accuracy": 0.6804078817367554, + "num_tokens": 515721818.0, + "step": 19930 + }, + { + "epoch": 2.188776630792884, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4083621501922607, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6919174194335938, + "num_tokens": 515754030.0, + "step": 19931 + }, + { + "epoch": 2.1888864484954973, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5794618129730225, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7113862633705139, + "num_tokens": 515780601.0, + "step": 19932 + }, + { + "epoch": 2.188996266198111, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6159555912017822, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7163457870483398, + "num_tokens": 515803712.0, + "step": 19933 + }, + { + "epoch": 2.189106083900725, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.628037929534912, + "learning_rate": 1e-06, + "loss": 0.9292, + "mean_token_accuracy": 0.7277088165283203, + "num_tokens": 515826421.0, + "step": 19934 + }, + { + "epoch": 2.1892159016033386, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6758968830108643, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.693615198135376, + "num_tokens": 515852185.0, + "step": 19935 + }, + { + "epoch": 2.1893257193059523, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.564434766769409, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6975948214530945, + "num_tokens": 515880345.0, + "step": 19936 + }, + { + "epoch": 2.1894355370085656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.800476551055908, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.706379234790802, + "num_tokens": 515909406.0, + "step": 19937 + }, + { + "epoch": 2.1895453547111794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.484863042831421, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.7154802680015564, + "num_tokens": 515938440.0, + "step": 19938 + }, + { + "epoch": 2.189655172413793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8388776779174805, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7062304615974426, + "num_tokens": 515960058.0, + "step": 19939 + }, + { + "epoch": 2.189764990116407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.906585931777954, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7130908966064453, + "num_tokens": 515980139.0, + "step": 19940 + }, + { + "epoch": 2.18987480781902, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7672743797302246, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7139614820480347, + "num_tokens": 516000716.0, + "step": 19941 + }, + { + "epoch": 2.189984625521634, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1722731590270996, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7206493020057678, + "num_tokens": 516019694.0, + "step": 19942 + }, + { + "epoch": 2.1900944432242477, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5266506671905518, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6934802532196045, + "num_tokens": 516049134.0, + "step": 19943 + }, + { + "epoch": 2.1902042609268615, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505600929260254, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7004393339157104, + "num_tokens": 516076668.0, + "step": 19944 + }, + { + "epoch": 2.1903140786294752, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.67146372795105, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.711319088935852, + "num_tokens": 516101344.0, + "step": 19945 + }, + { + "epoch": 2.1904238963320886, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.711369037628174, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6937465667724609, + "num_tokens": 516127699.0, + "step": 19946 + }, + { + "epoch": 2.1905337140347023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.677330255508423, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7215811014175415, + "num_tokens": 516153496.0, + "step": 19947 + }, + { + "epoch": 2.190643531737316, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6957929134368896, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6883590817451477, + "num_tokens": 516179007.0, + "step": 19948 + }, + { + "epoch": 2.19075334943993, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.529033899307251, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7066727876663208, + "num_tokens": 516203928.0, + "step": 19949 + }, + { + "epoch": 2.1908631671425436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.769102096557617, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7172939777374268, + "num_tokens": 516227878.0, + "step": 19950 + }, + { + "epoch": 2.190972984845157, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.666484832763672, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7015880346298218, + "num_tokens": 516254967.0, + "step": 19951 + }, + { + "epoch": 2.1910828025477707, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6532418727874756, + "learning_rate": 1e-06, + "loss": 0.9198, + "mean_token_accuracy": 0.7310366630554199, + "num_tokens": 516277432.0, + "step": 19952 + }, + { + "epoch": 2.1911926202503844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0309174060821533, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.6982436180114746, + "num_tokens": 516300329.0, + "step": 19953 + }, + { + "epoch": 2.191302437952998, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6000328063964844, + "learning_rate": 1e-06, + "loss": 1.082, + "mean_token_accuracy": 0.69003826379776, + "num_tokens": 516328649.0, + "step": 19954 + }, + { + "epoch": 2.1914122556556115, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6462337970733643, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.7018678188323975, + "num_tokens": 516353113.0, + "step": 19955 + }, + { + "epoch": 2.1915220733582252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.332770347595215, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7183504104614258, + "num_tokens": 516383463.0, + "step": 19956 + }, + { + "epoch": 2.191631891060839, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5484607219696045, + "learning_rate": 1e-06, + "loss": 0.9198, + "mean_token_accuracy": 0.7291434407234192, + "num_tokens": 516407754.0, + "step": 19957 + }, + { + "epoch": 2.1917417087634528, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6855721473693848, + "learning_rate": 1e-06, + "loss": 0.9341, + "mean_token_accuracy": 0.7237322330474854, + "num_tokens": 516428975.0, + "step": 19958 + }, + { + "epoch": 2.1918515264660665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 4.63739013671875, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7241644859313965, + "num_tokens": 516448000.0, + "step": 19959 + }, + { + "epoch": 2.19196134416868, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5856499671936035, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.690913200378418, + "num_tokens": 516475023.0, + "step": 19960 + }, + { + "epoch": 2.1920711618712936, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.670938491821289, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.716702938079834, + "num_tokens": 516501229.0, + "step": 19961 + }, + { + "epoch": 2.1921809795739073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617496967315674, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7131516933441162, + "num_tokens": 516532095.0, + "step": 19962 + }, + { + "epoch": 2.192290797276521, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7337820529937744, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7080579996109009, + "num_tokens": 516555590.0, + "step": 19963 + }, + { + "epoch": 2.1924006149791344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4257192611694336, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.7012829184532166, + "num_tokens": 516586018.0, + "step": 19964 + }, + { + "epoch": 2.192510432681748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6759743690490723, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.6976225972175598, + "num_tokens": 516610101.0, + "step": 19965 + }, + { + "epoch": 2.192620250384362, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7220780849456787, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.710168182849884, + "num_tokens": 516634371.0, + "step": 19966 + }, + { + "epoch": 2.1927300680869757, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.623602867126465, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7252553105354309, + "num_tokens": 516659537.0, + "step": 19967 + }, + { + "epoch": 2.1928398857895894, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.627581834793091, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7157554626464844, + "num_tokens": 516684554.0, + "step": 19968 + }, + { + "epoch": 2.1929497034922028, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.714341402053833, + "learning_rate": 1e-06, + "loss": 0.9099, + "mean_token_accuracy": 0.7285706400871277, + "num_tokens": 516706127.0, + "step": 19969 + }, + { + "epoch": 2.1930595211948165, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.672981023788452, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7130423784255981, + "num_tokens": 516729918.0, + "step": 19970 + }, + { + "epoch": 2.1931693388974303, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.542534589767456, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7252388000488281, + "num_tokens": 516754609.0, + "step": 19971 + }, + { + "epoch": 2.193279156600044, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.939445972442627, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6944072246551514, + "num_tokens": 516779618.0, + "step": 19972 + }, + { + "epoch": 2.193388974302658, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.6063363552093506, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.718503475189209, + "num_tokens": 516803725.0, + "step": 19973 + }, + { + "epoch": 2.193498792005271, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.646543025970459, + "learning_rate": 1e-06, + "loss": 0.9176, + "mean_token_accuracy": 0.7308294773101807, + "num_tokens": 516829951.0, + "step": 19974 + }, + { + "epoch": 2.193608609707885, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4325802326202393, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6838035583496094, + "num_tokens": 516861823.0, + "step": 19975 + }, + { + "epoch": 2.1937184274104986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7580060958862305, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7005778551101685, + "num_tokens": 516888694.0, + "step": 19976 + }, + { + "epoch": 2.1938282451131124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5350661277770996, + "learning_rate": 1e-06, + "loss": 0.907, + "mean_token_accuracy": 0.7314736843109131, + "num_tokens": 516913859.0, + "step": 19977 + }, + { + "epoch": 2.193938062815726, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9301655292510986, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7087873816490173, + "num_tokens": 516934106.0, + "step": 19978 + }, + { + "epoch": 2.1940478805183394, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.483259439468384, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7198010087013245, + "num_tokens": 516962220.0, + "step": 19979 + }, + { + "epoch": 2.194157698220953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.470189332962036, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7037347555160522, + "num_tokens": 516992097.0, + "step": 19980 + }, + { + "epoch": 2.194267515923567, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2829482555389404, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7114982604980469, + "num_tokens": 517024985.0, + "step": 19981 + }, + { + "epoch": 2.1943773336261807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6295361518859863, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7061409950256348, + "num_tokens": 517049686.0, + "step": 19982 + }, + { + "epoch": 2.194487151328794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7454540729522705, + "learning_rate": 1e-06, + "loss": 0.977, + "mean_token_accuracy": 0.7099012136459351, + "num_tokens": 517073345.0, + "step": 19983 + }, + { + "epoch": 2.1945969690314078, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4365832805633545, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7129387855529785, + "num_tokens": 517099659.0, + "step": 19984 + }, + { + "epoch": 2.1947067867340215, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4408023357391357, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.700823187828064, + "num_tokens": 517127778.0, + "step": 19985 + }, + { + "epoch": 2.1948166044366353, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.771691083908081, + "learning_rate": 1e-06, + "loss": 0.8588, + "mean_token_accuracy": 0.7391160726547241, + "num_tokens": 517147665.0, + "step": 19986 + }, + { + "epoch": 2.194926422139249, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.806788206100464, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6978867650032043, + "num_tokens": 517173900.0, + "step": 19987 + }, + { + "epoch": 2.1950362398418624, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6316518783569336, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7165619730949402, + "num_tokens": 517201170.0, + "step": 19988 + }, + { + "epoch": 2.195146057544476, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4894216060638428, + "learning_rate": 1e-06, + "loss": 1.1237, + "mean_token_accuracy": 0.6717474460601807, + "num_tokens": 517230427.0, + "step": 19989 + }, + { + "epoch": 2.19525587524709, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.48507022857666, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.694271981716156, + "num_tokens": 517265792.0, + "step": 19990 + }, + { + "epoch": 2.1953656929497036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.825259208679199, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7143954038619995, + "num_tokens": 517287592.0, + "step": 19991 + }, + { + "epoch": 2.195475510652317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0049662590026855, + "learning_rate": 1e-06, + "loss": 0.9223, + "mean_token_accuracy": 0.7246800661087036, + "num_tokens": 517310886.0, + "step": 19992 + }, + { + "epoch": 2.1955853283549307, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.869832992553711, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.721583366394043, + "num_tokens": 517332572.0, + "step": 19993 + }, + { + "epoch": 2.1956951460575445, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.865279197692871, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7104553580284119, + "num_tokens": 517355634.0, + "step": 19994 + }, + { + "epoch": 2.195804963760158, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6206419467926025, + "learning_rate": 1e-06, + "loss": 1.0959, + "mean_token_accuracy": 0.6805282831192017, + "num_tokens": 517385468.0, + "step": 19995 + }, + { + "epoch": 2.195914781462772, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.332252025604248, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.7018272876739502, + "num_tokens": 517417204.0, + "step": 19996 + }, + { + "epoch": 2.1960245991653853, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7303576469421387, + "learning_rate": 1e-06, + "loss": 0.8908, + "mean_token_accuracy": 0.7366251945495605, + "num_tokens": 517441033.0, + "step": 19997 + }, + { + "epoch": 2.196134416867999, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.531695604324341, + "learning_rate": 1e-06, + "loss": 1.0849, + "mean_token_accuracy": 0.6892904043197632, + "num_tokens": 517469258.0, + "step": 19998 + }, + { + "epoch": 2.196244234570613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.69392991065979, + "learning_rate": 1e-06, + "loss": 1.1061, + "mean_token_accuracy": 0.6861956715583801, + "num_tokens": 517495697.0, + "step": 19999 + }, + { + "epoch": 2.1963540522732266, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.541004180908203, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7117398977279663, + "num_tokens": 517526328.0, + "step": 20000 + }, + { + "epoch": 2.1964638699758403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6029984951019287, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.684088945388794, + "num_tokens": 517556477.0, + "step": 20001 + }, + { + "epoch": 2.1965736876784536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5628113746643066, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.7035274505615234, + "num_tokens": 517583923.0, + "step": 20002 + }, + { + "epoch": 2.1966835053810674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.394033193588257, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6911916732788086, + "num_tokens": 517618071.0, + "step": 20003 + }, + { + "epoch": 2.196793323083681, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8891777992248535, + "learning_rate": 1e-06, + "loss": 0.9189, + "mean_token_accuracy": 0.7396491169929504, + "num_tokens": 517640507.0, + "step": 20004 + }, + { + "epoch": 2.196903140786295, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8545801639556885, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7177482843399048, + "num_tokens": 517665619.0, + "step": 20005 + }, + { + "epoch": 2.197012958488908, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.825904369354248, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7224157452583313, + "num_tokens": 517693362.0, + "step": 20006 + }, + { + "epoch": 2.197122776191522, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.897202968597412, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7088427543640137, + "num_tokens": 517716288.0, + "step": 20007 + }, + { + "epoch": 2.1972325938941357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.991225242614746, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6927983164787292, + "num_tokens": 517741774.0, + "step": 20008 + }, + { + "epoch": 2.1973424115967495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.467019557952881, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.693770170211792, + "num_tokens": 517773664.0, + "step": 20009 + }, + { + "epoch": 2.1974522292993632, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.979027509689331, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7175305485725403, + "num_tokens": 517793926.0, + "step": 20010 + }, + { + "epoch": 2.1975620470019765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6089255809783936, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6972615718841553, + "num_tokens": 517823769.0, + "step": 20011 + }, + { + "epoch": 2.1976718647045903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.696136474609375, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7106454968452454, + "num_tokens": 517849202.0, + "step": 20012 + }, + { + "epoch": 2.197781682407204, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520261764526367, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.7006811499595642, + "num_tokens": 517877265.0, + "step": 20013 + }, + { + "epoch": 2.197891500109818, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.59904408454895, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7122378349304199, + "num_tokens": 517900377.0, + "step": 20014 + }, + { + "epoch": 2.198001317812431, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.490799903869629, + "learning_rate": 1e-06, + "loss": 0.8893, + "mean_token_accuracy": 0.734145998954773, + "num_tokens": 517927383.0, + "step": 20015 + }, + { + "epoch": 2.198111135515045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.46695876121521, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6875112056732178, + "num_tokens": 517957042.0, + "step": 20016 + }, + { + "epoch": 2.1982209532176586, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5049245357513428, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7001534104347229, + "num_tokens": 517982093.0, + "step": 20017 + }, + { + "epoch": 2.1983307709202724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.567413806915283, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7297037839889526, + "num_tokens": 518006726.0, + "step": 20018 + }, + { + "epoch": 2.198440588622886, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.375108003616333, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.7059957981109619, + "num_tokens": 518038182.0, + "step": 20019 + }, + { + "epoch": 2.1985504063254995, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6898932456970215, + "learning_rate": 1e-06, + "loss": 0.8986, + "mean_token_accuracy": 0.7359618544578552, + "num_tokens": 518061637.0, + "step": 20020 + }, + { + "epoch": 2.1986602240281132, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.431370973587036, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7005677223205566, + "num_tokens": 518091519.0, + "step": 20021 + }, + { + "epoch": 2.198770041730727, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.09407901763916, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6866986155509949, + "num_tokens": 518116665.0, + "step": 20022 + }, + { + "epoch": 2.1988798594333407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5053775310516357, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.6970383524894714, + "num_tokens": 518151433.0, + "step": 20023 + }, + { + "epoch": 2.1989896771359545, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0158650875091553, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6899261474609375, + "num_tokens": 518172320.0, + "step": 20024 + }, + { + "epoch": 2.199099494838568, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8644070625305176, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.740394115447998, + "num_tokens": 518192686.0, + "step": 20025 + }, + { + "epoch": 2.1992093125411816, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6878247261047363, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6953511238098145, + "num_tokens": 518219581.0, + "step": 20026 + }, + { + "epoch": 2.1993191302437953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.478851795196533, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7069343328475952, + "num_tokens": 518247489.0, + "step": 20027 + }, + { + "epoch": 2.199428947946409, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2102854251861572, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7161932587623596, + "num_tokens": 518269276.0, + "step": 20028 + }, + { + "epoch": 2.199538765649023, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4843177795410156, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.72549968957901, + "num_tokens": 518297160.0, + "step": 20029 + }, + { + "epoch": 2.199648583351636, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5828099250793457, + "learning_rate": 1e-06, + "loss": 1.0776, + "mean_token_accuracy": 0.69216388463974, + "num_tokens": 518324758.0, + "step": 20030 + }, + { + "epoch": 2.19975840105425, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4740726947784424, + "learning_rate": 1e-06, + "loss": 0.917, + "mean_token_accuracy": 0.7224681377410889, + "num_tokens": 518351349.0, + "step": 20031 + }, + { + "epoch": 2.1998682187568637, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.516700267791748, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7191869616508484, + "num_tokens": 518379043.0, + "step": 20032 + }, + { + "epoch": 2.1999780364594774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.637693166732788, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7031729221343994, + "num_tokens": 518408717.0, + "step": 20033 + }, + { + "epoch": 2.2000878541620907, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7461283206939697, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7029595375061035, + "num_tokens": 518433853.0, + "step": 20034 + }, + { + "epoch": 2.2001976718647045, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.803380012512207, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.692084789276123, + "num_tokens": 518457724.0, + "step": 20035 + }, + { + "epoch": 2.2003074895673183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.634134531021118, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7323093414306641, + "num_tokens": 518485575.0, + "step": 20036 + }, + { + "epoch": 2.200417307269932, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.708279609680176, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6927086710929871, + "num_tokens": 518517005.0, + "step": 20037 + }, + { + "epoch": 2.2005271249725458, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8776767253875732, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6940251588821411, + "num_tokens": 518541236.0, + "step": 20038 + }, + { + "epoch": 2.200636942675159, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.689211130142212, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7149409055709839, + "num_tokens": 518565424.0, + "step": 20039 + }, + { + "epoch": 2.200746760377773, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7038235664367676, + "learning_rate": 1e-06, + "loss": 0.9257, + "mean_token_accuracy": 0.7301531434059143, + "num_tokens": 518589389.0, + "step": 20040 + }, + { + "epoch": 2.2008565780803866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.547027587890625, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7064707279205322, + "num_tokens": 518614095.0, + "step": 20041 + }, + { + "epoch": 2.2009663957830004, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7059969902038574, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7080789804458618, + "num_tokens": 518638191.0, + "step": 20042 + }, + { + "epoch": 2.2010762134856137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.726924180984497, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7120802402496338, + "num_tokens": 518660484.0, + "step": 20043 + }, + { + "epoch": 2.2011860311882274, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.580038070678711, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.692368745803833, + "num_tokens": 518687926.0, + "step": 20044 + }, + { + "epoch": 2.201295848890841, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.404552936553955, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6934516429901123, + "num_tokens": 518719279.0, + "step": 20045 + }, + { + "epoch": 2.201405666593455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.397742509841919, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6983476877212524, + "num_tokens": 518753491.0, + "step": 20046 + }, + { + "epoch": 2.2015154842960687, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5110321044921875, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6997891664505005, + "num_tokens": 518780926.0, + "step": 20047 + }, + { + "epoch": 2.201625301998682, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.801851272583008, + "learning_rate": 1e-06, + "loss": 1.0928, + "mean_token_accuracy": 0.6860044002532959, + "num_tokens": 518805478.0, + "step": 20048 + }, + { + "epoch": 2.2017351197012958, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.943469285964966, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7122671604156494, + "num_tokens": 518830675.0, + "step": 20049 + }, + { + "epoch": 2.2018449374039095, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6230530738830566, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7274372577667236, + "num_tokens": 518854521.0, + "step": 20050 + }, + { + "epoch": 2.2019547551065233, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5288071632385254, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.700423002243042, + "num_tokens": 518882182.0, + "step": 20051 + }, + { + "epoch": 2.202064572809137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.588871479034424, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7086325883865356, + "num_tokens": 518911153.0, + "step": 20052 + }, + { + "epoch": 2.2021743905117503, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6814515590667725, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6991797089576721, + "num_tokens": 518938259.0, + "step": 20053 + }, + { + "epoch": 2.202284208214364, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.594012975692749, + "learning_rate": 1e-06, + "loss": 0.8634, + "mean_token_accuracy": 0.7451852560043335, + "num_tokens": 518960702.0, + "step": 20054 + }, + { + "epoch": 2.202394025916978, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.523892879486084, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.69206702709198, + "num_tokens": 518989298.0, + "step": 20055 + }, + { + "epoch": 2.2025038436195916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6978931427001953, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6894968152046204, + "num_tokens": 519014972.0, + "step": 20056 + }, + { + "epoch": 2.2026136613222054, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.847783088684082, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.716459333896637, + "num_tokens": 519038925.0, + "step": 20057 + }, + { + "epoch": 2.2027234790248187, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.680048704147339, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7078623175621033, + "num_tokens": 519064807.0, + "step": 20058 + }, + { + "epoch": 2.2028332967274324, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.717036724090576, + "learning_rate": 1e-06, + "loss": 0.8848, + "mean_token_accuracy": 0.7369643449783325, + "num_tokens": 519086927.0, + "step": 20059 + }, + { + "epoch": 2.202943114430046, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.098032236099243, + "learning_rate": 1e-06, + "loss": 0.9035, + "mean_token_accuracy": 0.7361985445022583, + "num_tokens": 519111766.0, + "step": 20060 + }, + { + "epoch": 2.20305293213266, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.522662401199341, + "learning_rate": 1e-06, + "loss": 1.0479, + "mean_token_accuracy": 0.694908082485199, + "num_tokens": 519140981.0, + "step": 20061 + }, + { + "epoch": 2.2031627498352733, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.641145706176758, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6943529844284058, + "num_tokens": 519165772.0, + "step": 20062 + }, + { + "epoch": 2.203272567537887, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.609351634979248, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.725087583065033, + "num_tokens": 519191829.0, + "step": 20063 + }, + { + "epoch": 2.203382385240501, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7653913497924805, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6908660531044006, + "num_tokens": 519216179.0, + "step": 20064 + }, + { + "epoch": 2.2034922029431145, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3839128017425537, + "learning_rate": 1e-06, + "loss": 0.912, + "mean_token_accuracy": 0.7308673858642578, + "num_tokens": 519245075.0, + "step": 20065 + }, + { + "epoch": 2.2036020206457283, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5814759731292725, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7119525671005249, + "num_tokens": 519272624.0, + "step": 20066 + }, + { + "epoch": 2.2037118383483416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5810725688934326, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7096808552742004, + "num_tokens": 519297407.0, + "step": 20067 + }, + { + "epoch": 2.2038216560509554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.285705089569092, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.685110867023468, + "num_tokens": 519335419.0, + "step": 20068 + }, + { + "epoch": 2.203931473753569, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.89158034324646, + "learning_rate": 1e-06, + "loss": 0.9248, + "mean_token_accuracy": 0.7282014489173889, + "num_tokens": 519355629.0, + "step": 20069 + }, + { + "epoch": 2.204041291456183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.563361406326294, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7116918563842773, + "num_tokens": 519381439.0, + "step": 20070 + }, + { + "epoch": 2.204151109158796, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.836829423904419, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7009671330451965, + "num_tokens": 519404641.0, + "step": 20071 + }, + { + "epoch": 2.20426092686141, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5035746097564697, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7006710767745972, + "num_tokens": 519432333.0, + "step": 20072 + }, + { + "epoch": 2.2043707445640237, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.720453977584839, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.6997153759002686, + "num_tokens": 519462964.0, + "step": 20073 + }, + { + "epoch": 2.2044805622666375, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.53590989112854, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7140176892280579, + "num_tokens": 519487752.0, + "step": 20074 + }, + { + "epoch": 2.2045903799692512, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4515933990478516, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.7040865421295166, + "num_tokens": 519516482.0, + "step": 20075 + }, + { + "epoch": 2.2047001976718645, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7568721771240234, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.728429913520813, + "num_tokens": 519540401.0, + "step": 20076 + }, + { + "epoch": 2.2048100153744783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5699734687805176, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7003934383392334, + "num_tokens": 519564891.0, + "step": 20077 + }, + { + "epoch": 2.204919833077092, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.449016571044922, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7011823654174805, + "num_tokens": 519595185.0, + "step": 20078 + }, + { + "epoch": 2.205029650779706, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5736420154571533, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.70392906665802, + "num_tokens": 519620293.0, + "step": 20079 + }, + { + "epoch": 2.2051394684823196, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.821359872817993, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7176486253738403, + "num_tokens": 519643636.0, + "step": 20080 + }, + { + "epoch": 2.205249286184933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565016984939575, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6940156817436218, + "num_tokens": 519671611.0, + "step": 20081 + }, + { + "epoch": 2.2053591038875466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.175219774246216, + "learning_rate": 1e-06, + "loss": 1.0654, + "mean_token_accuracy": 0.695993959903717, + "num_tokens": 519706511.0, + "step": 20082 + }, + { + "epoch": 2.2054689215901604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.807955026626587, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7193117141723633, + "num_tokens": 519729225.0, + "step": 20083 + }, + { + "epoch": 2.205578739292774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6197338104248047, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.7075701951980591, + "num_tokens": 519757241.0, + "step": 20084 + }, + { + "epoch": 2.2056885569953875, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6901097297668457, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7153297066688538, + "num_tokens": 519783692.0, + "step": 20085 + }, + { + "epoch": 2.205798374698001, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.808105945587158, + "learning_rate": 1e-06, + "loss": 0.9287, + "mean_token_accuracy": 0.7311023473739624, + "num_tokens": 519807339.0, + "step": 20086 + }, + { + "epoch": 2.205908192400615, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7327516078948975, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.7202559113502502, + "num_tokens": 519830355.0, + "step": 20087 + }, + { + "epoch": 2.2060180101032287, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9198930263519287, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7080087661743164, + "num_tokens": 519859643.0, + "step": 20088 + }, + { + "epoch": 2.2061278278058425, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7649292945861816, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7193562984466553, + "num_tokens": 519883210.0, + "step": 20089 + }, + { + "epoch": 2.206237645508456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8883490562438965, + "learning_rate": 1e-06, + "loss": 0.9074, + "mean_token_accuracy": 0.7332743406295776, + "num_tokens": 519903342.0, + "step": 20090 + }, + { + "epoch": 2.2063474632110696, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.452336549758911, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7128487825393677, + "num_tokens": 519930300.0, + "step": 20091 + }, + { + "epoch": 2.2064572809136833, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.425506591796875, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.697464108467102, + "num_tokens": 519960474.0, + "step": 20092 + }, + { + "epoch": 2.206567098616297, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.492875337600708, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.71897953748703, + "num_tokens": 519986070.0, + "step": 20093 + }, + { + "epoch": 2.2066769163189104, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6287901401519775, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7195255756378174, + "num_tokens": 520011594.0, + "step": 20094 + }, + { + "epoch": 2.206786734021524, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6131696701049805, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6932158470153809, + "num_tokens": 520037705.0, + "step": 20095 + }, + { + "epoch": 2.206896551724138, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9391772747039795, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7153702974319458, + "num_tokens": 520056695.0, + "step": 20096 + }, + { + "epoch": 2.2070063694267517, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.770836114883423, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7191746830940247, + "num_tokens": 520079533.0, + "step": 20097 + }, + { + "epoch": 2.2071161871293654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.680563449859619, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7062627077102661, + "num_tokens": 520104959.0, + "step": 20098 + }, + { + "epoch": 2.2072260048319787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.74249267578125, + "learning_rate": 1e-06, + "loss": 0.9222, + "mean_token_accuracy": 0.7303287386894226, + "num_tokens": 520126584.0, + "step": 20099 + }, + { + "epoch": 2.2073358225345925, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.848085641860962, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7133443355560303, + "num_tokens": 520149976.0, + "step": 20100 + }, + { + "epoch": 2.2074456402372062, + "ewc_loss": 1.0788440704345703e-05, + "grad_norm": 3.3420774936676025, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7142165899276733, + "num_tokens": 520177916.0, + "step": 20101 + }, + { + "epoch": 2.20755545793982, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5584659576416016, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7039271593093872, + "num_tokens": 520205584.0, + "step": 20102 + }, + { + "epoch": 2.2076652756424338, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5233564376831055, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.701019287109375, + "num_tokens": 520234122.0, + "step": 20103 + }, + { + "epoch": 2.207775093345047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.665903329849243, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6864361763000488, + "num_tokens": 520258717.0, + "step": 20104 + }, + { + "epoch": 2.207884911047661, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.54022216796875, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7201782464981079, + "num_tokens": 520284535.0, + "step": 20105 + }, + { + "epoch": 2.2079947287502746, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5061750411987305, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7308750748634338, + "num_tokens": 520310647.0, + "step": 20106 + }, + { + "epoch": 2.2081045464528883, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.405100107192993, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.691335141658783, + "num_tokens": 520344625.0, + "step": 20107 + }, + { + "epoch": 2.208214364155502, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.837435483932495, + "learning_rate": 1e-06, + "loss": 0.9636, + "mean_token_accuracy": 0.7142505645751953, + "num_tokens": 520368818.0, + "step": 20108 + }, + { + "epoch": 2.2083241818581154, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7450764179229736, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7111491560935974, + "num_tokens": 520396062.0, + "step": 20109 + }, + { + "epoch": 2.208433999560729, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6487035751342773, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7080532312393188, + "num_tokens": 520423499.0, + "step": 20110 + }, + { + "epoch": 2.208543817263343, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4277074337005615, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6964989304542542, + "num_tokens": 520460909.0, + "step": 20111 + }, + { + "epoch": 2.2086536349659567, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5103349685668945, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7171838283538818, + "num_tokens": 520487116.0, + "step": 20112 + }, + { + "epoch": 2.20876345266857, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.985220432281494, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6897286176681519, + "num_tokens": 520516152.0, + "step": 20113 + }, + { + "epoch": 2.2088732703711838, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.684852123260498, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7219951152801514, + "num_tokens": 520539944.0, + "step": 20114 + }, + { + "epoch": 2.2089830880737975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4254367351531982, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7142559289932251, + "num_tokens": 520567663.0, + "step": 20115 + }, + { + "epoch": 2.2090929057764113, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8837482929229736, + "learning_rate": 1e-06, + "loss": 0.914, + "mean_token_accuracy": 0.7303557991981506, + "num_tokens": 520590056.0, + "step": 20116 + }, + { + "epoch": 2.209202723479025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.71279239654541, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7216199636459351, + "num_tokens": 520612078.0, + "step": 20117 + }, + { + "epoch": 2.2093125411816383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0138192176818848, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7133110165596008, + "num_tokens": 520631927.0, + "step": 20118 + }, + { + "epoch": 2.209422358884252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5631749629974365, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7068668603897095, + "num_tokens": 520658703.0, + "step": 20119 + }, + { + "epoch": 2.209532176586866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0231473445892334, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7181959748268127, + "num_tokens": 520677388.0, + "step": 20120 + }, + { + "epoch": 2.2096419942894796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.766026258468628, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6939057111740112, + "num_tokens": 520706453.0, + "step": 20121 + }, + { + "epoch": 2.209751811992093, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5575239658355713, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7105149030685425, + "num_tokens": 520736876.0, + "step": 20122 + }, + { + "epoch": 2.2098616296947067, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6292130947113037, + "learning_rate": 1e-06, + "loss": 1.1285, + "mean_token_accuracy": 0.6736156940460205, + "num_tokens": 520762241.0, + "step": 20123 + }, + { + "epoch": 2.2099714473973204, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561086893081665, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7140549421310425, + "num_tokens": 520785181.0, + "step": 20124 + }, + { + "epoch": 2.210081265099934, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 5.411225318908691, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.7047973871231079, + "num_tokens": 520812759.0, + "step": 20125 + }, + { + "epoch": 2.210191082802548, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.618218183517456, + "learning_rate": 1e-06, + "loss": 0.9457, + "mean_token_accuracy": 0.7235945463180542, + "num_tokens": 520839324.0, + "step": 20126 + }, + { + "epoch": 2.2103009005051613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4796197414398193, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7145581245422363, + "num_tokens": 520867844.0, + "step": 20127 + }, + { + "epoch": 2.210410718207775, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6101372241973877, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6868448853492737, + "num_tokens": 520897337.0, + "step": 20128 + }, + { + "epoch": 2.2105205359103888, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.582407236099243, + "learning_rate": 1e-06, + "loss": 0.9223, + "mean_token_accuracy": 0.7299692630767822, + "num_tokens": 520922710.0, + "step": 20129 + }, + { + "epoch": 2.2106303536130025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6825287342071533, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6937475204467773, + "num_tokens": 520948579.0, + "step": 20130 + }, + { + "epoch": 2.2107401713156163, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5628724098205566, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6975405216217041, + "num_tokens": 520976304.0, + "step": 20131 + }, + { + "epoch": 2.2108499890182296, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6767919063568115, + "learning_rate": 1e-06, + "loss": 1.118, + "mean_token_accuracy": 0.6733442544937134, + "num_tokens": 521005547.0, + "step": 20132 + }, + { + "epoch": 2.2109598067208434, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8248767852783203, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.706753134727478, + "num_tokens": 521032555.0, + "step": 20133 + }, + { + "epoch": 2.211069624423457, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4723901748657227, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6910355091094971, + "num_tokens": 521059832.0, + "step": 20134 + }, + { + "epoch": 2.211179442126071, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4789421558380127, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6978792548179626, + "num_tokens": 521088278.0, + "step": 20135 + }, + { + "epoch": 2.211289259828684, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.01053524017334, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.6958907246589661, + "num_tokens": 521111519.0, + "step": 20136 + }, + { + "epoch": 2.211399077531298, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.475543260574341, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7151268720626831, + "num_tokens": 521139870.0, + "step": 20137 + }, + { + "epoch": 2.2115088952339117, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5521483421325684, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7104694843292236, + "num_tokens": 521167763.0, + "step": 20138 + }, + { + "epoch": 2.2116187129365255, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7912232875823975, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7061647772789001, + "num_tokens": 521193950.0, + "step": 20139 + }, + { + "epoch": 2.211728530639139, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.308256149291992, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7025043964385986, + "num_tokens": 521227519.0, + "step": 20140 + }, + { + "epoch": 2.2118383483417525, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9829375743865967, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7158219814300537, + "num_tokens": 521247804.0, + "step": 20141 + }, + { + "epoch": 2.2119481660443663, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9358298778533936, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7013071179389954, + "num_tokens": 521269010.0, + "step": 20142 + }, + { + "epoch": 2.21205798374698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6481144428253174, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.7000976800918579, + "num_tokens": 521293769.0, + "step": 20143 + }, + { + "epoch": 2.212167801449594, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.649970054626465, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7145857810974121, + "num_tokens": 521318842.0, + "step": 20144 + }, + { + "epoch": 2.212277619152207, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5626509189605713, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7093901634216309, + "num_tokens": 521346698.0, + "step": 20145 + }, + { + "epoch": 2.212387436854821, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7026052474975586, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7190178036689758, + "num_tokens": 521371578.0, + "step": 20146 + }, + { + "epoch": 2.2124972545574346, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.735182523727417, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7248454689979553, + "num_tokens": 521397721.0, + "step": 20147 + }, + { + "epoch": 2.2126070722600484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.904233932495117, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7065610885620117, + "num_tokens": 521418959.0, + "step": 20148 + }, + { + "epoch": 2.212716889962662, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.510369300842285, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7028405666351318, + "num_tokens": 521447973.0, + "step": 20149 + }, + { + "epoch": 2.2128267076652754, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4585368633270264, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7078880071640015, + "num_tokens": 521475234.0, + "step": 20150 + }, + { + "epoch": 2.212936525367889, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1415274143218994, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7212020754814148, + "num_tokens": 521499190.0, + "step": 20151 + }, + { + "epoch": 2.213046343070503, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6481690406799316, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7104198932647705, + "num_tokens": 521525491.0, + "step": 20152 + }, + { + "epoch": 2.2131561607731167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7503583431243896, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7015237808227539, + "num_tokens": 521548624.0, + "step": 20153 + }, + { + "epoch": 2.2132659784757305, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3652536869049072, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.691325306892395, + "num_tokens": 521579307.0, + "step": 20154 + }, + { + "epoch": 2.213375796178344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.58735728263855, + "learning_rate": 1e-06, + "loss": 0.9012, + "mean_token_accuracy": 0.7327625751495361, + "num_tokens": 521602114.0, + "step": 20155 + }, + { + "epoch": 2.2134856138809575, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.703106641769409, + "learning_rate": 1e-06, + "loss": 0.9079, + "mean_token_accuracy": 0.730955958366394, + "num_tokens": 521625068.0, + "step": 20156 + }, + { + "epoch": 2.2135954315835713, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.611154794692993, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7216277718544006, + "num_tokens": 521649794.0, + "step": 20157 + }, + { + "epoch": 2.213705249286185, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9391415119171143, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.706640362739563, + "num_tokens": 521669433.0, + "step": 20158 + }, + { + "epoch": 2.213815066988799, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7839672565460205, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7036711573600769, + "num_tokens": 521691755.0, + "step": 20159 + }, + { + "epoch": 2.213924884691412, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7469418048858643, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7105189561843872, + "num_tokens": 521717300.0, + "step": 20160 + }, + { + "epoch": 2.214034702394026, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.503293037414551, + "learning_rate": 1e-06, + "loss": 1.128, + "mean_token_accuracy": 0.6770366430282593, + "num_tokens": 521746826.0, + "step": 20161 + }, + { + "epoch": 2.2141445200966396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8046152591705322, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7217978239059448, + "num_tokens": 521767377.0, + "step": 20162 + }, + { + "epoch": 2.2142543377992534, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.591028928756714, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.711698591709137, + "num_tokens": 521792448.0, + "step": 20163 + }, + { + "epoch": 2.2143641555018667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5282106399536133, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7098507881164551, + "num_tokens": 521817314.0, + "step": 20164 + }, + { + "epoch": 2.2144739732044805, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5993316173553467, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6997497081756592, + "num_tokens": 521844677.0, + "step": 20165 + }, + { + "epoch": 2.2145837909070942, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8169734477996826, + "learning_rate": 1e-06, + "loss": 0.9241, + "mean_token_accuracy": 0.7382969856262207, + "num_tokens": 521865563.0, + "step": 20166 + }, + { + "epoch": 2.214693608609708, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8084733486175537, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.720176100730896, + "num_tokens": 521885758.0, + "step": 20167 + }, + { + "epoch": 2.2148034263123217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.3946216106414795, + "learning_rate": 1e-06, + "loss": 0.8917, + "mean_token_accuracy": 0.7416057586669922, + "num_tokens": 521907417.0, + "step": 20168 + }, + { + "epoch": 2.214913244014935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5683255195617676, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7058969736099243, + "num_tokens": 521934415.0, + "step": 20169 + }, + { + "epoch": 2.215023061717549, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.941969156265259, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.7015048861503601, + "num_tokens": 521956542.0, + "step": 20170 + }, + { + "epoch": 2.2151328794201626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5484158992767334, + "learning_rate": 1e-06, + "loss": 0.9231, + "mean_token_accuracy": 0.7261647582054138, + "num_tokens": 521981406.0, + "step": 20171 + }, + { + "epoch": 2.2152426971227763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.580880880355835, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7050594687461853, + "num_tokens": 522006913.0, + "step": 20172 + }, + { + "epoch": 2.2153525148253896, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5524051189422607, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6981337070465088, + "num_tokens": 522032649.0, + "step": 20173 + }, + { + "epoch": 2.2154623325280034, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.934816837310791, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7155748009681702, + "num_tokens": 522054575.0, + "step": 20174 + }, + { + "epoch": 2.215572150230617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2678701877593994, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6841654777526855, + "num_tokens": 522088974.0, + "step": 20175 + }, + { + "epoch": 2.215681967933231, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5905697345733643, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7065045833587646, + "num_tokens": 522112488.0, + "step": 20176 + }, + { + "epoch": 2.2157917856358447, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8584327697753906, + "learning_rate": 1e-06, + "loss": 0.8949, + "mean_token_accuracy": 0.7331149578094482, + "num_tokens": 522133063.0, + "step": 20177 + }, + { + "epoch": 2.215901603338458, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.492945432662964, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6999198198318481, + "num_tokens": 522161195.0, + "step": 20178 + }, + { + "epoch": 2.2160114210410717, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7159268856048584, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7119389176368713, + "num_tokens": 522188462.0, + "step": 20179 + }, + { + "epoch": 2.2161212387436855, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.404906749725342, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6948667168617249, + "num_tokens": 522218385.0, + "step": 20180 + }, + { + "epoch": 2.2162310564462993, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3863110542297363, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7165321111679077, + "num_tokens": 522250015.0, + "step": 20181 + }, + { + "epoch": 2.216340874148913, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5776307582855225, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7071112990379333, + "num_tokens": 522276403.0, + "step": 20182 + }, + { + "epoch": 2.2164506918515263, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.434426784515381, + "learning_rate": 1e-06, + "loss": 1.1115, + "mean_token_accuracy": 0.6943032741546631, + "num_tokens": 522309096.0, + "step": 20183 + }, + { + "epoch": 2.21656050955414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.546536684036255, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.688650369644165, + "num_tokens": 522337840.0, + "step": 20184 + }, + { + "epoch": 2.216670327256754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.664362907409668, + "learning_rate": 1e-06, + "loss": 0.8625, + "mean_token_accuracy": 0.7415785789489746, + "num_tokens": 522361322.0, + "step": 20185 + }, + { + "epoch": 2.2167801449593676, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8447978496551514, + "learning_rate": 1e-06, + "loss": 0.934, + "mean_token_accuracy": 0.7279147505760193, + "num_tokens": 522382718.0, + "step": 20186 + }, + { + "epoch": 2.216889962661981, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.683154821395874, + "learning_rate": 1e-06, + "loss": 0.9111, + "mean_token_accuracy": 0.7310622930526733, + "num_tokens": 522407048.0, + "step": 20187 + }, + { + "epoch": 2.2169997803645947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5419044494628906, + "learning_rate": 1e-06, + "loss": 0.8534, + "mean_token_accuracy": 0.7506009936332703, + "num_tokens": 522432975.0, + "step": 20188 + }, + { + "epoch": 2.2171095980672084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6318721771240234, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7132464051246643, + "num_tokens": 522460947.0, + "step": 20189 + }, + { + "epoch": 2.217219415769822, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3082640171051025, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7161707878112793, + "num_tokens": 522491659.0, + "step": 20190 + }, + { + "epoch": 2.217329233472436, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.812387704849243, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7136889696121216, + "num_tokens": 522513786.0, + "step": 20191 + }, + { + "epoch": 2.2174390511750492, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6628754138946533, + "learning_rate": 1e-06, + "loss": 0.9726, + "mean_token_accuracy": 0.7153615355491638, + "num_tokens": 522540420.0, + "step": 20192 + }, + { + "epoch": 2.217548868877663, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9767420291900635, + "learning_rate": 1e-06, + "loss": 0.9308, + "mean_token_accuracy": 0.728083610534668, + "num_tokens": 522560474.0, + "step": 20193 + }, + { + "epoch": 2.2176586865802768, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7994635105133057, + "learning_rate": 1e-06, + "loss": 0.9393, + "mean_token_accuracy": 0.7226166129112244, + "num_tokens": 522583465.0, + "step": 20194 + }, + { + "epoch": 2.2177685042828905, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7593846321105957, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7116609811782837, + "num_tokens": 522609422.0, + "step": 20195 + }, + { + "epoch": 2.217878321985504, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.604912519454956, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.707926869392395, + "num_tokens": 522634318.0, + "step": 20196 + }, + { + "epoch": 2.2179881396881176, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6685776710510254, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7242923974990845, + "num_tokens": 522658245.0, + "step": 20197 + }, + { + "epoch": 2.2180979573907313, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.655844211578369, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7097611427307129, + "num_tokens": 522682109.0, + "step": 20198 + }, + { + "epoch": 2.218207775093345, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3529372215270996, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7074328660964966, + "num_tokens": 522713748.0, + "step": 20199 + }, + { + "epoch": 2.218317592795959, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.532362699508667, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6828169822692871, + "num_tokens": 522742450.0, + "step": 20200 + }, + { + "epoch": 2.218427410498572, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4594857692718506, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7034835815429688, + "num_tokens": 522771817.0, + "step": 20201 + }, + { + "epoch": 2.218537228201186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.449500560760498, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6929389238357544, + "num_tokens": 522799701.0, + "step": 20202 + }, + { + "epoch": 2.2186470459037997, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.637453079223633, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7107365131378174, + "num_tokens": 522824493.0, + "step": 20203 + }, + { + "epoch": 2.2187568636064134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7472922801971436, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7002474069595337, + "num_tokens": 522846289.0, + "step": 20204 + }, + { + "epoch": 2.218866681309027, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.885183095932007, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7256768345832825, + "num_tokens": 522867454.0, + "step": 20205 + }, + { + "epoch": 2.2189764990116405, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8961076736450195, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7090697288513184, + "num_tokens": 522889945.0, + "step": 20206 + }, + { + "epoch": 2.2190863167142543, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.359177827835083, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7139946222305298, + "num_tokens": 522923535.0, + "step": 20207 + }, + { + "epoch": 2.219196134416868, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.2704927921295166, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7072311639785767, + "num_tokens": 522945158.0, + "step": 20208 + }, + { + "epoch": 2.219305952119482, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.43786883354187, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7101927399635315, + "num_tokens": 522975291.0, + "step": 20209 + }, + { + "epoch": 2.2194157698220955, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5539047718048096, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7150458097457886, + "num_tokens": 523002590.0, + "step": 20210 + }, + { + "epoch": 2.219525587524709, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.99126935005188, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7010093927383423, + "num_tokens": 523024099.0, + "step": 20211 + }, + { + "epoch": 2.2196354052273226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7506155967712402, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7066972255706787, + "num_tokens": 523046434.0, + "step": 20212 + }, + { + "epoch": 2.2197452229299364, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.825643301010132, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7177289128303528, + "num_tokens": 523069020.0, + "step": 20213 + }, + { + "epoch": 2.21985504063255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.860173225402832, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.717501163482666, + "num_tokens": 523092677.0, + "step": 20214 + }, + { + "epoch": 2.2199648583351634, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.619691848754883, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.7008825540542603, + "num_tokens": 523125967.0, + "step": 20215 + }, + { + "epoch": 2.220074676037777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.665283679962158, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7115375399589539, + "num_tokens": 523152192.0, + "step": 20216 + }, + { + "epoch": 2.220184493740391, + "ewc_loss": 1.0967254638671875e-05, + "grad_norm": 80.54727172851562, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7042413949966431, + "num_tokens": 523174814.0, + "step": 20217 + }, + { + "epoch": 2.2202943114430047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.757594585418701, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6993658542633057, + "num_tokens": 523199753.0, + "step": 20218 + }, + { + "epoch": 2.2204041291456185, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6243977546691895, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6987354755401611, + "num_tokens": 523228109.0, + "step": 20219 + }, + { + "epoch": 2.220513946848232, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.546452760696411, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7017194032669067, + "num_tokens": 523257988.0, + "step": 20220 + }, + { + "epoch": 2.2206237645508455, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9036858081817627, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7169471979141235, + "num_tokens": 523277505.0, + "step": 20221 + }, + { + "epoch": 2.2207335822534593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5265347957611084, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7023496627807617, + "num_tokens": 523304940.0, + "step": 20222 + }, + { + "epoch": 2.220843399956073, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.470459461212158, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6974767446517944, + "num_tokens": 523332632.0, + "step": 20223 + }, + { + "epoch": 2.2209532176586864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.54531192779541, + "learning_rate": 1e-06, + "loss": 1.103, + "mean_token_accuracy": 0.6827049851417542, + "num_tokens": 523361444.0, + "step": 20224 + }, + { + "epoch": 2.2210630353613, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.867051362991333, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7111301422119141, + "num_tokens": 523383465.0, + "step": 20225 + }, + { + "epoch": 2.221172853063914, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.785181999206543, + "learning_rate": 1e-06, + "loss": 0.9141, + "mean_token_accuracy": 0.7285359501838684, + "num_tokens": 523406351.0, + "step": 20226 + }, + { + "epoch": 2.2212826707665276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.618757486343384, + "learning_rate": 1e-06, + "loss": 0.9279, + "mean_token_accuracy": 0.7262256741523743, + "num_tokens": 523432018.0, + "step": 20227 + }, + { + "epoch": 2.2213924884691414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6021275520324707, + "learning_rate": 1e-06, + "loss": 1.0716, + "mean_token_accuracy": 0.692552924156189, + "num_tokens": 523463555.0, + "step": 20228 + }, + { + "epoch": 2.2215023061717547, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0133144855499268, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7136671543121338, + "num_tokens": 523485185.0, + "step": 20229 + }, + { + "epoch": 2.2216121238743685, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.747185468673706, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6874457597732544, + "num_tokens": 523509089.0, + "step": 20230 + }, + { + "epoch": 2.221721941576982, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7004055976867676, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7123613357543945, + "num_tokens": 523534084.0, + "step": 20231 + }, + { + "epoch": 2.221831759279596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.445701837539673, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6944596171379089, + "num_tokens": 523564309.0, + "step": 20232 + }, + { + "epoch": 2.2219415769822097, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.539031744003296, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7220577597618103, + "num_tokens": 523588865.0, + "step": 20233 + }, + { + "epoch": 2.222051394684823, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4322245121002197, + "learning_rate": 1e-06, + "loss": 1.0774, + "mean_token_accuracy": 0.6862386465072632, + "num_tokens": 523620490.0, + "step": 20234 + }, + { + "epoch": 2.222161212387437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7801294326782227, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.720024824142456, + "num_tokens": 523645323.0, + "step": 20235 + }, + { + "epoch": 2.2222710300900506, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7488772869110107, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6970573663711548, + "num_tokens": 523668337.0, + "step": 20236 + }, + { + "epoch": 2.2223808477926643, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.51507568359375, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.715896487236023, + "num_tokens": 523692874.0, + "step": 20237 + }, + { + "epoch": 2.222490665495278, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7161972522735596, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7291257977485657, + "num_tokens": 523714459.0, + "step": 20238 + }, + { + "epoch": 2.2226004831978914, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7596261501312256, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7148429751396179, + "num_tokens": 523736494.0, + "step": 20239 + }, + { + "epoch": 2.222710300900505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6059751510620117, + "learning_rate": 1e-06, + "loss": 1.0845, + "mean_token_accuracy": 0.6971315741539001, + "num_tokens": 523766414.0, + "step": 20240 + }, + { + "epoch": 2.222820118603119, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7104601860046387, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7001326084136963, + "num_tokens": 523794278.0, + "step": 20241 + }, + { + "epoch": 2.2229299363057327, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.266916513442993, + "learning_rate": 1e-06, + "loss": 0.9314, + "mean_token_accuracy": 0.7261812686920166, + "num_tokens": 523829092.0, + "step": 20242 + }, + { + "epoch": 2.223039754008346, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2131783962249756, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7193940877914429, + "num_tokens": 523847786.0, + "step": 20243 + }, + { + "epoch": 2.2231495717109597, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1421642303466797, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6951518058776855, + "num_tokens": 523873224.0, + "step": 20244 + }, + { + "epoch": 2.2232593894135735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.756337881088257, + "learning_rate": 1e-06, + "loss": 0.9574, + "mean_token_accuracy": 0.7210471630096436, + "num_tokens": 523899560.0, + "step": 20245 + }, + { + "epoch": 2.2233692071161872, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4819560050964355, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7028400897979736, + "num_tokens": 523926135.0, + "step": 20246 + }, + { + "epoch": 2.223479024818801, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.901824951171875, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7168165445327759, + "num_tokens": 523947372.0, + "step": 20247 + }, + { + "epoch": 2.2235888425214143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6034064292907715, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7049091458320618, + "num_tokens": 523972720.0, + "step": 20248 + }, + { + "epoch": 2.223698660224028, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1107754707336426, + "learning_rate": 1e-06, + "loss": 0.8749, + "mean_token_accuracy": 0.7398402690887451, + "num_tokens": 523990408.0, + "step": 20249 + }, + { + "epoch": 2.223808477926642, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.497188091278076, + "learning_rate": 1e-06, + "loss": 0.9606, + "mean_token_accuracy": 0.7146056294441223, + "num_tokens": 524018298.0, + "step": 20250 + }, + { + "epoch": 2.2239182956292556, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5025172233581543, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7027609348297119, + "num_tokens": 524047897.0, + "step": 20251 + }, + { + "epoch": 2.224028113331869, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7871198654174805, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6870074272155762, + "num_tokens": 524072670.0, + "step": 20252 + }, + { + "epoch": 2.2241379310344827, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.793419599533081, + "learning_rate": 1e-06, + "loss": 0.9299, + "mean_token_accuracy": 0.7282342314720154, + "num_tokens": 524094237.0, + "step": 20253 + }, + { + "epoch": 2.2242477487370964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8793084621429443, + "learning_rate": 1e-06, + "loss": 1.098, + "mean_token_accuracy": 0.6900449991226196, + "num_tokens": 524119778.0, + "step": 20254 + }, + { + "epoch": 2.22435756643971, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5044164657592773, + "learning_rate": 1e-06, + "loss": 0.9991, + "mean_token_accuracy": 0.7083182334899902, + "num_tokens": 524147937.0, + "step": 20255 + }, + { + "epoch": 2.224467384142324, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.260915756225586, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7105709314346313, + "num_tokens": 524182165.0, + "step": 20256 + }, + { + "epoch": 2.2245772018449372, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9568092823028564, + "learning_rate": 1e-06, + "loss": 0.917, + "mean_token_accuracy": 0.731641411781311, + "num_tokens": 524201340.0, + "step": 20257 + }, + { + "epoch": 2.224687019547551, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6072561740875244, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7071096301078796, + "num_tokens": 524226267.0, + "step": 20258 + }, + { + "epoch": 2.2247968372501647, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7029759883880615, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7209890484809875, + "num_tokens": 524249983.0, + "step": 20259 + }, + { + "epoch": 2.2249066549527785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5293357372283936, + "learning_rate": 1e-06, + "loss": 1.0609, + "mean_token_accuracy": 0.6872947216033936, + "num_tokens": 524277426.0, + "step": 20260 + }, + { + "epoch": 2.2250164726553923, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6158571243286133, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6964680552482605, + "num_tokens": 524302820.0, + "step": 20261 + }, + { + "epoch": 2.2251262903580056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5728583335876465, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7084152698516846, + "num_tokens": 524331358.0, + "step": 20262 + }, + { + "epoch": 2.2252361080606193, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.469688892364502, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7035763263702393, + "num_tokens": 524358418.0, + "step": 20263 + }, + { + "epoch": 2.225345925763233, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8207991123199463, + "learning_rate": 1e-06, + "loss": 0.9062, + "mean_token_accuracy": 0.740206778049469, + "num_tokens": 524379393.0, + "step": 20264 + }, + { + "epoch": 2.225455743465847, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4881858825683594, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.721336305141449, + "num_tokens": 524406388.0, + "step": 20265 + }, + { + "epoch": 2.22556556116846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4660587310791016, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.691803514957428, + "num_tokens": 524434241.0, + "step": 20266 + }, + { + "epoch": 2.225675378871074, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.440436363220215, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6975700855255127, + "num_tokens": 524461854.0, + "step": 20267 + }, + { + "epoch": 2.2257851965736877, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.337223529815674, + "learning_rate": 1e-06, + "loss": 1.1379, + "mean_token_accuracy": 0.6742950081825256, + "num_tokens": 524496859.0, + "step": 20268 + }, + { + "epoch": 2.2258950142763014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.642000913619995, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.693794310092926, + "num_tokens": 524523319.0, + "step": 20269 + }, + { + "epoch": 2.226004831978915, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.815032720565796, + "learning_rate": 1e-06, + "loss": 0.9245, + "mean_token_accuracy": 0.7308844327926636, + "num_tokens": 524543772.0, + "step": 20270 + }, + { + "epoch": 2.2261146496815285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.867379903793335, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6940534114837646, + "num_tokens": 524570526.0, + "step": 20271 + }, + { + "epoch": 2.2262244673841423, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9861392974853516, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7177586555480957, + "num_tokens": 524591893.0, + "step": 20272 + }, + { + "epoch": 2.226334285086756, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5936248302459717, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.711129367351532, + "num_tokens": 524617959.0, + "step": 20273 + }, + { + "epoch": 2.2264441027893698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4927306175231934, + "learning_rate": 1e-06, + "loss": 0.9537, + "mean_token_accuracy": 0.7180591821670532, + "num_tokens": 524650018.0, + "step": 20274 + }, + { + "epoch": 2.226553920491983, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.594594955444336, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7048735618591309, + "num_tokens": 524677642.0, + "step": 20275 + }, + { + "epoch": 2.226663738194597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.798866033554077, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7069187164306641, + "num_tokens": 524700131.0, + "step": 20276 + }, + { + "epoch": 2.2267735558972106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7426741123199463, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7263904809951782, + "num_tokens": 524723625.0, + "step": 20277 + }, + { + "epoch": 2.2268833735998244, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5508811473846436, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7014027237892151, + "num_tokens": 524749654.0, + "step": 20278 + }, + { + "epoch": 2.226993191302438, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.77349853515625, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7211989760398865, + "num_tokens": 524770999.0, + "step": 20279 + }, + { + "epoch": 2.2271030090050514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4910728931427, + "learning_rate": 1e-06, + "loss": 0.9307, + "mean_token_accuracy": 0.7242051959037781, + "num_tokens": 524798476.0, + "step": 20280 + }, + { + "epoch": 2.227212826707665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3745155334472656, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6980871558189392, + "num_tokens": 524830257.0, + "step": 20281 + }, + { + "epoch": 2.227322644410279, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3349437713623047, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7149832248687744, + "num_tokens": 524860125.0, + "step": 20282 + }, + { + "epoch": 2.2274324621128927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.485051155090332, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7118308544158936, + "num_tokens": 524886906.0, + "step": 20283 + }, + { + "epoch": 2.2275422798155065, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5323944091796875, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7085049748420715, + "num_tokens": 524913202.0, + "step": 20284 + }, + { + "epoch": 2.2276520975181198, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.927795886993408, + "learning_rate": 1e-06, + "loss": 0.9411, + "mean_token_accuracy": 0.730936586856842, + "num_tokens": 524934388.0, + "step": 20285 + }, + { + "epoch": 2.2277619152207335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9171249866485596, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7178997993469238, + "num_tokens": 524954701.0, + "step": 20286 + }, + { + "epoch": 2.2278717329233473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.700044631958008, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.7221031188964844, + "num_tokens": 524982373.0, + "step": 20287 + }, + { + "epoch": 2.227981550625961, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8643832206726074, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7206780314445496, + "num_tokens": 525006604.0, + "step": 20288 + }, + { + "epoch": 2.228091368328575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8292744159698486, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.725521445274353, + "num_tokens": 525030729.0, + "step": 20289 + }, + { + "epoch": 2.228201186031188, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.715567111968994, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.6985145807266235, + "num_tokens": 525057137.0, + "step": 20290 + }, + { + "epoch": 2.228311003733802, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.596501350402832, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7056477069854736, + "num_tokens": 525083368.0, + "step": 20291 + }, + { + "epoch": 2.2284208214364156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6570122241973877, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7043960094451904, + "num_tokens": 525106564.0, + "step": 20292 + }, + { + "epoch": 2.2285306391390294, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5207107067108154, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7127726674079895, + "num_tokens": 525131346.0, + "step": 20293 + }, + { + "epoch": 2.2286404568416427, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.701340913772583, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7043508291244507, + "num_tokens": 525153578.0, + "step": 20294 + }, + { + "epoch": 2.2287502745442564, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.467040777206421, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7145304679870605, + "num_tokens": 525180802.0, + "step": 20295 + }, + { + "epoch": 2.22886009224687, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5620405673980713, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.6915731430053711, + "num_tokens": 525207531.0, + "step": 20296 + }, + { + "epoch": 2.228969909949484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5507712364196777, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.71895432472229, + "num_tokens": 525235532.0, + "step": 20297 + }, + { + "epoch": 2.2290797276520977, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7605576515197754, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.6885532140731812, + "num_tokens": 525260748.0, + "step": 20298 + }, + { + "epoch": 2.229189545354711, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.485529899597168, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.7054063081741333, + "num_tokens": 525290637.0, + "step": 20299 + }, + { + "epoch": 2.229299363057325, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9941389560699463, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7047094106674194, + "num_tokens": 525311476.0, + "step": 20300 + }, + { + "epoch": 2.2294091807599385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6665256023406982, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7167124152183533, + "num_tokens": 525336307.0, + "step": 20301 + }, + { + "epoch": 2.2295189984625523, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.745661735534668, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7091790437698364, + "num_tokens": 525360801.0, + "step": 20302 + }, + { + "epoch": 2.2296288161651656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2358670234680176, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7056492567062378, + "num_tokens": 525381272.0, + "step": 20303 + }, + { + "epoch": 2.2297386338677794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6883790493011475, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6893479824066162, + "num_tokens": 525405143.0, + "step": 20304 + }, + { + "epoch": 2.229848451570393, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5459628105163574, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7001072764396667, + "num_tokens": 525432417.0, + "step": 20305 + }, + { + "epoch": 2.229958269273007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.53606915473938, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.704820454120636, + "num_tokens": 525458740.0, + "step": 20306 + }, + { + "epoch": 2.2300680869756206, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7493650913238525, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7059318423271179, + "num_tokens": 525483970.0, + "step": 20307 + }, + { + "epoch": 2.230177904678234, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5755348205566406, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7114511728286743, + "num_tokens": 525511566.0, + "step": 20308 + }, + { + "epoch": 2.2302877223808477, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.675499439239502, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7286041975021362, + "num_tokens": 525534404.0, + "step": 20309 + }, + { + "epoch": 2.2303975400834615, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7197189331054688, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7053481340408325, + "num_tokens": 525559670.0, + "step": 20310 + }, + { + "epoch": 2.2305073577860752, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.075291395187378, + "learning_rate": 1e-06, + "loss": 0.947, + "mean_token_accuracy": 0.7213698625564575, + "num_tokens": 525582828.0, + "step": 20311 + }, + { + "epoch": 2.230617175488689, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8656599521636963, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7102314829826355, + "num_tokens": 525605437.0, + "step": 20312 + }, + { + "epoch": 2.2307269931913023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.808525562286377, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7274249196052551, + "num_tokens": 525628750.0, + "step": 20313 + }, + { + "epoch": 2.230836810893916, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6309986114501953, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.710394024848938, + "num_tokens": 525652681.0, + "step": 20314 + }, + { + "epoch": 2.23094662859653, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7000203132629395, + "learning_rate": 1e-06, + "loss": 0.9411, + "mean_token_accuracy": 0.7198227643966675, + "num_tokens": 525675705.0, + "step": 20315 + }, + { + "epoch": 2.2310564462991436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.483508348464966, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7151798009872437, + "num_tokens": 525702444.0, + "step": 20316 + }, + { + "epoch": 2.231166264001757, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.760352611541748, + "learning_rate": 1e-06, + "loss": 0.9566, + "mean_token_accuracy": 0.7255716323852539, + "num_tokens": 525725438.0, + "step": 20317 + }, + { + "epoch": 2.2312760817043706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4420628547668457, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7100197076797485, + "num_tokens": 525755652.0, + "step": 20318 + }, + { + "epoch": 2.2313858994069844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5713467597961426, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7160252332687378, + "num_tokens": 525781826.0, + "step": 20319 + }, + { + "epoch": 2.231495717109598, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4227874279022217, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.687635064125061, + "num_tokens": 525820208.0, + "step": 20320 + }, + { + "epoch": 2.231605534812212, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.656545400619507, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7187085151672363, + "num_tokens": 525841822.0, + "step": 20321 + }, + { + "epoch": 2.2317153525148252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.378012180328369, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.6821414232254028, + "num_tokens": 525873699.0, + "step": 20322 + }, + { + "epoch": 2.231825170217439, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9083337783813477, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6940538883209229, + "num_tokens": 525896504.0, + "step": 20323 + }, + { + "epoch": 2.2319349879200527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6237895488739014, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7268296480178833, + "num_tokens": 525923052.0, + "step": 20324 + }, + { + "epoch": 2.2320448056226665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.519268751144409, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.7000539302825928, + "num_tokens": 525950202.0, + "step": 20325 + }, + { + "epoch": 2.23215462332528, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3084371089935303, + "learning_rate": 1e-06, + "loss": 1.1361, + "mean_token_accuracy": 0.6813706159591675, + "num_tokens": 525983447.0, + "step": 20326 + }, + { + "epoch": 2.2322644410278936, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6355628967285156, + "learning_rate": 1e-06, + "loss": 1.0959, + "mean_token_accuracy": 0.6814525127410889, + "num_tokens": 526010645.0, + "step": 20327 + }, + { + "epoch": 2.2323742587305073, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8134355545043945, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.70102459192276, + "num_tokens": 526033189.0, + "step": 20328 + }, + { + "epoch": 2.232484076433121, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.820708990097046, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6996698379516602, + "num_tokens": 526057076.0, + "step": 20329 + }, + { + "epoch": 2.232593894135735, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5313658714294434, + "learning_rate": 1e-06, + "loss": 1.0925, + "mean_token_accuracy": 0.6871675848960876, + "num_tokens": 526088194.0, + "step": 20330 + }, + { + "epoch": 2.232703711838348, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.617769718170166, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6875482201576233, + "num_tokens": 526112889.0, + "step": 20331 + }, + { + "epoch": 2.232813529540962, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.424021005630493, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7115978598594666, + "num_tokens": 526140702.0, + "step": 20332 + }, + { + "epoch": 2.2329233472435757, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.330343246459961, + "learning_rate": 1e-06, + "loss": 1.1105, + "mean_token_accuracy": 0.6745290160179138, + "num_tokens": 526172674.0, + "step": 20333 + }, + { + "epoch": 2.2330331649461894, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.836359739303589, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7269428968429565, + "num_tokens": 526194778.0, + "step": 20334 + }, + { + "epoch": 2.233142982648803, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6868770122528076, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7058572173118591, + "num_tokens": 526221856.0, + "step": 20335 + }, + { + "epoch": 2.2332528003514165, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6319823265075684, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.7168316841125488, + "num_tokens": 526245796.0, + "step": 20336 + }, + { + "epoch": 2.2333626180540302, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559563159942627, + "learning_rate": 1e-06, + "loss": 0.9391, + "mean_token_accuracy": 0.7196947336196899, + "num_tokens": 526269914.0, + "step": 20337 + }, + { + "epoch": 2.233472435756644, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7288143634796143, + "learning_rate": 1e-06, + "loss": 0.9042, + "mean_token_accuracy": 0.740467369556427, + "num_tokens": 526292531.0, + "step": 20338 + }, + { + "epoch": 2.2335822534592578, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0316860675811768, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7175160050392151, + "num_tokens": 526311733.0, + "step": 20339 + }, + { + "epoch": 2.2336920711618715, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.537285089492798, + "learning_rate": 1e-06, + "loss": 0.9021, + "mean_token_accuracy": 0.7338019609451294, + "num_tokens": 526338515.0, + "step": 20340 + }, + { + "epoch": 2.233801888864485, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.49055814743042, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7275522351264954, + "num_tokens": 526371474.0, + "step": 20341 + }, + { + "epoch": 2.2339117065670986, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0645925998687744, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7163918614387512, + "num_tokens": 526390129.0, + "step": 20342 + }, + { + "epoch": 2.2340215242697123, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.521484375, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7158948183059692, + "num_tokens": 526417402.0, + "step": 20343 + }, + { + "epoch": 2.234131341972326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6397664546966553, + "learning_rate": 1e-06, + "loss": 0.9206, + "mean_token_accuracy": 0.7362412214279175, + "num_tokens": 526440645.0, + "step": 20344 + }, + { + "epoch": 2.2342411596749394, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.451436996459961, + "learning_rate": 1e-06, + "loss": 0.9231, + "mean_token_accuracy": 0.7363483309745789, + "num_tokens": 526466874.0, + "step": 20345 + }, + { + "epoch": 2.234350977377553, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.707745313644409, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.7079037427902222, + "num_tokens": 526490122.0, + "step": 20346 + }, + { + "epoch": 2.234460795080167, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3529398441314697, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6896811127662659, + "num_tokens": 526517745.0, + "step": 20347 + }, + { + "epoch": 2.2345706127827807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8083488941192627, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7238504886627197, + "num_tokens": 526542473.0, + "step": 20348 + }, + { + "epoch": 2.2346804304853944, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.677971839904785, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6811537742614746, + "num_tokens": 526567896.0, + "step": 20349 + }, + { + "epoch": 2.2347902481880078, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8113789558410645, + "learning_rate": 1e-06, + "loss": 0.9212, + "mean_token_accuracy": 0.727066159248352, + "num_tokens": 526588527.0, + "step": 20350 + }, + { + "epoch": 2.2349000658906215, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.779994487762451, + "learning_rate": 1e-06, + "loss": 0.9276, + "mean_token_accuracy": 0.7287989854812622, + "num_tokens": 526610904.0, + "step": 20351 + }, + { + "epoch": 2.2350098835932353, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4678611755371094, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.7136116623878479, + "num_tokens": 526638307.0, + "step": 20352 + }, + { + "epoch": 2.235119701295849, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8011159896850586, + "learning_rate": 1e-06, + "loss": 0.9143, + "mean_token_accuracy": 0.7251030206680298, + "num_tokens": 526657833.0, + "step": 20353 + }, + { + "epoch": 2.2352295189984623, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3907134532928467, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6942042112350464, + "num_tokens": 526688820.0, + "step": 20354 + }, + { + "epoch": 2.235339336701076, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.789621114730835, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7092285752296448, + "num_tokens": 526712067.0, + "step": 20355 + }, + { + "epoch": 2.23544915440369, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7992701530456543, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7167363166809082, + "num_tokens": 526734605.0, + "step": 20356 + }, + { + "epoch": 2.2355589721063036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5695836544036865, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6974159479141235, + "num_tokens": 526763488.0, + "step": 20357 + }, + { + "epoch": 2.2356687898089174, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.60837459564209, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7081819772720337, + "num_tokens": 526788452.0, + "step": 20358 + }, + { + "epoch": 2.2357786075115307, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.633678913116455, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7083367109298706, + "num_tokens": 526812108.0, + "step": 20359 + }, + { + "epoch": 2.2358884252141444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6617350578308105, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.716648280620575, + "num_tokens": 526836705.0, + "step": 20360 + }, + { + "epoch": 2.235998242916758, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6112332344055176, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7277675867080688, + "num_tokens": 526860268.0, + "step": 20361 + }, + { + "epoch": 2.236108060619372, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.493597984313965, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7082868814468384, + "num_tokens": 526888612.0, + "step": 20362 + }, + { + "epoch": 2.2362178783219857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3607230186462402, + "learning_rate": 1e-06, + "loss": 0.8928, + "mean_token_accuracy": 0.739337146282196, + "num_tokens": 526918780.0, + "step": 20363 + }, + { + "epoch": 2.236327696024599, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4940366744995117, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7208142876625061, + "num_tokens": 526947039.0, + "step": 20364 + }, + { + "epoch": 2.236437513727213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.709426164627075, + "learning_rate": 1e-06, + "loss": 0.8755, + "mean_token_accuracy": 0.7404533624649048, + "num_tokens": 526967088.0, + "step": 20365 + }, + { + "epoch": 2.2365473314298265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.823272228240967, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7074779272079468, + "num_tokens": 526988400.0, + "step": 20366 + }, + { + "epoch": 2.2366571491324403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7418336868286133, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.696663498878479, + "num_tokens": 527015572.0, + "step": 20367 + }, + { + "epoch": 2.2367669668350536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5558860301971436, + "learning_rate": 1e-06, + "loss": 0.9429, + "mean_token_accuracy": 0.7240228652954102, + "num_tokens": 527041790.0, + "step": 20368 + }, + { + "epoch": 2.2368767845376674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.794938564300537, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7150338888168335, + "num_tokens": 527066764.0, + "step": 20369 + }, + { + "epoch": 2.236986602240281, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7207632064819336, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7107947468757629, + "num_tokens": 527092131.0, + "step": 20370 + }, + { + "epoch": 2.237096419942895, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.812809944152832, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.6950305700302124, + "num_tokens": 527115857.0, + "step": 20371 + }, + { + "epoch": 2.2372062376455086, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6316006183624268, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6975781321525574, + "num_tokens": 527146876.0, + "step": 20372 + }, + { + "epoch": 2.237316055348122, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9032511711120605, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7200760841369629, + "num_tokens": 527167614.0, + "step": 20373 + }, + { + "epoch": 2.2374258730507357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8062491416931152, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.708971381187439, + "num_tokens": 527191952.0, + "step": 20374 + }, + { + "epoch": 2.2375356907533495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8485772609710693, + "learning_rate": 1e-06, + "loss": 0.8987, + "mean_token_accuracy": 0.7368134260177612, + "num_tokens": 527211500.0, + "step": 20375 + }, + { + "epoch": 2.237645508455963, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.583993911743164, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7189860343933105, + "num_tokens": 527236686.0, + "step": 20376 + }, + { + "epoch": 2.2377553261585765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.820094347000122, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.719598650932312, + "num_tokens": 527260395.0, + "step": 20377 + }, + { + "epoch": 2.2378651438611903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8086447715759277, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7070952653884888, + "num_tokens": 527282151.0, + "step": 20378 + }, + { + "epoch": 2.237974961563804, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.756178140640259, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7053492665290833, + "num_tokens": 527306096.0, + "step": 20379 + }, + { + "epoch": 2.238084779266418, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.141200065612793, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7088937759399414, + "num_tokens": 527325235.0, + "step": 20380 + }, + { + "epoch": 2.2381945969690316, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7119462490081787, + "learning_rate": 1e-06, + "loss": 0.9218, + "mean_token_accuracy": 0.7281110882759094, + "num_tokens": 527347389.0, + "step": 20381 + }, + { + "epoch": 2.238304414671645, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7192559242248535, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6907499432563782, + "num_tokens": 527373511.0, + "step": 20382 + }, + { + "epoch": 2.2384142323742586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4529662132263184, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7083860039710999, + "num_tokens": 527402166.0, + "step": 20383 + }, + { + "epoch": 2.2385240500768724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8643903732299805, + "learning_rate": 1e-06, + "loss": 0.979, + "mean_token_accuracy": 0.7148283123970032, + "num_tokens": 527422959.0, + "step": 20384 + }, + { + "epoch": 2.238633867779486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6518208980560303, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.7005375027656555, + "num_tokens": 527450587.0, + "step": 20385 + }, + { + "epoch": 2.2387436854821, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5105648040771484, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7218402624130249, + "num_tokens": 527476126.0, + "step": 20386 + }, + { + "epoch": 2.238853503184713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8732497692108154, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7088149785995483, + "num_tokens": 527496509.0, + "step": 20387 + }, + { + "epoch": 2.238963320887327, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7790510654449463, + "learning_rate": 1e-06, + "loss": 0.9666, + "mean_token_accuracy": 0.7191424369812012, + "num_tokens": 527518868.0, + "step": 20388 + }, + { + "epoch": 2.2390731385899407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6325695514678955, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6904279589653015, + "num_tokens": 527547509.0, + "step": 20389 + }, + { + "epoch": 2.2391829562925545, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4266695976257324, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7064846754074097, + "num_tokens": 527579173.0, + "step": 20390 + }, + { + "epoch": 2.2392927739951682, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0801808834075928, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7340742945671082, + "num_tokens": 527599196.0, + "step": 20391 + }, + { + "epoch": 2.2394025916977816, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4037868976593018, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.7162680625915527, + "num_tokens": 527629669.0, + "step": 20392 + }, + { + "epoch": 2.2395124094003953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.397249221801758, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.7097364664077759, + "num_tokens": 527660922.0, + "step": 20393 + }, + { + "epoch": 2.239622227103009, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.119072675704956, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7236471772193909, + "num_tokens": 527680675.0, + "step": 20394 + }, + { + "epoch": 2.239732044805623, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.629671573638916, + "learning_rate": 1e-06, + "loss": 0.9223, + "mean_token_accuracy": 0.7205089330673218, + "num_tokens": 527703669.0, + "step": 20395 + }, + { + "epoch": 2.239841862508236, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.512059450149536, + "learning_rate": 1e-06, + "loss": 1.1208, + "mean_token_accuracy": 0.6802457571029663, + "num_tokens": 527735082.0, + "step": 20396 + }, + { + "epoch": 2.23995168021085, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.611316680908203, + "learning_rate": 1e-06, + "loss": 0.9253, + "mean_token_accuracy": 0.7235620021820068, + "num_tokens": 527759437.0, + "step": 20397 + }, + { + "epoch": 2.2400614979134637, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4542503356933594, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6842290163040161, + "num_tokens": 527789906.0, + "step": 20398 + }, + { + "epoch": 2.2401713156160774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.666687250137329, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.721981942653656, + "num_tokens": 527812118.0, + "step": 20399 + }, + { + "epoch": 2.240281133318691, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.733704090118408, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6999332904815674, + "num_tokens": 527837353.0, + "step": 20400 + }, + { + "epoch": 2.2403909510213045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2826249599456787, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6946310997009277, + "num_tokens": 527868683.0, + "step": 20401 + }, + { + "epoch": 2.2405007687239182, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6185848712921143, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7205483913421631, + "num_tokens": 527892039.0, + "step": 20402 + }, + { + "epoch": 2.240610586426532, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5320894718170166, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7022290825843811, + "num_tokens": 527918095.0, + "step": 20403 + }, + { + "epoch": 2.2407204041291457, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.87496280670166, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.6960120797157288, + "num_tokens": 527938217.0, + "step": 20404 + }, + { + "epoch": 2.240830221831759, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5182087421417236, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7021971940994263, + "num_tokens": 527965025.0, + "step": 20405 + }, + { + "epoch": 2.240940039534373, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7131354808807373, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6899933815002441, + "num_tokens": 527989612.0, + "step": 20406 + }, + { + "epoch": 2.2410498572369866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9594216346740723, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7336708307266235, + "num_tokens": 528009851.0, + "step": 20407 + }, + { + "epoch": 2.2411596749396003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4624547958374023, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6924793124198914, + "num_tokens": 528043087.0, + "step": 20408 + }, + { + "epoch": 2.241269492642214, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.610184669494629, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6887037754058838, + "num_tokens": 528072194.0, + "step": 20409 + }, + { + "epoch": 2.2413793103448274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6402206420898438, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.692285418510437, + "num_tokens": 528098834.0, + "step": 20410 + }, + { + "epoch": 2.241489128047441, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.691464424133301, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7194010615348816, + "num_tokens": 528122187.0, + "step": 20411 + }, + { + "epoch": 2.241598945750055, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3317787647247314, + "learning_rate": 1e-06, + "loss": 1.1161, + "mean_token_accuracy": 0.6894366145133972, + "num_tokens": 528154044.0, + "step": 20412 + }, + { + "epoch": 2.2417087634526687, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.691904067993164, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7075573205947876, + "num_tokens": 528175938.0, + "step": 20413 + }, + { + "epoch": 2.2418185811552824, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5855979919433594, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6847466230392456, + "num_tokens": 528204258.0, + "step": 20414 + }, + { + "epoch": 2.2419283988578957, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.449885129928589, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.6994947791099548, + "num_tokens": 528237244.0, + "step": 20415 + }, + { + "epoch": 2.2420382165605095, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617696762084961, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6947042942047119, + "num_tokens": 528264249.0, + "step": 20416 + }, + { + "epoch": 2.2421480342631233, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.909632921218872, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7140663862228394, + "num_tokens": 528284839.0, + "step": 20417 + }, + { + "epoch": 2.242257851965737, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6853106021881104, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6933854222297668, + "num_tokens": 528310191.0, + "step": 20418 + }, + { + "epoch": 2.2423676696683508, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.510972261428833, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7072240114212036, + "num_tokens": 528338473.0, + "step": 20419 + }, + { + "epoch": 2.242477487370964, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.477937936782837, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7029151916503906, + "num_tokens": 528368942.0, + "step": 20420 + }, + { + "epoch": 2.242587305073578, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.819056749343872, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6947596073150635, + "num_tokens": 528394289.0, + "step": 20421 + }, + { + "epoch": 2.2426971227761916, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7541961669921875, + "learning_rate": 1e-06, + "loss": 1.1526, + "mean_token_accuracy": 0.6745280027389526, + "num_tokens": 528424723.0, + "step": 20422 + }, + { + "epoch": 2.2428069404788054, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.83065128326416, + "learning_rate": 1e-06, + "loss": 1.1002, + "mean_token_accuracy": 0.6831885576248169, + "num_tokens": 528451787.0, + "step": 20423 + }, + { + "epoch": 2.2429167581814187, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.558227300643921, + "learning_rate": 1e-06, + "loss": 1.0718, + "mean_token_accuracy": 0.6934434175491333, + "num_tokens": 528484557.0, + "step": 20424 + }, + { + "epoch": 2.2430265758840324, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.672795295715332, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6948051452636719, + "num_tokens": 528511132.0, + "step": 20425 + }, + { + "epoch": 2.243136393586646, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.48197340965271, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.7083616256713867, + "num_tokens": 528539163.0, + "step": 20426 + }, + { + "epoch": 2.24324621128926, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5187342166900635, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7164021730422974, + "num_tokens": 528567797.0, + "step": 20427 + }, + { + "epoch": 2.2433560289918737, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4840352535247803, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6930091977119446, + "num_tokens": 528595182.0, + "step": 20428 + }, + { + "epoch": 2.243465846694487, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6150994300842285, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.720999002456665, + "num_tokens": 528620815.0, + "step": 20429 + }, + { + "epoch": 2.2435756643971008, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3019907474517822, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6953871846199036, + "num_tokens": 528653451.0, + "step": 20430 + }, + { + "epoch": 2.2436854820997145, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3848726749420166, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6985287666320801, + "num_tokens": 528682954.0, + "step": 20431 + }, + { + "epoch": 2.2437952998023283, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.50162410736084, + "learning_rate": 1e-06, + "loss": 1.0862, + "mean_token_accuracy": 0.6906063556671143, + "num_tokens": 528713588.0, + "step": 20432 + }, + { + "epoch": 2.2439051175049416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.487755298614502, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7222158312797546, + "num_tokens": 528738256.0, + "step": 20433 + }, + { + "epoch": 2.2440149352075554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.804075241088867, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7108780145645142, + "num_tokens": 528764360.0, + "step": 20434 + }, + { + "epoch": 2.244124752910169, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.619131565093994, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7039381265640259, + "num_tokens": 528791968.0, + "step": 20435 + }, + { + "epoch": 2.244234570612783, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 8.831783294677734, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6921266317367554, + "num_tokens": 528818306.0, + "step": 20436 + }, + { + "epoch": 2.2443443883153966, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9492061138153076, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.71921706199646, + "num_tokens": 528844044.0, + "step": 20437 + }, + { + "epoch": 2.24445420601801, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7091152667999268, + "learning_rate": 1e-06, + "loss": 0.9177, + "mean_token_accuracy": 0.7302685379981995, + "num_tokens": 528869218.0, + "step": 20438 + }, + { + "epoch": 2.2445640237206237, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5544567108154297, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7197175025939941, + "num_tokens": 528895694.0, + "step": 20439 + }, + { + "epoch": 2.2446738414232374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6196110248565674, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6931519508361816, + "num_tokens": 528924203.0, + "step": 20440 + }, + { + "epoch": 2.244783659125851, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5983757972717285, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.708332896232605, + "num_tokens": 528948436.0, + "step": 20441 + }, + { + "epoch": 2.244893476828465, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.846932888031006, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.7233024835586548, + "num_tokens": 528969233.0, + "step": 20442 + }, + { + "epoch": 2.2450032945310783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.606105089187622, + "learning_rate": 1e-06, + "loss": 0.8915, + "mean_token_accuracy": 0.7367679476737976, + "num_tokens": 528993458.0, + "step": 20443 + }, + { + "epoch": 2.245113112233692, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.648608684539795, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7121872901916504, + "num_tokens": 529019352.0, + "step": 20444 + }, + { + "epoch": 2.245222929936306, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6434152126312256, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6991665363311768, + "num_tokens": 529043241.0, + "step": 20445 + }, + { + "epoch": 2.2453327476389195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.502549171447754, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.697698712348938, + "num_tokens": 529071459.0, + "step": 20446 + }, + { + "epoch": 2.245442565341533, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.567188024520874, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7062458992004395, + "num_tokens": 529097966.0, + "step": 20447 + }, + { + "epoch": 2.2455523830441466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.600277900695801, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6963078379631042, + "num_tokens": 529123413.0, + "step": 20448 + }, + { + "epoch": 2.2456622007467604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.504795551300049, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.6883348226547241, + "num_tokens": 529151037.0, + "step": 20449 + }, + { + "epoch": 2.245772018449374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7824316024780273, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.7096105813980103, + "num_tokens": 529173773.0, + "step": 20450 + }, + { + "epoch": 2.245881836151988, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0254945755004883, + "learning_rate": 1e-06, + "loss": 0.9185, + "mean_token_accuracy": 0.7269473075866699, + "num_tokens": 529195203.0, + "step": 20451 + }, + { + "epoch": 2.245991653854601, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7187564373016357, + "learning_rate": 1e-06, + "loss": 0.9115, + "mean_token_accuracy": 0.7250373363494873, + "num_tokens": 529218317.0, + "step": 20452 + }, + { + "epoch": 2.246101471557215, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6083569526672363, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.722480058670044, + "num_tokens": 529242615.0, + "step": 20453 + }, + { + "epoch": 2.2462112892598287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.555739164352417, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7073872089385986, + "num_tokens": 529268984.0, + "step": 20454 + }, + { + "epoch": 2.2463211069624425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.574404001235962, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7145454287528992, + "num_tokens": 529293638.0, + "step": 20455 + }, + { + "epoch": 2.246430924665056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6767265796661377, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7094316482543945, + "num_tokens": 529317290.0, + "step": 20456 + }, + { + "epoch": 2.2465407423676695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9829654693603516, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7234082221984863, + "num_tokens": 529336431.0, + "step": 20457 + }, + { + "epoch": 2.2466505600702833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3778269290924072, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6825143098831177, + "num_tokens": 529366266.0, + "step": 20458 + }, + { + "epoch": 2.246760377772897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.032529354095459, + "learning_rate": 1e-06, + "loss": 0.9273, + "mean_token_accuracy": 0.7261554002761841, + "num_tokens": 529383851.0, + "step": 20459 + }, + { + "epoch": 2.246870195475511, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6173930168151855, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7130678296089172, + "num_tokens": 529409177.0, + "step": 20460 + }, + { + "epoch": 2.246980013178124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4101319313049316, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7074849605560303, + "num_tokens": 529439888.0, + "step": 20461 + }, + { + "epoch": 2.247089830880738, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7242660522460938, + "learning_rate": 1e-06, + "loss": 0.9171, + "mean_token_accuracy": 0.7279435992240906, + "num_tokens": 529462820.0, + "step": 20462 + }, + { + "epoch": 2.2471996485833516, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7731516361236572, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7114818692207336, + "num_tokens": 529486913.0, + "step": 20463 + }, + { + "epoch": 2.2473094662859654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5568783283233643, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7170407176017761, + "num_tokens": 529513992.0, + "step": 20464 + }, + { + "epoch": 2.247419283988579, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.65986967086792, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7183059453964233, + "num_tokens": 529540114.0, + "step": 20465 + }, + { + "epoch": 2.2475291016911925, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.461538791656494, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7144585847854614, + "num_tokens": 529566298.0, + "step": 20466 + }, + { + "epoch": 2.2476389193938062, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.623138666152954, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7060348391532898, + "num_tokens": 529591700.0, + "step": 20467 + }, + { + "epoch": 2.24774873709642, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9751739501953125, + "learning_rate": 1e-06, + "loss": 0.8526, + "mean_token_accuracy": 0.7485093474388123, + "num_tokens": 529612780.0, + "step": 20468 + }, + { + "epoch": 2.2478585547990337, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4794156551361084, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.695531964302063, + "num_tokens": 529641782.0, + "step": 20469 + }, + { + "epoch": 2.2479683725016475, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561230421066284, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7093693614006042, + "num_tokens": 529668437.0, + "step": 20470 + }, + { + "epoch": 2.248078190204261, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.687260389328003, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.6941934823989868, + "num_tokens": 529692229.0, + "step": 20471 + }, + { + "epoch": 2.2481880079068746, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.279210090637207, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7123998403549194, + "num_tokens": 529722955.0, + "step": 20472 + }, + { + "epoch": 2.2482978256094883, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.512986660003662, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7149598002433777, + "num_tokens": 529750229.0, + "step": 20473 + }, + { + "epoch": 2.248407643312102, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.489203691482544, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7006436586380005, + "num_tokens": 529777962.0, + "step": 20474 + }, + { + "epoch": 2.2485174610147154, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4903042316436768, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7096741199493408, + "num_tokens": 529805968.0, + "step": 20475 + }, + { + "epoch": 2.248627278717329, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7136147022247314, + "learning_rate": 1e-06, + "loss": 0.9271, + "mean_token_accuracy": 0.7299504280090332, + "num_tokens": 529828306.0, + "step": 20476 + }, + { + "epoch": 2.248737096419943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.540137767791748, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7009856104850769, + "num_tokens": 529855477.0, + "step": 20477 + }, + { + "epoch": 2.2488469141225567, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.759744882583618, + "learning_rate": 1e-06, + "loss": 0.9321, + "mean_token_accuracy": 0.7268332242965698, + "num_tokens": 529879914.0, + "step": 20478 + }, + { + "epoch": 2.2489567318251704, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6142663955688477, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7184597253799438, + "num_tokens": 529902911.0, + "step": 20479 + }, + { + "epoch": 2.2490665495277837, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.360941171646118, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6847403645515442, + "num_tokens": 529936411.0, + "step": 20480 + }, + { + "epoch": 2.2491763672303975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7044057846069336, + "learning_rate": 1e-06, + "loss": 0.9628, + "mean_token_accuracy": 0.7100906372070312, + "num_tokens": 529957626.0, + "step": 20481 + }, + { + "epoch": 2.2492861849330112, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.805682420730591, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7031570672988892, + "num_tokens": 529981233.0, + "step": 20482 + }, + { + "epoch": 2.249396002635625, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5609846115112305, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7107532024383545, + "num_tokens": 530008794.0, + "step": 20483 + }, + { + "epoch": 2.2495058203382383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5552010536193848, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6999145746231079, + "num_tokens": 530036435.0, + "step": 20484 + }, + { + "epoch": 2.249615638040852, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.625148296356201, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7130661010742188, + "num_tokens": 530060470.0, + "step": 20485 + }, + { + "epoch": 2.249725455743466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.435692548751831, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.7005485892295837, + "num_tokens": 530090529.0, + "step": 20486 + }, + { + "epoch": 2.2498352734460796, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.527395248413086, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7123929262161255, + "num_tokens": 530117324.0, + "step": 20487 + }, + { + "epoch": 2.2499450911486933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6529223918914795, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7171637415885925, + "num_tokens": 530139881.0, + "step": 20488 + }, + { + "epoch": 2.2500549088513067, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8099758625030518, + "learning_rate": 1e-06, + "loss": 0.8584, + "mean_token_accuracy": 0.7383596897125244, + "num_tokens": 530159955.0, + "step": 20489 + }, + { + "epoch": 2.2501647265539204, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7546324729919434, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7038525938987732, + "num_tokens": 530183815.0, + "step": 20490 + }, + { + "epoch": 2.250274544256534, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6033663749694824, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7090374827384949, + "num_tokens": 530209864.0, + "step": 20491 + }, + { + "epoch": 2.250384361959148, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6128475666046143, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7218062877655029, + "num_tokens": 530236326.0, + "step": 20492 + }, + { + "epoch": 2.2504941796617617, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.821211576461792, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7090906500816345, + "num_tokens": 530256498.0, + "step": 20493 + }, + { + "epoch": 2.250603997364375, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.678131341934204, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7080360651016235, + "num_tokens": 530280928.0, + "step": 20494 + }, + { + "epoch": 2.2507138150669888, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8381497859954834, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7094706296920776, + "num_tokens": 530301881.0, + "step": 20495 + }, + { + "epoch": 2.2508236327696025, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6042468547821045, + "learning_rate": 1e-06, + "loss": 1.1075, + "mean_token_accuracy": 0.6813589334487915, + "num_tokens": 530333726.0, + "step": 20496 + }, + { + "epoch": 2.2509334504722163, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536829710006714, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.722522497177124, + "num_tokens": 530360414.0, + "step": 20497 + }, + { + "epoch": 2.25104326817483, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.469027042388916, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6906452775001526, + "num_tokens": 530387789.0, + "step": 20498 + }, + { + "epoch": 2.2511530858774433, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.192662239074707, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7069966793060303, + "num_tokens": 530410661.0, + "step": 20499 + }, + { + "epoch": 2.251262903580057, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.56008243560791, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.704120397567749, + "num_tokens": 530438675.0, + "step": 20500 + }, + { + "epoch": 2.251372721282671, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4945290088653564, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7085947394371033, + "num_tokens": 530466771.0, + "step": 20501 + }, + { + "epoch": 2.2514825389852846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.639441967010498, + "learning_rate": 1e-06, + "loss": 0.8681, + "mean_token_accuracy": 0.7406495213508606, + "num_tokens": 530489353.0, + "step": 20502 + }, + { + "epoch": 2.251592356687898, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8082919120788574, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7034895420074463, + "num_tokens": 530514382.0, + "step": 20503 + }, + { + "epoch": 2.2517021743905117, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6087822914123535, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.69952791929245, + "num_tokens": 530539874.0, + "step": 20504 + }, + { + "epoch": 2.2518119920931254, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6798019409179688, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7130061388015747, + "num_tokens": 530564468.0, + "step": 20505 + }, + { + "epoch": 2.251921809795739, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.410914897918701, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.723468005657196, + "num_tokens": 530592630.0, + "step": 20506 + }, + { + "epoch": 2.2520316274983525, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.511014938354492, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6985454559326172, + "num_tokens": 530621981.0, + "step": 20507 + }, + { + "epoch": 2.2521414452009663, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.753571033477783, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7135809659957886, + "num_tokens": 530646013.0, + "step": 20508 + }, + { + "epoch": 2.25225126290358, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.729128837585449, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7232522368431091, + "num_tokens": 530669801.0, + "step": 20509 + }, + { + "epoch": 2.2523610806061938, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4787535667419434, + "learning_rate": 1e-06, + "loss": 0.8932, + "mean_token_accuracy": 0.7446408271789551, + "num_tokens": 530695794.0, + "step": 20510 + }, + { + "epoch": 2.2524708983088075, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5860652923583984, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7046593427658081, + "num_tokens": 530725142.0, + "step": 20511 + }, + { + "epoch": 2.252580716011421, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3940773010253906, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7188683748245239, + "num_tokens": 530754745.0, + "step": 20512 + }, + { + "epoch": 2.2526905337140346, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7268171310424805, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7079085111618042, + "num_tokens": 530778615.0, + "step": 20513 + }, + { + "epoch": 2.2528003514166484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4591257572174072, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7174904346466064, + "num_tokens": 530806303.0, + "step": 20514 + }, + { + "epoch": 2.252910169119262, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.295692205429077, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.709388017654419, + "num_tokens": 530836596.0, + "step": 20515 + }, + { + "epoch": 2.253019986821876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6223807334899902, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.705817699432373, + "num_tokens": 530862211.0, + "step": 20516 + }, + { + "epoch": 2.253129804524489, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4387474060058594, + "learning_rate": 1e-06, + "loss": 1.0958, + "mean_token_accuracy": 0.6865798234939575, + "num_tokens": 530892342.0, + "step": 20517 + }, + { + "epoch": 2.253239622227103, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.374225378036499, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6803109049797058, + "num_tokens": 530924070.0, + "step": 20518 + }, + { + "epoch": 2.2533494399297167, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5746662616729736, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6837829351425171, + "num_tokens": 530953116.0, + "step": 20519 + }, + { + "epoch": 2.2534592576323305, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.362774133682251, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7095140218734741, + "num_tokens": 530981887.0, + "step": 20520 + }, + { + "epoch": 2.253569075334944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.517946481704712, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.710130512714386, + "num_tokens": 531009953.0, + "step": 20521 + }, + { + "epoch": 2.2536788930375575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.32827091217041, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.70814049243927, + "num_tokens": 531042200.0, + "step": 20522 + }, + { + "epoch": 2.2537887107401713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.400552272796631, + "learning_rate": 1e-06, + "loss": 0.9231, + "mean_token_accuracy": 0.7297953367233276, + "num_tokens": 531071165.0, + "step": 20523 + }, + { + "epoch": 2.253898528442785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.655756711959839, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7205649614334106, + "num_tokens": 531095433.0, + "step": 20524 + }, + { + "epoch": 2.254008346145399, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6322503089904785, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7060694098472595, + "num_tokens": 531120481.0, + "step": 20525 + }, + { + "epoch": 2.254118163848012, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.844726085662842, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7196572422981262, + "num_tokens": 531143677.0, + "step": 20526 + }, + { + "epoch": 2.254227981550626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4483134746551514, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6953543424606323, + "num_tokens": 531173405.0, + "step": 20527 + }, + { + "epoch": 2.2543377992532396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6036548614501953, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6840816140174866, + "num_tokens": 531200292.0, + "step": 20528 + }, + { + "epoch": 2.2544476169558534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8044450283050537, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7116945385932922, + "num_tokens": 531223427.0, + "step": 20529 + }, + { + "epoch": 2.2545574346584667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.618295192718506, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6954474449157715, + "num_tokens": 531248597.0, + "step": 20530 + }, + { + "epoch": 2.2546672523610805, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.787529706954956, + "learning_rate": 1e-06, + "loss": 0.9403, + "mean_token_accuracy": 0.7210094928741455, + "num_tokens": 531269512.0, + "step": 20531 + }, + { + "epoch": 2.254777070063694, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5523693561553955, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6969622373580933, + "num_tokens": 531298178.0, + "step": 20532 + }, + { + "epoch": 2.254886887766308, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6007394790649414, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7166907787322998, + "num_tokens": 531325600.0, + "step": 20533 + }, + { + "epoch": 2.2549967054689217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6848907470703125, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.7043324112892151, + "num_tokens": 531351729.0, + "step": 20534 + }, + { + "epoch": 2.255106523171535, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.772716999053955, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7119868397712708, + "num_tokens": 531376139.0, + "step": 20535 + }, + { + "epoch": 2.255216340874149, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6868247985839844, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7001107931137085, + "num_tokens": 531402112.0, + "step": 20536 + }, + { + "epoch": 2.2553261585767626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4119930267333984, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6998406052589417, + "num_tokens": 531431035.0, + "step": 20537 + }, + { + "epoch": 2.2554359762793763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4985342025756836, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.7033670544624329, + "num_tokens": 531459968.0, + "step": 20538 + }, + { + "epoch": 2.25554579398199, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7746458053588867, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6973017454147339, + "num_tokens": 531482442.0, + "step": 20539 + }, + { + "epoch": 2.2556556116846034, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.633793354034424, + "learning_rate": 1e-06, + "loss": 0.941, + "mean_token_accuracy": 0.721737265586853, + "num_tokens": 531506980.0, + "step": 20540 + }, + { + "epoch": 2.255765429387217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.522792100906372, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7140461206436157, + "num_tokens": 531536863.0, + "step": 20541 + }, + { + "epoch": 2.255875247089831, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5896687507629395, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7151314616203308, + "num_tokens": 531561662.0, + "step": 20542 + }, + { + "epoch": 2.2559850647924446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7187414169311523, + "learning_rate": 1e-06, + "loss": 0.9265, + "mean_token_accuracy": 0.7269601225852966, + "num_tokens": 531583732.0, + "step": 20543 + }, + { + "epoch": 2.2560948824950584, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.51094651222229, + "learning_rate": 1e-06, + "loss": 0.8886, + "mean_token_accuracy": 0.7368428111076355, + "num_tokens": 531611244.0, + "step": 20544 + }, + { + "epoch": 2.2562047001976717, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.97491717338562, + "learning_rate": 1e-06, + "loss": 0.8751, + "mean_token_accuracy": 0.7378718256950378, + "num_tokens": 531629118.0, + "step": 20545 + }, + { + "epoch": 2.2563145179002855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4404454231262207, + "learning_rate": 1e-06, + "loss": 1.1379, + "mean_token_accuracy": 0.673774003982544, + "num_tokens": 531660232.0, + "step": 20546 + }, + { + "epoch": 2.2564243356028992, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.4360108375549316, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7141236066818237, + "num_tokens": 531691246.0, + "step": 20547 + }, + { + "epoch": 2.256534153305513, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.344031810760498, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6875758171081543, + "num_tokens": 531727836.0, + "step": 20548 + }, + { + "epoch": 2.2566439710081267, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9504613876342773, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7198573350906372, + "num_tokens": 531748360.0, + "step": 20549 + }, + { + "epoch": 2.25675378871074, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.341608762741089, + "learning_rate": 1e-06, + "loss": 1.1114, + "mean_token_accuracy": 0.6728223562240601, + "num_tokens": 531782369.0, + "step": 20550 + }, + { + "epoch": 2.256863606413354, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0626883506774902, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.6943238377571106, + "num_tokens": 531803350.0, + "step": 20551 + }, + { + "epoch": 2.2569734241159676, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.684400796890259, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7050610184669495, + "num_tokens": 531830245.0, + "step": 20552 + }, + { + "epoch": 2.2570832418185813, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.453037738800049, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7206361889839172, + "num_tokens": 531857075.0, + "step": 20553 + }, + { + "epoch": 2.2571930595211946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.348421812057495, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6989383101463318, + "num_tokens": 531887055.0, + "step": 20554 + }, + { + "epoch": 2.2573028772238084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5294759273529053, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6993756294250488, + "num_tokens": 531914390.0, + "step": 20555 + }, + { + "epoch": 2.257412694926422, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.556899309158325, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7093727588653564, + "num_tokens": 531944376.0, + "step": 20556 + }, + { + "epoch": 2.257522512629036, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.512248992919922, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7102442383766174, + "num_tokens": 531970531.0, + "step": 20557 + }, + { + "epoch": 2.2576323303316492, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.57658052444458, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7172799706459045, + "num_tokens": 531997450.0, + "step": 20558 + }, + { + "epoch": 2.257742148034263, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6221256256103516, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7163639068603516, + "num_tokens": 532023100.0, + "step": 20559 + }, + { + "epoch": 2.2578519657368767, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5075180530548096, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7064880728721619, + "num_tokens": 532049239.0, + "step": 20560 + }, + { + "epoch": 2.2579617834394905, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2963075637817383, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7078736424446106, + "num_tokens": 532083272.0, + "step": 20561 + }, + { + "epoch": 2.2580716011421043, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5913896560668945, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.707042932510376, + "num_tokens": 532112550.0, + "step": 20562 + }, + { + "epoch": 2.2581814188447176, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3412129878997803, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7060253620147705, + "num_tokens": 532141991.0, + "step": 20563 + }, + { + "epoch": 2.2582912365473313, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4290833473205566, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7171470522880554, + "num_tokens": 532169669.0, + "step": 20564 + }, + { + "epoch": 2.258401054249945, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7513678073883057, + "learning_rate": 1e-06, + "loss": 0.7991, + "mean_token_accuracy": 0.7622663974761963, + "num_tokens": 532189423.0, + "step": 20565 + }, + { + "epoch": 2.258510871952559, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.697770118713379, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7088606953620911, + "num_tokens": 532213736.0, + "step": 20566 + }, + { + "epoch": 2.2586206896551726, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4108705520629883, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.685494065284729, + "num_tokens": 532242845.0, + "step": 20567 + }, + { + "epoch": 2.258730507357786, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.621290683746338, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7248983383178711, + "num_tokens": 532269830.0, + "step": 20568 + }, + { + "epoch": 2.2588403250603997, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.803832769393921, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7222217321395874, + "num_tokens": 532294985.0, + "step": 20569 + }, + { + "epoch": 2.2589501427630134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5127649307250977, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7273224592208862, + "num_tokens": 532322489.0, + "step": 20570 + }, + { + "epoch": 2.259059960465627, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.70585560798645, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7066532373428345, + "num_tokens": 532346418.0, + "step": 20571 + }, + { + "epoch": 2.259169778168241, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7061331272125244, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7275854349136353, + "num_tokens": 532368992.0, + "step": 20572 + }, + { + "epoch": 2.2592795958708543, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.694796562194824, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6956978440284729, + "num_tokens": 532394273.0, + "step": 20573 + }, + { + "epoch": 2.259389413573468, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7755327224731445, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7032889127731323, + "num_tokens": 532416996.0, + "step": 20574 + }, + { + "epoch": 2.2594992312760818, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7108707427978516, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7123602628707886, + "num_tokens": 532442497.0, + "step": 20575 + }, + { + "epoch": 2.2596090489786955, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.678318500518799, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7001287937164307, + "num_tokens": 532469051.0, + "step": 20576 + }, + { + "epoch": 2.259718866681309, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7357969284057617, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7072643041610718, + "num_tokens": 532495620.0, + "step": 20577 + }, + { + "epoch": 2.2598286843839226, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7135519981384277, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.7168837189674377, + "num_tokens": 532518777.0, + "step": 20578 + }, + { + "epoch": 2.2599385020865363, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6244583129882812, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7198828458786011, + "num_tokens": 532543082.0, + "step": 20579 + }, + { + "epoch": 2.26004831978915, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.510788917541504, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7035503387451172, + "num_tokens": 532571941.0, + "step": 20580 + }, + { + "epoch": 2.260158137491764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.593188762664795, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.713353157043457, + "num_tokens": 532596864.0, + "step": 20581 + }, + { + "epoch": 2.260267955194377, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5709917545318604, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6954169273376465, + "num_tokens": 532624412.0, + "step": 20582 + }, + { + "epoch": 2.260377772896991, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.844839572906494, + "learning_rate": 1e-06, + "loss": 0.8974, + "mean_token_accuracy": 0.7315608263015747, + "num_tokens": 532645362.0, + "step": 20583 + }, + { + "epoch": 2.2604875905996047, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8450045585632324, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7102213501930237, + "num_tokens": 532665259.0, + "step": 20584 + }, + { + "epoch": 2.2605974083022184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7845265865325928, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6968448162078857, + "num_tokens": 532688492.0, + "step": 20585 + }, + { + "epoch": 2.2607072260048318, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7363078594207764, + "learning_rate": 1e-06, + "loss": 0.977, + "mean_token_accuracy": 0.7184653878211975, + "num_tokens": 532712808.0, + "step": 20586 + }, + { + "epoch": 2.2608170437074455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.509807825088501, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7173056602478027, + "num_tokens": 532741960.0, + "step": 20587 + }, + { + "epoch": 2.2609268614100593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7302615642547607, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.70159912109375, + "num_tokens": 532775010.0, + "step": 20588 + }, + { + "epoch": 2.261036679112673, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.670215368270874, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7073535919189453, + "num_tokens": 532799610.0, + "step": 20589 + }, + { + "epoch": 2.261146496815287, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.743354082107544, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7149062156677246, + "num_tokens": 532822318.0, + "step": 20590 + }, + { + "epoch": 2.2612563145179, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.631091833114624, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7185418605804443, + "num_tokens": 532848644.0, + "step": 20591 + }, + { + "epoch": 2.261366132220514, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6514859199523926, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6991926431655884, + "num_tokens": 532874857.0, + "step": 20592 + }, + { + "epoch": 2.2614759499231276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.723299980163574, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7106645703315735, + "num_tokens": 532901140.0, + "step": 20593 + }, + { + "epoch": 2.2615857676257414, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.451678991317749, + "learning_rate": 1e-06, + "loss": 1.1363, + "mean_token_accuracy": 0.6752382516860962, + "num_tokens": 532933633.0, + "step": 20594 + }, + { + "epoch": 2.261695585328355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5892105102539062, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.6982371211051941, + "num_tokens": 532960356.0, + "step": 20595 + }, + { + "epoch": 2.2618054030309684, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.939805507659912, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.6979782581329346, + "num_tokens": 532982102.0, + "step": 20596 + }, + { + "epoch": 2.261915220733582, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.557917356491089, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.716397762298584, + "num_tokens": 533008657.0, + "step": 20597 + }, + { + "epoch": 2.262025038436196, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3849005699157715, + "learning_rate": 1e-06, + "loss": 1.1019, + "mean_token_accuracy": 0.6939024925231934, + "num_tokens": 533040050.0, + "step": 20598 + }, + { + "epoch": 2.2621348561388097, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.63874888420105, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.692158579826355, + "num_tokens": 533065654.0, + "step": 20599 + }, + { + "epoch": 2.2622446738414235, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6952195167541504, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.723583459854126, + "num_tokens": 533089382.0, + "step": 20600 + }, + { + "epoch": 2.262354491544037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5880014896392822, + "learning_rate": 1e-06, + "loss": 1.2154, + "mean_token_accuracy": 0.6757068037986755, + "num_tokens": 533119669.0, + "step": 20601 + }, + { + "epoch": 2.2624643092466505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.913059711456299, + "learning_rate": 1e-06, + "loss": 0.8372, + "mean_token_accuracy": 0.7542265057563782, + "num_tokens": 533138406.0, + "step": 20602 + }, + { + "epoch": 2.2625741269492643, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.394524574279785, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7164648175239563, + "num_tokens": 533165108.0, + "step": 20603 + }, + { + "epoch": 2.262683944651878, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3733463287353516, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7086306214332581, + "num_tokens": 533196382.0, + "step": 20604 + }, + { + "epoch": 2.2627937623544914, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.708782434463501, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7121182680130005, + "num_tokens": 533221981.0, + "step": 20605 + }, + { + "epoch": 2.262903580057105, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7340784072875977, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6915867328643799, + "num_tokens": 533248532.0, + "step": 20606 + }, + { + "epoch": 2.263013397759719, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4227986335754395, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.701598584651947, + "num_tokens": 533276848.0, + "step": 20607 + }, + { + "epoch": 2.2631232154623326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2685012817382812, + "learning_rate": 1e-06, + "loss": 1.0837, + "mean_token_accuracy": 0.6833568811416626, + "num_tokens": 533309991.0, + "step": 20608 + }, + { + "epoch": 2.263233033164946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5944371223449707, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7220151424407959, + "num_tokens": 533337717.0, + "step": 20609 + }, + { + "epoch": 2.2633428508675597, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.619860887527466, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7047362923622131, + "num_tokens": 533361471.0, + "step": 20610 + }, + { + "epoch": 2.2634526685701735, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4195287227630615, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7158024907112122, + "num_tokens": 533390149.0, + "step": 20611 + }, + { + "epoch": 2.263562486272787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6472110748291016, + "learning_rate": 1e-06, + "loss": 0.8767, + "mean_token_accuracy": 0.7403868436813354, + "num_tokens": 533412367.0, + "step": 20612 + }, + { + "epoch": 2.263672303975401, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.549232244491577, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.712677001953125, + "num_tokens": 533439118.0, + "step": 20613 + }, + { + "epoch": 2.2637821216780143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.72739577293396, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6942644715309143, + "num_tokens": 533465411.0, + "step": 20614 + }, + { + "epoch": 2.263891939380628, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.70463490486145, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7047420740127563, + "num_tokens": 533491447.0, + "step": 20615 + }, + { + "epoch": 2.264001757083242, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0001380443573, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.7212116718292236, + "num_tokens": 533513943.0, + "step": 20616 + }, + { + "epoch": 2.2641115747858556, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6197404861450195, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6903141736984253, + "num_tokens": 533541238.0, + "step": 20617 + }, + { + "epoch": 2.2642213924884693, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.526146411895752, + "learning_rate": 1e-06, + "loss": 0.9431, + "mean_token_accuracy": 0.7215731143951416, + "num_tokens": 533565011.0, + "step": 20618 + }, + { + "epoch": 2.2643312101910826, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.279158592224121, + "learning_rate": 1e-06, + "loss": 0.9598, + "mean_token_accuracy": 0.721390962600708, + "num_tokens": 533598006.0, + "step": 20619 + }, + { + "epoch": 2.2644410278936964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6906840801239014, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6880043745040894, + "num_tokens": 533623268.0, + "step": 20620 + }, + { + "epoch": 2.26455084559631, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3145864009857178, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7117637991905212, + "num_tokens": 533652105.0, + "step": 20621 + }, + { + "epoch": 2.264660663298924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0762078762054443, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7143586874008179, + "num_tokens": 533670980.0, + "step": 20622 + }, + { + "epoch": 2.2647704810015377, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.566225051879883, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7020184993743896, + "num_tokens": 533697461.0, + "step": 20623 + }, + { + "epoch": 2.264880298704151, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7775373458862305, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7064757347106934, + "num_tokens": 533723770.0, + "step": 20624 + }, + { + "epoch": 2.2649901164067647, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4633562564849854, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7182670831680298, + "num_tokens": 533752403.0, + "step": 20625 + }, + { + "epoch": 2.2650999341093785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.67352294921875, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7160003781318665, + "num_tokens": 533777072.0, + "step": 20626 + }, + { + "epoch": 2.2652097518119922, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.666882276535034, + "learning_rate": 1e-06, + "loss": 0.9338, + "mean_token_accuracy": 0.7201128602027893, + "num_tokens": 533800909.0, + "step": 20627 + }, + { + "epoch": 2.265319569514606, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6752331256866455, + "learning_rate": 1e-06, + "loss": 0.9007, + "mean_token_accuracy": 0.7330519556999207, + "num_tokens": 533825942.0, + "step": 20628 + }, + { + "epoch": 2.2654293872172193, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6412038803100586, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7234907150268555, + "num_tokens": 533851246.0, + "step": 20629 + }, + { + "epoch": 2.265539204919833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8873071670532227, + "learning_rate": 1e-06, + "loss": 0.9143, + "mean_token_accuracy": 0.7277921438217163, + "num_tokens": 533871288.0, + "step": 20630 + }, + { + "epoch": 2.265649022622447, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6043028831481934, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.6963893175125122, + "num_tokens": 533897718.0, + "step": 20631 + }, + { + "epoch": 2.2657588403250606, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.75004506111145, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7176617980003357, + "num_tokens": 533919924.0, + "step": 20632 + }, + { + "epoch": 2.265868658027674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5920073986053467, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7001897692680359, + "num_tokens": 533945249.0, + "step": 20633 + }, + { + "epoch": 2.2659784757302877, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4732494354248047, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6881498098373413, + "num_tokens": 533973516.0, + "step": 20634 + }, + { + "epoch": 2.2660882934329014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.616042137145996, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7244812846183777, + "num_tokens": 533997573.0, + "step": 20635 + }, + { + "epoch": 2.266198111135515, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8871405124664307, + "learning_rate": 1e-06, + "loss": 0.8812, + "mean_token_accuracy": 0.7367376089096069, + "num_tokens": 534016343.0, + "step": 20636 + }, + { + "epoch": 2.2663079288381285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5206048488616943, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7015882134437561, + "num_tokens": 534046554.0, + "step": 20637 + }, + { + "epoch": 2.2664177465407422, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7136096954345703, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7155406475067139, + "num_tokens": 534070506.0, + "step": 20638 + }, + { + "epoch": 2.266527564243356, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.583632707595825, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.6946675181388855, + "num_tokens": 534098458.0, + "step": 20639 + }, + { + "epoch": 2.2666373819459698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9245035648345947, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7256745100021362, + "num_tokens": 534119676.0, + "step": 20640 + }, + { + "epoch": 2.2667471996485835, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.621626853942871, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7299199104309082, + "num_tokens": 534147125.0, + "step": 20641 + }, + { + "epoch": 2.266857017351197, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6016228199005127, + "learning_rate": 1e-06, + "loss": 1.1338, + "mean_token_accuracy": 0.6735513806343079, + "num_tokens": 534175831.0, + "step": 20642 + }, + { + "epoch": 2.2669668350538106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6133015155792236, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.724517822265625, + "num_tokens": 534202981.0, + "step": 20643 + }, + { + "epoch": 2.2670766527564243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.588351249694824, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6952903270721436, + "num_tokens": 534232202.0, + "step": 20644 + }, + { + "epoch": 2.267186470459038, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.600508451461792, + "learning_rate": 1e-06, + "loss": 0.9283, + "mean_token_accuracy": 0.7276235818862915, + "num_tokens": 534256887.0, + "step": 20645 + }, + { + "epoch": 2.267296288161652, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4394891262054443, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7241035103797913, + "num_tokens": 534283370.0, + "step": 20646 + }, + { + "epoch": 2.267406105864265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5760772228240967, + "learning_rate": 1e-06, + "loss": 0.9278, + "mean_token_accuracy": 0.7284421324729919, + "num_tokens": 534305855.0, + "step": 20647 + }, + { + "epoch": 2.267515923566879, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.951674222946167, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7071957588195801, + "num_tokens": 534324343.0, + "step": 20648 + }, + { + "epoch": 2.2676257412694927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8064191341400146, + "learning_rate": 1e-06, + "loss": 0.9405, + "mean_token_accuracy": 0.7187570333480835, + "num_tokens": 534349733.0, + "step": 20649 + }, + { + "epoch": 2.2677355589721064, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.591132164001465, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6981972455978394, + "num_tokens": 534376612.0, + "step": 20650 + }, + { + "epoch": 2.26784537667472, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.474958896636963, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7142499089241028, + "num_tokens": 534405067.0, + "step": 20651 + }, + { + "epoch": 2.2679551943773335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.50874662399292, + "learning_rate": 1e-06, + "loss": 1.0938, + "mean_token_accuracy": 0.6838627457618713, + "num_tokens": 534434465.0, + "step": 20652 + }, + { + "epoch": 2.2680650120799473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.738572120666504, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.6990211009979248, + "num_tokens": 534457313.0, + "step": 20653 + }, + { + "epoch": 2.268174829782561, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5855047702789307, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7214192152023315, + "num_tokens": 534483692.0, + "step": 20654 + }, + { + "epoch": 2.2682846474851748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9570322036743164, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7035160064697266, + "num_tokens": 534506119.0, + "step": 20655 + }, + { + "epoch": 2.268394465187788, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4427249431610107, + "learning_rate": 1e-06, + "loss": 1.1034, + "mean_token_accuracy": 0.6848434805870056, + "num_tokens": 534539490.0, + "step": 20656 + }, + { + "epoch": 2.268504282890402, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5931460857391357, + "learning_rate": 1e-06, + "loss": 0.9163, + "mean_token_accuracy": 0.735582709312439, + "num_tokens": 534566135.0, + "step": 20657 + }, + { + "epoch": 2.2686141005930156, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6593735218048096, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7218526601791382, + "num_tokens": 534591151.0, + "step": 20658 + }, + { + "epoch": 2.2687239182956294, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.388622760772705, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7052053809165955, + "num_tokens": 534620416.0, + "step": 20659 + }, + { + "epoch": 2.2688337359982427, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7735040187835693, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7169991731643677, + "num_tokens": 534643272.0, + "step": 20660 + }, + { + "epoch": 2.2689435537008564, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.443814277648926, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6946439743041992, + "num_tokens": 534674244.0, + "step": 20661 + }, + { + "epoch": 2.26905337140347, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.470396041870117, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7243515849113464, + "num_tokens": 534702621.0, + "step": 20662 + }, + { + "epoch": 2.269163189106084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.851508140563965, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7140306830406189, + "num_tokens": 534724554.0, + "step": 20663 + }, + { + "epoch": 2.2692730068086977, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 4.562774658203125, + "learning_rate": 1e-06, + "loss": 1.0682, + "mean_token_accuracy": 0.6923375129699707, + "num_tokens": 534754776.0, + "step": 20664 + }, + { + "epoch": 2.269382824511311, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6620450019836426, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.709060549736023, + "num_tokens": 534781638.0, + "step": 20665 + }, + { + "epoch": 2.2694926422139248, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8275418281555176, + "learning_rate": 1e-06, + "loss": 0.934, + "mean_token_accuracy": 0.7250908017158508, + "num_tokens": 534800711.0, + "step": 20666 + }, + { + "epoch": 2.2696024599165385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.440046787261963, + "learning_rate": 1e-06, + "loss": 1.0997, + "mean_token_accuracy": 0.6822617650032043, + "num_tokens": 534833402.0, + "step": 20667 + }, + { + "epoch": 2.2697122776191523, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.667293071746826, + "learning_rate": 1e-06, + "loss": 0.9213, + "mean_token_accuracy": 0.7242000102996826, + "num_tokens": 534858057.0, + "step": 20668 + }, + { + "epoch": 2.269822095321766, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.505455255508423, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6867361068725586, + "num_tokens": 534884852.0, + "step": 20669 + }, + { + "epoch": 2.2699319130243794, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7738752365112305, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7147426605224609, + "num_tokens": 534906785.0, + "step": 20670 + }, + { + "epoch": 2.270041730726993, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4111034870147705, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.7067849636077881, + "num_tokens": 534939871.0, + "step": 20671 + }, + { + "epoch": 2.270151548429607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.643946409225464, + "learning_rate": 1e-06, + "loss": 0.9143, + "mean_token_accuracy": 0.733279824256897, + "num_tokens": 534957889.0, + "step": 20672 + }, + { + "epoch": 2.2702613661322206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5649049282073975, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6935437321662903, + "num_tokens": 534988918.0, + "step": 20673 + }, + { + "epoch": 2.2703711838348344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5923445224761963, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6961301565170288, + "num_tokens": 535016145.0, + "step": 20674 + }, + { + "epoch": 2.2704810015374477, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.718621253967285, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7225356101989746, + "num_tokens": 535040159.0, + "step": 20675 + }, + { + "epoch": 2.2705908192400615, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.403931140899658, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.725141167640686, + "num_tokens": 535068420.0, + "step": 20676 + }, + { + "epoch": 2.270700636942675, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0193803310394287, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7065267562866211, + "num_tokens": 535086914.0, + "step": 20677 + }, + { + "epoch": 2.270810454645289, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.790591239929199, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7236476540565491, + "num_tokens": 535107860.0, + "step": 20678 + }, + { + "epoch": 2.2709202723479027, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5679287910461426, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7170901298522949, + "num_tokens": 535134240.0, + "step": 20679 + }, + { + "epoch": 2.271030090050516, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7613210678100586, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.691501796245575, + "num_tokens": 535159061.0, + "step": 20680 + }, + { + "epoch": 2.27113990775313, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.594825506210327, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7167718410491943, + "num_tokens": 535183824.0, + "step": 20681 + }, + { + "epoch": 2.2712497254557436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.660499095916748, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7051942944526672, + "num_tokens": 535209253.0, + "step": 20682 + }, + { + "epoch": 2.2713595431583573, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.837662935256958, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.6999592781066895, + "num_tokens": 535232310.0, + "step": 20683 + }, + { + "epoch": 2.2714693608609706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2927982807159424, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7177509069442749, + "num_tokens": 535254604.0, + "step": 20684 + }, + { + "epoch": 2.2715791785635844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.460636854171753, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6970659494400024, + "num_tokens": 535288578.0, + "step": 20685 + }, + { + "epoch": 2.271688996266198, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.662858724594116, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6974492073059082, + "num_tokens": 535313776.0, + "step": 20686 + }, + { + "epoch": 2.271798813968812, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3487906455993652, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7060941457748413, + "num_tokens": 535343650.0, + "step": 20687 + }, + { + "epoch": 2.271908631671425, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.995356559753418, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7233666181564331, + "num_tokens": 535363387.0, + "step": 20688 + }, + { + "epoch": 2.272018449374039, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7991442680358887, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7347467541694641, + "num_tokens": 535385358.0, + "step": 20689 + }, + { + "epoch": 2.2721282670766527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.65401029586792, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6762214303016663, + "num_tokens": 535416079.0, + "step": 20690 + }, + { + "epoch": 2.2722380847792665, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.623581886291504, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7175612449645996, + "num_tokens": 535440675.0, + "step": 20691 + }, + { + "epoch": 2.2723479024818802, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.593383550643921, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.6959205269813538, + "num_tokens": 535469567.0, + "step": 20692 + }, + { + "epoch": 2.2724577201844935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3901355266571045, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6841006278991699, + "num_tokens": 535503782.0, + "step": 20693 + }, + { + "epoch": 2.2725675378871073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.53855562210083, + "learning_rate": 1e-06, + "loss": 0.9265, + "mean_token_accuracy": 0.7266219854354858, + "num_tokens": 535529052.0, + "step": 20694 + }, + { + "epoch": 2.272677355589721, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8434581756591797, + "learning_rate": 1e-06, + "loss": 0.9177, + "mean_token_accuracy": 0.7277755737304688, + "num_tokens": 535550344.0, + "step": 20695 + }, + { + "epoch": 2.272787173292335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1789190769195557, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7063156962394714, + "num_tokens": 535574011.0, + "step": 20696 + }, + { + "epoch": 2.2728969909949486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6104276180267334, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7006388902664185, + "num_tokens": 535604376.0, + "step": 20697 + }, + { + "epoch": 2.273006808697562, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.525956153869629, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.704687774181366, + "num_tokens": 535633623.0, + "step": 20698 + }, + { + "epoch": 2.2731166264001756, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.517806053161621, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7034610509872437, + "num_tokens": 535658667.0, + "step": 20699 + }, + { + "epoch": 2.2732264441027894, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.552211284637451, + "learning_rate": 1e-06, + "loss": 0.8902, + "mean_token_accuracy": 0.7423791885375977, + "num_tokens": 535682680.0, + "step": 20700 + }, + { + "epoch": 2.273336261805403, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6807398796081543, + "learning_rate": 1e-06, + "loss": 0.9367, + "mean_token_accuracy": 0.729858934879303, + "num_tokens": 535704776.0, + "step": 20701 + }, + { + "epoch": 2.273446079508017, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5732064247131348, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7049466371536255, + "num_tokens": 535731321.0, + "step": 20702 + }, + { + "epoch": 2.2735558972106302, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9813992977142334, + "learning_rate": 1e-06, + "loss": 0.8863, + "mean_token_accuracy": 0.7283066511154175, + "num_tokens": 535749357.0, + "step": 20703 + }, + { + "epoch": 2.273665714913244, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.496756076812744, + "learning_rate": 1e-06, + "loss": 0.9443, + "mean_token_accuracy": 0.7253127098083496, + "num_tokens": 535776414.0, + "step": 20704 + }, + { + "epoch": 2.2737755326158577, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.800457000732422, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6883904337882996, + "num_tokens": 535800583.0, + "step": 20705 + }, + { + "epoch": 2.2738853503184715, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5954067707061768, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.7036559581756592, + "num_tokens": 535827948.0, + "step": 20706 + }, + { + "epoch": 2.273995168021085, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.607569694519043, + "learning_rate": 1e-06, + "loss": 0.911, + "mean_token_accuracy": 0.7272263765335083, + "num_tokens": 535855953.0, + "step": 20707 + }, + { + "epoch": 2.2741049857236986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6392807960510254, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7276254892349243, + "num_tokens": 535878268.0, + "step": 20708 + }, + { + "epoch": 2.2742148034263123, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5760676860809326, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7090705633163452, + "num_tokens": 535906332.0, + "step": 20709 + }, + { + "epoch": 2.274324621128926, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.789034605026245, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7080975770950317, + "num_tokens": 535930170.0, + "step": 20710 + }, + { + "epoch": 2.2744344388315394, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.726752996444702, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7072398662567139, + "num_tokens": 535956811.0, + "step": 20711 + }, + { + "epoch": 2.274544256534153, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.506397247314453, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7095808982849121, + "num_tokens": 535987062.0, + "step": 20712 + }, + { + "epoch": 2.274654074236767, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3922293186187744, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7233036756515503, + "num_tokens": 536016727.0, + "step": 20713 + }, + { + "epoch": 2.2747638919393807, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.762969493865967, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7051205635070801, + "num_tokens": 536040280.0, + "step": 20714 + }, + { + "epoch": 2.2748737096419944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.243429183959961, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.6954963207244873, + "num_tokens": 536067056.0, + "step": 20715 + }, + { + "epoch": 2.2749835273446077, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8240606784820557, + "learning_rate": 1e-06, + "loss": 1.1169, + "mean_token_accuracy": 0.6875247359275818, + "num_tokens": 536090216.0, + "step": 20716 + }, + { + "epoch": 2.2750933450472215, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.770742893218994, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7182551026344299, + "num_tokens": 536113676.0, + "step": 20717 + }, + { + "epoch": 2.2752031627498353, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.482510805130005, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7199894785881042, + "num_tokens": 536140983.0, + "step": 20718 + }, + { + "epoch": 2.275312980452449, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7510910034179688, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6876808404922485, + "num_tokens": 536164566.0, + "step": 20719 + }, + { + "epoch": 2.2754227981550628, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5711610317230225, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7230355143547058, + "num_tokens": 536189222.0, + "step": 20720 + }, + { + "epoch": 2.275532615857676, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.899827241897583, + "learning_rate": 1e-06, + "loss": 0.92, + "mean_token_accuracy": 0.7215570211410522, + "num_tokens": 536207842.0, + "step": 20721 + }, + { + "epoch": 2.27564243356029, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.399911880493164, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7156881093978882, + "num_tokens": 536238776.0, + "step": 20722 + }, + { + "epoch": 2.2757522512629036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.956357479095459, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7154749631881714, + "num_tokens": 536258384.0, + "step": 20723 + }, + { + "epoch": 2.2758620689655173, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718301773071289, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7009512782096863, + "num_tokens": 536284965.0, + "step": 20724 + }, + { + "epoch": 2.275971886668131, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.520195722579956, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7125950455665588, + "num_tokens": 536313428.0, + "step": 20725 + }, + { + "epoch": 2.2760817043707444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7619993686676025, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6939448118209839, + "num_tokens": 536339517.0, + "step": 20726 + }, + { + "epoch": 2.276191522073358, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.562683582305908, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7131878137588501, + "num_tokens": 536365847.0, + "step": 20727 + }, + { + "epoch": 2.276301339775972, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5536489486694336, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6890479922294617, + "num_tokens": 536393986.0, + "step": 20728 + }, + { + "epoch": 2.2764111574785857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4824509620666504, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6864178776741028, + "num_tokens": 536422355.0, + "step": 20729 + }, + { + "epoch": 2.2765209751811994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9131901264190674, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7047772407531738, + "num_tokens": 536447075.0, + "step": 20730 + }, + { + "epoch": 2.2766307928838128, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6329286098480225, + "learning_rate": 1e-06, + "loss": 1.1213, + "mean_token_accuracy": 0.6765977740287781, + "num_tokens": 536478409.0, + "step": 20731 + }, + { + "epoch": 2.2767406105864265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5511302947998047, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6964800357818604, + "num_tokens": 536505080.0, + "step": 20732 + }, + { + "epoch": 2.2768504282890403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2739174365997314, + "learning_rate": 1e-06, + "loss": 0.9308, + "mean_token_accuracy": 0.7273228168487549, + "num_tokens": 536529626.0, + "step": 20733 + }, + { + "epoch": 2.276960245991654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.565802574157715, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7255827188491821, + "num_tokens": 536556253.0, + "step": 20734 + }, + { + "epoch": 2.2770700636942673, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7201149463653564, + "learning_rate": 1e-06, + "loss": 0.8841, + "mean_token_accuracy": 0.7354156970977783, + "num_tokens": 536578185.0, + "step": 20735 + }, + { + "epoch": 2.277179881396881, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9442710876464844, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.704420804977417, + "num_tokens": 536601047.0, + "step": 20736 + }, + { + "epoch": 2.277289699099495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.475581407546997, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7219835519790649, + "num_tokens": 536631078.0, + "step": 20737 + }, + { + "epoch": 2.2773995168021086, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0889031887054443, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7044581174850464, + "num_tokens": 536649282.0, + "step": 20738 + }, + { + "epoch": 2.277509334504722, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5372133255004883, + "learning_rate": 1e-06, + "loss": 1.0492, + "mean_token_accuracy": 0.6932123899459839, + "num_tokens": 536676114.0, + "step": 20739 + }, + { + "epoch": 2.2776191522073357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.663037061691284, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.7172044515609741, + "num_tokens": 536701464.0, + "step": 20740 + }, + { + "epoch": 2.2777289699099494, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.639918327331543, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6985271573066711, + "num_tokens": 536733933.0, + "step": 20741 + }, + { + "epoch": 2.277838787612563, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3713200092315674, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7183948755264282, + "num_tokens": 536764702.0, + "step": 20742 + }, + { + "epoch": 2.277948605315177, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.648329496383667, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.692101240158081, + "num_tokens": 536789400.0, + "step": 20743 + }, + { + "epoch": 2.2780584230177903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9313724040985107, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7376289367675781, + "num_tokens": 536810565.0, + "step": 20744 + }, + { + "epoch": 2.278168240720404, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5257091522216797, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7074965834617615, + "num_tokens": 536835174.0, + "step": 20745 + }, + { + "epoch": 2.278278058423018, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.572683572769165, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7024907469749451, + "num_tokens": 536860085.0, + "step": 20746 + }, + { + "epoch": 2.2783878761256315, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8056845664978027, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7088286876678467, + "num_tokens": 536883060.0, + "step": 20747 + }, + { + "epoch": 2.2784976938282453, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6277894973754883, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.702419638633728, + "num_tokens": 536912050.0, + "step": 20748 + }, + { + "epoch": 2.2786075115308586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1992547512054443, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7194234132766724, + "num_tokens": 536934709.0, + "step": 20749 + }, + { + "epoch": 2.2787173292334724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3609111309051514, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.715424120426178, + "num_tokens": 536963123.0, + "step": 20750 + }, + { + "epoch": 2.278827146936086, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.383591413497925, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6941614151000977, + "num_tokens": 536994318.0, + "step": 20751 + }, + { + "epoch": 2.2789369646387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1409072875976562, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7137851119041443, + "num_tokens": 537013036.0, + "step": 20752 + }, + { + "epoch": 2.2790467823413136, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4822518825531006, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.685884952545166, + "num_tokens": 537042769.0, + "step": 20753 + }, + { + "epoch": 2.279156600043927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.650552749633789, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7149984836578369, + "num_tokens": 537070224.0, + "step": 20754 + }, + { + "epoch": 2.2792664177465407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.463646650314331, + "learning_rate": 1e-06, + "loss": 1.1366, + "mean_token_accuracy": 0.6799811124801636, + "num_tokens": 537100398.0, + "step": 20755 + }, + { + "epoch": 2.2793762354491545, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.655128240585327, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7134865522384644, + "num_tokens": 537124562.0, + "step": 20756 + }, + { + "epoch": 2.279486053151768, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.728111982345581, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.6988872289657593, + "num_tokens": 537148606.0, + "step": 20757 + }, + { + "epoch": 2.2795958708543815, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4909069538116455, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.703552782535553, + "num_tokens": 537177375.0, + "step": 20758 + }, + { + "epoch": 2.2797056885569953, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.745173454284668, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7290425300598145, + "num_tokens": 537198936.0, + "step": 20759 + }, + { + "epoch": 2.279815506259609, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.823607921600342, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7274607419967651, + "num_tokens": 537222002.0, + "step": 20760 + }, + { + "epoch": 2.279925323962223, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4008986949920654, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7026574611663818, + "num_tokens": 537252011.0, + "step": 20761 + }, + { + "epoch": 2.2800351416648366, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6170437335968018, + "learning_rate": 1e-06, + "loss": 0.941, + "mean_token_accuracy": 0.7225198149681091, + "num_tokens": 537275960.0, + "step": 20762 + }, + { + "epoch": 2.28014495936745, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.436789035797119, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7124065160751343, + "num_tokens": 537306553.0, + "step": 20763 + }, + { + "epoch": 2.2802547770700636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.850900888442993, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7208466529846191, + "num_tokens": 537329446.0, + "step": 20764 + }, + { + "epoch": 2.2803645947726774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7213850021362305, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7164653539657593, + "num_tokens": 537352940.0, + "step": 20765 + }, + { + "epoch": 2.280474412475291, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4967639446258545, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7183555364608765, + "num_tokens": 537381267.0, + "step": 20766 + }, + { + "epoch": 2.2805842301779045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6496779918670654, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7097351551055908, + "num_tokens": 537407946.0, + "step": 20767 + }, + { + "epoch": 2.280694047880518, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9623796939849854, + "learning_rate": 1e-06, + "loss": 0.9365, + "mean_token_accuracy": 0.7271442413330078, + "num_tokens": 537426873.0, + "step": 20768 + }, + { + "epoch": 2.280803865583132, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.521886110305786, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.7118216156959534, + "num_tokens": 537453441.0, + "step": 20769 + }, + { + "epoch": 2.2809136832857457, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7225828170776367, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7061460018157959, + "num_tokens": 537477762.0, + "step": 20770 + }, + { + "epoch": 2.2810235009883595, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.571800947189331, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.703304648399353, + "num_tokens": 537504972.0, + "step": 20771 + }, + { + "epoch": 2.281133318690973, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.330641031265259, + "learning_rate": 1e-06, + "loss": 1.1256, + "mean_token_accuracy": 0.6822733879089355, + "num_tokens": 537538606.0, + "step": 20772 + }, + { + "epoch": 2.2812431363935866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.361257791519165, + "learning_rate": 1e-06, + "loss": 1.0639, + "mean_token_accuracy": 0.6978272795677185, + "num_tokens": 537570654.0, + "step": 20773 + }, + { + "epoch": 2.2813529540962003, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.928992986679077, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7063472270965576, + "num_tokens": 537597100.0, + "step": 20774 + }, + { + "epoch": 2.281462771798814, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7159016132354736, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7175266146659851, + "num_tokens": 537621795.0, + "step": 20775 + }, + { + "epoch": 2.281572589501428, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7135813236236572, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7135292291641235, + "num_tokens": 537645966.0, + "step": 20776 + }, + { + "epoch": 2.281682407204041, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4217286109924316, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6925817728042603, + "num_tokens": 537675897.0, + "step": 20777 + }, + { + "epoch": 2.281792224906655, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.492969036102295, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7144508957862854, + "num_tokens": 537703630.0, + "step": 20778 + }, + { + "epoch": 2.2819020426092687, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6170923709869385, + "learning_rate": 1e-06, + "loss": 1.1132, + "mean_token_accuracy": 0.6784123778343201, + "num_tokens": 537731596.0, + "step": 20779 + }, + { + "epoch": 2.2820118603118824, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5428924560546875, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.705676257610321, + "num_tokens": 537759309.0, + "step": 20780 + }, + { + "epoch": 2.282121678014496, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.864952564239502, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7142961025238037, + "num_tokens": 537779453.0, + "step": 20781 + }, + { + "epoch": 2.2822314957171095, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.534339427947998, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7201604843139648, + "num_tokens": 537804432.0, + "step": 20782 + }, + { + "epoch": 2.2823413134197232, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3298840522766113, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.713799774646759, + "num_tokens": 537833522.0, + "step": 20783 + }, + { + "epoch": 2.282451131122337, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.426957607269287, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7059022188186646, + "num_tokens": 537861995.0, + "step": 20784 + }, + { + "epoch": 2.2825609488249508, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8417506217956543, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7272520065307617, + "num_tokens": 537883858.0, + "step": 20785 + }, + { + "epoch": 2.282670766527564, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6068413257598877, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6872550845146179, + "num_tokens": 537910691.0, + "step": 20786 + }, + { + "epoch": 2.282780584230178, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.977726459503174, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7087683081626892, + "num_tokens": 537932131.0, + "step": 20787 + }, + { + "epoch": 2.2828904019327916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0531527996063232, + "learning_rate": 1e-06, + "loss": 1.1174, + "mean_token_accuracy": 0.6918210983276367, + "num_tokens": 537959397.0, + "step": 20788 + }, + { + "epoch": 2.2830002196354053, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.969932794570923, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.6994423866271973, + "num_tokens": 537980465.0, + "step": 20789 + }, + { + "epoch": 2.2831100373380186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3396739959716797, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6976807117462158, + "num_tokens": 538012996.0, + "step": 20790 + }, + { + "epoch": 2.2832198550406324, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6686534881591797, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6891130805015564, + "num_tokens": 538042730.0, + "step": 20791 + }, + { + "epoch": 2.283329672743246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8352086544036865, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7118554711341858, + "num_tokens": 538064188.0, + "step": 20792 + }, + { + "epoch": 2.28343949044586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 4.110329627990723, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.7258481383323669, + "num_tokens": 538089270.0, + "step": 20793 + }, + { + "epoch": 2.2835493081484737, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1833083629608154, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7031887173652649, + "num_tokens": 538117966.0, + "step": 20794 + }, + { + "epoch": 2.283659125851087, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5215392112731934, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.723673403263092, + "num_tokens": 538146195.0, + "step": 20795 + }, + { + "epoch": 2.2837689435537007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6021859645843506, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.7104477882385254, + "num_tokens": 538171123.0, + "step": 20796 + }, + { + "epoch": 2.2838787612563145, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6470885276794434, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6894128918647766, + "num_tokens": 538195696.0, + "step": 20797 + }, + { + "epoch": 2.2839885789589283, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.374180793762207, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7050774097442627, + "num_tokens": 538224843.0, + "step": 20798 + }, + { + "epoch": 2.284098396661542, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.081341028213501, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7189220190048218, + "num_tokens": 538243669.0, + "step": 20799 + }, + { + "epoch": 2.2842082143641553, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584679126739502, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7151574492454529, + "num_tokens": 538271782.0, + "step": 20800 + }, + { + "epoch": 2.284318032066769, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6111481189727783, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7053237557411194, + "num_tokens": 538295796.0, + "step": 20801 + }, + { + "epoch": 2.284427849769383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7878577709198, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6953223347663879, + "num_tokens": 538319690.0, + "step": 20802 + }, + { + "epoch": 2.2845376674719966, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6004419326782227, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7035513520240784, + "num_tokens": 538345376.0, + "step": 20803 + }, + { + "epoch": 2.2846474851746104, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.686544895172119, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7180287837982178, + "num_tokens": 538368600.0, + "step": 20804 + }, + { + "epoch": 2.2847573028772237, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4855589866638184, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7091460824012756, + "num_tokens": 538396941.0, + "step": 20805 + }, + { + "epoch": 2.2848671205798374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.687070846557617, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6836089491844177, + "num_tokens": 538423386.0, + "step": 20806 + }, + { + "epoch": 2.284976938282451, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6845765113830566, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.7001731395721436, + "num_tokens": 538449073.0, + "step": 20807 + }, + { + "epoch": 2.285086755985065, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4481890201568604, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7104825973510742, + "num_tokens": 538481914.0, + "step": 20808 + }, + { + "epoch": 2.2851965736876787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8073835372924805, + "learning_rate": 1e-06, + "loss": 0.9139, + "mean_token_accuracy": 0.7276119589805603, + "num_tokens": 538502236.0, + "step": 20809 + }, + { + "epoch": 2.285306391390292, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.99271559715271, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7096697092056274, + "num_tokens": 538522773.0, + "step": 20810 + }, + { + "epoch": 2.2854162090929058, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6133744716644287, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7148836851119995, + "num_tokens": 538547524.0, + "step": 20811 + }, + { + "epoch": 2.2855260267955195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.779627799987793, + "learning_rate": 1e-06, + "loss": 0.9508, + "mean_token_accuracy": 0.7228979468345642, + "num_tokens": 538569996.0, + "step": 20812 + }, + { + "epoch": 2.2856358444981333, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.484724283218384, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7036766409873962, + "num_tokens": 538597151.0, + "step": 20813 + }, + { + "epoch": 2.2857456622007466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4099888801574707, + "learning_rate": 1e-06, + "loss": 1.0974, + "mean_token_accuracy": 0.6881179809570312, + "num_tokens": 538627382.0, + "step": 20814 + }, + { + "epoch": 2.2858554799033604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7540476322174072, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7062039375305176, + "num_tokens": 538652426.0, + "step": 20815 + }, + { + "epoch": 2.285965297605974, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5003318786621094, + "learning_rate": 1e-06, + "loss": 1.1393, + "mean_token_accuracy": 0.6677215099334717, + "num_tokens": 538681280.0, + "step": 20816 + }, + { + "epoch": 2.286075115308588, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545367479324341, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.6993893384933472, + "num_tokens": 538709258.0, + "step": 20817 + }, + { + "epoch": 2.286184933011201, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.752537727355957, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7113903760910034, + "num_tokens": 538734645.0, + "step": 20818 + }, + { + "epoch": 2.286294750713815, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7249534130096436, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.7177342772483826, + "num_tokens": 538759285.0, + "step": 20819 + }, + { + "epoch": 2.2864045684164287, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.388631820678711, + "learning_rate": 1e-06, + "loss": 1.1171, + "mean_token_accuracy": 0.6836628913879395, + "num_tokens": 538789988.0, + "step": 20820 + }, + { + "epoch": 2.2865143861190425, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6906707286834717, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7161847352981567, + "num_tokens": 538813693.0, + "step": 20821 + }, + { + "epoch": 2.286624203821656, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5668280124664307, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7193716168403625, + "num_tokens": 538839601.0, + "step": 20822 + }, + { + "epoch": 2.2867340215242695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.420316457748413, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7072627544403076, + "num_tokens": 538869111.0, + "step": 20823 + }, + { + "epoch": 2.2868438392268833, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.472353458404541, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7065558433532715, + "num_tokens": 538895327.0, + "step": 20824 + }, + { + "epoch": 2.286953656929497, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.919523000717163, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6878825426101685, + "num_tokens": 538917551.0, + "step": 20825 + }, + { + "epoch": 2.287063474632111, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.356166362762451, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6961147785186768, + "num_tokens": 538948810.0, + "step": 20826 + }, + { + "epoch": 2.2871732923347246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5838537216186523, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.686172604560852, + "num_tokens": 538977071.0, + "step": 20827 + }, + { + "epoch": 2.287283110037338, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4724478721618652, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7210190296173096, + "num_tokens": 539005221.0, + "step": 20828 + }, + { + "epoch": 2.2873929277399516, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3973147869110107, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6990039944648743, + "num_tokens": 539037391.0, + "step": 20829 + }, + { + "epoch": 2.2875027454425654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.194671630859375, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7193669080734253, + "num_tokens": 539054853.0, + "step": 20830 + }, + { + "epoch": 2.287612563145179, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.767810583114624, + "learning_rate": 1e-06, + "loss": 0.9404, + "mean_token_accuracy": 0.7266448140144348, + "num_tokens": 539076794.0, + "step": 20831 + }, + { + "epoch": 2.287722380847793, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.669041872024536, + "learning_rate": 1e-06, + "loss": 0.894, + "mean_token_accuracy": 0.7422579526901245, + "num_tokens": 539099218.0, + "step": 20832 + }, + { + "epoch": 2.287832198550406, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596658706665039, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7066942453384399, + "num_tokens": 539124379.0, + "step": 20833 + }, + { + "epoch": 2.28794201625302, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5859146118164062, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.7021248936653137, + "num_tokens": 539150985.0, + "step": 20834 + }, + { + "epoch": 2.2880518339556337, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9176604747772217, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.7219400405883789, + "num_tokens": 539171635.0, + "step": 20835 + }, + { + "epoch": 2.2881616516582475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.792431592941284, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7202763557434082, + "num_tokens": 539195741.0, + "step": 20836 + }, + { + "epoch": 2.288271469360861, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.711899995803833, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7129874229431152, + "num_tokens": 539224786.0, + "step": 20837 + }, + { + "epoch": 2.2883812870634745, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9964611530303955, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6886187791824341, + "num_tokens": 539249631.0, + "step": 20838 + }, + { + "epoch": 2.2884911047660883, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565589189529419, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.706246018409729, + "num_tokens": 539277615.0, + "step": 20839 + }, + { + "epoch": 2.288600922468702, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.841510772705078, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7088122367858887, + "num_tokens": 539300415.0, + "step": 20840 + }, + { + "epoch": 2.2887107401713154, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.632864236831665, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.7195262908935547, + "num_tokens": 539324690.0, + "step": 20841 + }, + { + "epoch": 2.288820557873929, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.480764150619507, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7207285165786743, + "num_tokens": 539353522.0, + "step": 20842 + }, + { + "epoch": 2.288930375576543, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5689170360565186, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7111538052558899, + "num_tokens": 539380717.0, + "step": 20843 + }, + { + "epoch": 2.2890401932791566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7106451988220215, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6849135160446167, + "num_tokens": 539406095.0, + "step": 20844 + }, + { + "epoch": 2.2891500109817704, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.94452166557312, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7169382572174072, + "num_tokens": 539428055.0, + "step": 20845 + }, + { + "epoch": 2.2892598286843837, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4905498027801514, + "learning_rate": 1e-06, + "loss": 0.9311, + "mean_token_accuracy": 0.722201406955719, + "num_tokens": 539453392.0, + "step": 20846 + }, + { + "epoch": 2.2893696463869975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.546247720718384, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7138252258300781, + "num_tokens": 539478657.0, + "step": 20847 + }, + { + "epoch": 2.2894794640896112, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6430482864379883, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7136767506599426, + "num_tokens": 539503239.0, + "step": 20848 + }, + { + "epoch": 2.289589281792225, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5525104999542236, + "learning_rate": 1e-06, + "loss": 0.9328, + "mean_token_accuracy": 0.7299906015396118, + "num_tokens": 539529013.0, + "step": 20849 + }, + { + "epoch": 2.2896990994948387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6387317180633545, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.7269062995910645, + "num_tokens": 539552767.0, + "step": 20850 + }, + { + "epoch": 2.289808917197452, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.572477102279663, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.7337256669998169, + "num_tokens": 539579488.0, + "step": 20851 + }, + { + "epoch": 2.289918734900066, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5779831409454346, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7005631327629089, + "num_tokens": 539606583.0, + "step": 20852 + }, + { + "epoch": 2.2900285526026796, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.277311325073242, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6931403875350952, + "num_tokens": 539631989.0, + "step": 20853 + }, + { + "epoch": 2.2901383703052933, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6813805103302, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.703732430934906, + "num_tokens": 539656172.0, + "step": 20854 + }, + { + "epoch": 2.290248188007907, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6144282817840576, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.7010195255279541, + "num_tokens": 539684394.0, + "step": 20855 + }, + { + "epoch": 2.2903580057105204, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.573801040649414, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.6827096343040466, + "num_tokens": 539712695.0, + "step": 20856 + }, + { + "epoch": 2.290467823413134, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5502352714538574, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6930855512619019, + "num_tokens": 539741458.0, + "step": 20857 + }, + { + "epoch": 2.290577641115748, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4300143718719482, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7067824006080627, + "num_tokens": 539770286.0, + "step": 20858 + }, + { + "epoch": 2.2906874588183617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.554469585418701, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.6943299770355225, + "num_tokens": 539797632.0, + "step": 20859 + }, + { + "epoch": 2.2907972765209754, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.345808506011963, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.696609377861023, + "num_tokens": 539831806.0, + "step": 20860 + }, + { + "epoch": 2.2909070942235887, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.433422327041626, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7037420272827148, + "num_tokens": 539859437.0, + "step": 20861 + }, + { + "epoch": 2.2910169119262025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8001718521118164, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.6986583471298218, + "num_tokens": 539880540.0, + "step": 20862 + }, + { + "epoch": 2.2911267296288162, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.059607744216919, + "learning_rate": 1e-06, + "loss": 0.9123, + "mean_token_accuracy": 0.7293241024017334, + "num_tokens": 539899973.0, + "step": 20863 + }, + { + "epoch": 2.29123654733143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2440218925476074, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.7020160555839539, + "num_tokens": 539934351.0, + "step": 20864 + }, + { + "epoch": 2.2913463650340433, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8650143146514893, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7218880653381348, + "num_tokens": 539956011.0, + "step": 20865 + }, + { + "epoch": 2.291456182736657, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7457220554351807, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.7206591963768005, + "num_tokens": 539976916.0, + "step": 20866 + }, + { + "epoch": 2.291566000439271, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6650993824005127, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6951066255569458, + "num_tokens": 540002841.0, + "step": 20867 + }, + { + "epoch": 2.2916758181418846, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6452972888946533, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.7073190212249756, + "num_tokens": 540028328.0, + "step": 20868 + }, + { + "epoch": 2.291785635844498, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5090606212615967, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7049633860588074, + "num_tokens": 540054231.0, + "step": 20869 + }, + { + "epoch": 2.2918954535471117, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.908714532852173, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7139624357223511, + "num_tokens": 540072985.0, + "step": 20870 + }, + { + "epoch": 2.2920052712497254, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9778358936309814, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7173696756362915, + "num_tokens": 540093628.0, + "step": 20871 + }, + { + "epoch": 2.292115088952339, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.782280206680298, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7110922336578369, + "num_tokens": 540115163.0, + "step": 20872 + }, + { + "epoch": 2.292224906654953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.720594882965088, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7275714874267578, + "num_tokens": 540137017.0, + "step": 20873 + }, + { + "epoch": 2.2923347243575662, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.77093505859375, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.704789400100708, + "num_tokens": 540158223.0, + "step": 20874 + }, + { + "epoch": 2.29244454206018, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6566107273101807, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6913022398948669, + "num_tokens": 540184447.0, + "step": 20875 + }, + { + "epoch": 2.2925543597627938, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 6.677761077880859, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6986107230186462, + "num_tokens": 540214098.0, + "step": 20876 + }, + { + "epoch": 2.2926641774654075, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9270899295806885, + "learning_rate": 1e-06, + "loss": 0.8686, + "mean_token_accuracy": 0.7425394654273987, + "num_tokens": 540236665.0, + "step": 20877 + }, + { + "epoch": 2.2927739951680213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6354470252990723, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7228595018386841, + "num_tokens": 540262127.0, + "step": 20878 + }, + { + "epoch": 2.2928838128706346, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8992748260498047, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.712834358215332, + "num_tokens": 540286459.0, + "step": 20879 + }, + { + "epoch": 2.2929936305732483, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7959883213043213, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.701202392578125, + "num_tokens": 540308637.0, + "step": 20880 + }, + { + "epoch": 2.293103448275862, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.734010934829712, + "learning_rate": 1e-06, + "loss": 0.9033, + "mean_token_accuracy": 0.729962170124054, + "num_tokens": 540330801.0, + "step": 20881 + }, + { + "epoch": 2.293213265978476, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8813159465789795, + "learning_rate": 1e-06, + "loss": 0.8769, + "mean_token_accuracy": 0.7358241677284241, + "num_tokens": 540350290.0, + "step": 20882 + }, + { + "epoch": 2.2933230836810896, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5854530334472656, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7178425788879395, + "num_tokens": 540375822.0, + "step": 20883 + }, + { + "epoch": 2.293432901383703, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.886908769607544, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.7042490243911743, + "num_tokens": 540400247.0, + "step": 20884 + }, + { + "epoch": 2.2935427190863167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0050346851348877, + "learning_rate": 1e-06, + "loss": 0.8554, + "mean_token_accuracy": 0.742678701877594, + "num_tokens": 540419036.0, + "step": 20885 + }, + { + "epoch": 2.2936525367889304, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.33200740814209, + "learning_rate": 1e-06, + "loss": 1.0994, + "mean_token_accuracy": 0.6772603392601013, + "num_tokens": 540454269.0, + "step": 20886 + }, + { + "epoch": 2.293762354491544, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5560524463653564, + "learning_rate": 1e-06, + "loss": 0.9606, + "mean_token_accuracy": 0.7186869978904724, + "num_tokens": 540478961.0, + "step": 20887 + }, + { + "epoch": 2.2938721721941575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4974491596221924, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7237439155578613, + "num_tokens": 540505132.0, + "step": 20888 + }, + { + "epoch": 2.2939819898967713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9780118465423584, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7131714224815369, + "num_tokens": 540524365.0, + "step": 20889 + }, + { + "epoch": 2.294091807599385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5149378776550293, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.7038170099258423, + "num_tokens": 540551399.0, + "step": 20890 + }, + { + "epoch": 2.294201625301999, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5643956661224365, + "learning_rate": 1e-06, + "loss": 0.9316, + "mean_token_accuracy": 0.7202543616294861, + "num_tokens": 540575806.0, + "step": 20891 + }, + { + "epoch": 2.294311443004612, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5715091228485107, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7077816128730774, + "num_tokens": 540599400.0, + "step": 20892 + }, + { + "epoch": 2.294421260707226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.490604877471924, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7133170366287231, + "num_tokens": 540629135.0, + "step": 20893 + }, + { + "epoch": 2.2945310784098396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4895052909851074, + "learning_rate": 1e-06, + "loss": 1.1058, + "mean_token_accuracy": 0.6784271001815796, + "num_tokens": 540659440.0, + "step": 20894 + }, + { + "epoch": 2.2946408961124534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7375447750091553, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7246402502059937, + "num_tokens": 540680501.0, + "step": 20895 + }, + { + "epoch": 2.294750713815067, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7620303630828857, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7129225730895996, + "num_tokens": 540703039.0, + "step": 20896 + }, + { + "epoch": 2.2948605315176804, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.406132936477661, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6964340209960938, + "num_tokens": 540732665.0, + "step": 20897 + }, + { + "epoch": 2.294970349220294, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5117955207824707, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.7027980089187622, + "num_tokens": 540759246.0, + "step": 20898 + }, + { + "epoch": 2.295080166922908, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7533276081085205, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.6958104372024536, + "num_tokens": 540788127.0, + "step": 20899 + }, + { + "epoch": 2.2951899846255217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.98690128326416, + "learning_rate": 1e-06, + "loss": 0.942, + "mean_token_accuracy": 0.7241922616958618, + "num_tokens": 540806931.0, + "step": 20900 + }, + { + "epoch": 2.2952998023281355, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0443711280822754, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7115681171417236, + "num_tokens": 540826294.0, + "step": 20901 + }, + { + "epoch": 2.2954096200307488, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5752065181732178, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.702203631401062, + "num_tokens": 540851938.0, + "step": 20902 + }, + { + "epoch": 2.2955194377333625, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5139081478118896, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6922941207885742, + "num_tokens": 540879803.0, + "step": 20903 + }, + { + "epoch": 2.2956292554359763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.562049150466919, + "learning_rate": 1e-06, + "loss": 1.0698, + "mean_token_accuracy": 0.6878634691238403, + "num_tokens": 540908534.0, + "step": 20904 + }, + { + "epoch": 2.29573907313859, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7153589725494385, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.6999812722206116, + "num_tokens": 540930373.0, + "step": 20905 + }, + { + "epoch": 2.295848890841204, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.548053026199341, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7177244424819946, + "num_tokens": 540954331.0, + "step": 20906 + }, + { + "epoch": 2.295958708543817, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7695109844207764, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7250181436538696, + "num_tokens": 540975078.0, + "step": 20907 + }, + { + "epoch": 2.296068526246431, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5830254554748535, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6948144435882568, + "num_tokens": 541000601.0, + "step": 20908 + }, + { + "epoch": 2.2961783439490446, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6576547622680664, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6994129419326782, + "num_tokens": 541028698.0, + "step": 20909 + }, + { + "epoch": 2.2962881616516584, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.381222724914551, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6851440072059631, + "num_tokens": 541060964.0, + "step": 20910 + }, + { + "epoch": 2.296397979354272, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.681852102279663, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7029962539672852, + "num_tokens": 541083273.0, + "step": 20911 + }, + { + "epoch": 2.2965077970568855, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.689342737197876, + "learning_rate": 1e-06, + "loss": 1.1195, + "mean_token_accuracy": 0.6859505772590637, + "num_tokens": 541107607.0, + "step": 20912 + }, + { + "epoch": 2.296617614759499, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6527605056762695, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.708397626876831, + "num_tokens": 541131056.0, + "step": 20913 + }, + { + "epoch": 2.296727432462113, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 4.23444128036499, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6994035840034485, + "num_tokens": 541158773.0, + "step": 20914 + }, + { + "epoch": 2.2968372501647267, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 9.75293254852295, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.717073917388916, + "num_tokens": 541181050.0, + "step": 20915 + }, + { + "epoch": 2.29694706786734, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6582860946655273, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.7029699087142944, + "num_tokens": 541210007.0, + "step": 20916 + }, + { + "epoch": 2.297056885569954, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5253403186798096, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.7008564472198486, + "num_tokens": 541235748.0, + "step": 20917 + }, + { + "epoch": 2.2971667032725676, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.480994939804077, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.702576220035553, + "num_tokens": 541267118.0, + "step": 20918 + }, + { + "epoch": 2.2972765209751813, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.769620180130005, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7129129767417908, + "num_tokens": 541290251.0, + "step": 20919 + }, + { + "epoch": 2.2973863386777946, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7585604190826416, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.6897006630897522, + "num_tokens": 541321159.0, + "step": 20920 + }, + { + "epoch": 2.2974961563804084, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4804444313049316, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6970593333244324, + "num_tokens": 541348450.0, + "step": 20921 + }, + { + "epoch": 2.297605974083022, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7348506450653076, + "learning_rate": 1e-06, + "loss": 0.9679, + "mean_token_accuracy": 0.7215172648429871, + "num_tokens": 541372135.0, + "step": 20922 + }, + { + "epoch": 2.297715791785636, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.432631015777588, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.6982884407043457, + "num_tokens": 541401150.0, + "step": 20923 + }, + { + "epoch": 2.2978256094882497, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.66871452331543, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6955443620681763, + "num_tokens": 541435158.0, + "step": 20924 + }, + { + "epoch": 2.297935427190863, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2353267669677734, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7087351083755493, + "num_tokens": 541453374.0, + "step": 20925 + }, + { + "epoch": 2.2980452448934767, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.782756805419922, + "learning_rate": 1e-06, + "loss": 0.8183, + "mean_token_accuracy": 0.7608811855316162, + "num_tokens": 541475660.0, + "step": 20926 + }, + { + "epoch": 2.2981550625960905, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.494553565979004, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7314828634262085, + "num_tokens": 541503522.0, + "step": 20927 + }, + { + "epoch": 2.2982648802987042, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6034960746765137, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7113826274871826, + "num_tokens": 541528360.0, + "step": 20928 + }, + { + "epoch": 2.298374698001318, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6214191913604736, + "learning_rate": 1e-06, + "loss": 1.0511, + "mean_token_accuracy": 0.6960808038711548, + "num_tokens": 541554598.0, + "step": 20929 + }, + { + "epoch": 2.2984845157039313, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6474757194519043, + "learning_rate": 1e-06, + "loss": 0.9019, + "mean_token_accuracy": 0.7339421510696411, + "num_tokens": 541577840.0, + "step": 20930 + }, + { + "epoch": 2.298594333406545, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4103264808654785, + "learning_rate": 1e-06, + "loss": 1.0731, + "mean_token_accuracy": 0.6915647983551025, + "num_tokens": 541608864.0, + "step": 20931 + }, + { + "epoch": 2.298704151109159, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.547373056411743, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6820883750915527, + "num_tokens": 541636803.0, + "step": 20932 + }, + { + "epoch": 2.2988139688117726, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.542813539505005, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7047812938690186, + "num_tokens": 541664820.0, + "step": 20933 + }, + { + "epoch": 2.2989237865143863, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.60090708732605, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7000637650489807, + "num_tokens": 541690577.0, + "step": 20934 + }, + { + "epoch": 2.2990336042169996, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.518068790435791, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7080966234207153, + "num_tokens": 541716375.0, + "step": 20935 + }, + { + "epoch": 2.2991434219196134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6375248432159424, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7228201627731323, + "num_tokens": 541742331.0, + "step": 20936 + }, + { + "epoch": 2.299253239622227, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.534832000732422, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7160242795944214, + "num_tokens": 541767319.0, + "step": 20937 + }, + { + "epoch": 2.299363057324841, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.72006893157959, + "learning_rate": 1e-06, + "loss": 1.1056, + "mean_token_accuracy": 0.6859025955200195, + "num_tokens": 541795125.0, + "step": 20938 + }, + { + "epoch": 2.2994728750274542, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2994649410247803, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6923831701278687, + "num_tokens": 541830697.0, + "step": 20939 + }, + { + "epoch": 2.299582692730068, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.70136260986328, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7149090766906738, + "num_tokens": 541856540.0, + "step": 20940 + }, + { + "epoch": 2.2996925104326817, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0787014961242676, + "learning_rate": 1e-06, + "loss": 0.9023, + "mean_token_accuracy": 0.7282554507255554, + "num_tokens": 541876615.0, + "step": 20941 + }, + { + "epoch": 2.2998023281352955, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0021324157714844, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7033287286758423, + "num_tokens": 541903735.0, + "step": 20942 + }, + { + "epoch": 2.2999121458379093, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.385585069656372, + "learning_rate": 1e-06, + "loss": 1.0681, + "mean_token_accuracy": 0.6920878887176514, + "num_tokens": 541935807.0, + "step": 20943 + }, + { + "epoch": 2.3000219635405226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.529592752456665, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7208937406539917, + "num_tokens": 541962688.0, + "step": 20944 + }, + { + "epoch": 2.3001317812431363, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5651915073394775, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7014179825782776, + "num_tokens": 541988938.0, + "step": 20945 + }, + { + "epoch": 2.30024159894575, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6991963386535645, + "learning_rate": 1e-06, + "loss": 0.881, + "mean_token_accuracy": 0.7383135557174683, + "num_tokens": 542010466.0, + "step": 20946 + }, + { + "epoch": 2.300351416648364, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6836400032043457, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.714942991733551, + "num_tokens": 542034293.0, + "step": 20947 + }, + { + "epoch": 2.300461234350977, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.727663993835449, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6920602321624756, + "num_tokens": 542059149.0, + "step": 20948 + }, + { + "epoch": 2.300571052053591, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.430356979370117, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.705807089805603, + "num_tokens": 542089035.0, + "step": 20949 + }, + { + "epoch": 2.3006808697562047, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.655925750732422, + "learning_rate": 1e-06, + "loss": 1.119, + "mean_token_accuracy": 0.6742687225341797, + "num_tokens": 542116054.0, + "step": 20950 + }, + { + "epoch": 2.3007906874588184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.497616767883301, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.6986196041107178, + "num_tokens": 542145231.0, + "step": 20951 + }, + { + "epoch": 2.300900505161432, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4819459915161133, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.702309250831604, + "num_tokens": 542173544.0, + "step": 20952 + }, + { + "epoch": 2.3010103228640455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.639880895614624, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6899086236953735, + "num_tokens": 542202205.0, + "step": 20953 + }, + { + "epoch": 2.3011201405666593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6485347747802734, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7034099698066711, + "num_tokens": 542227842.0, + "step": 20954 + }, + { + "epoch": 2.301229958269273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.412222146987915, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7043783664703369, + "num_tokens": 542255069.0, + "step": 20955 + }, + { + "epoch": 2.3013397759718868, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.547020673751831, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7090955972671509, + "num_tokens": 542282401.0, + "step": 20956 + }, + { + "epoch": 2.3014495936745005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5055487155914307, + "learning_rate": 1e-06, + "loss": 0.8902, + "mean_token_accuracy": 0.7348900437355042, + "num_tokens": 542308418.0, + "step": 20957 + }, + { + "epoch": 2.301559411377114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.707180976867676, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.7038382291793823, + "num_tokens": 542333647.0, + "step": 20958 + }, + { + "epoch": 2.3016692290797276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.913879871368408, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7117270827293396, + "num_tokens": 542359218.0, + "step": 20959 + }, + { + "epoch": 2.3017790467823414, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.636749505996704, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7162100076675415, + "num_tokens": 542388516.0, + "step": 20960 + }, + { + "epoch": 2.301888864484955, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5624287128448486, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6940227150917053, + "num_tokens": 542417076.0, + "step": 20961 + }, + { + "epoch": 2.301998682187569, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.875809669494629, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7073496580123901, + "num_tokens": 542439792.0, + "step": 20962 + }, + { + "epoch": 2.302108499890182, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.463625431060791, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7010698914527893, + "num_tokens": 542469589.0, + "step": 20963 + }, + { + "epoch": 2.302218317592796, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.699411153793335, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6986815929412842, + "num_tokens": 542494322.0, + "step": 20964 + }, + { + "epoch": 2.3023281352954097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.583556890487671, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7291053533554077, + "num_tokens": 542519882.0, + "step": 20965 + }, + { + "epoch": 2.3024379529980235, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4610016345977783, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.7024050354957581, + "num_tokens": 542548456.0, + "step": 20966 + }, + { + "epoch": 2.3025477707006368, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7687735557556152, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7077882289886475, + "num_tokens": 542573067.0, + "step": 20967 + }, + { + "epoch": 2.3026575884032505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5231261253356934, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.6989635229110718, + "num_tokens": 542602039.0, + "step": 20968 + }, + { + "epoch": 2.3027674061058643, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.806607484817505, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7259374856948853, + "num_tokens": 542622463.0, + "step": 20969 + }, + { + "epoch": 2.302877223808478, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.532942771911621, + "learning_rate": 1e-06, + "loss": 1.0786, + "mean_token_accuracy": 0.6831611394882202, + "num_tokens": 542649997.0, + "step": 20970 + }, + { + "epoch": 2.3029870415110913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3215389251708984, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7189860343933105, + "num_tokens": 542679946.0, + "step": 20971 + }, + { + "epoch": 2.303096859213705, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1808011531829834, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7130767107009888, + "num_tokens": 542704940.0, + "step": 20972 + }, + { + "epoch": 2.303206676916319, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.537025213241577, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6981106996536255, + "num_tokens": 542735386.0, + "step": 20973 + }, + { + "epoch": 2.3033164946189326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.556748867034912, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7184740900993347, + "num_tokens": 542760679.0, + "step": 20974 + }, + { + "epoch": 2.3034263123215464, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.350144386291504, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7202741503715515, + "num_tokens": 542789594.0, + "step": 20975 + }, + { + "epoch": 2.3035361300241597, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8486053943634033, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6984996795654297, + "num_tokens": 542817666.0, + "step": 20976 + }, + { + "epoch": 2.3036459477267734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.732084274291992, + "learning_rate": 1e-06, + "loss": 0.9395, + "mean_token_accuracy": 0.7277757525444031, + "num_tokens": 542841983.0, + "step": 20977 + }, + { + "epoch": 2.303755765429387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4429922103881836, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7136525511741638, + "num_tokens": 542869857.0, + "step": 20978 + }, + { + "epoch": 2.303865583132001, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3352341651916504, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7054091691970825, + "num_tokens": 542900501.0, + "step": 20979 + }, + { + "epoch": 2.3039754008346147, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7005209922790527, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7121385335922241, + "num_tokens": 542922771.0, + "step": 20980 + }, + { + "epoch": 2.304085218537228, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.637263059616089, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.7291936874389648, + "num_tokens": 542945980.0, + "step": 20981 + }, + { + "epoch": 2.304195036239842, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.718583345413208, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6957383155822754, + "num_tokens": 542969733.0, + "step": 20982 + }, + { + "epoch": 2.3043048539424555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6031014919281006, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.7231690287590027, + "num_tokens": 542997008.0, + "step": 20983 + }, + { + "epoch": 2.3044146716450693, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.691340446472168, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.713280439376831, + "num_tokens": 543021143.0, + "step": 20984 + }, + { + "epoch": 2.304524489347683, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.68619441986084, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7212775945663452, + "num_tokens": 543045313.0, + "step": 20985 + }, + { + "epoch": 2.3046343070502964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7640838623046875, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7166604995727539, + "num_tokens": 543068448.0, + "step": 20986 + }, + { + "epoch": 2.30474412475291, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.406686782836914, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7004287242889404, + "num_tokens": 543099282.0, + "step": 20987 + }, + { + "epoch": 2.304853942455524, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0433006286621094, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7120745182037354, + "num_tokens": 543120078.0, + "step": 20988 + }, + { + "epoch": 2.3049637601581376, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6487889289855957, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7116349339485168, + "num_tokens": 543143251.0, + "step": 20989 + }, + { + "epoch": 2.3050735778607514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.854121685028076, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6976239681243896, + "num_tokens": 543167052.0, + "step": 20990 + }, + { + "epoch": 2.3051833955633647, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9847824573516846, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7078206539154053, + "num_tokens": 543188857.0, + "step": 20991 + }, + { + "epoch": 2.3052932132659785, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.5259714126586914, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7117398977279663, + "num_tokens": 543215243.0, + "step": 20992 + }, + { + "epoch": 2.3054030309685922, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8137574195861816, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7025313377380371, + "num_tokens": 543239720.0, + "step": 20993 + }, + { + "epoch": 2.305512848671206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8893349170684814, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7138906121253967, + "num_tokens": 543261966.0, + "step": 20994 + }, + { + "epoch": 2.3056226663738193, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.766422271728516, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7230595350265503, + "num_tokens": 543281693.0, + "step": 20995 + }, + { + "epoch": 2.305732484076433, + "ewc_loss": 1.0967254638671875e-05, + "grad_norm": 80.05146789550781, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7199229598045349, + "num_tokens": 543303138.0, + "step": 20996 + }, + { + "epoch": 2.305842301779047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9513254165649414, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6839890480041504, + "num_tokens": 543330442.0, + "step": 20997 + }, + { + "epoch": 2.3059521194816606, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6551425457000732, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6922470331192017, + "num_tokens": 543362534.0, + "step": 20998 + }, + { + "epoch": 2.306061937184274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8725857734680176, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7162948250770569, + "num_tokens": 543382404.0, + "step": 20999 + }, + { + "epoch": 2.3061717548868876, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.640390634536743, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6952174305915833, + "num_tokens": 543407543.0, + "step": 21000 + }, + { + "epoch": 2.3062815725895014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5805797576904297, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7219012975692749, + "num_tokens": 543430168.0, + "step": 21001 + }, + { + "epoch": 2.306391390292115, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5880398750305176, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.7009981274604797, + "num_tokens": 543458488.0, + "step": 21002 + }, + { + "epoch": 2.306501207994729, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.939897298812866, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7191194891929626, + "num_tokens": 543481233.0, + "step": 21003 + }, + { + "epoch": 2.306611025697342, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8744490146636963, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7060319781303406, + "num_tokens": 543504692.0, + "step": 21004 + }, + { + "epoch": 2.306720843399956, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9480698108673096, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7140072584152222, + "num_tokens": 543526621.0, + "step": 21005 + }, + { + "epoch": 2.3068306611025697, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.443331480026245, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7059707641601562, + "num_tokens": 543554414.0, + "step": 21006 + }, + { + "epoch": 2.3069404788051835, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.341479539871216, + "learning_rate": 1e-06, + "loss": 1.0855, + "mean_token_accuracy": 0.686569333076477, + "num_tokens": 543587486.0, + "step": 21007 + }, + { + "epoch": 2.3070502965077972, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4916021823883057, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6918361186981201, + "num_tokens": 543615678.0, + "step": 21008 + }, + { + "epoch": 2.3071601142104106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.700486898422241, + "learning_rate": 1e-06, + "loss": 0.9042, + "mean_token_accuracy": 0.7318325638771057, + "num_tokens": 543639927.0, + "step": 21009 + }, + { + "epoch": 2.3072699319130243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6339492797851562, + "learning_rate": 1e-06, + "loss": 1.0524, + "mean_token_accuracy": 0.6912025809288025, + "num_tokens": 543668013.0, + "step": 21010 + }, + { + "epoch": 2.307379749615638, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8418049812316895, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7255293726921082, + "num_tokens": 543692129.0, + "step": 21011 + }, + { + "epoch": 2.307489567318252, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.776174545288086, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7220292091369629, + "num_tokens": 543714848.0, + "step": 21012 + }, + { + "epoch": 2.3075993850208656, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.675654172897339, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7049423456192017, + "num_tokens": 543739756.0, + "step": 21013 + }, + { + "epoch": 2.307709202723479, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.116950750350952, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7037942409515381, + "num_tokens": 543761497.0, + "step": 21014 + }, + { + "epoch": 2.3078190204260927, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8514509201049805, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.708861231803894, + "num_tokens": 543783793.0, + "step": 21015 + }, + { + "epoch": 2.3079288381287064, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1370630264282227, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6992865800857544, + "num_tokens": 543809473.0, + "step": 21016 + }, + { + "epoch": 2.30803865583132, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.848616361618042, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7130757570266724, + "num_tokens": 543832119.0, + "step": 21017 + }, + { + "epoch": 2.3081484735339335, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.656845808029175, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7018131613731384, + "num_tokens": 543858050.0, + "step": 21018 + }, + { + "epoch": 2.3082582912365472, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6349411010742188, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7094242572784424, + "num_tokens": 543883102.0, + "step": 21019 + }, + { + "epoch": 2.308368108939161, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.764976978302002, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.7283889055252075, + "num_tokens": 543904191.0, + "step": 21020 + }, + { + "epoch": 2.3084779266417748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8807373046875, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7118606567382812, + "num_tokens": 543925282.0, + "step": 21021 + }, + { + "epoch": 2.308587744344388, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7543182373046875, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7029050588607788, + "num_tokens": 543948199.0, + "step": 21022 + }, + { + "epoch": 2.308697562047002, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.268354654312134, + "learning_rate": 1e-06, + "loss": 0.9102, + "mean_token_accuracy": 0.7334077954292297, + "num_tokens": 543978822.0, + "step": 21023 + }, + { + "epoch": 2.3088073797496156, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.458080530166626, + "learning_rate": 1e-06, + "loss": 1.074, + "mean_token_accuracy": 0.6882659792900085, + "num_tokens": 544009634.0, + "step": 21024 + }, + { + "epoch": 2.3089171974522293, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2465641498565674, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6941825151443481, + "num_tokens": 544045825.0, + "step": 21025 + }, + { + "epoch": 2.309027015154843, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.693199872970581, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7167367339134216, + "num_tokens": 544066655.0, + "step": 21026 + }, + { + "epoch": 2.3091368328574564, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.464974880218506, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7002871632575989, + "num_tokens": 544088602.0, + "step": 21027 + }, + { + "epoch": 2.30924665056007, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.67535662651062, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7199574708938599, + "num_tokens": 544111969.0, + "step": 21028 + }, + { + "epoch": 2.309356468262684, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.694552183151245, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.7274158000946045, + "num_tokens": 544136419.0, + "step": 21029 + }, + { + "epoch": 2.3094662859652977, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.335909128189087, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.7016010284423828, + "num_tokens": 544167652.0, + "step": 21030 + }, + { + "epoch": 2.3095761036679114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.541844129562378, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7033352255821228, + "num_tokens": 544196676.0, + "step": 21031 + }, + { + "epoch": 2.3096859213705248, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.580488920211792, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7134159803390503, + "num_tokens": 544222160.0, + "step": 21032 + }, + { + "epoch": 2.3097957390731385, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5434086322784424, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.6940048933029175, + "num_tokens": 544250547.0, + "step": 21033 + }, + { + "epoch": 2.3099055567757523, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.652036190032959, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6931911706924438, + "num_tokens": 544275216.0, + "step": 21034 + }, + { + "epoch": 2.310015374478366, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.416252851486206, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7063794732093811, + "num_tokens": 544302957.0, + "step": 21035 + }, + { + "epoch": 2.31012519218098, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6149189472198486, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6973593235015869, + "num_tokens": 544330432.0, + "step": 21036 + }, + { + "epoch": 2.310235009883593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5576331615448, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7102487087249756, + "num_tokens": 544358067.0, + "step": 21037 + }, + { + "epoch": 2.310344827586207, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.846280574798584, + "learning_rate": 1e-06, + "loss": 0.795, + "mean_token_accuracy": 0.763603925704956, + "num_tokens": 544378540.0, + "step": 21038 + }, + { + "epoch": 2.3104546452888206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.652559280395508, + "learning_rate": 1e-06, + "loss": 0.9117, + "mean_token_accuracy": 0.7373524904251099, + "num_tokens": 544402832.0, + "step": 21039 + }, + { + "epoch": 2.3105644629914344, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.769489049911499, + "learning_rate": 1e-06, + "loss": 0.9203, + "mean_token_accuracy": 0.7334482669830322, + "num_tokens": 544424797.0, + "step": 21040 + }, + { + "epoch": 2.310674280694048, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.512624502182007, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.7025871276855469, + "num_tokens": 544457150.0, + "step": 21041 + }, + { + "epoch": 2.3107840983966614, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3302719593048096, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7111039161682129, + "num_tokens": 544488923.0, + "step": 21042 + }, + { + "epoch": 2.310893916099275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3478453159332275, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7131084203720093, + "num_tokens": 544518724.0, + "step": 21043 + }, + { + "epoch": 2.311003733801889, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.303227424621582, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7181273698806763, + "num_tokens": 544551721.0, + "step": 21044 + }, + { + "epoch": 2.3111135515045027, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.81099009513855, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6957919001579285, + "num_tokens": 544575492.0, + "step": 21045 + }, + { + "epoch": 2.311223369207116, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.535595417022705, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.6998854875564575, + "num_tokens": 544602952.0, + "step": 21046 + }, + { + "epoch": 2.3113331869097298, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5288357734680176, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7249152660369873, + "num_tokens": 544628679.0, + "step": 21047 + }, + { + "epoch": 2.3114430046123435, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545182704925537, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7108453512191772, + "num_tokens": 544655406.0, + "step": 21048 + }, + { + "epoch": 2.3115528223149573, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8253591060638428, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.6974416971206665, + "num_tokens": 544676038.0, + "step": 21049 + }, + { + "epoch": 2.3116626400175706, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.943737745285034, + "learning_rate": 1e-06, + "loss": 0.9219, + "mean_token_accuracy": 0.7313489317893982, + "num_tokens": 544696723.0, + "step": 21050 + }, + { + "epoch": 2.3117724577201844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7348077297210693, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7169608473777771, + "num_tokens": 544720954.0, + "step": 21051 + }, + { + "epoch": 2.311882275422798, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.932107448577881, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7057543992996216, + "num_tokens": 544748396.0, + "step": 21052 + }, + { + "epoch": 2.311992093125412, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7678873538970947, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.719191312789917, + "num_tokens": 544771867.0, + "step": 21053 + }, + { + "epoch": 2.3121019108280256, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6354923248291016, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7251685261726379, + "num_tokens": 544794970.0, + "step": 21054 + }, + { + "epoch": 2.312211728530639, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5299954414367676, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7243737578392029, + "num_tokens": 544821841.0, + "step": 21055 + }, + { + "epoch": 2.3123215462332527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.48598051071167, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.703723669052124, + "num_tokens": 544849110.0, + "step": 21056 + }, + { + "epoch": 2.3124313639358665, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5363996028900146, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.7042891979217529, + "num_tokens": 544876829.0, + "step": 21057 + }, + { + "epoch": 2.31254118163848, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5100176334381104, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7242053747177124, + "num_tokens": 544904983.0, + "step": 21058 + }, + { + "epoch": 2.312650999341094, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9691390991210938, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6956359148025513, + "num_tokens": 544926799.0, + "step": 21059 + }, + { + "epoch": 2.3127608170437073, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.527981996536255, + "learning_rate": 1e-06, + "loss": 1.1216, + "mean_token_accuracy": 0.6851043105125427, + "num_tokens": 544956576.0, + "step": 21060 + }, + { + "epoch": 2.312870634746321, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7684719562530518, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7296881675720215, + "num_tokens": 544977466.0, + "step": 21061 + }, + { + "epoch": 2.312980452448935, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6011221408843994, + "learning_rate": 1e-06, + "loss": 1.0946, + "mean_token_accuracy": 0.6904098987579346, + "num_tokens": 545006051.0, + "step": 21062 + }, + { + "epoch": 2.3130902701515486, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6110363006591797, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6940751075744629, + "num_tokens": 545032919.0, + "step": 21063 + }, + { + "epoch": 2.3132000878541623, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.556428909301758, + "learning_rate": 1e-06, + "loss": 1.0319, + "mean_token_accuracy": 0.70562744140625, + "num_tokens": 545058743.0, + "step": 21064 + }, + { + "epoch": 2.3133099055567756, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7343406677246094, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.7235133051872253, + "num_tokens": 545080941.0, + "step": 21065 + }, + { + "epoch": 2.3134197232593894, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4751553535461426, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.6979712247848511, + "num_tokens": 545111743.0, + "step": 21066 + }, + { + "epoch": 2.313529540962003, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6920595169067383, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.6954916715621948, + "num_tokens": 545137372.0, + "step": 21067 + }, + { + "epoch": 2.313639358664617, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8043651580810547, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7032797932624817, + "num_tokens": 545161184.0, + "step": 21068 + }, + { + "epoch": 2.31374917636723, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.854416847229004, + "learning_rate": 1e-06, + "loss": 0.882, + "mean_token_accuracy": 0.7419109344482422, + "num_tokens": 545184304.0, + "step": 21069 + }, + { + "epoch": 2.313858994069844, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.641744375228882, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7010747790336609, + "num_tokens": 545210656.0, + "step": 21070 + }, + { + "epoch": 2.3139688117724577, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6460866928100586, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.7323469519615173, + "num_tokens": 545235013.0, + "step": 21071 + }, + { + "epoch": 2.3140786294750715, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5073137283325195, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7043815851211548, + "num_tokens": 545261614.0, + "step": 21072 + }, + { + "epoch": 2.314188447177685, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4406356811523438, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7113499641418457, + "num_tokens": 545288530.0, + "step": 21073 + }, + { + "epoch": 2.3142982648802986, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.589587450027466, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6945257782936096, + "num_tokens": 545316653.0, + "step": 21074 + }, + { + "epoch": 2.3144080825829123, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5997567176818848, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7103421688079834, + "num_tokens": 545344921.0, + "step": 21075 + }, + { + "epoch": 2.314517900285526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.422567844390869, + "learning_rate": 1e-06, + "loss": 1.0848, + "mean_token_accuracy": 0.6866584420204163, + "num_tokens": 545375523.0, + "step": 21076 + }, + { + "epoch": 2.31462771798814, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.297029495239258, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7083674073219299, + "num_tokens": 545407587.0, + "step": 21077 + }, + { + "epoch": 2.314737535690753, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6412196159362793, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7204620838165283, + "num_tokens": 545432350.0, + "step": 21078 + }, + { + "epoch": 2.314847353393367, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4216647148132324, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7242988348007202, + "num_tokens": 545460827.0, + "step": 21079 + }, + { + "epoch": 2.3149571710959806, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6275246143341064, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7122768759727478, + "num_tokens": 545484607.0, + "step": 21080 + }, + { + "epoch": 2.3150669887985944, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7449657917022705, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7022596597671509, + "num_tokens": 545508854.0, + "step": 21081 + }, + { + "epoch": 2.315176806501208, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.648414373397827, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6819695830345154, + "num_tokens": 545536439.0, + "step": 21082 + }, + { + "epoch": 2.3152866242038215, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.54703950881958, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.70515376329422, + "num_tokens": 545562043.0, + "step": 21083 + }, + { + "epoch": 2.3153964419064352, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.402571678161621, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7110135555267334, + "num_tokens": 545589760.0, + "step": 21084 + }, + { + "epoch": 2.315506259609049, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.563610076904297, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7231079936027527, + "num_tokens": 545614440.0, + "step": 21085 + }, + { + "epoch": 2.3156160773116627, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.533566474914551, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7156257629394531, + "num_tokens": 545639962.0, + "step": 21086 + }, + { + "epoch": 2.3157258950142765, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9903793334960938, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6928479671478271, + "num_tokens": 545663324.0, + "step": 21087 + }, + { + "epoch": 2.31583571271689, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7203755378723145, + "learning_rate": 1e-06, + "loss": 0.9261, + "mean_token_accuracy": 0.7297338843345642, + "num_tokens": 545685668.0, + "step": 21088 + }, + { + "epoch": 2.3159455304195036, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6946754455566406, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.7046064138412476, + "num_tokens": 545709631.0, + "step": 21089 + }, + { + "epoch": 2.3160553481221173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7135846614837646, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7254269123077393, + "num_tokens": 545733222.0, + "step": 21090 + }, + { + "epoch": 2.316165165824731, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5984158515930176, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6971359252929688, + "num_tokens": 545759942.0, + "step": 21091 + }, + { + "epoch": 2.316274983527345, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.613550901412964, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7227891087532043, + "num_tokens": 545785642.0, + "step": 21092 + }, + { + "epoch": 2.316384801229958, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4691314697265625, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7170726656913757, + "num_tokens": 545814865.0, + "step": 21093 + }, + { + "epoch": 2.316494618932572, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6479525566101074, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.7007073163986206, + "num_tokens": 545838748.0, + "step": 21094 + }, + { + "epoch": 2.3166044366351857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.267763376235962, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6956965923309326, + "num_tokens": 545876189.0, + "step": 21095 + }, + { + "epoch": 2.3167142543377994, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.354525089263916, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6947271823883057, + "num_tokens": 545906308.0, + "step": 21096 + }, + { + "epoch": 2.3168240720404127, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6526105403900146, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7139765024185181, + "num_tokens": 545931183.0, + "step": 21097 + }, + { + "epoch": 2.3169338897430265, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.379235029220581, + "learning_rate": 1e-06, + "loss": 0.9359, + "mean_token_accuracy": 0.7284360527992249, + "num_tokens": 545958608.0, + "step": 21098 + }, + { + "epoch": 2.3170437074456403, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.975590467453003, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.703816294670105, + "num_tokens": 545988236.0, + "step": 21099 + }, + { + "epoch": 2.317153525148254, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4282286167144775, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7040868401527405, + "num_tokens": 546016863.0, + "step": 21100 + }, + { + "epoch": 2.3172633428508673, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7081761360168457, + "learning_rate": 1e-06, + "loss": 0.9069, + "mean_token_accuracy": 0.7307999134063721, + "num_tokens": 546038858.0, + "step": 21101 + }, + { + "epoch": 2.317373160553481, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.556715965270996, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7234604358673096, + "num_tokens": 546064374.0, + "step": 21102 + }, + { + "epoch": 2.317482978256095, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4939815998077393, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7202844619750977, + "num_tokens": 546091560.0, + "step": 21103 + }, + { + "epoch": 2.3175927959587086, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.388206958770752, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7138198018074036, + "num_tokens": 546121413.0, + "step": 21104 + }, + { + "epoch": 2.3177026136613224, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.64652156829834, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6869224309921265, + "num_tokens": 546146179.0, + "step": 21105 + }, + { + "epoch": 2.3178124313639357, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.65238618850708, + "learning_rate": 1e-06, + "loss": 0.9552, + "mean_token_accuracy": 0.728491485118866, + "num_tokens": 546169669.0, + "step": 21106 + }, + { + "epoch": 2.3179222490665494, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5622763633728027, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6872662305831909, + "num_tokens": 546196400.0, + "step": 21107 + }, + { + "epoch": 2.318032066769163, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7057251930236816, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.710725724697113, + "num_tokens": 546219794.0, + "step": 21108 + }, + { + "epoch": 2.318141884471777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2142879962921143, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.6937735676765442, + "num_tokens": 546252721.0, + "step": 21109 + }, + { + "epoch": 2.3182517021743907, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.706327199935913, + "learning_rate": 1e-06, + "loss": 0.908, + "mean_token_accuracy": 0.7376248240470886, + "num_tokens": 546275903.0, + "step": 21110 + }, + { + "epoch": 2.318361519877004, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.135563850402832, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7171564698219299, + "num_tokens": 546296258.0, + "step": 21111 + }, + { + "epoch": 2.3184713375796178, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9243383407592773, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7048733234405518, + "num_tokens": 546320414.0, + "step": 21112 + }, + { + "epoch": 2.3185811552822315, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8304495811462402, + "learning_rate": 1e-06, + "loss": 0.911, + "mean_token_accuracy": 0.7307613492012024, + "num_tokens": 546341684.0, + "step": 21113 + }, + { + "epoch": 2.3186909729848453, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9090123176574707, + "learning_rate": 1e-06, + "loss": 0.874, + "mean_token_accuracy": 0.7398656606674194, + "num_tokens": 546362874.0, + "step": 21114 + }, + { + "epoch": 2.318800790687459, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5422310829162598, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6962631940841675, + "num_tokens": 546390052.0, + "step": 21115 + }, + { + "epoch": 2.3189106083900723, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.416706085205078, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.7021557688713074, + "num_tokens": 546419092.0, + "step": 21116 + }, + { + "epoch": 2.319020426092686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8072128295898438, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7071151733398438, + "num_tokens": 546441330.0, + "step": 21117 + }, + { + "epoch": 2.3191302437953, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.024770498275757, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.711289644241333, + "num_tokens": 546462439.0, + "step": 21118 + }, + { + "epoch": 2.3192400614979136, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6880691051483154, + "learning_rate": 1e-06, + "loss": 0.8712, + "mean_token_accuracy": 0.737296462059021, + "num_tokens": 546487977.0, + "step": 21119 + }, + { + "epoch": 2.319349879200527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2100777626037598, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7274590730667114, + "num_tokens": 546505219.0, + "step": 21120 + }, + { + "epoch": 2.3194596969031407, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.495300769805908, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7099043130874634, + "num_tokens": 546533972.0, + "step": 21121 + }, + { + "epoch": 2.3195695146057544, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.635528802871704, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6967360377311707, + "num_tokens": 546559537.0, + "step": 21122 + }, + { + "epoch": 2.319679332308368, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6663265228271484, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.691839337348938, + "num_tokens": 546583319.0, + "step": 21123 + }, + { + "epoch": 2.319789150010982, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.535592555999756, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.7088804841041565, + "num_tokens": 546609774.0, + "step": 21124 + }, + { + "epoch": 2.3198989677135953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.602285861968994, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7094895839691162, + "num_tokens": 546637470.0, + "step": 21125 + }, + { + "epoch": 2.320008785416209, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.561558961868286, + "learning_rate": 1e-06, + "loss": 0.9321, + "mean_token_accuracy": 0.7253246903419495, + "num_tokens": 546663033.0, + "step": 21126 + }, + { + "epoch": 2.320118603118823, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.864414691925049, + "learning_rate": 1e-06, + "loss": 1.1326, + "mean_token_accuracy": 0.6716832518577576, + "num_tokens": 546686081.0, + "step": 21127 + }, + { + "epoch": 2.3202284208214365, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7128891944885254, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7207627296447754, + "num_tokens": 546711171.0, + "step": 21128 + }, + { + "epoch": 2.32033823852405, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.830249786376953, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6923074126243591, + "num_tokens": 546733502.0, + "step": 21129 + }, + { + "epoch": 2.3204480562266636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4275248050689697, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6964526176452637, + "num_tokens": 546762358.0, + "step": 21130 + }, + { + "epoch": 2.3205578739292774, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.707145929336548, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7074833512306213, + "num_tokens": 546787678.0, + "step": 21131 + }, + { + "epoch": 2.320667691631891, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4275381565093994, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.6995286345481873, + "num_tokens": 546818176.0, + "step": 21132 + }, + { + "epoch": 2.320777509334505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.687537908554077, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.710280179977417, + "num_tokens": 546842222.0, + "step": 21133 + }, + { + "epoch": 2.320887327037118, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6739206314086914, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7058649063110352, + "num_tokens": 546866707.0, + "step": 21134 + }, + { + "epoch": 2.320997144739732, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6457629203796387, + "learning_rate": 1e-06, + "loss": 0.9764, + "mean_token_accuracy": 0.7190301418304443, + "num_tokens": 546891770.0, + "step": 21135 + }, + { + "epoch": 2.3211069624423457, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.749851942062378, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7041906118392944, + "num_tokens": 546915119.0, + "step": 21136 + }, + { + "epoch": 2.3212167801449595, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.404154062271118, + "learning_rate": 1e-06, + "loss": 1.0917, + "mean_token_accuracy": 0.6915171146392822, + "num_tokens": 546946013.0, + "step": 21137 + }, + { + "epoch": 2.3213265978475732, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6024160385131836, + "learning_rate": 1e-06, + "loss": 0.9297, + "mean_token_accuracy": 0.7258056998252869, + "num_tokens": 546971462.0, + "step": 21138 + }, + { + "epoch": 2.3214364155501865, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.462101936340332, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7054029107093811, + "num_tokens": 547000981.0, + "step": 21139 + }, + { + "epoch": 2.3215462332528003, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7947747707366943, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7253451347351074, + "num_tokens": 547025222.0, + "step": 21140 + }, + { + "epoch": 2.321656050955414, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7658629417419434, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7049610614776611, + "num_tokens": 547049261.0, + "step": 21141 + }, + { + "epoch": 2.321765868658028, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.742694854736328, + "learning_rate": 1e-06, + "loss": 0.9555, + "mean_token_accuracy": 0.7238836288452148, + "num_tokens": 547071238.0, + "step": 21142 + }, + { + "epoch": 2.3218756863606416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.363168239593506, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.696168839931488, + "num_tokens": 547102348.0, + "step": 21143 + }, + { + "epoch": 2.321985504063255, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584226608276367, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.7024706602096558, + "num_tokens": 547130826.0, + "step": 21144 + }, + { + "epoch": 2.3220953217658686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9999172687530518, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7168668508529663, + "num_tokens": 547152390.0, + "step": 21145 + }, + { + "epoch": 2.3222051394684824, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8348472118377686, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7055736184120178, + "num_tokens": 547176962.0, + "step": 21146 + }, + { + "epoch": 2.322314957171096, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6345176696777344, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7055250406265259, + "num_tokens": 547201818.0, + "step": 21147 + }, + { + "epoch": 2.3224247748737095, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6841440200805664, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7168470621109009, + "num_tokens": 547227966.0, + "step": 21148 + }, + { + "epoch": 2.322534592576323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6651370525360107, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7234474420547485, + "num_tokens": 547254343.0, + "step": 21149 + }, + { + "epoch": 2.322644410278937, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6955623626708984, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7149286866188049, + "num_tokens": 547281730.0, + "step": 21150 + }, + { + "epoch": 2.3227542279815507, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4870197772979736, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7022308707237244, + "num_tokens": 547310001.0, + "step": 21151 + }, + { + "epoch": 2.322864045684164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.854200601577759, + "learning_rate": 1e-06, + "loss": 0.9388, + "mean_token_accuracy": 0.726003885269165, + "num_tokens": 547331369.0, + "step": 21152 + }, + { + "epoch": 2.322973863386778, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.64152193069458, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7086533904075623, + "num_tokens": 547357263.0, + "step": 21153 + }, + { + "epoch": 2.3230836810893916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.50297474861145, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6883841156959534, + "num_tokens": 547386568.0, + "step": 21154 + }, + { + "epoch": 2.3231934987920053, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3331689834594727, + "learning_rate": 1e-06, + "loss": 1.1625, + "mean_token_accuracy": 0.6678251028060913, + "num_tokens": 547422711.0, + "step": 21155 + }, + { + "epoch": 2.323303316494619, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4694783687591553, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.6898905038833618, + "num_tokens": 547448999.0, + "step": 21156 + }, + { + "epoch": 2.3234131341972324, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8520877361297607, + "learning_rate": 1e-06, + "loss": 0.9885, + "mean_token_accuracy": 0.7095056176185608, + "num_tokens": 547473272.0, + "step": 21157 + }, + { + "epoch": 2.323522951899846, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.04966402053833, + "learning_rate": 1e-06, + "loss": 0.9844, + "mean_token_accuracy": 0.7120948433876038, + "num_tokens": 547492875.0, + "step": 21158 + }, + { + "epoch": 2.32363276960246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.357398748397827, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6953630447387695, + "num_tokens": 547526202.0, + "step": 21159 + }, + { + "epoch": 2.3237425873050737, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7590668201446533, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7072923183441162, + "num_tokens": 547551740.0, + "step": 21160 + }, + { + "epoch": 2.3238524050076874, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.725196123123169, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.696701169013977, + "num_tokens": 547576042.0, + "step": 21161 + }, + { + "epoch": 2.3239622227103007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.459591865539551, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.6971126198768616, + "num_tokens": 547605542.0, + "step": 21162 + }, + { + "epoch": 2.3240720404129145, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2969915866851807, + "learning_rate": 1e-06, + "loss": 1.1045, + "mean_token_accuracy": 0.6831998229026794, + "num_tokens": 547636849.0, + "step": 21163 + }, + { + "epoch": 2.3241818581155282, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.702718734741211, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.6950925588607788, + "num_tokens": 547661953.0, + "step": 21164 + }, + { + "epoch": 2.324291675818142, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.690643310546875, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7199457883834839, + "num_tokens": 547685654.0, + "step": 21165 + }, + { + "epoch": 2.3244014935207558, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6120381355285645, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6935962438583374, + "num_tokens": 547712398.0, + "step": 21166 + }, + { + "epoch": 2.324511311223369, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.745861768722534, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7111092805862427, + "num_tokens": 547737811.0, + "step": 21167 + }, + { + "epoch": 2.324621128925983, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6981446743011475, + "learning_rate": 1e-06, + "loss": 1.0418, + "mean_token_accuracy": 0.7077709436416626, + "num_tokens": 547762265.0, + "step": 21168 + }, + { + "epoch": 2.3247309466285966, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.879978895187378, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7168611288070679, + "num_tokens": 547783120.0, + "step": 21169 + }, + { + "epoch": 2.3248407643312103, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565436840057373, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6915591955184937, + "num_tokens": 547814700.0, + "step": 21170 + }, + { + "epoch": 2.324950582033824, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.538907051086426, + "learning_rate": 1e-06, + "loss": 1.0922, + "mean_token_accuracy": 0.6906662583351135, + "num_tokens": 547842383.0, + "step": 21171 + }, + { + "epoch": 2.3250603997364374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.618290662765503, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.7001451849937439, + "num_tokens": 547868389.0, + "step": 21172 + }, + { + "epoch": 2.325170217439051, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.322772979736328, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.6988336443901062, + "num_tokens": 547898639.0, + "step": 21173 + }, + { + "epoch": 2.325280035141665, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.739532470703125, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.713651180267334, + "num_tokens": 547925825.0, + "step": 21174 + }, + { + "epoch": 2.3253898528442787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7887461185455322, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7084989547729492, + "num_tokens": 547949645.0, + "step": 21175 + }, + { + "epoch": 2.325499670546892, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.664402484893799, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6962136030197144, + "num_tokens": 547974075.0, + "step": 21176 + }, + { + "epoch": 2.3256094882495058, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5761282444000244, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7098586559295654, + "num_tokens": 548000066.0, + "step": 21177 + }, + { + "epoch": 2.3257193059521195, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.455653190612793, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.6910766363143921, + "num_tokens": 548028726.0, + "step": 21178 + }, + { + "epoch": 2.3258291236547333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.919140577316284, + "learning_rate": 1e-06, + "loss": 0.861, + "mean_token_accuracy": 0.735702633857727, + "num_tokens": 548050967.0, + "step": 21179 + }, + { + "epoch": 2.3259389413573466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.691929578781128, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7068474292755127, + "num_tokens": 548075738.0, + "step": 21180 + }, + { + "epoch": 2.3260487590599603, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.743771553039551, + "learning_rate": 1e-06, + "loss": 0.8701, + "mean_token_accuracy": 0.7430076599121094, + "num_tokens": 548097259.0, + "step": 21181 + }, + { + "epoch": 2.326158576762574, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4262518882751465, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.694240391254425, + "num_tokens": 548128313.0, + "step": 21182 + }, + { + "epoch": 2.326268394465188, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9200563430786133, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.700284481048584, + "num_tokens": 548155143.0, + "step": 21183 + }, + { + "epoch": 2.3263782121678016, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1612279415130615, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7123305797576904, + "num_tokens": 548176848.0, + "step": 21184 + }, + { + "epoch": 2.326488029870415, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.977945566177368, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7228584885597229, + "num_tokens": 548196503.0, + "step": 21185 + }, + { + "epoch": 2.3265978475730287, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.050086498260498, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.7199789881706238, + "num_tokens": 548214433.0, + "step": 21186 + }, + { + "epoch": 2.3267076652756424, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7109158039093018, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.693485677242279, + "num_tokens": 548242208.0, + "step": 21187 + }, + { + "epoch": 2.326817482978256, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.510183095932007, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7081530094146729, + "num_tokens": 548267871.0, + "step": 21188 + }, + { + "epoch": 2.32692730068087, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.171252965927124, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7144188284873962, + "num_tokens": 548285796.0, + "step": 21189 + }, + { + "epoch": 2.3270371183834833, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5490705966949463, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7050435543060303, + "num_tokens": 548314029.0, + "step": 21190 + }, + { + "epoch": 2.327146936086097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7474780082702637, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7339993715286255, + "num_tokens": 548339195.0, + "step": 21191 + }, + { + "epoch": 2.3272567537887108, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5627288818359375, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7034654021263123, + "num_tokens": 548363915.0, + "step": 21192 + }, + { + "epoch": 2.3273665714913245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3964483737945557, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7027905583381653, + "num_tokens": 548392975.0, + "step": 21193 + }, + { + "epoch": 2.3274763891939383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4598770141601562, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6838514804840088, + "num_tokens": 548423278.0, + "step": 21194 + }, + { + "epoch": 2.3275862068965516, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.461134195327759, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7099261283874512, + "num_tokens": 548451510.0, + "step": 21195 + }, + { + "epoch": 2.3276960245991654, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6231563091278076, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.716410219669342, + "num_tokens": 548477218.0, + "step": 21196 + }, + { + "epoch": 2.327805842301779, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7098658084869385, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7264591455459595, + "num_tokens": 548503026.0, + "step": 21197 + }, + { + "epoch": 2.327915660004393, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0522279739379883, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7146927118301392, + "num_tokens": 548525537.0, + "step": 21198 + }, + { + "epoch": 2.328025477707006, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7536683082580566, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6984540820121765, + "num_tokens": 548554097.0, + "step": 21199 + }, + { + "epoch": 2.32813529540962, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559255361557007, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.7122831344604492, + "num_tokens": 548578518.0, + "step": 21200 + }, + { + "epoch": 2.3282451131122337, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.712125062942505, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7083988189697266, + "num_tokens": 548603344.0, + "step": 21201 + }, + { + "epoch": 2.3283549308148475, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.569490909576416, + "learning_rate": 1e-06, + "loss": 0.9159, + "mean_token_accuracy": 0.7262662649154663, + "num_tokens": 548627826.0, + "step": 21202 + }, + { + "epoch": 2.3284647485174608, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.782608985900879, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7210026979446411, + "num_tokens": 548652117.0, + "step": 21203 + }, + { + "epoch": 2.3285745662200745, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6217141151428223, + "learning_rate": 1e-06, + "loss": 0.9785, + "mean_token_accuracy": 0.7244991660118103, + "num_tokens": 548677237.0, + "step": 21204 + }, + { + "epoch": 2.3286843839226883, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.392207622528076, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7049061059951782, + "num_tokens": 548709854.0, + "step": 21205 + }, + { + "epoch": 2.328794201625302, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.607725143432617, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.6983626484870911, + "num_tokens": 548736894.0, + "step": 21206 + }, + { + "epoch": 2.328904019327916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9854249954223633, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7213425636291504, + "num_tokens": 548758761.0, + "step": 21207 + }, + { + "epoch": 2.329013837030529, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.20668888092041, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.7016289234161377, + "num_tokens": 548783443.0, + "step": 21208 + }, + { + "epoch": 2.329123654733143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.632730484008789, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6951555013656616, + "num_tokens": 548809750.0, + "step": 21209 + }, + { + "epoch": 2.3292334724357566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.433995246887207, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7041435837745667, + "num_tokens": 548839854.0, + "step": 21210 + }, + { + "epoch": 2.3293432901383704, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.642702102661133, + "learning_rate": 1e-06, + "loss": 0.9103, + "mean_token_accuracy": 0.7333424091339111, + "num_tokens": 548864134.0, + "step": 21211 + }, + { + "epoch": 2.329453107840984, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.533396005630493, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6850165724754333, + "num_tokens": 548893718.0, + "step": 21212 + }, + { + "epoch": 2.3295629255435975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6137311458587646, + "learning_rate": 1e-06, + "loss": 1.0757, + "mean_token_accuracy": 0.6950175762176514, + "num_tokens": 548920479.0, + "step": 21213 + }, + { + "epoch": 2.329672743246211, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0881245136260986, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.6998400092124939, + "num_tokens": 548944177.0, + "step": 21214 + }, + { + "epoch": 2.329782560948825, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1073038578033447, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6917670965194702, + "num_tokens": 548969981.0, + "step": 21215 + }, + { + "epoch": 2.3298923786514387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.4642751216888428, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6933196187019348, + "num_tokens": 548991299.0, + "step": 21216 + }, + { + "epoch": 2.3300021963540525, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7945098876953125, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.711216151714325, + "num_tokens": 549013991.0, + "step": 21217 + }, + { + "epoch": 2.330112014056666, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6655642986297607, + "learning_rate": 1e-06, + "loss": 0.8907, + "mean_token_accuracy": 0.7389410138130188, + "num_tokens": 549038032.0, + "step": 21218 + }, + { + "epoch": 2.3302218317592795, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.749764919281006, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6814984679222107, + "num_tokens": 549064922.0, + "step": 21219 + }, + { + "epoch": 2.3303316494618933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3135156631469727, + "learning_rate": 1e-06, + "loss": 1.1207, + "mean_token_accuracy": 0.6799054145812988, + "num_tokens": 549100841.0, + "step": 21220 + }, + { + "epoch": 2.330441467164507, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.716914176940918, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.6962131261825562, + "num_tokens": 549125312.0, + "step": 21221 + }, + { + "epoch": 2.330551284867121, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.894544839859009, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7155749201774597, + "num_tokens": 549144977.0, + "step": 21222 + }, + { + "epoch": 2.330661102569734, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.467927932739258, + "learning_rate": 1e-06, + "loss": 1.1336, + "mean_token_accuracy": 0.6811860203742981, + "num_tokens": 549178168.0, + "step": 21223 + }, + { + "epoch": 2.330770920272348, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6551034450531006, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7128878235816956, + "num_tokens": 549203052.0, + "step": 21224 + }, + { + "epoch": 2.3308807379749616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.831076145172119, + "learning_rate": 1e-06, + "loss": 0.9205, + "mean_token_accuracy": 0.7305639982223511, + "num_tokens": 549226087.0, + "step": 21225 + }, + { + "epoch": 2.3309905556775754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6524078845977783, + "learning_rate": 1e-06, + "loss": 0.9035, + "mean_token_accuracy": 0.7309978604316711, + "num_tokens": 549251485.0, + "step": 21226 + }, + { + "epoch": 2.3311003733801887, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.770692825317383, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.6948513984680176, + "num_tokens": 549280757.0, + "step": 21227 + }, + { + "epoch": 2.3312101910828025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4790751934051514, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.7054076790809631, + "num_tokens": 549311095.0, + "step": 21228 + }, + { + "epoch": 2.3313200087854162, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5053272247314453, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.7139809131622314, + "num_tokens": 549336923.0, + "step": 21229 + }, + { + "epoch": 2.33142982648803, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6409990787506104, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6951522827148438, + "num_tokens": 549367414.0, + "step": 21230 + }, + { + "epoch": 2.3315396441906433, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8029589653015137, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7156714200973511, + "num_tokens": 549390980.0, + "step": 21231 + }, + { + "epoch": 2.331649461893257, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.680643320083618, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.716718316078186, + "num_tokens": 549415582.0, + "step": 21232 + }, + { + "epoch": 2.331759279595871, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5267701148986816, + "learning_rate": 1e-06, + "loss": 1.107, + "mean_token_accuracy": 0.6869932413101196, + "num_tokens": 549442023.0, + "step": 21233 + }, + { + "epoch": 2.3318690972984846, + "ewc_loss": 1.0788440704345703e-05, + "grad_norm": 3.62788462638855, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7193161845207214, + "num_tokens": 549468923.0, + "step": 21234 + }, + { + "epoch": 2.3319789150010983, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.870236873626709, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.690390408039093, + "num_tokens": 549493366.0, + "step": 21235 + }, + { + "epoch": 2.3320887327037116, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.470898151397705, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7038738131523132, + "num_tokens": 549521336.0, + "step": 21236 + }, + { + "epoch": 2.3321985504063254, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6821587085723877, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6932697892189026, + "num_tokens": 549547172.0, + "step": 21237 + }, + { + "epoch": 2.332308368108939, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1552300453186035, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7286632657051086, + "num_tokens": 549566702.0, + "step": 21238 + }, + { + "epoch": 2.332418185811553, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6439895629882812, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7258302569389343, + "num_tokens": 549591628.0, + "step": 21239 + }, + { + "epoch": 2.3325280035141667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.512603998184204, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6959951519966125, + "num_tokens": 549618880.0, + "step": 21240 + }, + { + "epoch": 2.33263782121678, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2759861946105957, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.7050401568412781, + "num_tokens": 549652810.0, + "step": 21241 + }, + { + "epoch": 2.3327476389193937, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9980177879333496, + "learning_rate": 1e-06, + "loss": 0.854, + "mean_token_accuracy": 0.7478783130645752, + "num_tokens": 549671444.0, + "step": 21242 + }, + { + "epoch": 2.3328574566220075, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.848266839981079, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.7297572493553162, + "num_tokens": 549691769.0, + "step": 21243 + }, + { + "epoch": 2.3329672743246213, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.71968412399292, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.709104597568512, + "num_tokens": 549715034.0, + "step": 21244 + }, + { + "epoch": 2.333077092027235, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536853075027466, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7148847579956055, + "num_tokens": 549740360.0, + "step": 21245 + }, + { + "epoch": 2.3331869097298483, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6322665214538574, + "learning_rate": 1e-06, + "loss": 0.8969, + "mean_token_accuracy": 0.733222484588623, + "num_tokens": 549766285.0, + "step": 21246 + }, + { + "epoch": 2.333296727432462, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4952895641326904, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6881264448165894, + "num_tokens": 549796952.0, + "step": 21247 + }, + { + "epoch": 2.333406545135076, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8536107540130615, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7275611758232117, + "num_tokens": 549816546.0, + "step": 21248 + }, + { + "epoch": 2.3335163628376896, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.655411958694458, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7131562232971191, + "num_tokens": 549840980.0, + "step": 21249 + }, + { + "epoch": 2.333626180540303, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7699482440948486, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6885817050933838, + "num_tokens": 549865581.0, + "step": 21250 + }, + { + "epoch": 2.3337359982429167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6646008491516113, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.7074363827705383, + "num_tokens": 549893505.0, + "step": 21251 + }, + { + "epoch": 2.3338458159455304, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9213674068450928, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.714368462562561, + "num_tokens": 549913166.0, + "step": 21252 + }, + { + "epoch": 2.333955633648144, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.754554271697998, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7089396119117737, + "num_tokens": 549936725.0, + "step": 21253 + }, + { + "epoch": 2.334065451350758, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.416098117828369, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7119276523590088, + "num_tokens": 549965484.0, + "step": 21254 + }, + { + "epoch": 2.3341752690533712, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584836959838867, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7113139033317566, + "num_tokens": 549990882.0, + "step": 21255 + }, + { + "epoch": 2.334285086755985, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.52384352684021, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6853336095809937, + "num_tokens": 550019734.0, + "step": 21256 + }, + { + "epoch": 2.3343949044585988, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.772590160369873, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.7234987020492554, + "num_tokens": 550040980.0, + "step": 21257 + }, + { + "epoch": 2.3345047221612125, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6386654376983643, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.6969268321990967, + "num_tokens": 550065783.0, + "step": 21258 + }, + { + "epoch": 2.334614539863826, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5839476585388184, + "learning_rate": 1e-06, + "loss": 0.9108, + "mean_token_accuracy": 0.7292078137397766, + "num_tokens": 550094010.0, + "step": 21259 + }, + { + "epoch": 2.3347243575664396, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.016864538192749, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7007016539573669, + "num_tokens": 550114344.0, + "step": 21260 + }, + { + "epoch": 2.3348341752690533, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8470211029052734, + "learning_rate": 1e-06, + "loss": 1.0204, + "mean_token_accuracy": 0.7024649977684021, + "num_tokens": 550142710.0, + "step": 21261 + }, + { + "epoch": 2.334943992971667, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.579171895980835, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.7165133953094482, + "num_tokens": 550168823.0, + "step": 21262 + }, + { + "epoch": 2.335053810674281, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.602973699569702, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7095910310745239, + "num_tokens": 550193547.0, + "step": 21263 + }, + { + "epoch": 2.335163628376894, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5492749214172363, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7091199159622192, + "num_tokens": 550219685.0, + "step": 21264 + }, + { + "epoch": 2.335273446079508, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.535186290740967, + "learning_rate": 1e-06, + "loss": 1.0853, + "mean_token_accuracy": 0.6920067667961121, + "num_tokens": 550248353.0, + "step": 21265 + }, + { + "epoch": 2.3353832637821217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5843422412872314, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7210325002670288, + "num_tokens": 550278798.0, + "step": 21266 + }, + { + "epoch": 2.3354930814847354, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6099328994750977, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.6930551528930664, + "num_tokens": 550306111.0, + "step": 21267 + }, + { + "epoch": 2.335602899187349, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.495198965072632, + "learning_rate": 1e-06, + "loss": 0.9491, + "mean_token_accuracy": 0.7208591103553772, + "num_tokens": 550333461.0, + "step": 21268 + }, + { + "epoch": 2.3357127168899625, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8678395748138428, + "learning_rate": 1e-06, + "loss": 0.8092, + "mean_token_accuracy": 0.7593343257904053, + "num_tokens": 550354430.0, + "step": 21269 + }, + { + "epoch": 2.3358225345925763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.714616060256958, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7057551741600037, + "num_tokens": 550379375.0, + "step": 21270 + }, + { + "epoch": 2.33593235229519, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5537095069885254, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6965266466140747, + "num_tokens": 550408939.0, + "step": 21271 + }, + { + "epoch": 2.336042169997804, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9568216800689697, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.714525043964386, + "num_tokens": 550430170.0, + "step": 21272 + }, + { + "epoch": 2.3361519877004175, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.800610065460205, + "learning_rate": 1e-06, + "loss": 0.9442, + "mean_token_accuracy": 0.7302467823028564, + "num_tokens": 550453836.0, + "step": 21273 + }, + { + "epoch": 2.336261805403031, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3716299533843994, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6998112201690674, + "num_tokens": 550485340.0, + "step": 21274 + }, + { + "epoch": 2.3363716231056446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4174423217773438, + "learning_rate": 1e-06, + "loss": 1.1184, + "mean_token_accuracy": 0.6830108165740967, + "num_tokens": 550520121.0, + "step": 21275 + }, + { + "epoch": 2.3364814408082584, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.955836057662964, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7157189846038818, + "num_tokens": 550545685.0, + "step": 21276 + }, + { + "epoch": 2.336591258510872, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.574636697769165, + "learning_rate": 1e-06, + "loss": 1.1386, + "mean_token_accuracy": 0.6744176149368286, + "num_tokens": 550576946.0, + "step": 21277 + }, + { + "epoch": 2.3367010762134854, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7589874267578125, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7280110120773315, + "num_tokens": 550598771.0, + "step": 21278 + }, + { + "epoch": 2.336810893916099, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5498714447021484, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.7003005743026733, + "num_tokens": 550625923.0, + "step": 21279 + }, + { + "epoch": 2.336920711618713, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.553255319595337, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.7017781138420105, + "num_tokens": 550654256.0, + "step": 21280 + }, + { + "epoch": 2.3370305293213267, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.502814531326294, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6952152252197266, + "num_tokens": 550684929.0, + "step": 21281 + }, + { + "epoch": 2.33714034702394, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4775447845458984, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.70005202293396, + "num_tokens": 550712460.0, + "step": 21282 + }, + { + "epoch": 2.337250164726554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.178781032562256, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7085430026054382, + "num_tokens": 550740829.0, + "step": 21283 + }, + { + "epoch": 2.3373599824291675, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6305079460144043, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7088410258293152, + "num_tokens": 550767310.0, + "step": 21284 + }, + { + "epoch": 2.3374698001317813, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545452356338501, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.721522867679596, + "num_tokens": 550793297.0, + "step": 21285 + }, + { + "epoch": 2.337579617834395, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.791659355163574, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7132225036621094, + "num_tokens": 550814235.0, + "step": 21286 + }, + { + "epoch": 2.3376894355370084, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5656208992004395, + "learning_rate": 1e-06, + "loss": 0.9309, + "mean_token_accuracy": 0.7339372038841248, + "num_tokens": 550838545.0, + "step": 21287 + }, + { + "epoch": 2.337799253239622, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.885598659515381, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7035335898399353, + "num_tokens": 550861019.0, + "step": 21288 + }, + { + "epoch": 2.337909070942236, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.795819044113159, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7081599831581116, + "num_tokens": 550884329.0, + "step": 21289 + }, + { + "epoch": 2.3380188886448496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.772679328918457, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6968311071395874, + "num_tokens": 550914916.0, + "step": 21290 + }, + { + "epoch": 2.3381287063474634, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8819124698638916, + "learning_rate": 1e-06, + "loss": 0.9597, + "mean_token_accuracy": 0.7186353206634521, + "num_tokens": 550937653.0, + "step": 21291 + }, + { + "epoch": 2.3382385240500767, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5741758346557617, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6977189183235168, + "num_tokens": 550964748.0, + "step": 21292 + }, + { + "epoch": 2.3383483417526905, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6777241230010986, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7120379209518433, + "num_tokens": 550988746.0, + "step": 21293 + }, + { + "epoch": 2.338458159455304, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5947511196136475, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.722037672996521, + "num_tokens": 551012690.0, + "step": 21294 + }, + { + "epoch": 2.338567977157918, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4126250743865967, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7138298749923706, + "num_tokens": 551046199.0, + "step": 21295 + }, + { + "epoch": 2.3386777948605317, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5805490016937256, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6986358165740967, + "num_tokens": 551072392.0, + "step": 21296 + }, + { + "epoch": 2.338787612563145, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5694432258605957, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.6958414316177368, + "num_tokens": 551096181.0, + "step": 21297 + }, + { + "epoch": 2.338897430265759, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.517239809036255, + "learning_rate": 1e-06, + "loss": 1.1322, + "mean_token_accuracy": 0.6770464181900024, + "num_tokens": 551126784.0, + "step": 21298 + }, + { + "epoch": 2.3390072479683726, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4613866806030273, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6956122517585754, + "num_tokens": 551157975.0, + "step": 21299 + }, + { + "epoch": 2.3391170656709863, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4027180671691895, + "learning_rate": 1e-06, + "loss": 1.0469, + "mean_token_accuracy": 0.6941730976104736, + "num_tokens": 551188296.0, + "step": 21300 + }, + { + "epoch": 2.3392268833735996, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8089840412139893, + "learning_rate": 1e-06, + "loss": 0.9072, + "mean_token_accuracy": 0.7262108325958252, + "num_tokens": 551209808.0, + "step": 21301 + }, + { + "epoch": 2.3393367010762134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6503477096557617, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7090223431587219, + "num_tokens": 551237982.0, + "step": 21302 + }, + { + "epoch": 2.339446518778827, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5677943229675293, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7141784429550171, + "num_tokens": 551263064.0, + "step": 21303 + }, + { + "epoch": 2.339556336481441, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8455233573913574, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7164510488510132, + "num_tokens": 551290330.0, + "step": 21304 + }, + { + "epoch": 2.3396661541840547, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4079105854034424, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7200671434402466, + "num_tokens": 551324588.0, + "step": 21305 + }, + { + "epoch": 2.339775971886668, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.812450885772705, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7084810733795166, + "num_tokens": 551346696.0, + "step": 21306 + }, + { + "epoch": 2.3398857895892817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5193474292755127, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6858499050140381, + "num_tokens": 551374637.0, + "step": 21307 + }, + { + "epoch": 2.3399956072918955, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3947784900665283, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7027561664581299, + "num_tokens": 551402293.0, + "step": 21308 + }, + { + "epoch": 2.3401054249945092, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.629671573638916, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.695568323135376, + "num_tokens": 551430867.0, + "step": 21309 + }, + { + "epoch": 2.3402152426971226, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.620568037033081, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.690697193145752, + "num_tokens": 551456102.0, + "step": 21310 + }, + { + "epoch": 2.3403250603997363, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8245463371276855, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7058480978012085, + "num_tokens": 551479576.0, + "step": 21311 + }, + { + "epoch": 2.34043487810235, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.415505886077881, + "learning_rate": 1e-06, + "loss": 0.8904, + "mean_token_accuracy": 0.7430588603019714, + "num_tokens": 551508248.0, + "step": 21312 + }, + { + "epoch": 2.340544695804964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8382163047790527, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7043308019638062, + "num_tokens": 551529376.0, + "step": 21313 + }, + { + "epoch": 2.3406545135075776, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5389163494110107, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.7147151231765747, + "num_tokens": 551554610.0, + "step": 21314 + }, + { + "epoch": 2.340764331210191, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.994418144226074, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7088651657104492, + "num_tokens": 551579653.0, + "step": 21315 + }, + { + "epoch": 2.3408741489128047, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6515231132507324, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7161832451820374, + "num_tokens": 551603025.0, + "step": 21316 + }, + { + "epoch": 2.3409839666154184, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 17.05360221862793, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7071981430053711, + "num_tokens": 551631007.0, + "step": 21317 + }, + { + "epoch": 2.341093784318032, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.446193218231201, + "learning_rate": 1e-06, + "loss": 1.1237, + "mean_token_accuracy": 0.6742968559265137, + "num_tokens": 551662782.0, + "step": 21318 + }, + { + "epoch": 2.341203602020646, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0976009368896484, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7068345546722412, + "num_tokens": 551683837.0, + "step": 21319 + }, + { + "epoch": 2.3413134197232592, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.724174737930298, + "learning_rate": 1e-06, + "loss": 0.924, + "mean_token_accuracy": 0.7350099086761475, + "num_tokens": 551705496.0, + "step": 21320 + }, + { + "epoch": 2.341423237425873, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.781733274459839, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7027889490127563, + "num_tokens": 551728394.0, + "step": 21321 + }, + { + "epoch": 2.3415330551284868, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.709259033203125, + "learning_rate": 1e-06, + "loss": 0.9519, + "mean_token_accuracy": 0.7243852615356445, + "num_tokens": 551748347.0, + "step": 21322 + }, + { + "epoch": 2.3416428728311005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7108733654022217, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7207876443862915, + "num_tokens": 551771301.0, + "step": 21323 + }, + { + "epoch": 2.3417526905337143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7380881309509277, + "learning_rate": 1e-06, + "loss": 0.9341, + "mean_token_accuracy": 0.7206910848617554, + "num_tokens": 551791860.0, + "step": 21324 + }, + { + "epoch": 2.3418625082363276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.470590114593506, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.709176778793335, + "num_tokens": 551820556.0, + "step": 21325 + }, + { + "epoch": 2.3419723259389413, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5916383266448975, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7188740968704224, + "num_tokens": 551845083.0, + "step": 21326 + }, + { + "epoch": 2.342082143641555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7476987838745117, + "learning_rate": 1e-06, + "loss": 0.8924, + "mean_token_accuracy": 0.7354441285133362, + "num_tokens": 551866820.0, + "step": 21327 + }, + { + "epoch": 2.342191961344169, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.655062198638916, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.71793133020401, + "num_tokens": 551890731.0, + "step": 21328 + }, + { + "epoch": 2.342301779046782, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5185601711273193, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.7037286758422852, + "num_tokens": 551919352.0, + "step": 21329 + }, + { + "epoch": 2.342411596749396, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.723339796066284, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.699996829032898, + "num_tokens": 551946592.0, + "step": 21330 + }, + { + "epoch": 2.3425214144520097, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5692291259765625, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6965774297714233, + "num_tokens": 551972924.0, + "step": 21331 + }, + { + "epoch": 2.3426312321546234, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.825941324234009, + "learning_rate": 1e-06, + "loss": 0.9127, + "mean_token_accuracy": 0.734085202217102, + "num_tokens": 551994666.0, + "step": 21332 + }, + { + "epoch": 2.3427410498572367, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.644866466522217, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6918683648109436, + "num_tokens": 552021422.0, + "step": 21333 + }, + { + "epoch": 2.3428508675598505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.419192314147949, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7024621963500977, + "num_tokens": 552051876.0, + "step": 21334 + }, + { + "epoch": 2.3429606852624643, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.6117732524871826, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7198566198348999, + "num_tokens": 552074281.0, + "step": 21335 + }, + { + "epoch": 2.343070502965078, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.824233293533325, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.712583065032959, + "num_tokens": 552101939.0, + "step": 21336 + }, + { + "epoch": 2.3431803206676918, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4077820777893066, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7145509719848633, + "num_tokens": 552133240.0, + "step": 21337 + }, + { + "epoch": 2.343290138370305, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5401387214660645, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7239788770675659, + "num_tokens": 552160652.0, + "step": 21338 + }, + { + "epoch": 2.343399956072919, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5032756328582764, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.696689784526825, + "num_tokens": 552188326.0, + "step": 21339 + }, + { + "epoch": 2.3435097737755326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6502795219421387, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7283450365066528, + "num_tokens": 552211141.0, + "step": 21340 + }, + { + "epoch": 2.3436195914781464, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.765068531036377, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.7067239284515381, + "num_tokens": 552235120.0, + "step": 21341 + }, + { + "epoch": 2.34372940918076, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7805426120758057, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7207980155944824, + "num_tokens": 552257858.0, + "step": 21342 + }, + { + "epoch": 2.3438392268833734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5054965019226074, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7180736064910889, + "num_tokens": 552286546.0, + "step": 21343 + }, + { + "epoch": 2.343949044585987, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7119765281677246, + "learning_rate": 1e-06, + "loss": 0.9019, + "mean_token_accuracy": 0.7347414493560791, + "num_tokens": 552308667.0, + "step": 21344 + }, + { + "epoch": 2.344058862288601, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5079009532928467, + "learning_rate": 1e-06, + "loss": 1.1318, + "mean_token_accuracy": 0.6692798137664795, + "num_tokens": 552342909.0, + "step": 21345 + }, + { + "epoch": 2.3441686799912147, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5573060512542725, + "learning_rate": 1e-06, + "loss": 0.9124, + "mean_token_accuracy": 0.7282849550247192, + "num_tokens": 552368696.0, + "step": 21346 + }, + { + "epoch": 2.3442784976938285, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.90596604347229, + "learning_rate": 1e-06, + "loss": 1.0797, + "mean_token_accuracy": 0.68649822473526, + "num_tokens": 552395639.0, + "step": 21347 + }, + { + "epoch": 2.3443883153964418, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9964771270751953, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6951768398284912, + "num_tokens": 552420544.0, + "step": 21348 + }, + { + "epoch": 2.3444981330990555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0474236011505127, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7188398838043213, + "num_tokens": 552439038.0, + "step": 21349 + }, + { + "epoch": 2.3446079508016693, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.628396511077881, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7143670320510864, + "num_tokens": 552464579.0, + "step": 21350 + }, + { + "epoch": 2.344717768504283, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7238144874572754, + "learning_rate": 1e-06, + "loss": 1.0338, + "mean_token_accuracy": 0.7026959657669067, + "num_tokens": 552495383.0, + "step": 21351 + }, + { + "epoch": 2.344827586206897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.769897937774658, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7018553614616394, + "num_tokens": 552520115.0, + "step": 21352 + }, + { + "epoch": 2.34493740390951, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7280008792877197, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7142181396484375, + "num_tokens": 552541490.0, + "step": 21353 + }, + { + "epoch": 2.345047221612124, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.684260606765747, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.701233446598053, + "num_tokens": 552571728.0, + "step": 21354 + }, + { + "epoch": 2.3451570393147376, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4863274097442627, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7142499685287476, + "num_tokens": 552599496.0, + "step": 21355 + }, + { + "epoch": 2.3452668570173514, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4943206310272217, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6972897052764893, + "num_tokens": 552627444.0, + "step": 21356 + }, + { + "epoch": 2.3453766747199647, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.71598744392395, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7094192504882812, + "num_tokens": 552651759.0, + "step": 21357 + }, + { + "epoch": 2.3454864924225785, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3536994457244873, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7119210958480835, + "num_tokens": 552684943.0, + "step": 21358 + }, + { + "epoch": 2.345596310125192, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9376168251037598, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7046970725059509, + "num_tokens": 552708085.0, + "step": 21359 + }, + { + "epoch": 2.345706127827806, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5346713066101074, + "learning_rate": 1e-06, + "loss": 1.0729, + "mean_token_accuracy": 0.6946731805801392, + "num_tokens": 552738251.0, + "step": 21360 + }, + { + "epoch": 2.3458159455304193, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4531233310699463, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7149301767349243, + "num_tokens": 552770263.0, + "step": 21361 + }, + { + "epoch": 2.345925763233033, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4946022033691406, + "learning_rate": 1e-06, + "loss": 1.0986, + "mean_token_accuracy": 0.691472053527832, + "num_tokens": 552799588.0, + "step": 21362 + }, + { + "epoch": 2.346035580935647, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.531985282897949, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7055584192276001, + "num_tokens": 552825885.0, + "step": 21363 + }, + { + "epoch": 2.3461453986382605, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.840651512145996, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7002149820327759, + "num_tokens": 552849550.0, + "step": 21364 + }, + { + "epoch": 2.3462552163408743, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.150768280029297, + "learning_rate": 1e-06, + "loss": 0.9253, + "mean_token_accuracy": 0.7236624360084534, + "num_tokens": 552870271.0, + "step": 21365 + }, + { + "epoch": 2.3463650340434876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4191412925720215, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7209643125534058, + "num_tokens": 552897487.0, + "step": 21366 + }, + { + "epoch": 2.3464748517461014, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5948071479797363, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6916047930717468, + "num_tokens": 552931433.0, + "step": 21367 + }, + { + "epoch": 2.346584669448715, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.113185405731201, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7142077684402466, + "num_tokens": 552952389.0, + "step": 21368 + }, + { + "epoch": 2.346694487151329, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6671485900878906, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7097785472869873, + "num_tokens": 552980525.0, + "step": 21369 + }, + { + "epoch": 2.3468043048539426, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.650707244873047, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7066599130630493, + "num_tokens": 553006900.0, + "step": 21370 + }, + { + "epoch": 2.346914122556556, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.562415599822998, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.698578953742981, + "num_tokens": 553032119.0, + "step": 21371 + }, + { + "epoch": 2.3470239402591697, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9010775089263916, + "learning_rate": 1e-06, + "loss": 0.9247, + "mean_token_accuracy": 0.7311926484107971, + "num_tokens": 553051757.0, + "step": 21372 + }, + { + "epoch": 2.3471337579617835, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.279463291168213, + "learning_rate": 1e-06, + "loss": 0.9067, + "mean_token_accuracy": 0.7314124703407288, + "num_tokens": 553068911.0, + "step": 21373 + }, + { + "epoch": 2.3472435756643972, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7067575454711914, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.7189552783966064, + "num_tokens": 553091522.0, + "step": 21374 + }, + { + "epoch": 2.347353393367011, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7513864040374756, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7009987831115723, + "num_tokens": 553114261.0, + "step": 21375 + }, + { + "epoch": 2.3474632110696243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5795886516571045, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7005932331085205, + "num_tokens": 553138696.0, + "step": 21376 + }, + { + "epoch": 2.347573028772238, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6740849018096924, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6914058923721313, + "num_tokens": 553164347.0, + "step": 21377 + }, + { + "epoch": 2.347682846474852, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3700573444366455, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6865122318267822, + "num_tokens": 553196295.0, + "step": 21378 + }, + { + "epoch": 2.3477926641774656, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.35591197013855, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.6996743083000183, + "num_tokens": 553224688.0, + "step": 21379 + }, + { + "epoch": 2.347902481880079, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.567714214324951, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7102299928665161, + "num_tokens": 553248386.0, + "step": 21380 + }, + { + "epoch": 2.3480122995826926, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5763537883758545, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.695099949836731, + "num_tokens": 553275638.0, + "step": 21381 + }, + { + "epoch": 2.3481221172853064, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.950472354888916, + "learning_rate": 1e-06, + "loss": 0.8728, + "mean_token_accuracy": 0.7328858375549316, + "num_tokens": 553294839.0, + "step": 21382 + }, + { + "epoch": 2.34823193498792, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8597142696380615, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7295154333114624, + "num_tokens": 553316714.0, + "step": 21383 + }, + { + "epoch": 2.3483417526905335, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4265506267547607, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.6987736821174622, + "num_tokens": 553346341.0, + "step": 21384 + }, + { + "epoch": 2.3484515703931472, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8808720111846924, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7010501623153687, + "num_tokens": 553368820.0, + "step": 21385 + }, + { + "epoch": 2.348561388095761, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.824842691421509, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.7293962240219116, + "num_tokens": 553390215.0, + "step": 21386 + }, + { + "epoch": 2.3486712057983747, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5441997051239014, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.7026766538619995, + "num_tokens": 553416577.0, + "step": 21387 + }, + { + "epoch": 2.3487810235009885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.528761386871338, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7242522239685059, + "num_tokens": 553443738.0, + "step": 21388 + }, + { + "epoch": 2.348890841203602, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7736294269561768, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.700106143951416, + "num_tokens": 553465023.0, + "step": 21389 + }, + { + "epoch": 2.3490006589062156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.715790033340454, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6874697208404541, + "num_tokens": 553489967.0, + "step": 21390 + }, + { + "epoch": 2.3491104766088293, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5509183406829834, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.718684196472168, + "num_tokens": 553516728.0, + "step": 21391 + }, + { + "epoch": 2.349220294311443, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.785712718963623, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7197182178497314, + "num_tokens": 553538784.0, + "step": 21392 + }, + { + "epoch": 2.349330112014057, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4385011196136475, + "learning_rate": 1e-06, + "loss": 1.1167, + "mean_token_accuracy": 0.6724883317947388, + "num_tokens": 553568410.0, + "step": 21393 + }, + { + "epoch": 2.34943992971667, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4628090858459473, + "learning_rate": 1e-06, + "loss": 0.9456, + "mean_token_accuracy": 0.7281159162521362, + "num_tokens": 553597423.0, + "step": 21394 + }, + { + "epoch": 2.349549747419284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9464006423950195, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7129555344581604, + "num_tokens": 553619223.0, + "step": 21395 + }, + { + "epoch": 2.3496595651218977, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3649790287017822, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6924554705619812, + "num_tokens": 553648604.0, + "step": 21396 + }, + { + "epoch": 2.3497693828245114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.852956771850586, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7186952829360962, + "num_tokens": 553668209.0, + "step": 21397 + }, + { + "epoch": 2.349879200527125, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6013214588165283, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7098255157470703, + "num_tokens": 553693394.0, + "step": 21398 + }, + { + "epoch": 2.3499890182297385, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8915750980377197, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7094746828079224, + "num_tokens": 553714622.0, + "step": 21399 + }, + { + "epoch": 2.3500988359323522, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.737015962600708, + "learning_rate": 1e-06, + "loss": 0.899, + "mean_token_accuracy": 0.7338458299636841, + "num_tokens": 553738769.0, + "step": 21400 + }, + { + "epoch": 2.350208653634966, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.782991647720337, + "learning_rate": 1e-06, + "loss": 1.099, + "mean_token_accuracy": 0.682282030582428, + "num_tokens": 553762917.0, + "step": 21401 + }, + { + "epoch": 2.3503184713375798, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.1943747997283936, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7031755447387695, + "num_tokens": 553798724.0, + "step": 21402 + }, + { + "epoch": 2.3504282890401935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2650482654571533, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7107371091842651, + "num_tokens": 553819757.0, + "step": 21403 + }, + { + "epoch": 2.350538106742807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5589237213134766, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7142918109893799, + "num_tokens": 553846996.0, + "step": 21404 + }, + { + "epoch": 2.3506479244454206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.736527442932129, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6995679140090942, + "num_tokens": 553871149.0, + "step": 21405 + }, + { + "epoch": 2.3507577421480343, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.447650909423828, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6929811239242554, + "num_tokens": 553904582.0, + "step": 21406 + }, + { + "epoch": 2.350867559850648, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.4344961643218994, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7245700359344482, + "num_tokens": 553925932.0, + "step": 21407 + }, + { + "epoch": 2.3509773775532614, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5927135944366455, + "learning_rate": 1e-06, + "loss": 0.9234, + "mean_token_accuracy": 0.7247322797775269, + "num_tokens": 553950760.0, + "step": 21408 + }, + { + "epoch": 2.351087195255875, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.463813304901123, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6908669471740723, + "num_tokens": 553980795.0, + "step": 21409 + }, + { + "epoch": 2.351197012958489, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4818031787872314, + "learning_rate": 1e-06, + "loss": 0.8954, + "mean_token_accuracy": 0.7333100438117981, + "num_tokens": 554007473.0, + "step": 21410 + }, + { + "epoch": 2.3513068306611027, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.508298635482788, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.6943743228912354, + "num_tokens": 554036706.0, + "step": 21411 + }, + { + "epoch": 2.351416648363716, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0309736728668213, + "learning_rate": 1e-06, + "loss": 0.8793, + "mean_token_accuracy": 0.744189977645874, + "num_tokens": 554055256.0, + "step": 21412 + }, + { + "epoch": 2.3515264660663298, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7882144451141357, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7102179527282715, + "num_tokens": 554081341.0, + "step": 21413 + }, + { + "epoch": 2.3516362837689435, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.419686794281006, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7064838409423828, + "num_tokens": 554110544.0, + "step": 21414 + }, + { + "epoch": 2.3517461014715573, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.572402238845825, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7069716453552246, + "num_tokens": 554136803.0, + "step": 21415 + }, + { + "epoch": 2.351855919174171, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.577075242996216, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7033517956733704, + "num_tokens": 554164323.0, + "step": 21416 + }, + { + "epoch": 2.3519657368767843, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.643361806869507, + "learning_rate": 1e-06, + "loss": 0.9356, + "mean_token_accuracy": 0.7254046201705933, + "num_tokens": 554188916.0, + "step": 21417 + }, + { + "epoch": 2.352075554579398, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7793214321136475, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7144608497619629, + "num_tokens": 554210815.0, + "step": 21418 + }, + { + "epoch": 2.352185372282012, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.751620054244995, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7099744081497192, + "num_tokens": 554233278.0, + "step": 21419 + }, + { + "epoch": 2.3522951899846256, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7571523189544678, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7038259506225586, + "num_tokens": 554256319.0, + "step": 21420 + }, + { + "epoch": 2.3524050076872394, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.06095290184021, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.6983192563056946, + "num_tokens": 554275103.0, + "step": 21421 + }, + { + "epoch": 2.3525148253898527, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.605532646179199, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7004513740539551, + "num_tokens": 554302972.0, + "step": 21422 + }, + { + "epoch": 2.3526246430924664, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.710155963897705, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7046865224838257, + "num_tokens": 554326597.0, + "step": 21423 + }, + { + "epoch": 2.35273446079508, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718174934387207, + "learning_rate": 1e-06, + "loss": 1.1126, + "mean_token_accuracy": 0.6912041306495667, + "num_tokens": 554350438.0, + "step": 21424 + }, + { + "epoch": 2.352844278497694, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8371758460998535, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7079344391822815, + "num_tokens": 554372166.0, + "step": 21425 + }, + { + "epoch": 2.3529540962003077, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.687438488006592, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7198480367660522, + "num_tokens": 554394105.0, + "step": 21426 + }, + { + "epoch": 2.353063913902921, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8407280445098877, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.6985776424407959, + "num_tokens": 554419285.0, + "step": 21427 + }, + { + "epoch": 2.353173731605535, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7050139904022217, + "learning_rate": 1e-06, + "loss": 0.9536, + "mean_token_accuracy": 0.7162582874298096, + "num_tokens": 554442757.0, + "step": 21428 + }, + { + "epoch": 2.3532835493081485, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.603431224822998, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.690039873123169, + "num_tokens": 554468867.0, + "step": 21429 + }, + { + "epoch": 2.3533933670107623, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.459477424621582, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7060761451721191, + "num_tokens": 554499879.0, + "step": 21430 + }, + { + "epoch": 2.3535031847133756, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7361042499542236, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7124652862548828, + "num_tokens": 554524143.0, + "step": 21431 + }, + { + "epoch": 2.3536130024159894, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.768446207046509, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7115869522094727, + "num_tokens": 554546452.0, + "step": 21432 + }, + { + "epoch": 2.353722820118603, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5672624111175537, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7186621427536011, + "num_tokens": 554571337.0, + "step": 21433 + }, + { + "epoch": 2.353832637821217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0414559841156006, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7095186114311218, + "num_tokens": 554595132.0, + "step": 21434 + }, + { + "epoch": 2.3539424555238306, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.563410520553589, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7019525170326233, + "num_tokens": 554622249.0, + "step": 21435 + }, + { + "epoch": 2.354052273226444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4615070819854736, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6972118616104126, + "num_tokens": 554654011.0, + "step": 21436 + }, + { + "epoch": 2.3541620909290577, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.474313974380493, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7104939222335815, + "num_tokens": 554681248.0, + "step": 21437 + }, + { + "epoch": 2.3542719086316715, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6080026626586914, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7222363948822021, + "num_tokens": 554704839.0, + "step": 21438 + }, + { + "epoch": 2.354381726334285, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3454861640930176, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7054556608200073, + "num_tokens": 554736566.0, + "step": 21439 + }, + { + "epoch": 2.3544915440368985, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7616727352142334, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7077466249465942, + "num_tokens": 554759453.0, + "step": 21440 + }, + { + "epoch": 2.3546013617395123, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545053720474243, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.7086625099182129, + "num_tokens": 554785923.0, + "step": 21441 + }, + { + "epoch": 2.354711179442126, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8607819080352783, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.701216459274292, + "num_tokens": 554807141.0, + "step": 21442 + }, + { + "epoch": 2.35482099714474, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6670870780944824, + "learning_rate": 1e-06, + "loss": 0.8743, + "mean_token_accuracy": 0.7455445528030396, + "num_tokens": 554830120.0, + "step": 21443 + }, + { + "epoch": 2.3549308148473536, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6599249839782715, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7097451090812683, + "num_tokens": 554854882.0, + "step": 21444 + }, + { + "epoch": 2.355040632549967, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6420788764953613, + "learning_rate": 1e-06, + "loss": 1.1113, + "mean_token_accuracy": 0.6816482543945312, + "num_tokens": 554879697.0, + "step": 21445 + }, + { + "epoch": 2.3551504502525806, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.715557336807251, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7295043468475342, + "num_tokens": 554901742.0, + "step": 21446 + }, + { + "epoch": 2.3552602679551944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.631218194961548, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7033017873764038, + "num_tokens": 554927477.0, + "step": 21447 + }, + { + "epoch": 2.355370085657808, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9658501148223877, + "learning_rate": 1e-06, + "loss": 0.8961, + "mean_token_accuracy": 0.7415372729301453, + "num_tokens": 554945121.0, + "step": 21448 + }, + { + "epoch": 2.355479903360422, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5866549015045166, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7023728489875793, + "num_tokens": 554971482.0, + "step": 21449 + }, + { + "epoch": 2.355589721063035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7715911865234375, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.705296516418457, + "num_tokens": 554991702.0, + "step": 21450 + }, + { + "epoch": 2.355699538765649, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.841060161590576, + "learning_rate": 1e-06, + "loss": 0.9158, + "mean_token_accuracy": 0.7278257608413696, + "num_tokens": 555012775.0, + "step": 21451 + }, + { + "epoch": 2.3558093564682627, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.439638137817383, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7004910111427307, + "num_tokens": 555042151.0, + "step": 21452 + }, + { + "epoch": 2.3559191741708765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.550673246383667, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.711770236492157, + "num_tokens": 555069862.0, + "step": 21453 + }, + { + "epoch": 2.3560289918734902, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.845515012741089, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.722191333770752, + "num_tokens": 555092922.0, + "step": 21454 + }, + { + "epoch": 2.3561388095761036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4634172916412354, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7142139673233032, + "num_tokens": 555119169.0, + "step": 21455 + }, + { + "epoch": 2.3562486272787173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5593605041503906, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7054938077926636, + "num_tokens": 555145956.0, + "step": 21456 + }, + { + "epoch": 2.356358444981331, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.529954433441162, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.6962141394615173, + "num_tokens": 555170733.0, + "step": 21457 + }, + { + "epoch": 2.356468262683945, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.919687032699585, + "learning_rate": 1e-06, + "loss": 0.9071, + "mean_token_accuracy": 0.7326866984367371, + "num_tokens": 555190957.0, + "step": 21458 + }, + { + "epoch": 2.356578080386558, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0863239765167236, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7109900712966919, + "num_tokens": 555211908.0, + "step": 21459 + }, + { + "epoch": 2.356687898089172, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8408050537109375, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.7024370431900024, + "num_tokens": 555236327.0, + "step": 21460 + }, + { + "epoch": 2.3567977157917857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.710869073867798, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7135167121887207, + "num_tokens": 555260806.0, + "step": 21461 + }, + { + "epoch": 2.3569075334943994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5229456424713135, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6996967792510986, + "num_tokens": 555290160.0, + "step": 21462 + }, + { + "epoch": 2.3570173511970127, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.942244291305542, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7132160663604736, + "num_tokens": 555310957.0, + "step": 21463 + }, + { + "epoch": 2.3571271688996265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.635193109512329, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7038880586624146, + "num_tokens": 555335161.0, + "step": 21464 + }, + { + "epoch": 2.3572369866022402, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5222883224487305, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7011834383010864, + "num_tokens": 555362095.0, + "step": 21465 + }, + { + "epoch": 2.357346804304854, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.672429323196411, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7226380109786987, + "num_tokens": 555384986.0, + "step": 21466 + }, + { + "epoch": 2.3574566220074678, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7157037258148193, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.71187824010849, + "num_tokens": 555410007.0, + "step": 21467 + }, + { + "epoch": 2.357566439710081, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4362056255340576, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.699745774269104, + "num_tokens": 555441338.0, + "step": 21468 + }, + { + "epoch": 2.357676257412695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.443770408630371, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6916024684906006, + "num_tokens": 555470582.0, + "step": 21469 + }, + { + "epoch": 2.3577860751153086, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3131825923919678, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7099568843841553, + "num_tokens": 555503886.0, + "step": 21470 + }, + { + "epoch": 2.3578958928179223, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6762917041778564, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7114222645759583, + "num_tokens": 555528617.0, + "step": 21471 + }, + { + "epoch": 2.358005710520536, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6368632316589355, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.718535304069519, + "num_tokens": 555555610.0, + "step": 21472 + }, + { + "epoch": 2.3581155282231494, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596620798110962, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6821193695068359, + "num_tokens": 555581989.0, + "step": 21473 + }, + { + "epoch": 2.358225345925763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.859203577041626, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7129067182540894, + "num_tokens": 555604084.0, + "step": 21474 + }, + { + "epoch": 2.358335163628377, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6063575744628906, + "learning_rate": 1e-06, + "loss": 0.932, + "mean_token_accuracy": 0.7223641872406006, + "num_tokens": 555631310.0, + "step": 21475 + }, + { + "epoch": 2.3584449813309907, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.495861530303955, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6975170373916626, + "num_tokens": 555661471.0, + "step": 21476 + }, + { + "epoch": 2.3585547990336044, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.621614933013916, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7272933721542358, + "num_tokens": 555686611.0, + "step": 21477 + }, + { + "epoch": 2.3586646167362177, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.821442127227783, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7222796678543091, + "num_tokens": 555706852.0, + "step": 21478 + }, + { + "epoch": 2.3587744344388315, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.668844699859619, + "learning_rate": 1e-06, + "loss": 0.917, + "mean_token_accuracy": 0.7305741310119629, + "num_tokens": 555731573.0, + "step": 21479 + }, + { + "epoch": 2.3588842521414453, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7365329265594482, + "learning_rate": 1e-06, + "loss": 0.9056, + "mean_token_accuracy": 0.7306556701660156, + "num_tokens": 555754399.0, + "step": 21480 + }, + { + "epoch": 2.358994069844059, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8390698432922363, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6952702403068542, + "num_tokens": 555779355.0, + "step": 21481 + }, + { + "epoch": 2.3591038875466728, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.814621925354004, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.718887209892273, + "num_tokens": 555801736.0, + "step": 21482 + }, + { + "epoch": 2.359213705249286, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5835368633270264, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7070912718772888, + "num_tokens": 555828488.0, + "step": 21483 + }, + { + "epoch": 2.3593235229519, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6443088054656982, + "learning_rate": 1e-06, + "loss": 0.9219, + "mean_token_accuracy": 0.7335458993911743, + "num_tokens": 555853675.0, + "step": 21484 + }, + { + "epoch": 2.3594333406545136, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.660984516143799, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7217473387718201, + "num_tokens": 555877499.0, + "step": 21485 + }, + { + "epoch": 2.3595431583571274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.501366376876831, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7027853727340698, + "num_tokens": 555904816.0, + "step": 21486 + }, + { + "epoch": 2.3596529760597407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6693546772003174, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7181904315948486, + "num_tokens": 555930791.0, + "step": 21487 + }, + { + "epoch": 2.3597627937623544, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8143653869628906, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.6993987560272217, + "num_tokens": 555954763.0, + "step": 21488 + }, + { + "epoch": 2.359872611464968, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7202835083007812, + "learning_rate": 1e-06, + "loss": 0.9008, + "mean_token_accuracy": 0.7377982139587402, + "num_tokens": 555978907.0, + "step": 21489 + }, + { + "epoch": 2.359982429167582, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6270081996917725, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7051904201507568, + "num_tokens": 556004729.0, + "step": 21490 + }, + { + "epoch": 2.3600922468701953, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.495837926864624, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.6977120637893677, + "num_tokens": 556031476.0, + "step": 21491 + }, + { + "epoch": 2.360202064572809, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.692254066467285, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7103768587112427, + "num_tokens": 556054552.0, + "step": 21492 + }, + { + "epoch": 2.3603118822754228, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8112218379974365, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.6910878419876099, + "num_tokens": 556079224.0, + "step": 21493 + }, + { + "epoch": 2.3604216999780365, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.986211061477661, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7188606858253479, + "num_tokens": 556099862.0, + "step": 21494 + }, + { + "epoch": 2.3605315176806503, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3981330394744873, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6935135722160339, + "num_tokens": 556131139.0, + "step": 21495 + }, + { + "epoch": 2.3606413353832636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.5351972579956055, + "learning_rate": 1e-06, + "loss": 0.9526, + "mean_token_accuracy": 0.7200809121131897, + "num_tokens": 556154231.0, + "step": 21496 + }, + { + "epoch": 2.3607511530858774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4093430042266846, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.7036210298538208, + "num_tokens": 556185699.0, + "step": 21497 + }, + { + "epoch": 2.360860970788491, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3690755367279053, + "learning_rate": 1e-06, + "loss": 0.9478, + "mean_token_accuracy": 0.7209346890449524, + "num_tokens": 556213295.0, + "step": 21498 + }, + { + "epoch": 2.360970788491105, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6963887214660645, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.694464385509491, + "num_tokens": 556236909.0, + "step": 21499 + }, + { + "epoch": 2.3610806061937186, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8983945846557617, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.711743950843811, + "num_tokens": 556260911.0, + "step": 21500 + }, + { + "epoch": 2.361190423896332, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6530778408050537, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7259437441825867, + "num_tokens": 556284155.0, + "step": 21501 + }, + { + "epoch": 2.3613002415989457, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.612436294555664, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6992839574813843, + "num_tokens": 556311007.0, + "step": 21502 + }, + { + "epoch": 2.3614100593015594, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1123404502868652, + "learning_rate": 1e-06, + "loss": 0.8383, + "mean_token_accuracy": 0.7485024929046631, + "num_tokens": 556327874.0, + "step": 21503 + }, + { + "epoch": 2.361519877004173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5152828693389893, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7153106331825256, + "num_tokens": 556354663.0, + "step": 21504 + }, + { + "epoch": 2.361629694706787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596407413482666, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7387192249298096, + "num_tokens": 556380131.0, + "step": 21505 + }, + { + "epoch": 2.3617395124094003, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.304067611694336, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.710294246673584, + "num_tokens": 556412535.0, + "step": 21506 + }, + { + "epoch": 2.361849330112014, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7350525856018066, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.7050961256027222, + "num_tokens": 556438374.0, + "step": 21507 + }, + { + "epoch": 2.361959147814628, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.725396156311035, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7063980102539062, + "num_tokens": 556461616.0, + "step": 21508 + }, + { + "epoch": 2.3620689655172415, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6376760005950928, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7172976732254028, + "num_tokens": 556483612.0, + "step": 21509 + }, + { + "epoch": 2.362178783219855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5758347511291504, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7191990613937378, + "num_tokens": 556508354.0, + "step": 21510 + }, + { + "epoch": 2.3622886009224686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.663195848464966, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7149943113327026, + "num_tokens": 556533953.0, + "step": 21511 + }, + { + "epoch": 2.3623984186250824, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8928606510162354, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7063109874725342, + "num_tokens": 556559964.0, + "step": 21512 + }, + { + "epoch": 2.362508236327696, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.491041421890259, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6960748434066772, + "num_tokens": 556591345.0, + "step": 21513 + }, + { + "epoch": 2.3626180540303094, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9526147842407227, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7141153216362, + "num_tokens": 556610964.0, + "step": 21514 + }, + { + "epoch": 2.362727871732923, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6261963844299316, + "learning_rate": 1e-06, + "loss": 0.9291, + "mean_token_accuracy": 0.7251731157302856, + "num_tokens": 556635293.0, + "step": 21515 + }, + { + "epoch": 2.362837689435537, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.538339138031006, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6889400482177734, + "num_tokens": 556663735.0, + "step": 21516 + }, + { + "epoch": 2.3629475071381507, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7568552494049072, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7087666988372803, + "num_tokens": 556686187.0, + "step": 21517 + }, + { + "epoch": 2.3630573248407645, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.595571756362915, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6996111869812012, + "num_tokens": 556714669.0, + "step": 21518 + }, + { + "epoch": 2.363167142543378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5449206829071045, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7061250805854797, + "num_tokens": 556741696.0, + "step": 21519 + }, + { + "epoch": 2.3632769602459915, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.775346040725708, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7178929448127747, + "num_tokens": 556763765.0, + "step": 21520 + }, + { + "epoch": 2.3633867779486053, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.689260721206665, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.702771782875061, + "num_tokens": 556790514.0, + "step": 21521 + }, + { + "epoch": 2.363496595651219, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7514331340789795, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7074961066246033, + "num_tokens": 556813619.0, + "step": 21522 + }, + { + "epoch": 2.363606413353833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.479726552963257, + "learning_rate": 1e-06, + "loss": 0.9534, + "mean_token_accuracy": 0.7182517647743225, + "num_tokens": 556841250.0, + "step": 21523 + }, + { + "epoch": 2.363716231056446, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.703418493270874, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7251532673835754, + "num_tokens": 556863528.0, + "step": 21524 + }, + { + "epoch": 2.36382604875906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.614274501800537, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.703362226486206, + "num_tokens": 556892962.0, + "step": 21525 + }, + { + "epoch": 2.3639358664616736, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5841286182403564, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7009573578834534, + "num_tokens": 556920673.0, + "step": 21526 + }, + { + "epoch": 2.3640456841642874, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.655550718307495, + "learning_rate": 1e-06, + "loss": 0.8899, + "mean_token_accuracy": 0.7370742559432983, + "num_tokens": 556943449.0, + "step": 21527 + }, + { + "epoch": 2.364155501866901, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.503239154815674, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.695357084274292, + "num_tokens": 556972204.0, + "step": 21528 + }, + { + "epoch": 2.3642653195695145, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5186545848846436, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6842461824417114, + "num_tokens": 557002786.0, + "step": 21529 + }, + { + "epoch": 2.3643751372721282, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8659915924072266, + "learning_rate": 1e-06, + "loss": 1.0789, + "mean_token_accuracy": 0.6859642863273621, + "num_tokens": 557028432.0, + "step": 21530 + }, + { + "epoch": 2.364484954974742, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.920987844467163, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6884819865226746, + "num_tokens": 557053635.0, + "step": 21531 + }, + { + "epoch": 2.3645947726773557, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.640618324279785, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7136543393135071, + "num_tokens": 557080238.0, + "step": 21532 + }, + { + "epoch": 2.3647045903799695, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.784363269805908, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7162028551101685, + "num_tokens": 557102780.0, + "step": 21533 + }, + { + "epoch": 2.364814408082583, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8249545097351074, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6976860761642456, + "num_tokens": 557131178.0, + "step": 21534 + }, + { + "epoch": 2.3649242257851966, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.883739948272705, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.722025990486145, + "num_tokens": 557153394.0, + "step": 21535 + }, + { + "epoch": 2.3650340434878103, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.397212505340576, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7102463841438293, + "num_tokens": 557180490.0, + "step": 21536 + }, + { + "epoch": 2.365143861190424, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3107333183288574, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7075443863868713, + "num_tokens": 557213714.0, + "step": 21537 + }, + { + "epoch": 2.3652536788930374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6097800731658936, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7163093090057373, + "num_tokens": 557238581.0, + "step": 21538 + }, + { + "epoch": 2.365363496595651, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.703117609024048, + "learning_rate": 1e-06, + "loss": 1.0594, + "mean_token_accuracy": 0.6962326765060425, + "num_tokens": 557264846.0, + "step": 21539 + }, + { + "epoch": 2.365473314298265, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.551316976547241, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7036279439926147, + "num_tokens": 557291147.0, + "step": 21540 + }, + { + "epoch": 2.3655831320008787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8158838748931885, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7095032930374146, + "num_tokens": 557312834.0, + "step": 21541 + }, + { + "epoch": 2.365692949703492, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7541518211364746, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.721235990524292, + "num_tokens": 557339964.0, + "step": 21542 + }, + { + "epoch": 2.3658027674061057, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.619166135787964, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.723931074142456, + "num_tokens": 557366144.0, + "step": 21543 + }, + { + "epoch": 2.3659125851087195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6620936393737793, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.7000116109848022, + "num_tokens": 557391270.0, + "step": 21544 + }, + { + "epoch": 2.3660224028113332, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6161892414093018, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7080868482589722, + "num_tokens": 557417639.0, + "step": 21545 + }, + { + "epoch": 2.366132220513947, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.525758981704712, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.711732804775238, + "num_tokens": 557444990.0, + "step": 21546 + }, + { + "epoch": 2.3662420382165603, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.804885149002075, + "learning_rate": 1e-06, + "loss": 1.0927, + "mean_token_accuracy": 0.6834051609039307, + "num_tokens": 557467022.0, + "step": 21547 + }, + { + "epoch": 2.366351855919174, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5239179134368896, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6990879774093628, + "num_tokens": 557495147.0, + "step": 21548 + }, + { + "epoch": 2.366461673621788, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7599804401397705, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7162758708000183, + "num_tokens": 557519017.0, + "step": 21549 + }, + { + "epoch": 2.3665714913244016, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.789203643798828, + "learning_rate": 1e-06, + "loss": 0.8787, + "mean_token_accuracy": 0.7453387975692749, + "num_tokens": 557543080.0, + "step": 21550 + }, + { + "epoch": 2.3666813090270153, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.640181303024292, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7077829837799072, + "num_tokens": 557569638.0, + "step": 21551 + }, + { + "epoch": 2.3667911267296287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4847450256347656, + "learning_rate": 1e-06, + "loss": 0.9443, + "mean_token_accuracy": 0.7199223041534424, + "num_tokens": 557598536.0, + "step": 21552 + }, + { + "epoch": 2.3669009444322424, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3281736373901367, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6967439651489258, + "num_tokens": 557634436.0, + "step": 21553 + }, + { + "epoch": 2.367010762134856, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.920048236846924, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7279917001724243, + "num_tokens": 557654703.0, + "step": 21554 + }, + { + "epoch": 2.36712057983747, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5651445388793945, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7056822776794434, + "num_tokens": 557683104.0, + "step": 21555 + }, + { + "epoch": 2.3672303975400837, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.470660924911499, + "learning_rate": 1e-06, + "loss": 1.0765, + "mean_token_accuracy": 0.6929062604904175, + "num_tokens": 557715197.0, + "step": 21556 + }, + { + "epoch": 2.367340215242697, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.623579263687134, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7107794880867004, + "num_tokens": 557742148.0, + "step": 21557 + }, + { + "epoch": 2.3674500329453108, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.43210768699646, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7063046097755432, + "num_tokens": 557769927.0, + "step": 21558 + }, + { + "epoch": 2.3675598506479245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.736720323562622, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7117432951927185, + "num_tokens": 557793071.0, + "step": 21559 + }, + { + "epoch": 2.3676696683505383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.876128911972046, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7239379286766052, + "num_tokens": 557822869.0, + "step": 21560 + }, + { + "epoch": 2.3677794860531516, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.446639060974121, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7112073302268982, + "num_tokens": 557852209.0, + "step": 21561 + }, + { + "epoch": 2.3678893037557653, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.621389389038086, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.6979018449783325, + "num_tokens": 557876654.0, + "step": 21562 + }, + { + "epoch": 2.367999121458379, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6583049297332764, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7168778777122498, + "num_tokens": 557906940.0, + "step": 21563 + }, + { + "epoch": 2.368108939160993, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5198991298675537, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7000933885574341, + "num_tokens": 557938472.0, + "step": 21564 + }, + { + "epoch": 2.368218756863606, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.805041551589966, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7056306004524231, + "num_tokens": 557960421.0, + "step": 21565 + }, + { + "epoch": 2.36832857456622, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.820683002471924, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7202692627906799, + "num_tokens": 557981998.0, + "step": 21566 + }, + { + "epoch": 2.3684383922688337, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6894149780273438, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7188202142715454, + "num_tokens": 558004848.0, + "step": 21567 + }, + { + "epoch": 2.3685482099714474, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7383203506469727, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7180158495903015, + "num_tokens": 558028237.0, + "step": 21568 + }, + { + "epoch": 2.368658027674061, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4198572635650635, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7050083875656128, + "num_tokens": 558058594.0, + "step": 21569 + }, + { + "epoch": 2.3687678453766745, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5980732440948486, + "learning_rate": 1e-06, + "loss": 0.9264, + "mean_token_accuracy": 0.7291872501373291, + "num_tokens": 558084242.0, + "step": 21570 + }, + { + "epoch": 2.3688776630792883, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.8304171562194824, + "learning_rate": 1e-06, + "loss": 1.117, + "mean_token_accuracy": 0.6801732182502747, + "num_tokens": 558115959.0, + "step": 21571 + }, + { + "epoch": 2.368987480781902, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.725844383239746, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.6977161765098572, + "num_tokens": 558141050.0, + "step": 21572 + }, + { + "epoch": 2.369097298484516, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5071754455566406, + "learning_rate": 1e-06, + "loss": 1.142, + "mean_token_accuracy": 0.6742473244667053, + "num_tokens": 558173423.0, + "step": 21573 + }, + { + "epoch": 2.3692071161871295, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5714290142059326, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7111175060272217, + "num_tokens": 558197401.0, + "step": 21574 + }, + { + "epoch": 2.369316933889743, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.840409994125366, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.715539813041687, + "num_tokens": 558220872.0, + "step": 21575 + }, + { + "epoch": 2.3694267515923566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.734529733657837, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7106367349624634, + "num_tokens": 558245340.0, + "step": 21576 + }, + { + "epoch": 2.3695365692949704, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.631837844848633, + "learning_rate": 1e-06, + "loss": 0.9174, + "mean_token_accuracy": 0.7318660020828247, + "num_tokens": 558269052.0, + "step": 21577 + }, + { + "epoch": 2.369646386997584, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5696096420288086, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7147325277328491, + "num_tokens": 558296135.0, + "step": 21578 + }, + { + "epoch": 2.369756204700198, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7667489051818848, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.7017560601234436, + "num_tokens": 558319289.0, + "step": 21579 + }, + { + "epoch": 2.369866022402811, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7039968967437744, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.721108615398407, + "num_tokens": 558341681.0, + "step": 21580 + }, + { + "epoch": 2.369975840105425, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5311992168426514, + "learning_rate": 1e-06, + "loss": 1.0924, + "mean_token_accuracy": 0.6888716220855713, + "num_tokens": 558369106.0, + "step": 21581 + }, + { + "epoch": 2.3700856578080387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4172677993774414, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.7104745507240295, + "num_tokens": 558398945.0, + "step": 21582 + }, + { + "epoch": 2.3701954755106525, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8865256309509277, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7186225652694702, + "num_tokens": 558422580.0, + "step": 21583 + }, + { + "epoch": 2.370305293213266, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3943045139312744, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.7171175479888916, + "num_tokens": 558450847.0, + "step": 21584 + }, + { + "epoch": 2.3704151109158795, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5219051837921143, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7117394804954529, + "num_tokens": 558475667.0, + "step": 21585 + }, + { + "epoch": 2.3705249286184933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.494225025177002, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7200102806091309, + "num_tokens": 558504062.0, + "step": 21586 + }, + { + "epoch": 2.370634746321107, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6890649795532227, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7184061408042908, + "num_tokens": 558527171.0, + "step": 21587 + }, + { + "epoch": 2.370744564023721, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9159207344055176, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7177959680557251, + "num_tokens": 558546295.0, + "step": 21588 + }, + { + "epoch": 2.370854381726334, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8089072704315186, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.6970365047454834, + "num_tokens": 558568236.0, + "step": 21589 + }, + { + "epoch": 2.370964199428948, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5403177738189697, + "learning_rate": 1e-06, + "loss": 1.1128, + "mean_token_accuracy": 0.6834635734558105, + "num_tokens": 558598369.0, + "step": 21590 + }, + { + "epoch": 2.3710740171315616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.765293598175049, + "learning_rate": 1e-06, + "loss": 0.9075, + "mean_token_accuracy": 0.7387437224388123, + "num_tokens": 558620381.0, + "step": 21591 + }, + { + "epoch": 2.3711838348341754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.822659969329834, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7165522575378418, + "num_tokens": 558642173.0, + "step": 21592 + }, + { + "epoch": 2.3712936525367887, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5791738033294678, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.6998858451843262, + "num_tokens": 558669174.0, + "step": 21593 + }, + { + "epoch": 2.3714034702394025, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.971952199935913, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7157952785491943, + "num_tokens": 558691184.0, + "step": 21594 + }, + { + "epoch": 2.371513287942016, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.178403377532959, + "learning_rate": 1e-06, + "loss": 0.8869, + "mean_token_accuracy": 0.732641339302063, + "num_tokens": 558707598.0, + "step": 21595 + }, + { + "epoch": 2.37162310564463, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.52353572845459, + "learning_rate": 1e-06, + "loss": 1.1404, + "mean_token_accuracy": 0.6727414131164551, + "num_tokens": 558744396.0, + "step": 21596 + }, + { + "epoch": 2.3717329233472437, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.694324016571045, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.719035804271698, + "num_tokens": 558769673.0, + "step": 21597 + }, + { + "epoch": 2.371842741049857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.672515869140625, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7115463018417358, + "num_tokens": 558795749.0, + "step": 21598 + }, + { + "epoch": 2.371952558752471, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4597818851470947, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7243480682373047, + "num_tokens": 558826193.0, + "step": 21599 + }, + { + "epoch": 2.3720623764550846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.778893232345581, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7292132377624512, + "num_tokens": 558850118.0, + "step": 21600 + }, + { + "epoch": 2.3721721941576983, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0584561824798584, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7086223363876343, + "num_tokens": 558870899.0, + "step": 21601 + }, + { + "epoch": 2.372282011860312, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617995262145996, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.7020910382270813, + "num_tokens": 558897874.0, + "step": 21602 + }, + { + "epoch": 2.3723918295629254, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.613732099533081, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.6996506452560425, + "num_tokens": 558925023.0, + "step": 21603 + }, + { + "epoch": 2.372501647265539, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.655682325363159, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7139574885368347, + "num_tokens": 558950567.0, + "step": 21604 + }, + { + "epoch": 2.372611464968153, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5913591384887695, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7193983197212219, + "num_tokens": 558974871.0, + "step": 21605 + }, + { + "epoch": 2.3727212826707667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.675194501876831, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6988221406936646, + "num_tokens": 559005912.0, + "step": 21606 + }, + { + "epoch": 2.3728311003733804, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.494239091873169, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7066497206687927, + "num_tokens": 559032318.0, + "step": 21607 + }, + { + "epoch": 2.3729409180759937, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.497525453567505, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7297854423522949, + "num_tokens": 559050923.0, + "step": 21608 + }, + { + "epoch": 2.3730507357786075, + "ewc_loss": 1.0967254638671875e-05, + "grad_norm": 80.54205322265625, + "learning_rate": 1e-06, + "loss": 0.9069, + "mean_token_accuracy": 0.7321262359619141, + "num_tokens": 559081779.0, + "step": 21609 + }, + { + "epoch": 2.3731605534812212, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.82052230834961, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.6993268728256226, + "num_tokens": 559101853.0, + "step": 21610 + }, + { + "epoch": 2.373270371183835, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2142655849456787, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6872893571853638, + "num_tokens": 559125924.0, + "step": 21611 + }, + { + "epoch": 2.3733801888864483, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7391703128814697, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7219530940055847, + "num_tokens": 559155653.0, + "step": 21612 + }, + { + "epoch": 2.373490006589062, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.89555025100708, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7116358876228333, + "num_tokens": 559179068.0, + "step": 21613 + }, + { + "epoch": 2.373599824291676, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.580967664718628, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6980466842651367, + "num_tokens": 559206875.0, + "step": 21614 + }, + { + "epoch": 2.3737096419942896, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.539133071899414, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7056945562362671, + "num_tokens": 559235853.0, + "step": 21615 + }, + { + "epoch": 2.3738194596969033, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9990041255950928, + "learning_rate": 1e-06, + "loss": 0.8775, + "mean_token_accuracy": 0.7379140853881836, + "num_tokens": 559256191.0, + "step": 21616 + }, + { + "epoch": 2.3739292773995166, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.816493272781372, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6925957202911377, + "num_tokens": 559278183.0, + "step": 21617 + }, + { + "epoch": 2.3740390951021304, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7254018783569336, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7158108353614807, + "num_tokens": 559299964.0, + "step": 21618 + }, + { + "epoch": 2.374148912804744, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.692408323287964, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.6963491439819336, + "num_tokens": 559324445.0, + "step": 21619 + }, + { + "epoch": 2.374258730507358, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7509703636169434, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7062704563140869, + "num_tokens": 559349323.0, + "step": 21620 + }, + { + "epoch": 2.3743685482099712, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7360684871673584, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7008192539215088, + "num_tokens": 559378716.0, + "step": 21621 + }, + { + "epoch": 2.374478365912585, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.296271324157715, + "learning_rate": 1e-06, + "loss": 1.0921, + "mean_token_accuracy": 0.6853536367416382, + "num_tokens": 559412640.0, + "step": 21622 + }, + { + "epoch": 2.3745881836151987, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.904186487197876, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7214162349700928, + "num_tokens": 559431472.0, + "step": 21623 + }, + { + "epoch": 2.3746980013178125, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0413734912872314, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7232038974761963, + "num_tokens": 559453465.0, + "step": 21624 + }, + { + "epoch": 2.3748078190204263, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8530683517456055, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6955028772354126, + "num_tokens": 559479660.0, + "step": 21625 + }, + { + "epoch": 2.3749176367230396, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4600653648376465, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7198132276535034, + "num_tokens": 559509927.0, + "step": 21626 + }, + { + "epoch": 2.3750274544256533, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7228457927703857, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.7057376503944397, + "num_tokens": 559535535.0, + "step": 21627 + }, + { + "epoch": 2.375137272128267, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5551822185516357, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7175233364105225, + "num_tokens": 559563278.0, + "step": 21628 + }, + { + "epoch": 2.375247089830881, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.50016450881958, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7063762545585632, + "num_tokens": 559588563.0, + "step": 21629 + }, + { + "epoch": 2.3753569075334946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4101195335388184, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6909413933753967, + "num_tokens": 559619051.0, + "step": 21630 + }, + { + "epoch": 2.375466725236108, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9297120571136475, + "learning_rate": 1e-06, + "loss": 0.8962, + "mean_token_accuracy": 0.7334082126617432, + "num_tokens": 559639925.0, + "step": 21631 + }, + { + "epoch": 2.3755765429387217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3836557865142822, + "learning_rate": 1e-06, + "loss": 0.9236, + "mean_token_accuracy": 0.7305291891098022, + "num_tokens": 559671210.0, + "step": 21632 + }, + { + "epoch": 2.3756863606413354, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8366425037384033, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7030900716781616, + "num_tokens": 559694236.0, + "step": 21633 + }, + { + "epoch": 2.375796178343949, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9861998558044434, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7001873254776001, + "num_tokens": 559714548.0, + "step": 21634 + }, + { + "epoch": 2.375905996046563, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.5546975135803223, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6953949928283691, + "num_tokens": 559744762.0, + "step": 21635 + }, + { + "epoch": 2.3760158137491763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7597641944885254, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7080614566802979, + "num_tokens": 559770635.0, + "step": 21636 + }, + { + "epoch": 2.37612563145179, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2670159339904785, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7039984464645386, + "num_tokens": 559791610.0, + "step": 21637 + }, + { + "epoch": 2.3762354491544038, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.666679620742798, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7070364356040955, + "num_tokens": 559814186.0, + "step": 21638 + }, + { + "epoch": 2.3763452668570175, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.717491388320923, + "learning_rate": 1e-06, + "loss": 1.0769, + "mean_token_accuracy": 0.6974440813064575, + "num_tokens": 559835190.0, + "step": 21639 + }, + { + "epoch": 2.376455084559631, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.37624454498291, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7015091180801392, + "num_tokens": 559866682.0, + "step": 21640 + }, + { + "epoch": 2.3765649022622446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.566023826599121, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7009643316268921, + "num_tokens": 559895962.0, + "step": 21641 + }, + { + "epoch": 2.3766747199648584, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7579565048217773, + "learning_rate": 1e-06, + "loss": 0.9335, + "mean_token_accuracy": 0.7288522720336914, + "num_tokens": 559918730.0, + "step": 21642 + }, + { + "epoch": 2.376784537667472, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6977803707122803, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.7056854963302612, + "num_tokens": 559942407.0, + "step": 21643 + }, + { + "epoch": 2.3768943553700854, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7297139167785645, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.6884613633155823, + "num_tokens": 559969006.0, + "step": 21644 + }, + { + "epoch": 2.377004173072699, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.693734884262085, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7127473950386047, + "num_tokens": 559992947.0, + "step": 21645 + }, + { + "epoch": 2.377113990775313, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.591346025466919, + "learning_rate": 1e-06, + "loss": 1.0991, + "mean_token_accuracy": 0.6843945384025574, + "num_tokens": 560020395.0, + "step": 21646 + }, + { + "epoch": 2.3772238084779267, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.509819984436035, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7161736488342285, + "num_tokens": 560048467.0, + "step": 21647 + }, + { + "epoch": 2.3773336261805404, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 8.0047607421875, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7097300291061401, + "num_tokens": 560070449.0, + "step": 21648 + }, + { + "epoch": 2.3774434438831538, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4206337928771973, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6872798204421997, + "num_tokens": 560102206.0, + "step": 21649 + }, + { + "epoch": 2.3775532615857675, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.695756435394287, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7046383023262024, + "num_tokens": 560124865.0, + "step": 21650 + }, + { + "epoch": 2.3776630792883813, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7419235706329346, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7178348302841187, + "num_tokens": 560146976.0, + "step": 21651 + }, + { + "epoch": 2.377772896990995, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9530646800994873, + "learning_rate": 1e-06, + "loss": 0.9297, + "mean_token_accuracy": 0.7213715314865112, + "num_tokens": 560166722.0, + "step": 21652 + }, + { + "epoch": 2.377882714693609, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.675429344177246, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7141383290290833, + "num_tokens": 560189172.0, + "step": 21653 + }, + { + "epoch": 2.377992532396222, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7392380237579346, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.7008823156356812, + "num_tokens": 560214066.0, + "step": 21654 + }, + { + "epoch": 2.378102350098836, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5347845554351807, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7140407562255859, + "num_tokens": 560246232.0, + "step": 21655 + }, + { + "epoch": 2.3782121678014496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5547826290130615, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7092888355255127, + "num_tokens": 560272804.0, + "step": 21656 + }, + { + "epoch": 2.3783219855040634, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6714415550231934, + "learning_rate": 1e-06, + "loss": 0.972, + "mean_token_accuracy": 0.7113696336746216, + "num_tokens": 560296297.0, + "step": 21657 + }, + { + "epoch": 2.378431803206677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.598533868789673, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6963127851486206, + "num_tokens": 560319923.0, + "step": 21658 + }, + { + "epoch": 2.3785416209092904, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.95436429977417, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.7002127766609192, + "num_tokens": 560345057.0, + "step": 21659 + }, + { + "epoch": 2.378651438611904, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0635921955108643, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7101351022720337, + "num_tokens": 560365551.0, + "step": 21660 + }, + { + "epoch": 2.378761256314518, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6498732566833496, + "learning_rate": 1e-06, + "loss": 0.9112, + "mean_token_accuracy": 0.7327998280525208, + "num_tokens": 560390724.0, + "step": 21661 + }, + { + "epoch": 2.3788710740171317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.155261516571045, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.6961817145347595, + "num_tokens": 560412951.0, + "step": 21662 + }, + { + "epoch": 2.3789808917197455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.687187433242798, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.7053084373474121, + "num_tokens": 560437061.0, + "step": 21663 + }, + { + "epoch": 2.379090709422359, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5796611309051514, + "learning_rate": 1e-06, + "loss": 0.9006, + "mean_token_accuracy": 0.7364277839660645, + "num_tokens": 560459323.0, + "step": 21664 + }, + { + "epoch": 2.3792005271249725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8518402576446533, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7007743120193481, + "num_tokens": 560480023.0, + "step": 21665 + }, + { + "epoch": 2.3793103448275863, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7096097469329834, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6977553367614746, + "num_tokens": 560507057.0, + "step": 21666 + }, + { + "epoch": 2.3794201625302, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.730295419692993, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6972659826278687, + "num_tokens": 560531806.0, + "step": 21667 + }, + { + "epoch": 2.3795299802328134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3492681980133057, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7137879133224487, + "num_tokens": 560562967.0, + "step": 21668 + }, + { + "epoch": 2.379639797935427, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7448973655700684, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7230809926986694, + "num_tokens": 560587734.0, + "step": 21669 + }, + { + "epoch": 2.379749615638041, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6829683780670166, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.6990026831626892, + "num_tokens": 560612771.0, + "step": 21670 + }, + { + "epoch": 2.3798594333406546, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9480998516082764, + "learning_rate": 1e-06, + "loss": 0.8334, + "mean_token_accuracy": 0.7605084776878357, + "num_tokens": 560631954.0, + "step": 21671 + }, + { + "epoch": 2.379969251043268, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5161311626434326, + "learning_rate": 1e-06, + "loss": 1.1095, + "mean_token_accuracy": 0.6782968044281006, + "num_tokens": 560664443.0, + "step": 21672 + }, + { + "epoch": 2.3800790687458817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.699636697769165, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7110599279403687, + "num_tokens": 560689391.0, + "step": 21673 + }, + { + "epoch": 2.3801888864484955, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3775746822357178, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7021551132202148, + "num_tokens": 560717244.0, + "step": 21674 + }, + { + "epoch": 2.3802987041511092, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.594719171524048, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7271000742912292, + "num_tokens": 560738910.0, + "step": 21675 + }, + { + "epoch": 2.380408521853723, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3384857177734375, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.705997109413147, + "num_tokens": 560769723.0, + "step": 21676 + }, + { + "epoch": 2.3805183395563363, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2810120582580566, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6998986005783081, + "num_tokens": 560806610.0, + "step": 21677 + }, + { + "epoch": 2.38062815725895, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7589054107666016, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.7024301290512085, + "num_tokens": 560829502.0, + "step": 21678 + }, + { + "epoch": 2.380737974961564, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.604281425476074, + "learning_rate": 1e-06, + "loss": 0.8988, + "mean_token_accuracy": 0.7316759824752808, + "num_tokens": 560855250.0, + "step": 21679 + }, + { + "epoch": 2.3808477926641776, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.778529167175293, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7264952659606934, + "num_tokens": 560876357.0, + "step": 21680 + }, + { + "epoch": 2.3809576103667913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.814854860305786, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6893618106842041, + "num_tokens": 560902717.0, + "step": 21681 + }, + { + "epoch": 2.3810674280694046, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5231308937072754, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7179856300354004, + "num_tokens": 560930540.0, + "step": 21682 + }, + { + "epoch": 2.3811772457720184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.713380813598633, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7181171178817749, + "num_tokens": 560954786.0, + "step": 21683 + }, + { + "epoch": 2.381287063474632, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6214640140533447, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7103296518325806, + "num_tokens": 560981734.0, + "step": 21684 + }, + { + "epoch": 2.381396881177246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6688077449798584, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7180763483047485, + "num_tokens": 561004489.0, + "step": 21685 + }, + { + "epoch": 2.3815066988798597, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8373398780822754, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7246745824813843, + "num_tokens": 561024652.0, + "step": 21686 + }, + { + "epoch": 2.381616516582473, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5925657749176025, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7176653742790222, + "num_tokens": 561049401.0, + "step": 21687 + }, + { + "epoch": 2.3817263342850867, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.778434991836548, + "learning_rate": 1e-06, + "loss": 0.9301, + "mean_token_accuracy": 0.7226715087890625, + "num_tokens": 561071127.0, + "step": 21688 + }, + { + "epoch": 2.3818361519877005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5625548362731934, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6909980177879333, + "num_tokens": 561095712.0, + "step": 21689 + }, + { + "epoch": 2.3819459696903142, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6537861824035645, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7198712825775146, + "num_tokens": 561118285.0, + "step": 21690 + }, + { + "epoch": 2.3820557873929276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.42691969871521, + "learning_rate": 1e-06, + "loss": 1.0978, + "mean_token_accuracy": 0.690967857837677, + "num_tokens": 561147901.0, + "step": 21691 + }, + { + "epoch": 2.3821656050955413, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.403627395629883, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7027861475944519, + "num_tokens": 561176893.0, + "step": 21692 + }, + { + "epoch": 2.382275422798155, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4858529567718506, + "learning_rate": 1e-06, + "loss": 1.1474, + "mean_token_accuracy": 0.6685819625854492, + "num_tokens": 561207332.0, + "step": 21693 + }, + { + "epoch": 2.382385240500769, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1092312335968018, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.71523517370224, + "num_tokens": 561228511.0, + "step": 21694 + }, + { + "epoch": 2.382495058203382, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.688065767288208, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.7032231688499451, + "num_tokens": 561256955.0, + "step": 21695 + }, + { + "epoch": 2.382604875905996, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.701504945755005, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7126918435096741, + "num_tokens": 561279513.0, + "step": 21696 + }, + { + "epoch": 2.3827146936086097, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6783838272094727, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6986871957778931, + "num_tokens": 561307164.0, + "step": 21697 + }, + { + "epoch": 2.3828245113112234, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4662418365478516, + "learning_rate": 1e-06, + "loss": 1.1572, + "mean_token_accuracy": 0.6720684766769409, + "num_tokens": 561337569.0, + "step": 21698 + }, + { + "epoch": 2.382934329013837, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4789934158325195, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6980350017547607, + "num_tokens": 561367097.0, + "step": 21699 + }, + { + "epoch": 2.3830441467164505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.516467809677124, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7016474604606628, + "num_tokens": 561394499.0, + "step": 21700 + }, + { + "epoch": 2.3831539644190642, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.532024383544922, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7049623727798462, + "num_tokens": 561422943.0, + "step": 21701 + }, + { + "epoch": 2.383263782121678, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.592529773712158, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7210433483123779, + "num_tokens": 561446362.0, + "step": 21702 + }, + { + "epoch": 2.3833735998242918, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6194283962249756, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6975065469741821, + "num_tokens": 561471673.0, + "step": 21703 + }, + { + "epoch": 2.3834834175269055, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.949666976928711, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7133746147155762, + "num_tokens": 561491792.0, + "step": 21704 + }, + { + "epoch": 2.383593235229519, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4396743774414062, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.6845523118972778, + "num_tokens": 561523490.0, + "step": 21705 + }, + { + "epoch": 2.3837030529321326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6983962059020996, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7095621824264526, + "num_tokens": 561546471.0, + "step": 21706 + }, + { + "epoch": 2.3838128706347463, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.69669771194458, + "learning_rate": 1e-06, + "loss": 0.9135, + "mean_token_accuracy": 0.7279461026191711, + "num_tokens": 561568689.0, + "step": 21707 + }, + { + "epoch": 2.38392268833736, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.583176374435425, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7024171352386475, + "num_tokens": 561595060.0, + "step": 21708 + }, + { + "epoch": 2.384032506039974, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6513659954071045, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7239763736724854, + "num_tokens": 561619795.0, + "step": 21709 + }, + { + "epoch": 2.384142323742587, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.3752856254577637, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7225576043128967, + "num_tokens": 561636235.0, + "step": 21710 + }, + { + "epoch": 2.384252141445201, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3532915115356445, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7281848192214966, + "num_tokens": 561665324.0, + "step": 21711 + }, + { + "epoch": 2.3843619591478147, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7557971477508545, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6919025182723999, + "num_tokens": 561686989.0, + "step": 21712 + }, + { + "epoch": 2.3844717768504284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.474365234375, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.7003963589668274, + "num_tokens": 561713475.0, + "step": 21713 + }, + { + "epoch": 2.384581594553042, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4608101844787598, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7090198397636414, + "num_tokens": 561740807.0, + "step": 21714 + }, + { + "epoch": 2.3846914122556555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7688868045806885, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7214446663856506, + "num_tokens": 561764139.0, + "step": 21715 + }, + { + "epoch": 2.3848012299582693, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.741032361984253, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7225354909896851, + "num_tokens": 561791526.0, + "step": 21716 + }, + { + "epoch": 2.384911047660883, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.837014675140381, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7106810808181763, + "num_tokens": 561813069.0, + "step": 21717 + }, + { + "epoch": 2.385020865363497, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6298909187316895, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.718783438205719, + "num_tokens": 561838720.0, + "step": 21718 + }, + { + "epoch": 2.38513068306611, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596318244934082, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.714788019657135, + "num_tokens": 561864126.0, + "step": 21719 + }, + { + "epoch": 2.385240500768724, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6390833854675293, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7089216709136963, + "num_tokens": 561889744.0, + "step": 21720 + }, + { + "epoch": 2.3853503184713376, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5117809772491455, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7090084552764893, + "num_tokens": 561914889.0, + "step": 21721 + }, + { + "epoch": 2.3854601361739514, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6403143405914307, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.705703616142273, + "num_tokens": 561939297.0, + "step": 21722 + }, + { + "epoch": 2.3855699538765647, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5543437004089355, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7099751234054565, + "num_tokens": 561965447.0, + "step": 21723 + }, + { + "epoch": 2.3856797715791784, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7090790271759033, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6780033707618713, + "num_tokens": 561994561.0, + "step": 21724 + }, + { + "epoch": 2.385789589281792, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6708359718322754, + "learning_rate": 1e-06, + "loss": 0.9155, + "mean_token_accuracy": 0.7223013043403625, + "num_tokens": 562028541.0, + "step": 21725 + }, + { + "epoch": 2.385899406984406, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.394749641418457, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7139493227005005, + "num_tokens": 562059217.0, + "step": 21726 + }, + { + "epoch": 2.3860092246870197, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4946751594543457, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.717167854309082, + "num_tokens": 562085241.0, + "step": 21727 + }, + { + "epoch": 2.386119042389633, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.643965721130371, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.7259630560874939, + "num_tokens": 562109433.0, + "step": 21728 + }, + { + "epoch": 2.3862288600922468, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.985724925994873, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7119501829147339, + "num_tokens": 562130464.0, + "step": 21729 + }, + { + "epoch": 2.3863386777948605, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7442703247070312, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.718735933303833, + "num_tokens": 562154334.0, + "step": 21730 + }, + { + "epoch": 2.3864484954974743, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6596662998199463, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.7267522811889648, + "num_tokens": 562179353.0, + "step": 21731 + }, + { + "epoch": 2.386558313200088, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0135295391082764, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7069398164749146, + "num_tokens": 562209174.0, + "step": 21732 + }, + { + "epoch": 2.3866681309027014, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.492201566696167, + "learning_rate": 1e-06, + "loss": 0.9475, + "mean_token_accuracy": 0.7235068678855896, + "num_tokens": 562238300.0, + "step": 21733 + }, + { + "epoch": 2.386777948605315, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.471534490585327, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.699425220489502, + "num_tokens": 562267018.0, + "step": 21734 + }, + { + "epoch": 2.386887766307929, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.449960708618164, + "learning_rate": 1e-06, + "loss": 0.9169, + "mean_token_accuracy": 0.7374119758605957, + "num_tokens": 562296651.0, + "step": 21735 + }, + { + "epoch": 2.3869975840105426, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6608474254608154, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.6962823271751404, + "num_tokens": 562322156.0, + "step": 21736 + }, + { + "epoch": 2.3871074017131564, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.749159812927246, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7147791385650635, + "num_tokens": 562344786.0, + "step": 21737 + }, + { + "epoch": 2.3872172194157697, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7006499767303467, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7106726169586182, + "num_tokens": 562368680.0, + "step": 21738 + }, + { + "epoch": 2.3873270371183835, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4589502811431885, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.6876140832901001, + "num_tokens": 562396086.0, + "step": 21739 + }, + { + "epoch": 2.387436854820997, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.388796806335449, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7092108726501465, + "num_tokens": 562424322.0, + "step": 21740 + }, + { + "epoch": 2.387546672523611, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5647716522216797, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.682152509689331, + "num_tokens": 562449818.0, + "step": 21741 + }, + { + "epoch": 2.3876564902262243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4612298011779785, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7056652307510376, + "num_tokens": 562479687.0, + "step": 21742 + }, + { + "epoch": 2.387766307928838, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8871889114379883, + "learning_rate": 1e-06, + "loss": 0.918, + "mean_token_accuracy": 0.7304173707962036, + "num_tokens": 562500708.0, + "step": 21743 + }, + { + "epoch": 2.387876125631452, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.761291980743408, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6947414875030518, + "num_tokens": 562524041.0, + "step": 21744 + }, + { + "epoch": 2.3879859433340656, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8096954822540283, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.7047064304351807, + "num_tokens": 562551157.0, + "step": 21745 + }, + { + "epoch": 2.388095761036679, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.51196026802063, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7179054021835327, + "num_tokens": 562583952.0, + "step": 21746 + }, + { + "epoch": 2.3882055787392926, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2235257625579834, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7154404520988464, + "num_tokens": 562606927.0, + "step": 21747 + }, + { + "epoch": 2.3883153964419064, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.840277910232544, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7076551914215088, + "num_tokens": 562627104.0, + "step": 21748 + }, + { + "epoch": 2.38842521414452, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7210025787353516, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7032853364944458, + "num_tokens": 562649867.0, + "step": 21749 + }, + { + "epoch": 2.388535031847134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.907170534133911, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6858488917350769, + "num_tokens": 562672739.0, + "step": 21750 + }, + { + "epoch": 2.388644849549747, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7124011516571045, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6913677453994751, + "num_tokens": 562699666.0, + "step": 21751 + }, + { + "epoch": 2.388754667252361, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9203858375549316, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7231117486953735, + "num_tokens": 562719402.0, + "step": 21752 + }, + { + "epoch": 2.3888644849549747, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7485549449920654, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.6971257925033569, + "num_tokens": 562743460.0, + "step": 21753 + }, + { + "epoch": 2.3889743026575885, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8695709705352783, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.7013271450996399, + "num_tokens": 562766855.0, + "step": 21754 + }, + { + "epoch": 2.3890841203602022, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.90789532661438, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7165865898132324, + "num_tokens": 562789578.0, + "step": 21755 + }, + { + "epoch": 2.3891939380628155, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505896806716919, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7168538570404053, + "num_tokens": 562816256.0, + "step": 21756 + }, + { + "epoch": 2.3893037557654293, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.670449733734131, + "learning_rate": 1e-06, + "loss": 1.134, + "mean_token_accuracy": 0.676817774772644, + "num_tokens": 562842188.0, + "step": 21757 + }, + { + "epoch": 2.389413573468043, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5032312870025635, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.709996223449707, + "num_tokens": 562870493.0, + "step": 21758 + }, + { + "epoch": 2.389523391170657, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.716451406478882, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7086454629898071, + "num_tokens": 562896126.0, + "step": 21759 + }, + { + "epoch": 2.3896332088732706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8652400970458984, + "learning_rate": 1e-06, + "loss": 0.95, + "mean_token_accuracy": 0.7248666286468506, + "num_tokens": 562916736.0, + "step": 21760 + }, + { + "epoch": 2.389743026575884, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4664618968963623, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6977559328079224, + "num_tokens": 562948370.0, + "step": 21761 + }, + { + "epoch": 2.3898528442784976, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.63484525680542, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7228661775588989, + "num_tokens": 562974389.0, + "step": 21762 + }, + { + "epoch": 2.3899626619811114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.514211416244507, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7103367447853088, + "num_tokens": 563001419.0, + "step": 21763 + }, + { + "epoch": 2.390072479683725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.573112726211548, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6892528533935547, + "num_tokens": 563028210.0, + "step": 21764 + }, + { + "epoch": 2.390182297386339, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.541411876678467, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.714250385761261, + "num_tokens": 563052099.0, + "step": 21765 + }, + { + "epoch": 2.3902921150889522, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.064146041870117, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6996806859970093, + "num_tokens": 563074501.0, + "step": 21766 + }, + { + "epoch": 2.390401932791566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.408024787902832, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.707679271697998, + "num_tokens": 563105438.0, + "step": 21767 + }, + { + "epoch": 2.3905117504941797, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7161309719085693, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7135884165763855, + "num_tokens": 563127829.0, + "step": 21768 + }, + { + "epoch": 2.3906215681967935, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7471415996551514, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7106960415840149, + "num_tokens": 563150649.0, + "step": 21769 + }, + { + "epoch": 2.390731385899407, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.043215036392212, + "learning_rate": 1e-06, + "loss": 1.0735, + "mean_token_accuracy": 0.6957535743713379, + "num_tokens": 563174548.0, + "step": 21770 + }, + { + "epoch": 2.3908412036020206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2891359329223633, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6907089948654175, + "num_tokens": 563211815.0, + "step": 21771 + }, + { + "epoch": 2.3909510213046343, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4714739322662354, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6936689615249634, + "num_tokens": 563240838.0, + "step": 21772 + }, + { + "epoch": 2.391060839007248, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.608449697494507, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7095210552215576, + "num_tokens": 563266701.0, + "step": 21773 + }, + { + "epoch": 2.3911706567098614, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5857181549072266, + "learning_rate": 1e-06, + "loss": 0.9087, + "mean_token_accuracy": 0.7317917346954346, + "num_tokens": 563290876.0, + "step": 21774 + }, + { + "epoch": 2.391280474412475, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9045045375823975, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.7249857783317566, + "num_tokens": 563310841.0, + "step": 21775 + }, + { + "epoch": 2.391390292115089, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6647961139678955, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7283843159675598, + "num_tokens": 563335458.0, + "step": 21776 + }, + { + "epoch": 2.3915001098177027, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.517082691192627, + "learning_rate": 1e-06, + "loss": 0.9206, + "mean_token_accuracy": 0.7285705804824829, + "num_tokens": 563362466.0, + "step": 21777 + }, + { + "epoch": 2.3916099275203164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4302103519439697, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7113939523696899, + "num_tokens": 563391871.0, + "step": 21778 + }, + { + "epoch": 2.3917197452229297, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6450607776641846, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.6998055577278137, + "num_tokens": 563417280.0, + "step": 21779 + }, + { + "epoch": 2.3918295629255435, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7548465728759766, + "learning_rate": 1e-06, + "loss": 1.0765, + "mean_token_accuracy": 0.6913049221038818, + "num_tokens": 563446860.0, + "step": 21780 + }, + { + "epoch": 2.3919393806281573, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.499154806137085, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7072097063064575, + "num_tokens": 563476694.0, + "step": 21781 + }, + { + "epoch": 2.392049198330771, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5878548622131348, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.723372757434845, + "num_tokens": 563501967.0, + "step": 21782 + }, + { + "epoch": 2.3921590160333848, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3624799251556396, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.7015623450279236, + "num_tokens": 563533006.0, + "step": 21783 + }, + { + "epoch": 2.392268833735998, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6695237159729004, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7321233153343201, + "num_tokens": 563556085.0, + "step": 21784 + }, + { + "epoch": 2.392378651438612, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.483870267868042, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.709986686706543, + "num_tokens": 563583818.0, + "step": 21785 + }, + { + "epoch": 2.3924884691412256, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.645026922225952, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.700498104095459, + "num_tokens": 563611700.0, + "step": 21786 + }, + { + "epoch": 2.3925982868438394, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8264107704162598, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.7073963284492493, + "num_tokens": 563635091.0, + "step": 21787 + }, + { + "epoch": 2.392708104546453, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.40596079826355, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7101529836654663, + "num_tokens": 563664243.0, + "step": 21788 + }, + { + "epoch": 2.3928179222490664, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7792203426361084, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7138078212738037, + "num_tokens": 563687430.0, + "step": 21789 + }, + { + "epoch": 2.39292773995168, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.462595224380493, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7064261436462402, + "num_tokens": 563716757.0, + "step": 21790 + }, + { + "epoch": 2.393037557654294, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7266879081726074, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7275084257125854, + "num_tokens": 563739681.0, + "step": 21791 + }, + { + "epoch": 2.3931473753569077, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6269309520721436, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7181239128112793, + "num_tokens": 563763013.0, + "step": 21792 + }, + { + "epoch": 2.393257193059521, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6637680530548096, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6886705756187439, + "num_tokens": 563789203.0, + "step": 21793 + }, + { + "epoch": 2.3933670107621348, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.684567451477051, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.708764910697937, + "num_tokens": 563813546.0, + "step": 21794 + }, + { + "epoch": 2.3934768284647485, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.537321090698242, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.6999614238739014, + "num_tokens": 563840301.0, + "step": 21795 + }, + { + "epoch": 2.3935866461673623, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.549410104751587, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.7032658457756042, + "num_tokens": 563867125.0, + "step": 21796 + }, + { + "epoch": 2.393696463869976, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.821880578994751, + "learning_rate": 1e-06, + "loss": 0.9977, + "mean_token_accuracy": 0.7082558870315552, + "num_tokens": 563888995.0, + "step": 21797 + }, + { + "epoch": 2.3938062815725893, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5747146606445312, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7152928709983826, + "num_tokens": 563914370.0, + "step": 21798 + }, + { + "epoch": 2.393916099275203, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5881214141845703, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7225905656814575, + "num_tokens": 563941515.0, + "step": 21799 + }, + { + "epoch": 2.394025916977817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.663249969482422, + "learning_rate": 1e-06, + "loss": 0.9485, + "mean_token_accuracy": 0.726311206817627, + "num_tokens": 563967176.0, + "step": 21800 + }, + { + "epoch": 2.3941357346804306, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7361345291137695, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7215182185173035, + "num_tokens": 563989351.0, + "step": 21801 + }, + { + "epoch": 2.394245552383044, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4764609336853027, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7152758836746216, + "num_tokens": 564015982.0, + "step": 21802 + }, + { + "epoch": 2.3943553700856577, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.469245672225952, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7213504314422607, + "num_tokens": 564043171.0, + "step": 21803 + }, + { + "epoch": 2.3944651877882714, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.432269811630249, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.698309063911438, + "num_tokens": 564073133.0, + "step": 21804 + }, + { + "epoch": 2.394575005490885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.440950393676758, + "learning_rate": 1e-06, + "loss": 0.9089, + "mean_token_accuracy": 0.7417504787445068, + "num_tokens": 564099305.0, + "step": 21805 + }, + { + "epoch": 2.394684823193499, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.548943281173706, + "learning_rate": 1e-06, + "loss": 0.9543, + "mean_token_accuracy": 0.7218447923660278, + "num_tokens": 564124842.0, + "step": 21806 + }, + { + "epoch": 2.3947946408961123, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4010281562805176, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6888017654418945, + "num_tokens": 564158542.0, + "step": 21807 + }, + { + "epoch": 2.394904458598726, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.642885208129883, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7074365615844727, + "num_tokens": 564185746.0, + "step": 21808 + }, + { + "epoch": 2.39501427630134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6950416564941406, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7140729427337646, + "num_tokens": 564208325.0, + "step": 21809 + }, + { + "epoch": 2.3951240940039535, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3817715644836426, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.708206057548523, + "num_tokens": 564237384.0, + "step": 21810 + }, + { + "epoch": 2.3952339117065673, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6879703998565674, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6954669952392578, + "num_tokens": 564263098.0, + "step": 21811 + }, + { + "epoch": 2.3953437294091806, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7032437324523926, + "learning_rate": 1e-06, + "loss": 1.0483, + "mean_token_accuracy": 0.6897838711738586, + "num_tokens": 564287499.0, + "step": 21812 + }, + { + "epoch": 2.3954535471117944, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9933691024780273, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7109101414680481, + "num_tokens": 564307598.0, + "step": 21813 + }, + { + "epoch": 2.395563364814408, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.511589527130127, + "learning_rate": 1e-06, + "loss": 1.0983, + "mean_token_accuracy": 0.687706470489502, + "num_tokens": 564338375.0, + "step": 21814 + }, + { + "epoch": 2.395673182517022, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6759889125823975, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7118183374404907, + "num_tokens": 564361749.0, + "step": 21815 + }, + { + "epoch": 2.3957830002196356, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.640512704849243, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7012186050415039, + "num_tokens": 564387269.0, + "step": 21816 + }, + { + "epoch": 2.395892817922249, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.635124683380127, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6958041787147522, + "num_tokens": 564411993.0, + "step": 21817 + }, + { + "epoch": 2.3960026356248627, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.464940071105957, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7181111574172974, + "num_tokens": 564437303.0, + "step": 21818 + }, + { + "epoch": 2.3961124533274765, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.446331024169922, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7099729180335999, + "num_tokens": 564467197.0, + "step": 21819 + }, + { + "epoch": 2.3962222710300902, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.1976804733276367, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.7001618146896362, + "num_tokens": 564501039.0, + "step": 21820 + }, + { + "epoch": 2.3963320887327035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.45926570892334, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.701400101184845, + "num_tokens": 564529656.0, + "step": 21821 + }, + { + "epoch": 2.3964419064353173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3384103775024414, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.711315929889679, + "num_tokens": 564560411.0, + "step": 21822 + }, + { + "epoch": 2.396551724137931, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.363250970840454, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7131545543670654, + "num_tokens": 564589332.0, + "step": 21823 + }, + { + "epoch": 2.396661541840545, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.478196859359741, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7208216190338135, + "num_tokens": 564617369.0, + "step": 21824 + }, + { + "epoch": 2.396771359543158, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.728444814682007, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7018653154373169, + "num_tokens": 564640388.0, + "step": 21825 + }, + { + "epoch": 2.396881177245772, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.600675106048584, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6910117268562317, + "num_tokens": 564664886.0, + "step": 21826 + }, + { + "epoch": 2.3969909949483856, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.612394094467163, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.6978639364242554, + "num_tokens": 564693525.0, + "step": 21827 + }, + { + "epoch": 2.3971008126509994, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.680295944213867, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7011063098907471, + "num_tokens": 564721609.0, + "step": 21828 + }, + { + "epoch": 2.397210630353613, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.638516902923584, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7101829051971436, + "num_tokens": 564747590.0, + "step": 21829 + }, + { + "epoch": 2.3973204480562265, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3688466548919678, + "learning_rate": 1e-06, + "loss": 1.1029, + "mean_token_accuracy": 0.6781383752822876, + "num_tokens": 564778965.0, + "step": 21830 + }, + { + "epoch": 2.39743026575884, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.956403970718384, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7183468341827393, + "num_tokens": 564800943.0, + "step": 21831 + }, + { + "epoch": 2.397540083461454, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7120566368103027, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6881011724472046, + "num_tokens": 564826490.0, + "step": 21832 + }, + { + "epoch": 2.3976499011640677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.590416669845581, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.6947664022445679, + "num_tokens": 564852336.0, + "step": 21833 + }, + { + "epoch": 2.3977597188666815, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.576056480407715, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.714792013168335, + "num_tokens": 564876595.0, + "step": 21834 + }, + { + "epoch": 2.397869536569295, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5453264713287354, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7051694393157959, + "num_tokens": 564906207.0, + "step": 21835 + }, + { + "epoch": 2.3979793542719086, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0964367389678955, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7206277847290039, + "num_tokens": 564925607.0, + "step": 21836 + }, + { + "epoch": 2.3980891719745223, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.859203338623047, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7236171960830688, + "num_tokens": 564948981.0, + "step": 21837 + }, + { + "epoch": 2.398198989677136, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4487011432647705, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.695512056350708, + "num_tokens": 564980853.0, + "step": 21838 + }, + { + "epoch": 2.39830880737975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.518549919128418, + "learning_rate": 1e-06, + "loss": 1.0611, + "mean_token_accuracy": 0.6902449131011963, + "num_tokens": 565010521.0, + "step": 21839 + }, + { + "epoch": 2.398418625082363, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8967766761779785, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7294108867645264, + "num_tokens": 565030037.0, + "step": 21840 + }, + { + "epoch": 2.398528442784977, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8112590312957764, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7025437951087952, + "num_tokens": 565056268.0, + "step": 21841 + }, + { + "epoch": 2.3986382604875907, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.788403034210205, + "learning_rate": 1e-06, + "loss": 1.0873, + "mean_token_accuracy": 0.6942389011383057, + "num_tokens": 565082378.0, + "step": 21842 + }, + { + "epoch": 2.3987480781902044, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2151169776916504, + "learning_rate": 1e-06, + "loss": 0.9295, + "mean_token_accuracy": 0.7205397486686707, + "num_tokens": 565100237.0, + "step": 21843 + }, + { + "epoch": 2.398857895892818, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6404058933258057, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.6936757564544678, + "num_tokens": 565125361.0, + "step": 21844 + }, + { + "epoch": 2.3989677135954315, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8871352672576904, + "learning_rate": 1e-06, + "loss": 0.8725, + "mean_token_accuracy": 0.7376846671104431, + "num_tokens": 565146362.0, + "step": 21845 + }, + { + "epoch": 2.3990775312980452, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7530364990234375, + "learning_rate": 1e-06, + "loss": 0.9111, + "mean_token_accuracy": 0.7282428741455078, + "num_tokens": 565168609.0, + "step": 21846 + }, + { + "epoch": 2.399187349000659, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.679840564727783, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.6978161334991455, + "num_tokens": 565192456.0, + "step": 21847 + }, + { + "epoch": 2.3992971667032728, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5823910236358643, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6952410936355591, + "num_tokens": 565221728.0, + "step": 21848 + }, + { + "epoch": 2.399406984405886, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.420506238937378, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7178182601928711, + "num_tokens": 565253334.0, + "step": 21849 + }, + { + "epoch": 2.3995168021085, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5789005756378174, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6961046457290649, + "num_tokens": 565280534.0, + "step": 21850 + }, + { + "epoch": 2.3996266198111136, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.634800672531128, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.70819091796875, + "num_tokens": 565307008.0, + "step": 21851 + }, + { + "epoch": 2.3997364375137273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7000787258148193, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7161412835121155, + "num_tokens": 565329287.0, + "step": 21852 + }, + { + "epoch": 2.3998462552163407, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.389403820037842, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7217097282409668, + "num_tokens": 565358958.0, + "step": 21853 + }, + { + "epoch": 2.3999560729189544, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3921000957489014, + "learning_rate": 1e-06, + "loss": 0.9454, + "mean_token_accuracy": 0.7242276668548584, + "num_tokens": 565388515.0, + "step": 21854 + }, + { + "epoch": 2.400065890621568, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7856597900390625, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7110973000526428, + "num_tokens": 565411671.0, + "step": 21855 + }, + { + "epoch": 2.400175708324182, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.662036418914795, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7220746874809265, + "num_tokens": 565434986.0, + "step": 21856 + }, + { + "epoch": 2.4002855260267957, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 20.431264877319336, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.702742874622345, + "num_tokens": 565460261.0, + "step": 21857 + }, + { + "epoch": 2.400395343729409, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8870320320129395, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6949496269226074, + "num_tokens": 565489292.0, + "step": 21858 + }, + { + "epoch": 2.4005051614320227, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6368162631988525, + "learning_rate": 1e-06, + "loss": 0.9261, + "mean_token_accuracy": 0.7342714071273804, + "num_tokens": 565512923.0, + "step": 21859 + }, + { + "epoch": 2.4006149791346365, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5479447841644287, + "learning_rate": 1e-06, + "loss": 1.11, + "mean_token_accuracy": 0.6798214316368103, + "num_tokens": 565542048.0, + "step": 21860 + }, + { + "epoch": 2.4007247968372503, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4646406173706055, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6968196034431458, + "num_tokens": 565571858.0, + "step": 21861 + }, + { + "epoch": 2.400834614539864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7078402042388916, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6955763101577759, + "num_tokens": 565597292.0, + "step": 21862 + }, + { + "epoch": 2.4009444322424773, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5354089736938477, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.71747225522995, + "num_tokens": 565622850.0, + "step": 21863 + }, + { + "epoch": 2.401054249945091, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7203872203826904, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.6992576122283936, + "num_tokens": 565645293.0, + "step": 21864 + }, + { + "epoch": 2.401164067647705, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5274927616119385, + "learning_rate": 1e-06, + "loss": 0.9003, + "mean_token_accuracy": 0.7452639937400818, + "num_tokens": 565673669.0, + "step": 21865 + }, + { + "epoch": 2.4012738853503186, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3715052604675293, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.6857354640960693, + "num_tokens": 565703567.0, + "step": 21866 + }, + { + "epoch": 2.4013837030529324, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.593212842941284, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6976720094680786, + "num_tokens": 565728942.0, + "step": 21867 + }, + { + "epoch": 2.4014935207555457, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.369990825653076, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7099356651306152, + "num_tokens": 565760420.0, + "step": 21868 + }, + { + "epoch": 2.4016033384581594, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.968085527420044, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7187157869338989, + "num_tokens": 565782604.0, + "step": 21869 + }, + { + "epoch": 2.401713156160773, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.842985153198242, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6987902522087097, + "num_tokens": 565803210.0, + "step": 21870 + }, + { + "epoch": 2.401822973863387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.483396530151367, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7058305144309998, + "num_tokens": 565830695.0, + "step": 21871 + }, + { + "epoch": 2.4019327915660003, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565544843673706, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7048735618591309, + "num_tokens": 565855945.0, + "step": 21872 + }, + { + "epoch": 2.402042609268614, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 8.067316055297852, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7221497893333435, + "num_tokens": 565875323.0, + "step": 21873 + }, + { + "epoch": 2.4021524269712278, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.597999334335327, + "learning_rate": 1e-06, + "loss": 1.116, + "mean_token_accuracy": 0.6840766668319702, + "num_tokens": 565904005.0, + "step": 21874 + }, + { + "epoch": 2.4022622446738415, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.68942928314209, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.698596715927124, + "num_tokens": 565933325.0, + "step": 21875 + }, + { + "epoch": 2.402372062376455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9853663444519043, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7090803980827332, + "num_tokens": 565952176.0, + "step": 21876 + }, + { + "epoch": 2.4024818800790686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8906965255737305, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7190823554992676, + "num_tokens": 565975579.0, + "step": 21877 + }, + { + "epoch": 2.4025916977816824, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.437659502029419, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6953938007354736, + "num_tokens": 566003823.0, + "step": 21878 + }, + { + "epoch": 2.402701515484296, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.503018856048584, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7146492004394531, + "num_tokens": 566031231.0, + "step": 21879 + }, + { + "epoch": 2.40281133318691, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8977198600769043, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7094246745109558, + "num_tokens": 566053359.0, + "step": 21880 + }, + { + "epoch": 2.402921150889523, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5996570587158203, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7190261483192444, + "num_tokens": 566081184.0, + "step": 21881 + }, + { + "epoch": 2.403030968592137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.729297161102295, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6980994939804077, + "num_tokens": 566107003.0, + "step": 21882 + }, + { + "epoch": 2.4031407862947507, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.639521598815918, + "learning_rate": 1e-06, + "loss": 1.0293, + "mean_token_accuracy": 0.7023004293441772, + "num_tokens": 566131996.0, + "step": 21883 + }, + { + "epoch": 2.4032506039973645, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5747885704040527, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7168393135070801, + "num_tokens": 566156993.0, + "step": 21884 + }, + { + "epoch": 2.403360421699978, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.508039712905884, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6955459713935852, + "num_tokens": 566181627.0, + "step": 21885 + }, + { + "epoch": 2.4034702394025915, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.660346746444702, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7032985687255859, + "num_tokens": 566207643.0, + "step": 21886 + }, + { + "epoch": 2.4035800571052053, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.891340494155884, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7059760689735413, + "num_tokens": 566232623.0, + "step": 21887 + }, + { + "epoch": 2.403689874807819, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.54363751411438, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6968176364898682, + "num_tokens": 566259132.0, + "step": 21888 + }, + { + "epoch": 2.403799692510433, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.716184139251709, + "learning_rate": 1e-06, + "loss": 0.9233, + "mean_token_accuracy": 0.7271113395690918, + "num_tokens": 566283076.0, + "step": 21889 + }, + { + "epoch": 2.4039095102130466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.891064167022705, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7154418230056763, + "num_tokens": 566307400.0, + "step": 21890 + }, + { + "epoch": 2.40401932791566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0849602222442627, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7141525745391846, + "num_tokens": 566326041.0, + "step": 21891 + }, + { + "epoch": 2.4041291456182736, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.630582332611084, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7092416286468506, + "num_tokens": 566355322.0, + "step": 21892 + }, + { + "epoch": 2.4042389633208874, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.572807788848877, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7076764702796936, + "num_tokens": 566381855.0, + "step": 21893 + }, + { + "epoch": 2.404348781023501, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7043044567108154, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7241604328155518, + "num_tokens": 566405157.0, + "step": 21894 + }, + { + "epoch": 2.404458598726115, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6816084384918213, + "learning_rate": 1e-06, + "loss": 0.9114, + "mean_token_accuracy": 0.7294414043426514, + "num_tokens": 566432532.0, + "step": 21895 + }, + { + "epoch": 2.404568416428728, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9796528816223145, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7219802141189575, + "num_tokens": 566452057.0, + "step": 21896 + }, + { + "epoch": 2.404678234131342, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8853306770324707, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7193621397018433, + "num_tokens": 566472312.0, + "step": 21897 + }, + { + "epoch": 2.4047880518339557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6001460552215576, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7151140570640564, + "num_tokens": 566498063.0, + "step": 21898 + }, + { + "epoch": 2.4048978695365695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9301223754882812, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7116385698318481, + "num_tokens": 566520570.0, + "step": 21899 + }, + { + "epoch": 2.405007687239183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.530207872390747, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.704590380191803, + "num_tokens": 566547892.0, + "step": 21900 + }, + { + "epoch": 2.4051175049417965, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7768638134002686, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7173449993133545, + "num_tokens": 566569457.0, + "step": 21901 + }, + { + "epoch": 2.4052273226444103, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2408206462860107, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7043943405151367, + "num_tokens": 566601355.0, + "step": 21902 + }, + { + "epoch": 2.405337140347024, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6594491004943848, + "learning_rate": 1e-06, + "loss": 0.9234, + "mean_token_accuracy": 0.7236852645874023, + "num_tokens": 566622944.0, + "step": 21903 + }, + { + "epoch": 2.4054469580496374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.920697212219238, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6970189213752747, + "num_tokens": 566649726.0, + "step": 21904 + }, + { + "epoch": 2.405556775752251, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.743673086166382, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.6916552782058716, + "num_tokens": 566676527.0, + "step": 21905 + }, + { + "epoch": 2.405666593454865, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.545159339904785, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7139822244644165, + "num_tokens": 566703545.0, + "step": 21906 + }, + { + "epoch": 2.4057764111574786, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.564704656600952, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.7255128026008606, + "num_tokens": 566730234.0, + "step": 21907 + }, + { + "epoch": 2.4058862288600924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.537581205368042, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7090312242507935, + "num_tokens": 566755985.0, + "step": 21908 + }, + { + "epoch": 2.4059960465627057, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.61696195602417, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7028640508651733, + "num_tokens": 566780924.0, + "step": 21909 + }, + { + "epoch": 2.4061058642653195, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.632009744644165, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.7230660915374756, + "num_tokens": 566805480.0, + "step": 21910 + }, + { + "epoch": 2.4062156819679332, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7189486026763916, + "learning_rate": 1e-06, + "loss": 1.1382, + "mean_token_accuracy": 0.6759551167488098, + "num_tokens": 566832750.0, + "step": 21911 + }, + { + "epoch": 2.406325499670547, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6969387531280518, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7196974754333496, + "num_tokens": 566855160.0, + "step": 21912 + }, + { + "epoch": 2.4064353173731607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7455737590789795, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7185553908348083, + "num_tokens": 566876672.0, + "step": 21913 + }, + { + "epoch": 2.406545135075774, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.497642993927002, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6943942904472351, + "num_tokens": 566906330.0, + "step": 21914 + }, + { + "epoch": 2.406654952778388, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.595057964324951, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6918796300888062, + "num_tokens": 566930922.0, + "step": 21915 + }, + { + "epoch": 2.4067647704810016, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559074878692627, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.7247904539108276, + "num_tokens": 566957596.0, + "step": 21916 + }, + { + "epoch": 2.4068745881836153, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6939949989318848, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7053657174110413, + "num_tokens": 566982957.0, + "step": 21917 + }, + { + "epoch": 2.406984405886229, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.408937454223633, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7089146375656128, + "num_tokens": 567006482.0, + "step": 21918 + }, + { + "epoch": 2.4070942235888424, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5969889163970947, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.715730607509613, + "num_tokens": 567032029.0, + "step": 21919 + }, + { + "epoch": 2.407204041291456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.652740716934204, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.7087811231613159, + "num_tokens": 567057629.0, + "step": 21920 + }, + { + "epoch": 2.40731385899407, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.568854570388794, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7026283144950867, + "num_tokens": 567085534.0, + "step": 21921 + }, + { + "epoch": 2.4074236766966837, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3417232036590576, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6962913274765015, + "num_tokens": 567118239.0, + "step": 21922 + }, + { + "epoch": 2.407533494399297, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7757930755615234, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.722808837890625, + "num_tokens": 567141682.0, + "step": 21923 + }, + { + "epoch": 2.4076433121019107, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3308684825897217, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7078657150268555, + "num_tokens": 567174548.0, + "step": 21924 + }, + { + "epoch": 2.4077531298045245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.714695930480957, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7232041358947754, + "num_tokens": 567197996.0, + "step": 21925 + }, + { + "epoch": 2.4078629475071383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3308780193328857, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.6995090246200562, + "num_tokens": 567231224.0, + "step": 21926 + }, + { + "epoch": 2.4079727652097516, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.634127140045166, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7045370936393738, + "num_tokens": 567257919.0, + "step": 21927 + }, + { + "epoch": 2.4080825829123653, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6157896518707275, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7019233107566833, + "num_tokens": 567283142.0, + "step": 21928 + }, + { + "epoch": 2.408192400614979, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584024667739868, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.690529465675354, + "num_tokens": 567308445.0, + "step": 21929 + }, + { + "epoch": 2.408302218317593, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4912590980529785, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7096102237701416, + "num_tokens": 567337026.0, + "step": 21930 + }, + { + "epoch": 2.4084120360202066, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6939027309417725, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6890978217124939, + "num_tokens": 567362751.0, + "step": 21931 + }, + { + "epoch": 2.40852185372282, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559880018234253, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7135723233222961, + "num_tokens": 567389259.0, + "step": 21932 + }, + { + "epoch": 2.4086316714254337, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6689939498901367, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7128614187240601, + "num_tokens": 567413614.0, + "step": 21933 + }, + { + "epoch": 2.4087414891280474, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.24558162689209, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7106844186782837, + "num_tokens": 567449268.0, + "step": 21934 + }, + { + "epoch": 2.408851306830661, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8741214275360107, + "learning_rate": 1e-06, + "loss": 0.9188, + "mean_token_accuracy": 0.7270383834838867, + "num_tokens": 567469816.0, + "step": 21935 + }, + { + "epoch": 2.408961124533275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5690720081329346, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6985045671463013, + "num_tokens": 567495710.0, + "step": 21936 + }, + { + "epoch": 2.4090709422358882, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.807093858718872, + "learning_rate": 1e-06, + "loss": 0.9457, + "mean_token_accuracy": 0.7198880910873413, + "num_tokens": 567518594.0, + "step": 21937 + }, + { + "epoch": 2.409180759938502, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.74045729637146, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7127236127853394, + "num_tokens": 567545070.0, + "step": 21938 + }, + { + "epoch": 2.4092905776411158, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5485925674438477, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7179616689682007, + "num_tokens": 567572442.0, + "step": 21939 + }, + { + "epoch": 2.4094003953437295, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6437008380889893, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.7011560201644897, + "num_tokens": 567597535.0, + "step": 21940 + }, + { + "epoch": 2.4095102130463433, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.568378448486328, + "learning_rate": 1e-06, + "loss": 1.1113, + "mean_token_accuracy": 0.6814290285110474, + "num_tokens": 567625135.0, + "step": 21941 + }, + { + "epoch": 2.4096200307489566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.685904026031494, + "learning_rate": 1e-06, + "loss": 0.9152, + "mean_token_accuracy": 0.7218515276908875, + "num_tokens": 567647853.0, + "step": 21942 + }, + { + "epoch": 2.4097298484515703, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.614783763885498, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6968618631362915, + "num_tokens": 567676905.0, + "step": 21943 + }, + { + "epoch": 2.409839666154184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7207908630371094, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7297898530960083, + "num_tokens": 567698642.0, + "step": 21944 + }, + { + "epoch": 2.409949483856798, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5439655780792236, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7166137099266052, + "num_tokens": 567729030.0, + "step": 21945 + }, + { + "epoch": 2.4100593015594116, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.811807870864868, + "learning_rate": 1e-06, + "loss": 0.9054, + "mean_token_accuracy": 0.7321763038635254, + "num_tokens": 567747945.0, + "step": 21946 + }, + { + "epoch": 2.410169119262025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5849008560180664, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.705516517162323, + "num_tokens": 567773337.0, + "step": 21947 + }, + { + "epoch": 2.4102789369646387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8804612159729004, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.6924120783805847, + "num_tokens": 567798204.0, + "step": 21948 + }, + { + "epoch": 2.4103887546672524, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.786870002746582, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.7015731334686279, + "num_tokens": 567824427.0, + "step": 21949 + }, + { + "epoch": 2.410498572369866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 4.297786712646484, + "learning_rate": 1e-06, + "loss": 0.9197, + "mean_token_accuracy": 0.7302865386009216, + "num_tokens": 567853385.0, + "step": 21950 + }, + { + "epoch": 2.4106083900724795, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9898250102996826, + "learning_rate": 1e-06, + "loss": 0.9734, + "mean_token_accuracy": 0.7129395604133606, + "num_tokens": 567874455.0, + "step": 21951 + }, + { + "epoch": 2.4107182077750933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3420870304107666, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7100287675857544, + "num_tokens": 567904011.0, + "step": 21952 + }, + { + "epoch": 2.410828025477707, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.638120651245117, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6924808025360107, + "num_tokens": 567929546.0, + "step": 21953 + }, + { + "epoch": 2.410937843180321, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4751572608947754, + "learning_rate": 1e-06, + "loss": 0.8843, + "mean_token_accuracy": 0.743929922580719, + "num_tokens": 567955058.0, + "step": 21954 + }, + { + "epoch": 2.411047660882934, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5417022705078125, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7233498096466064, + "num_tokens": 567982081.0, + "step": 21955 + }, + { + "epoch": 2.411157478585548, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7307732105255127, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7080428600311279, + "num_tokens": 568007640.0, + "step": 21956 + }, + { + "epoch": 2.4112672962881616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6608502864837646, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.7010600566864014, + "num_tokens": 568031630.0, + "step": 21957 + }, + { + "epoch": 2.4113771139907754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.633405923843384, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.709679126739502, + "num_tokens": 568054579.0, + "step": 21958 + }, + { + "epoch": 2.411486931693389, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.62430477142334, + "learning_rate": 1e-06, + "loss": 0.9444, + "mean_token_accuracy": 0.7242576479911804, + "num_tokens": 568078160.0, + "step": 21959 + }, + { + "epoch": 2.4115967493960024, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.63008713722229, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.6996756196022034, + "num_tokens": 568102673.0, + "step": 21960 + }, + { + "epoch": 2.411706567098616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4572694301605225, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7108237147331238, + "num_tokens": 568129707.0, + "step": 21961 + }, + { + "epoch": 2.41181638480123, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.503025531768799, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7124525904655457, + "num_tokens": 568159400.0, + "step": 21962 + }, + { + "epoch": 2.4119262025038437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5282585620880127, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.7037829756736755, + "num_tokens": 568184122.0, + "step": 21963 + }, + { + "epoch": 2.4120360202064575, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7161624431610107, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6941195726394653, + "num_tokens": 568207931.0, + "step": 21964 + }, + { + "epoch": 2.412145837909071, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2868709564208984, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7105154991149902, + "num_tokens": 568239815.0, + "step": 21965 + }, + { + "epoch": 2.4122556556116845, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2553601264953613, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6933801770210266, + "num_tokens": 568277237.0, + "step": 21966 + }, + { + "epoch": 2.4123654733142983, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2471842765808105, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6990002393722534, + "num_tokens": 568311920.0, + "step": 21967 + }, + { + "epoch": 2.412475291016912, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9333813190460205, + "learning_rate": 1e-06, + "loss": 0.8196, + "mean_token_accuracy": 0.750025749206543, + "num_tokens": 568330760.0, + "step": 21968 + }, + { + "epoch": 2.412585108719526, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.670766830444336, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.6942365765571594, + "num_tokens": 568355507.0, + "step": 21969 + }, + { + "epoch": 2.412694926422139, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8814938068389893, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6978797912597656, + "num_tokens": 568382024.0, + "step": 21970 + }, + { + "epoch": 2.412804744124753, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.742701768875122, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7048139572143555, + "num_tokens": 568405147.0, + "step": 21971 + }, + { + "epoch": 2.4129145618273666, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3934669494628906, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7013416290283203, + "num_tokens": 568433274.0, + "step": 21972 + }, + { + "epoch": 2.4130243795299804, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.570526361465454, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7092429399490356, + "num_tokens": 568457178.0, + "step": 21973 + }, + { + "epoch": 2.4131341972325937, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4409706592559814, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7250744104385376, + "num_tokens": 568487821.0, + "step": 21974 + }, + { + "epoch": 2.4132440149352075, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.732793092727661, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7008453011512756, + "num_tokens": 568511891.0, + "step": 21975 + }, + { + "epoch": 2.413353832637821, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7354118824005127, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7045332193374634, + "num_tokens": 568533065.0, + "step": 21976 + }, + { + "epoch": 2.413463650340435, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5155935287475586, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6867747902870178, + "num_tokens": 568562977.0, + "step": 21977 + }, + { + "epoch": 2.4135734680430487, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.526542901992798, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6984150409698486, + "num_tokens": 568588649.0, + "step": 21978 + }, + { + "epoch": 2.413683285745662, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9135780334472656, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7086220979690552, + "num_tokens": 568609408.0, + "step": 21979 + }, + { + "epoch": 2.413793103448276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.498525619506836, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6925775408744812, + "num_tokens": 568635159.0, + "step": 21980 + }, + { + "epoch": 2.4139029211508896, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.487992525100708, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.689210057258606, + "num_tokens": 568662805.0, + "step": 21981 + }, + { + "epoch": 2.4140127388535033, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.718987464904785, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7144694328308105, + "num_tokens": 568686176.0, + "step": 21982 + }, + { + "epoch": 2.4141225565561166, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9016053676605225, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.6971709728240967, + "num_tokens": 568706853.0, + "step": 21983 + }, + { + "epoch": 2.4142323742587304, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.404487133026123, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6951891779899597, + "num_tokens": 568736656.0, + "step": 21984 + }, + { + "epoch": 2.414342191961344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 20.679668426513672, + "learning_rate": 1e-06, + "loss": 0.8889, + "mean_token_accuracy": 0.7365449070930481, + "num_tokens": 568758166.0, + "step": 21985 + }, + { + "epoch": 2.414452009663958, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8361659049987793, + "learning_rate": 1e-06, + "loss": 0.9491, + "mean_token_accuracy": 0.7160965204238892, + "num_tokens": 568778611.0, + "step": 21986 + }, + { + "epoch": 2.4145618273665717, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.954110622406006, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7122997045516968, + "num_tokens": 568799349.0, + "step": 21987 + }, + { + "epoch": 2.414671645069185, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3810083866119385, + "learning_rate": 1e-06, + "loss": 1.1244, + "mean_token_accuracy": 0.6910351514816284, + "num_tokens": 568830941.0, + "step": 21988 + }, + { + "epoch": 2.4147814627717987, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2890467643737793, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6921683549880981, + "num_tokens": 568861571.0, + "step": 21989 + }, + { + "epoch": 2.4148912804744125, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4842793941497803, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7136640548706055, + "num_tokens": 568889511.0, + "step": 21990 + }, + { + "epoch": 2.4150010981770262, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5176010131835938, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7201827764511108, + "num_tokens": 568916393.0, + "step": 21991 + }, + { + "epoch": 2.41511091587964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6693367958068848, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.6992356777191162, + "num_tokens": 568943616.0, + "step": 21992 + }, + { + "epoch": 2.4152207335822533, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8848929405212402, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7179853916168213, + "num_tokens": 568966047.0, + "step": 21993 + }, + { + "epoch": 2.415330551284867, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8996922969818115, + "learning_rate": 1e-06, + "loss": 0.8992, + "mean_token_accuracy": 0.7324599027633667, + "num_tokens": 568987775.0, + "step": 21994 + }, + { + "epoch": 2.415440368987481, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5033652782440186, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7073575258255005, + "num_tokens": 569017181.0, + "step": 21995 + }, + { + "epoch": 2.4155501866900946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.534290075302124, + "learning_rate": 1e-06, + "loss": 0.922, + "mean_token_accuracy": 0.7265958786010742, + "num_tokens": 569043843.0, + "step": 21996 + }, + { + "epoch": 2.4156600043927083, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6706833839416504, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7037883400917053, + "num_tokens": 569070037.0, + "step": 21997 + }, + { + "epoch": 2.4157698220953217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9828858375549316, + "learning_rate": 1e-06, + "loss": 0.9328, + "mean_token_accuracy": 0.7169277667999268, + "num_tokens": 569090111.0, + "step": 21998 + }, + { + "epoch": 2.4158796397979354, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1461310386657715, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6921017169952393, + "num_tokens": 569108485.0, + "step": 21999 + }, + { + "epoch": 2.415989457500549, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6244308948516846, + "learning_rate": 1e-06, + "loss": 1.0486, + "mean_token_accuracy": 0.6911686658859253, + "num_tokens": 569134939.0, + "step": 22000 + }, + { + "epoch": 2.416099275203163, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.731562376022339, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7172378897666931, + "num_tokens": 569163020.0, + "step": 22001 + }, + { + "epoch": 2.4162090929057762, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.673391342163086, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.709819495677948, + "num_tokens": 569188176.0, + "step": 22002 + }, + { + "epoch": 2.41631891060839, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3642170429229736, + "learning_rate": 1e-06, + "loss": 0.9418, + "mean_token_accuracy": 0.7227590084075928, + "num_tokens": 569216999.0, + "step": 22003 + }, + { + "epoch": 2.4164287283110037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9630963802337646, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7117391228675842, + "num_tokens": 569247007.0, + "step": 22004 + }, + { + "epoch": 2.4165385460136175, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9464192390441895, + "learning_rate": 1e-06, + "loss": 0.9248, + "mean_token_accuracy": 0.723443329334259, + "num_tokens": 569264811.0, + "step": 22005 + }, + { + "epoch": 2.416648363716231, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7876288890838623, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.7005101442337036, + "num_tokens": 569289177.0, + "step": 22006 + }, + { + "epoch": 2.4167581814188446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4990627765655518, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6949137449264526, + "num_tokens": 569320467.0, + "step": 22007 + }, + { + "epoch": 2.4168679991214583, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7772064208984375, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7073284983634949, + "num_tokens": 569344687.0, + "step": 22008 + }, + { + "epoch": 2.416977816824072, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3677713871002197, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.701565146446228, + "num_tokens": 569376539.0, + "step": 22009 + }, + { + "epoch": 2.417087634526686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6083426475524902, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7300417423248291, + "num_tokens": 569399611.0, + "step": 22010 + }, + { + "epoch": 2.417197452229299, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5743792057037354, + "learning_rate": 1e-06, + "loss": 1.0986, + "mean_token_accuracy": 0.6844830513000488, + "num_tokens": 569425410.0, + "step": 22011 + }, + { + "epoch": 2.417307269931913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.907407760620117, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.716867208480835, + "num_tokens": 569447821.0, + "step": 22012 + }, + { + "epoch": 2.4174170876345267, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6892924308776855, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6994946002960205, + "num_tokens": 569471760.0, + "step": 22013 + }, + { + "epoch": 2.4175269053371404, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.81400465965271, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.712978720664978, + "num_tokens": 569495548.0, + "step": 22014 + }, + { + "epoch": 2.417636723039754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9004364013671875, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7064403295516968, + "num_tokens": 569518643.0, + "step": 22015 + }, + { + "epoch": 2.4177465407423675, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.909526824951172, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.7136424779891968, + "num_tokens": 569538571.0, + "step": 22016 + }, + { + "epoch": 2.4178563584449813, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4766509532928467, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6987404823303223, + "num_tokens": 569569174.0, + "step": 22017 + }, + { + "epoch": 2.417966176147595, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9606728553771973, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6922481060028076, + "num_tokens": 569595496.0, + "step": 22018 + }, + { + "epoch": 2.4180759938502088, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.351351022720337, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6886584162712097, + "num_tokens": 569629327.0, + "step": 22019 + }, + { + "epoch": 2.4181858115528225, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5800766944885254, + "learning_rate": 1e-06, + "loss": 0.9008, + "mean_token_accuracy": 0.7298154830932617, + "num_tokens": 569654413.0, + "step": 22020 + }, + { + "epoch": 2.418295629255436, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.931337833404541, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7246032357215881, + "num_tokens": 569676924.0, + "step": 22021 + }, + { + "epoch": 2.4184054469580496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4759130477905273, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7092618942260742, + "num_tokens": 569706225.0, + "step": 22022 + }, + { + "epoch": 2.4185152646606634, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6123545169830322, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7125047445297241, + "num_tokens": 569729987.0, + "step": 22023 + }, + { + "epoch": 2.418625082363277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.821777582168579, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.695279061794281, + "num_tokens": 569751708.0, + "step": 22024 + }, + { + "epoch": 2.418734900065891, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.760751485824585, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7078123092651367, + "num_tokens": 569775611.0, + "step": 22025 + }, + { + "epoch": 2.418844717768504, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4751532077789307, + "learning_rate": 1e-06, + "loss": 1.1927, + "mean_token_accuracy": 0.674979031085968, + "num_tokens": 569811089.0, + "step": 22026 + }, + { + "epoch": 2.418954535471118, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.503211259841919, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7056208252906799, + "num_tokens": 569836572.0, + "step": 22027 + }, + { + "epoch": 2.4190643531737317, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7934048175811768, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.717782735824585, + "num_tokens": 569856953.0, + "step": 22028 + }, + { + "epoch": 2.4191741708763455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5078794956207275, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7259870767593384, + "num_tokens": 569883376.0, + "step": 22029 + }, + { + "epoch": 2.4192839885789588, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.697389841079712, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7040979266166687, + "num_tokens": 569907999.0, + "step": 22030 + }, + { + "epoch": 2.4193938062815725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4372894763946533, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7049315571784973, + "num_tokens": 569938111.0, + "step": 22031 + }, + { + "epoch": 2.4195036239841863, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.460188150405884, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6946107149124146, + "num_tokens": 569966184.0, + "step": 22032 + }, + { + "epoch": 2.4196134416868, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.037626028060913, + "learning_rate": 1e-06, + "loss": 0.8661, + "mean_token_accuracy": 0.7438306212425232, + "num_tokens": 569985005.0, + "step": 22033 + }, + { + "epoch": 2.4197232593894134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.497807025909424, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7012303471565247, + "num_tokens": 570010300.0, + "step": 22034 + }, + { + "epoch": 2.419833077092027, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.774043560028076, + "learning_rate": 1e-06, + "loss": 0.9554, + "mean_token_accuracy": 0.7191952466964722, + "num_tokens": 570032860.0, + "step": 22035 + }, + { + "epoch": 2.419942894794641, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5973198413848877, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.7259777784347534, + "num_tokens": 570061059.0, + "step": 22036 + }, + { + "epoch": 2.4200527124972546, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6418371200561523, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7197952270507812, + "num_tokens": 570087969.0, + "step": 22037 + }, + { + "epoch": 2.4201625301998684, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.629981517791748, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.6877583265304565, + "num_tokens": 570115084.0, + "step": 22038 + }, + { + "epoch": 2.4202723479024817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.345731496810913, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7068061828613281, + "num_tokens": 570143869.0, + "step": 22039 + }, + { + "epoch": 2.4203821656050954, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4638237953186035, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7047224044799805, + "num_tokens": 570173988.0, + "step": 22040 + }, + { + "epoch": 2.420491983307709, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.455681085586548, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.6919727325439453, + "num_tokens": 570202500.0, + "step": 22041 + }, + { + "epoch": 2.420601801010323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.714171886444092, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7106769680976868, + "num_tokens": 570226143.0, + "step": 22042 + }, + { + "epoch": 2.4207116187129367, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.823861598968506, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7209558486938477, + "num_tokens": 570246966.0, + "step": 22043 + }, + { + "epoch": 2.42082143641555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3973195552825928, + "learning_rate": 1e-06, + "loss": 0.9221, + "mean_token_accuracy": 0.7273870706558228, + "num_tokens": 570276122.0, + "step": 22044 + }, + { + "epoch": 2.420931254118164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.752183675765991, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7019990682601929, + "num_tokens": 570299470.0, + "step": 22045 + }, + { + "epoch": 2.4210410718207775, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.61453914642334, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7101132273674011, + "num_tokens": 570326235.0, + "step": 22046 + }, + { + "epoch": 2.4211508895233913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8199031352996826, + "learning_rate": 1e-06, + "loss": 0.9217, + "mean_token_accuracy": 0.7213801145553589, + "num_tokens": 570346722.0, + "step": 22047 + }, + { + "epoch": 2.421260707226005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.862632989883423, + "learning_rate": 1e-06, + "loss": 1.0725, + "mean_token_accuracy": 0.6950350999832153, + "num_tokens": 570368090.0, + "step": 22048 + }, + { + "epoch": 2.4213705249286184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9548354148864746, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7107168436050415, + "num_tokens": 570391613.0, + "step": 22049 + }, + { + "epoch": 2.421480342631232, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.3752121925354004, + "learning_rate": 1e-06, + "loss": 0.9494, + "mean_token_accuracy": 0.7258433103561401, + "num_tokens": 570411032.0, + "step": 22050 + }, + { + "epoch": 2.421590160333846, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8747730255126953, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6990523338317871, + "num_tokens": 570435697.0, + "step": 22051 + }, + { + "epoch": 2.4216999780364596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8758203983306885, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.6989399790763855, + "num_tokens": 570457632.0, + "step": 22052 + }, + { + "epoch": 2.421809795739073, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6378798484802246, + "learning_rate": 1e-06, + "loss": 0.9267, + "mean_token_accuracy": 0.7314538359642029, + "num_tokens": 570482577.0, + "step": 22053 + }, + { + "epoch": 2.4219196134416867, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6016814708709717, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6986017227172852, + "num_tokens": 570515113.0, + "step": 22054 + }, + { + "epoch": 2.4220294311443005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3707778453826904, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.7004325985908508, + "num_tokens": 570546724.0, + "step": 22055 + }, + { + "epoch": 2.4221392488469142, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 5.437514305114746, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7113451957702637, + "num_tokens": 570572622.0, + "step": 22056 + }, + { + "epoch": 2.4222490665495275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.931204080581665, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.7354598045349121, + "num_tokens": 570592893.0, + "step": 22057 + }, + { + "epoch": 2.4223588842521413, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.995570182800293, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.7010444402694702, + "num_tokens": 570617906.0, + "step": 22058 + }, + { + "epoch": 2.422468701954755, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7303218841552734, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7076511979103088, + "num_tokens": 570640269.0, + "step": 22059 + }, + { + "epoch": 2.422578519657369, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.937861680984497, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.720832109451294, + "num_tokens": 570661526.0, + "step": 22060 + }, + { + "epoch": 2.4226883373599826, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5589630603790283, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.6943959593772888, + "num_tokens": 570692755.0, + "step": 22061 + }, + { + "epoch": 2.422798155062596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.992217540740967, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7148580551147461, + "num_tokens": 570715163.0, + "step": 22062 + }, + { + "epoch": 2.4229079727652096, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3674798011779785, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.691493809223175, + "num_tokens": 570746436.0, + "step": 22063 + }, + { + "epoch": 2.4230177904678234, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.827418088912964, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6982864737510681, + "num_tokens": 570768185.0, + "step": 22064 + }, + { + "epoch": 2.423127608170437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6013588905334473, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7083215117454529, + "num_tokens": 570793782.0, + "step": 22065 + }, + { + "epoch": 2.423237425873051, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.571974754333496, + "learning_rate": 1e-06, + "loss": 0.9373, + "mean_token_accuracy": 0.7282791137695312, + "num_tokens": 570817523.0, + "step": 22066 + }, + { + "epoch": 2.4233472435756642, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4913179874420166, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.704096794128418, + "num_tokens": 570847349.0, + "step": 22067 + }, + { + "epoch": 2.423457061278278, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.659024953842163, + "learning_rate": 1e-06, + "loss": 0.9416, + "mean_token_accuracy": 0.7219758033752441, + "num_tokens": 570874899.0, + "step": 22068 + }, + { + "epoch": 2.4235668789808917, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6487138271331787, + "learning_rate": 1e-06, + "loss": 0.8167, + "mean_token_accuracy": 0.7529280185699463, + "num_tokens": 570898312.0, + "step": 22069 + }, + { + "epoch": 2.4236766966835055, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5539257526397705, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7054373621940613, + "num_tokens": 570923526.0, + "step": 22070 + }, + { + "epoch": 2.4237865143861193, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.323964834213257, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7105281352996826, + "num_tokens": 570955857.0, + "step": 22071 + }, + { + "epoch": 2.4238963320887326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4349777698516846, + "learning_rate": 1e-06, + "loss": 0.9286, + "mean_token_accuracy": 0.7241862416267395, + "num_tokens": 570982408.0, + "step": 22072 + }, + { + "epoch": 2.4240061497913463, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4483187198638916, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7064394354820251, + "num_tokens": 571012542.0, + "step": 22073 + }, + { + "epoch": 2.42411596749396, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5550708770751953, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6944647431373596, + "num_tokens": 571038782.0, + "step": 22074 + }, + { + "epoch": 2.424225785196574, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.349895715713501, + "learning_rate": 1e-06, + "loss": 0.9376, + "mean_token_accuracy": 0.725025475025177, + "num_tokens": 571071042.0, + "step": 22075 + }, + { + "epoch": 2.4243356028991876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6019034385681152, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7069724798202515, + "num_tokens": 571097481.0, + "step": 22076 + }, + { + "epoch": 2.424445420601801, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.623344898223877, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7103856205940247, + "num_tokens": 571123289.0, + "step": 22077 + }, + { + "epoch": 2.4245552383044147, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.919816255569458, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7108404040336609, + "num_tokens": 571150458.0, + "step": 22078 + }, + { + "epoch": 2.4246650560070284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.76068115234375, + "learning_rate": 1e-06, + "loss": 0.9301, + "mean_token_accuracy": 0.7264083623886108, + "num_tokens": 571174379.0, + "step": 22079 + }, + { + "epoch": 2.424774873709642, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.629178285598755, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7114907503128052, + "num_tokens": 571203623.0, + "step": 22080 + }, + { + "epoch": 2.4248846914122555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.433760166168213, + "learning_rate": 1e-06, + "loss": 1.0857, + "mean_token_accuracy": 0.6865906715393066, + "num_tokens": 571236688.0, + "step": 22081 + }, + { + "epoch": 2.4249945091148692, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.407824754714966, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.7016626596450806, + "num_tokens": 571266124.0, + "step": 22082 + }, + { + "epoch": 2.425104326817483, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8100955486297607, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.698639988899231, + "num_tokens": 571289727.0, + "step": 22083 + }, + { + "epoch": 2.4252141445200968, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7370805740356445, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7119622826576233, + "num_tokens": 571317444.0, + "step": 22084 + }, + { + "epoch": 2.42532396222271, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.525825262069702, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6982259154319763, + "num_tokens": 571346632.0, + "step": 22085 + }, + { + "epoch": 2.425433779925324, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5495734214782715, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7018353939056396, + "num_tokens": 571374381.0, + "step": 22086 + }, + { + "epoch": 2.4255435976279376, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5624401569366455, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6978702545166016, + "num_tokens": 571400742.0, + "step": 22087 + }, + { + "epoch": 2.4256534153305513, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7474405765533447, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.7268482446670532, + "num_tokens": 571427455.0, + "step": 22088 + }, + { + "epoch": 2.425763233033165, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.1264195442199707, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.6938847303390503, + "num_tokens": 571452419.0, + "step": 22089 + }, + { + "epoch": 2.4258730507357784, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.657522678375244, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7068658471107483, + "num_tokens": 571479041.0, + "step": 22090 + }, + { + "epoch": 2.425982868438392, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6432456970214844, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.7204683423042297, + "num_tokens": 571501627.0, + "step": 22091 + }, + { + "epoch": 2.426092686141006, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.555324077606201, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7077776789665222, + "num_tokens": 571527922.0, + "step": 22092 + }, + { + "epoch": 2.4262025038436197, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6244421005249023, + "learning_rate": 1e-06, + "loss": 0.9385, + "mean_token_accuracy": 0.7312002778053284, + "num_tokens": 571552399.0, + "step": 22093 + }, + { + "epoch": 2.4263123215462334, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5568645000457764, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7107962369918823, + "num_tokens": 571577848.0, + "step": 22094 + }, + { + "epoch": 2.4264221392488468, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4706125259399414, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.6976988315582275, + "num_tokens": 571607660.0, + "step": 22095 + }, + { + "epoch": 2.4265319569514605, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.642331838607788, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7140229940414429, + "num_tokens": 571633207.0, + "step": 22096 + }, + { + "epoch": 2.4266417746540743, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.44279146194458, + "learning_rate": 1e-06, + "loss": 1.0725, + "mean_token_accuracy": 0.696727991104126, + "num_tokens": 571660053.0, + "step": 22097 + }, + { + "epoch": 2.426751592356688, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.57157301902771, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7098349332809448, + "num_tokens": 571689648.0, + "step": 22098 + }, + { + "epoch": 2.426861410059302, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.548997163772583, + "learning_rate": 1e-06, + "loss": 1.1, + "mean_token_accuracy": 0.6858603954315186, + "num_tokens": 571718070.0, + "step": 22099 + }, + { + "epoch": 2.426971227761915, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2588870525360107, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7102457880973816, + "num_tokens": 571755031.0, + "step": 22100 + }, + { + "epoch": 2.427081045464529, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.558558225631714, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7106218338012695, + "num_tokens": 571779782.0, + "step": 22101 + }, + { + "epoch": 2.4271908631671426, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7960166931152344, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7275137901306152, + "num_tokens": 571800827.0, + "step": 22102 + }, + { + "epoch": 2.4273006808697564, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6542959213256836, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7012640237808228, + "num_tokens": 571823948.0, + "step": 22103 + }, + { + "epoch": 2.4274104985723697, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9186835289001465, + "learning_rate": 1e-06, + "loss": 0.9655, + "mean_token_accuracy": 0.7221453785896301, + "num_tokens": 571844500.0, + "step": 22104 + }, + { + "epoch": 2.4275203162749834, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.532904863357544, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7005235552787781, + "num_tokens": 571871878.0, + "step": 22105 + }, + { + "epoch": 2.427630133977597, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.601435899734497, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6963214874267578, + "num_tokens": 571899440.0, + "step": 22106 + }, + { + "epoch": 2.427739951680211, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.570147752761841, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7109177708625793, + "num_tokens": 571930948.0, + "step": 22107 + }, + { + "epoch": 2.4278497693828243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3700578212738037, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.7001065611839294, + "num_tokens": 571960027.0, + "step": 22108 + }, + { + "epoch": 2.427959587085438, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.572638511657715, + "learning_rate": 1e-06, + "loss": 1.122, + "mean_token_accuracy": 0.6768724918365479, + "num_tokens": 571989873.0, + "step": 22109 + }, + { + "epoch": 2.4280694047880518, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.632617235183716, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7033355236053467, + "num_tokens": 572017896.0, + "step": 22110 + }, + { + "epoch": 2.4281792224906655, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7671008110046387, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.7030510902404785, + "num_tokens": 572042412.0, + "step": 22111 + }, + { + "epoch": 2.4282890401932793, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4844882488250732, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.7242100238800049, + "num_tokens": 572068675.0, + "step": 22112 + }, + { + "epoch": 2.4283988578958926, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7126388549804688, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7060952186584473, + "num_tokens": 572092858.0, + "step": 22113 + }, + { + "epoch": 2.4285086755985064, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.521118402481079, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7157971858978271, + "num_tokens": 572122554.0, + "step": 22114 + }, + { + "epoch": 2.42861849330112, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.713365316390991, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.7026130557060242, + "num_tokens": 572145919.0, + "step": 22115 + }, + { + "epoch": 2.428728311003734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4003045558929443, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.6992592811584473, + "num_tokens": 572174593.0, + "step": 22116 + }, + { + "epoch": 2.4288381287063476, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9657156467437744, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7132063508033752, + "num_tokens": 572194188.0, + "step": 22117 + }, + { + "epoch": 2.428947946408961, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9380671977996826, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7154558300971985, + "num_tokens": 572213760.0, + "step": 22118 + }, + { + "epoch": 2.4290577641115747, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7624619007110596, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.6986131072044373, + "num_tokens": 572240846.0, + "step": 22119 + }, + { + "epoch": 2.4291675818141885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9950387477874756, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7120469212532043, + "num_tokens": 572260410.0, + "step": 22120 + }, + { + "epoch": 2.429277399516802, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.818098306655884, + "learning_rate": 1e-06, + "loss": 1.072, + "mean_token_accuracy": 0.6899262070655823, + "num_tokens": 572288503.0, + "step": 22121 + }, + { + "epoch": 2.429387217219416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6116762161254883, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7028770446777344, + "num_tokens": 572314762.0, + "step": 22122 + }, + { + "epoch": 2.4294970349220293, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6387040615081787, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7063882946968079, + "num_tokens": 572338449.0, + "step": 22123 + }, + { + "epoch": 2.429606852624643, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4737331867218018, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7032344937324524, + "num_tokens": 572371814.0, + "step": 22124 + }, + { + "epoch": 2.429716670327257, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.583400249481201, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6941852569580078, + "num_tokens": 572401330.0, + "step": 22125 + }, + { + "epoch": 2.4298264880298706, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.304072618484497, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7204240560531616, + "num_tokens": 572431993.0, + "step": 22126 + }, + { + "epoch": 2.4299363057324843, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.295400381088257, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7278994917869568, + "num_tokens": 572461968.0, + "step": 22127 + }, + { + "epoch": 2.4300461234350976, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5757620334625244, + "learning_rate": 1e-06, + "loss": 1.0949, + "mean_token_accuracy": 0.6878179907798767, + "num_tokens": 572487597.0, + "step": 22128 + }, + { + "epoch": 2.4301559411377114, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.546982765197754, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7165695428848267, + "num_tokens": 572513688.0, + "step": 22129 + }, + { + "epoch": 2.430265758840325, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7550177574157715, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7023383378982544, + "num_tokens": 572537892.0, + "step": 22130 + }, + { + "epoch": 2.430375576542939, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5956969261169434, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7194768190383911, + "num_tokens": 572562504.0, + "step": 22131 + }, + { + "epoch": 2.430485394245552, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0253000259399414, + "learning_rate": 1e-06, + "loss": 0.8554, + "mean_token_accuracy": 0.7431918978691101, + "num_tokens": 572585240.0, + "step": 22132 + }, + { + "epoch": 2.430595211948166, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7486789226531982, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7017205357551575, + "num_tokens": 572608644.0, + "step": 22133 + }, + { + "epoch": 2.4307050296507797, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.015326976776123, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7232018709182739, + "num_tokens": 572634180.0, + "step": 22134 + }, + { + "epoch": 2.4308148473533935, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9041543006896973, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7083368301391602, + "num_tokens": 572655918.0, + "step": 22135 + }, + { + "epoch": 2.430924665056007, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5412116050720215, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7197712659835815, + "num_tokens": 572683866.0, + "step": 22136 + }, + { + "epoch": 2.4310344827586206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.748195171356201, + "learning_rate": 1e-06, + "loss": 0.9186, + "mean_token_accuracy": 0.7280156016349792, + "num_tokens": 572706081.0, + "step": 22137 + }, + { + "epoch": 2.4311443004612343, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.769343137741089, + "learning_rate": 1e-06, + "loss": 0.9688, + "mean_token_accuracy": 0.7227843999862671, + "num_tokens": 572729135.0, + "step": 22138 + }, + { + "epoch": 2.431254118163848, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.754389762878418, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7342872023582458, + "num_tokens": 572752702.0, + "step": 22139 + }, + { + "epoch": 2.431363935866462, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6509511470794678, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7162526249885559, + "num_tokens": 572778951.0, + "step": 22140 + }, + { + "epoch": 2.431473753569075, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9043567180633545, + "learning_rate": 1e-06, + "loss": 0.8581, + "mean_token_accuracy": 0.7517369985580444, + "num_tokens": 572797376.0, + "step": 22141 + }, + { + "epoch": 2.431583571271689, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6042683124542236, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7036006450653076, + "num_tokens": 572821225.0, + "step": 22142 + }, + { + "epoch": 2.4316933889743026, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4326329231262207, + "learning_rate": 1e-06, + "loss": 0.9241, + "mean_token_accuracy": 0.7235535383224487, + "num_tokens": 572851640.0, + "step": 22143 + }, + { + "epoch": 2.4318032066769164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.543145179748535, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.699353814125061, + "num_tokens": 572876784.0, + "step": 22144 + }, + { + "epoch": 2.43191302437953, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.88550066947937, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7153567671775818, + "num_tokens": 572901633.0, + "step": 22145 + }, + { + "epoch": 2.4320228420821435, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7251076698303223, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6979448199272156, + "num_tokens": 572925554.0, + "step": 22146 + }, + { + "epoch": 2.4321326597847572, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0155460834503174, + "learning_rate": 1e-06, + "loss": 0.9177, + "mean_token_accuracy": 0.7307198643684387, + "num_tokens": 572945689.0, + "step": 22147 + }, + { + "epoch": 2.432242477487371, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.579306125640869, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7136859893798828, + "num_tokens": 572969929.0, + "step": 22148 + }, + { + "epoch": 2.4323522951899847, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.754077911376953, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7238990068435669, + "num_tokens": 572996094.0, + "step": 22149 + }, + { + "epoch": 2.4324621128925985, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.6731207370758057, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7193300127983093, + "num_tokens": 573013678.0, + "step": 22150 + }, + { + "epoch": 2.432571930595212, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.504985809326172, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.701483964920044, + "num_tokens": 573044076.0, + "step": 22151 + }, + { + "epoch": 2.4326817482978256, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5544073581695557, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6966273784637451, + "num_tokens": 573075911.0, + "step": 22152 + }, + { + "epoch": 2.4327915660004393, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.491346836090088, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.6963250041007996, + "num_tokens": 573104044.0, + "step": 22153 + }, + { + "epoch": 2.432901383703053, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.633321523666382, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7004753947257996, + "num_tokens": 573128751.0, + "step": 22154 + }, + { + "epoch": 2.4330112014056664, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3576536178588867, + "learning_rate": 1e-06, + "loss": 0.9055, + "mean_token_accuracy": 0.7321681380271912, + "num_tokens": 573161004.0, + "step": 22155 + }, + { + "epoch": 2.43312101910828, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7933261394500732, + "learning_rate": 1e-06, + "loss": 0.9351, + "mean_token_accuracy": 0.7239529490470886, + "num_tokens": 573183890.0, + "step": 22156 + }, + { + "epoch": 2.433230836810894, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.732106924057007, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6884243488311768, + "num_tokens": 573214899.0, + "step": 22157 + }, + { + "epoch": 2.4333406545135077, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3989925384521484, + "learning_rate": 1e-06, + "loss": 1.0418, + "mean_token_accuracy": 0.7034907937049866, + "num_tokens": 573244349.0, + "step": 22158 + }, + { + "epoch": 2.4334504722161214, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7057437896728516, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7104215621948242, + "num_tokens": 573266804.0, + "step": 22159 + }, + { + "epoch": 2.4335602899187347, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9679977893829346, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7135373950004578, + "num_tokens": 573285840.0, + "step": 22160 + }, + { + "epoch": 2.4336701076213485, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.50339937210083, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.7028772830963135, + "num_tokens": 573313675.0, + "step": 22161 + }, + { + "epoch": 2.4337799253239623, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4514052867889404, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7068512439727783, + "num_tokens": 573341634.0, + "step": 22162 + }, + { + "epoch": 2.433889743026576, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.329163074493408, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6914618015289307, + "num_tokens": 573372386.0, + "step": 22163 + }, + { + "epoch": 2.4339995607291893, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.741251230239868, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.726547122001648, + "num_tokens": 573393531.0, + "step": 22164 + }, + { + "epoch": 2.434109378431803, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.559497833251953, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7167873382568359, + "num_tokens": 573421354.0, + "step": 22165 + }, + { + "epoch": 2.434219196134417, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.410691976547241, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6919062733650208, + "num_tokens": 573450837.0, + "step": 22166 + }, + { + "epoch": 2.4343290138370306, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.696176528930664, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7149430513381958, + "num_tokens": 573472620.0, + "step": 22167 + }, + { + "epoch": 2.4344388315396444, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5294241905212402, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6937917470932007, + "num_tokens": 573499618.0, + "step": 22168 + }, + { + "epoch": 2.4345486492422577, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5294456481933594, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7188729643821716, + "num_tokens": 573525270.0, + "step": 22169 + }, + { + "epoch": 2.4346584669448714, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4542720317840576, + "learning_rate": 1e-06, + "loss": 1.128, + "mean_token_accuracy": 0.6764863729476929, + "num_tokens": 573558375.0, + "step": 22170 + }, + { + "epoch": 2.434768284647485, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.684173822402954, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7031965255737305, + "num_tokens": 573580819.0, + "step": 22171 + }, + { + "epoch": 2.434878102350099, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8405566215515137, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.697754442691803, + "num_tokens": 573603032.0, + "step": 22172 + }, + { + "epoch": 2.4349879200527127, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.736100673675537, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7264155149459839, + "num_tokens": 573624413.0, + "step": 22173 + }, + { + "epoch": 2.435097737755326, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.65354323387146, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7089349627494812, + "num_tokens": 573647899.0, + "step": 22174 + }, + { + "epoch": 2.4352075554579398, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8250627517700195, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6828590631484985, + "num_tokens": 573672654.0, + "step": 22175 + }, + { + "epoch": 2.4353173731605535, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.882054090499878, + "learning_rate": 1e-06, + "loss": 0.8852, + "mean_token_accuracy": 0.7426784038543701, + "num_tokens": 573692131.0, + "step": 22176 + }, + { + "epoch": 2.4354271908631673, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0427303314208984, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7011264562606812, + "num_tokens": 573712666.0, + "step": 22177 + }, + { + "epoch": 2.435537008565781, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.290233612060547, + "learning_rate": 1e-06, + "loss": 0.9142, + "mean_token_accuracy": 0.7288107872009277, + "num_tokens": 573732616.0, + "step": 22178 + }, + { + "epoch": 2.4356468262683943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.745907783508301, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7126248478889465, + "num_tokens": 573755985.0, + "step": 22179 + }, + { + "epoch": 2.435756643971008, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.576747179031372, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6908097267150879, + "num_tokens": 573781635.0, + "step": 22180 + }, + { + "epoch": 2.435866461673622, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8518645763397217, + "learning_rate": 1e-06, + "loss": 0.9217, + "mean_token_accuracy": 0.7338570356369019, + "num_tokens": 573803827.0, + "step": 22181 + }, + { + "epoch": 2.4359762793762356, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5362040996551514, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7291521430015564, + "num_tokens": 573828707.0, + "step": 22182 + }, + { + "epoch": 2.436086097078849, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.340362548828125, + "learning_rate": 1e-06, + "loss": 0.9168, + "mean_token_accuracy": 0.7304396629333496, + "num_tokens": 573858685.0, + "step": 22183 + }, + { + "epoch": 2.4361959147814627, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6774253845214844, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.708220362663269, + "num_tokens": 573885476.0, + "step": 22184 + }, + { + "epoch": 2.4363057324840764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5619544982910156, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7092661261558533, + "num_tokens": 573910411.0, + "step": 22185 + }, + { + "epoch": 2.43641555018669, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.367784023284912, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7180335521697998, + "num_tokens": 573940680.0, + "step": 22186 + }, + { + "epoch": 2.4365253678893035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.246506452560425, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7194885015487671, + "num_tokens": 573974782.0, + "step": 22187 + }, + { + "epoch": 2.4366351855919173, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0385849475860596, + "learning_rate": 1e-06, + "loss": 1.125, + "mean_token_accuracy": 0.6807325482368469, + "num_tokens": 574001647.0, + "step": 22188 + }, + { + "epoch": 2.436745003294531, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5690510272979736, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7023399472236633, + "num_tokens": 574027374.0, + "step": 22189 + }, + { + "epoch": 2.436854820997145, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.3117761611938477, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7148120403289795, + "num_tokens": 574047848.0, + "step": 22190 + }, + { + "epoch": 2.4369646386997585, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5710647106170654, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7186996936798096, + "num_tokens": 574071840.0, + "step": 22191 + }, + { + "epoch": 2.437074456402372, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6735823154449463, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7009979486465454, + "num_tokens": 574096632.0, + "step": 22192 + }, + { + "epoch": 2.4371842741049856, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6924984455108643, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7062692642211914, + "num_tokens": 574121632.0, + "step": 22193 + }, + { + "epoch": 2.4372940918075994, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7466859817504883, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.719536542892456, + "num_tokens": 574143569.0, + "step": 22194 + }, + { + "epoch": 2.437403909510213, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8046209812164307, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7191846966743469, + "num_tokens": 574165403.0, + "step": 22195 + }, + { + "epoch": 2.437513727212827, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.872346878051758, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7011190056800842, + "num_tokens": 574192232.0, + "step": 22196 + }, + { + "epoch": 2.43762354491544, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.554619312286377, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.716314435005188, + "num_tokens": 574218583.0, + "step": 22197 + }, + { + "epoch": 2.437733362618054, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.700362205505371, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7100194096565247, + "num_tokens": 574247284.0, + "step": 22198 + }, + { + "epoch": 2.4378431803206677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6465442180633545, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7065662145614624, + "num_tokens": 574272439.0, + "step": 22199 + }, + { + "epoch": 2.4379529980232815, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5142409801483154, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7132833003997803, + "num_tokens": 574300708.0, + "step": 22200 + }, + { + "epoch": 2.4380628157258952, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4682116508483887, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.7017274498939514, + "num_tokens": 574329878.0, + "step": 22201 + }, + { + "epoch": 2.4381726334285085, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3305816650390625, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7081016302108765, + "num_tokens": 574362905.0, + "step": 22202 + }, + { + "epoch": 2.4382824511311223, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.937126398086548, + "learning_rate": 1e-06, + "loss": 1.1322, + "mean_token_accuracy": 0.6778230667114258, + "num_tokens": 574389659.0, + "step": 22203 + }, + { + "epoch": 2.438392268833736, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.614629030227661, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7020535469055176, + "num_tokens": 574415611.0, + "step": 22204 + }, + { + "epoch": 2.43850208653635, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.424919605255127, + "learning_rate": 1e-06, + "loss": 1.0715, + "mean_token_accuracy": 0.6949688196182251, + "num_tokens": 574444974.0, + "step": 22205 + }, + { + "epoch": 2.4386119042389636, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.679018020629883, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.720285177230835, + "num_tokens": 574467654.0, + "step": 22206 + }, + { + "epoch": 2.438721721941577, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8835606575012207, + "learning_rate": 1e-06, + "loss": 0.9241, + "mean_token_accuracy": 0.7291333675384521, + "num_tokens": 574487929.0, + "step": 22207 + }, + { + "epoch": 2.4388315396441906, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7194571495056152, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7105538845062256, + "num_tokens": 574510732.0, + "step": 22208 + }, + { + "epoch": 2.4389413573468044, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8831775188446045, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7035937905311584, + "num_tokens": 574536138.0, + "step": 22209 + }, + { + "epoch": 2.439051175049418, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4409101009368896, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.719039797782898, + "num_tokens": 574563499.0, + "step": 22210 + }, + { + "epoch": 2.4391609927520315, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4756767749786377, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.699838399887085, + "num_tokens": 574589829.0, + "step": 22211 + }, + { + "epoch": 2.439270810454645, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8566033840179443, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7129789590835571, + "num_tokens": 574611065.0, + "step": 22212 + }, + { + "epoch": 2.439380628157259, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5481321811676025, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6898173093795776, + "num_tokens": 574637669.0, + "step": 22213 + }, + { + "epoch": 2.4394904458598727, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1012909412384033, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7316223382949829, + "num_tokens": 574656869.0, + "step": 22214 + }, + { + "epoch": 2.439600263562486, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6539292335510254, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7081671953201294, + "num_tokens": 574684604.0, + "step": 22215 + }, + { + "epoch": 2.4397100812651, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7475192546844482, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6928832530975342, + "num_tokens": 574708139.0, + "step": 22216 + }, + { + "epoch": 2.4398198989677136, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.540644884109497, + "learning_rate": 1e-06, + "loss": 0.8693, + "mean_token_accuracy": 0.7425678372383118, + "num_tokens": 574732374.0, + "step": 22217 + }, + { + "epoch": 2.4399297166703273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.70664119720459, + "learning_rate": 1e-06, + "loss": 1.1251, + "mean_token_accuracy": 0.6847509145736694, + "num_tokens": 574760582.0, + "step": 22218 + }, + { + "epoch": 2.440039534372941, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8980841636657715, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7139701843261719, + "num_tokens": 574786011.0, + "step": 22219 + }, + { + "epoch": 2.4401493520755544, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.68699049949646, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7053502798080444, + "num_tokens": 574812478.0, + "step": 22220 + }, + { + "epoch": 2.440259169778168, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4947962760925293, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.716835081577301, + "num_tokens": 574838868.0, + "step": 22221 + }, + { + "epoch": 2.440368987480782, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5193405151367188, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6914049386978149, + "num_tokens": 574867871.0, + "step": 22222 + }, + { + "epoch": 2.4404788051833957, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.68593168258667, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.708825945854187, + "num_tokens": 574890716.0, + "step": 22223 + }, + { + "epoch": 2.4405886228860094, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5881752967834473, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7108469605445862, + "num_tokens": 574916637.0, + "step": 22224 + }, + { + "epoch": 2.4406984405886227, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5510356426239014, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7150486707687378, + "num_tokens": 574943381.0, + "step": 22225 + }, + { + "epoch": 2.4408082582912365, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.433365821838379, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7037667036056519, + "num_tokens": 574972351.0, + "step": 22226 + }, + { + "epoch": 2.4409180759938502, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3595693111419678, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7179093956947327, + "num_tokens": 575000134.0, + "step": 22227 + }, + { + "epoch": 2.441027893696464, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.30257511138916, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.708837628364563, + "num_tokens": 575035356.0, + "step": 22228 + }, + { + "epoch": 2.4411377113990778, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8429887294769287, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7093730568885803, + "num_tokens": 575057505.0, + "step": 22229 + }, + { + "epoch": 2.441247529101691, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.650238513946533, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7193257808685303, + "num_tokens": 575081014.0, + "step": 22230 + }, + { + "epoch": 2.441357346804305, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.416973352432251, + "learning_rate": 1e-06, + "loss": 1.1264, + "mean_token_accuracy": 0.678878903388977, + "num_tokens": 575115581.0, + "step": 22231 + }, + { + "epoch": 2.4414671645069186, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.69580078125, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7004618644714355, + "num_tokens": 575139436.0, + "step": 22232 + }, + { + "epoch": 2.4415769822095323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4041998386383057, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6909205913543701, + "num_tokens": 575171593.0, + "step": 22233 + }, + { + "epoch": 2.4416867999121457, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.535784959793091, + "learning_rate": 1e-06, + "loss": 0.9453, + "mean_token_accuracy": 0.7221863269805908, + "num_tokens": 575197302.0, + "step": 22234 + }, + { + "epoch": 2.4417966176147594, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3769469261169434, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7083796858787537, + "num_tokens": 575225667.0, + "step": 22235 + }, + { + "epoch": 2.441906435317373, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8866872787475586, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7105029821395874, + "num_tokens": 575250392.0, + "step": 22236 + }, + { + "epoch": 2.442016253019987, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9463634490966797, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.707877516746521, + "num_tokens": 575270867.0, + "step": 22237 + }, + { + "epoch": 2.4421260707226002, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2936739921569824, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6943010687828064, + "num_tokens": 575303943.0, + "step": 22238 + }, + { + "epoch": 2.442235888425214, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8198587894439697, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7120453119277954, + "num_tokens": 575325586.0, + "step": 22239 + }, + { + "epoch": 2.4423457061278278, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.16036319732666, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.714108943939209, + "num_tokens": 575345010.0, + "step": 22240 + }, + { + "epoch": 2.4424555238304415, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.62444806098938, + "learning_rate": 1e-06, + "loss": 1.0944, + "mean_token_accuracy": 0.6822992563247681, + "num_tokens": 575371827.0, + "step": 22241 + }, + { + "epoch": 2.4425653415330553, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.580592155456543, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7196055054664612, + "num_tokens": 575397851.0, + "step": 22242 + }, + { + "epoch": 2.4426751592356686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8151917457580566, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7018446922302246, + "num_tokens": 575420803.0, + "step": 22243 + }, + { + "epoch": 2.4427849769382823, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8394317626953125, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7214193344116211, + "num_tokens": 575441226.0, + "step": 22244 + }, + { + "epoch": 2.442894794640896, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545917510986328, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7053616046905518, + "num_tokens": 575467323.0, + "step": 22245 + }, + { + "epoch": 2.44300461234351, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.616966962814331, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7081400752067566, + "num_tokens": 575492186.0, + "step": 22246 + }, + { + "epoch": 2.4431144300461236, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8684911727905273, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.712294340133667, + "num_tokens": 575512859.0, + "step": 22247 + }, + { + "epoch": 2.443224247748737, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.658992290496826, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7216383814811707, + "num_tokens": 575534950.0, + "step": 22248 + }, + { + "epoch": 2.4433340654513507, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.598515272140503, + "learning_rate": 1e-06, + "loss": 0.9303, + "mean_token_accuracy": 0.7262052297592163, + "num_tokens": 575560010.0, + "step": 22249 + }, + { + "epoch": 2.4434438831539644, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.600764036178589, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.6912335157394409, + "num_tokens": 575587780.0, + "step": 22250 + }, + { + "epoch": 2.443553700856578, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6691994667053223, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6987630128860474, + "num_tokens": 575613729.0, + "step": 22251 + }, + { + "epoch": 2.443663518559192, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.562774419784546, + "learning_rate": 1e-06, + "loss": 1.1373, + "mean_token_accuracy": 0.67033851146698, + "num_tokens": 575641897.0, + "step": 22252 + }, + { + "epoch": 2.4437733362618053, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6608262062072754, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7149640321731567, + "num_tokens": 575663939.0, + "step": 22253 + }, + { + "epoch": 2.443883153964419, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.648550271987915, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6982476711273193, + "num_tokens": 575693398.0, + "step": 22254 + }, + { + "epoch": 2.4439929716670328, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8019285202026367, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7100292444229126, + "num_tokens": 575716003.0, + "step": 22255 + }, + { + "epoch": 2.4441027893696465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.835216522216797, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7121137380599976, + "num_tokens": 575737645.0, + "step": 22256 + }, + { + "epoch": 2.4442126070722603, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5534090995788574, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7124744057655334, + "num_tokens": 575764482.0, + "step": 22257 + }, + { + "epoch": 2.4443224247748736, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8523404598236084, + "learning_rate": 1e-06, + "loss": 1.0428, + "mean_token_accuracy": 0.6972663998603821, + "num_tokens": 575785339.0, + "step": 22258 + }, + { + "epoch": 2.4444322424774874, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.280444860458374, + "learning_rate": 1e-06, + "loss": 0.9164, + "mean_token_accuracy": 0.735718309879303, + "num_tokens": 575810353.0, + "step": 22259 + }, + { + "epoch": 2.444542060180101, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.01789927482605, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7043296694755554, + "num_tokens": 575837686.0, + "step": 22260 + }, + { + "epoch": 2.444651877882715, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5087077617645264, + "learning_rate": 1e-06, + "loss": 1.0703, + "mean_token_accuracy": 0.6849302649497986, + "num_tokens": 575873131.0, + "step": 22261 + }, + { + "epoch": 2.444761695585328, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.547166585922241, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.688725471496582, + "num_tokens": 575901282.0, + "step": 22262 + }, + { + "epoch": 2.444871513287942, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.840451955795288, + "learning_rate": 1e-06, + "loss": 0.9607, + "mean_token_accuracy": 0.7116923928260803, + "num_tokens": 575923327.0, + "step": 22263 + }, + { + "epoch": 2.4449813309905557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6866393089294434, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6962116360664368, + "num_tokens": 575955089.0, + "step": 22264 + }, + { + "epoch": 2.4450911486931695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5436220169067383, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6850268840789795, + "num_tokens": 575982716.0, + "step": 22265 + }, + { + "epoch": 2.4452009663957828, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4326391220092773, + "learning_rate": 1e-06, + "loss": 1.0974, + "mean_token_accuracy": 0.6858612895011902, + "num_tokens": 576012552.0, + "step": 22266 + }, + { + "epoch": 2.4453107840983965, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.918454170227051, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.7157353758811951, + "num_tokens": 576033064.0, + "step": 22267 + }, + { + "epoch": 2.4454206018010103, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4176783561706543, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6923043131828308, + "num_tokens": 576063360.0, + "step": 22268 + }, + { + "epoch": 2.445530419503624, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4986960887908936, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.7256066799163818, + "num_tokens": 576088138.0, + "step": 22269 + }, + { + "epoch": 2.445640237206238, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3171651363372803, + "learning_rate": 1e-06, + "loss": 1.0762, + "mean_token_accuracy": 0.6857420802116394, + "num_tokens": 576120804.0, + "step": 22270 + }, + { + "epoch": 2.445750054908851, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4842770099639893, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.711142897605896, + "num_tokens": 576146306.0, + "step": 22271 + }, + { + "epoch": 2.445859872611465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.668045997619629, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.7015045881271362, + "num_tokens": 576171014.0, + "step": 22272 + }, + { + "epoch": 2.4459696903140786, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6385631561279297, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.6979517340660095, + "num_tokens": 576197866.0, + "step": 22273 + }, + { + "epoch": 2.4460795080166924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.677747964859009, + "learning_rate": 1e-06, + "loss": 1.1249, + "mean_token_accuracy": 0.6754810214042664, + "num_tokens": 576225642.0, + "step": 22274 + }, + { + "epoch": 2.446189325719306, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.438786268234253, + "learning_rate": 1e-06, + "loss": 1.0023, + "mean_token_accuracy": 0.7120067477226257, + "num_tokens": 576253010.0, + "step": 22275 + }, + { + "epoch": 2.4462991434219195, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6003897190093994, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7008579969406128, + "num_tokens": 576279767.0, + "step": 22276 + }, + { + "epoch": 2.446408961124533, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.453927516937256, + "learning_rate": 1e-06, + "loss": 1.0742, + "mean_token_accuracy": 0.6867716312408447, + "num_tokens": 576310196.0, + "step": 22277 + }, + { + "epoch": 2.446518778827147, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6290905475616455, + "learning_rate": 1e-06, + "loss": 1.1106, + "mean_token_accuracy": 0.6823550462722778, + "num_tokens": 576336773.0, + "step": 22278 + }, + { + "epoch": 2.4466285965297607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9789247512817383, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7081853151321411, + "num_tokens": 576358569.0, + "step": 22279 + }, + { + "epoch": 2.4467384142323745, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8518686294555664, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.711540937423706, + "num_tokens": 576386669.0, + "step": 22280 + }, + { + "epoch": 2.446848231934988, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6765942573547363, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7189982533454895, + "num_tokens": 576409761.0, + "step": 22281 + }, + { + "epoch": 2.4469580496376016, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.846083164215088, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7192212343215942, + "num_tokens": 576432019.0, + "step": 22282 + }, + { + "epoch": 2.4470678673402153, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6727070808410645, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7096824049949646, + "num_tokens": 576459040.0, + "step": 22283 + }, + { + "epoch": 2.447177685042829, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5195882320404053, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7269193530082703, + "num_tokens": 576485630.0, + "step": 22284 + }, + { + "epoch": 2.4472875027454424, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6200296878814697, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7212245464324951, + "num_tokens": 576510185.0, + "step": 22285 + }, + { + "epoch": 2.447397320448056, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.676068067550659, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.707327127456665, + "num_tokens": 576533097.0, + "step": 22286 + }, + { + "epoch": 2.44750713815067, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9716906547546387, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7131327390670776, + "num_tokens": 576552424.0, + "step": 22287 + }, + { + "epoch": 2.4476169558532836, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7467093467712402, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.724446177482605, + "num_tokens": 576572851.0, + "step": 22288 + }, + { + "epoch": 2.447726773555897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5746281147003174, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7167134284973145, + "num_tokens": 576599655.0, + "step": 22289 + }, + { + "epoch": 2.4478365912585107, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4380593299865723, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.7071354389190674, + "num_tokens": 576628544.0, + "step": 22290 + }, + { + "epoch": 2.4479464089611245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4007720947265625, + "learning_rate": 1e-06, + "loss": 0.9507, + "mean_token_accuracy": 0.7232465744018555, + "num_tokens": 576657072.0, + "step": 22291 + }, + { + "epoch": 2.4480562266637382, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.448563814163208, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7042795419692993, + "num_tokens": 576683784.0, + "step": 22292 + }, + { + "epoch": 2.448166044366352, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7825100421905518, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7222694158554077, + "num_tokens": 576706149.0, + "step": 22293 + }, + { + "epoch": 2.4482758620689653, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.728381395339966, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7029211521148682, + "num_tokens": 576727558.0, + "step": 22294 + }, + { + "epoch": 2.448385679771579, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.51993465423584, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7110034823417664, + "num_tokens": 576754030.0, + "step": 22295 + }, + { + "epoch": 2.448495497474193, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.44985032081604, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7112696170806885, + "num_tokens": 576782912.0, + "step": 22296 + }, + { + "epoch": 2.4486053151768066, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.515908718109131, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7109801173210144, + "num_tokens": 576807119.0, + "step": 22297 + }, + { + "epoch": 2.4487151328794203, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6907525062561035, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.704853892326355, + "num_tokens": 576830037.0, + "step": 22298 + }, + { + "epoch": 2.4488249505820336, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7148139476776123, + "learning_rate": 1e-06, + "loss": 0.8686, + "mean_token_accuracy": 0.7422109246253967, + "num_tokens": 576851476.0, + "step": 22299 + }, + { + "epoch": 2.4489347682846474, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9898018836975098, + "learning_rate": 1e-06, + "loss": 1.0928, + "mean_token_accuracy": 0.6954923868179321, + "num_tokens": 576875256.0, + "step": 22300 + }, + { + "epoch": 2.449044585987261, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.858309268951416, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7077769637107849, + "num_tokens": 576905309.0, + "step": 22301 + }, + { + "epoch": 2.449154403689875, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6245980262756348, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.701110303401947, + "num_tokens": 576930590.0, + "step": 22302 + }, + { + "epoch": 2.4492642213924887, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7674050331115723, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7114894986152649, + "num_tokens": 576952543.0, + "step": 22303 + }, + { + "epoch": 2.449374039095102, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5551984310150146, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.6994832158088684, + "num_tokens": 576978843.0, + "step": 22304 + }, + { + "epoch": 2.4494838567977157, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9181416034698486, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6990177631378174, + "num_tokens": 577004525.0, + "step": 22305 + }, + { + "epoch": 2.4495936745003295, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6685891151428223, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7103002071380615, + "num_tokens": 577028989.0, + "step": 22306 + }, + { + "epoch": 2.4497034922029433, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.758934736251831, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.7292494177818298, + "num_tokens": 577049797.0, + "step": 22307 + }, + { + "epoch": 2.449813309905557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7287702560424805, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7155938148498535, + "num_tokens": 577072989.0, + "step": 22308 + }, + { + "epoch": 2.4499231276081703, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.677527666091919, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7067131400108337, + "num_tokens": 577097721.0, + "step": 22309 + }, + { + "epoch": 2.450032945310784, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.436960220336914, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.686710774898529, + "num_tokens": 577125235.0, + "step": 22310 + }, + { + "epoch": 2.450142763013398, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.721402406692505, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.7234787940979004, + "num_tokens": 577146621.0, + "step": 22311 + }, + { + "epoch": 2.4502525807160116, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.013202428817749, + "learning_rate": 1e-06, + "loss": 0.9361, + "mean_token_accuracy": 0.7252978086471558, + "num_tokens": 577165943.0, + "step": 22312 + }, + { + "epoch": 2.450362398418625, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.992117166519165, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.6960192918777466, + "num_tokens": 577187725.0, + "step": 22313 + }, + { + "epoch": 2.4504722161212387, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5855798721313477, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7196156978607178, + "num_tokens": 577214425.0, + "step": 22314 + }, + { + "epoch": 2.4505820338238524, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4707932472229004, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7134486436843872, + "num_tokens": 577241342.0, + "step": 22315 + }, + { + "epoch": 2.450691851526466, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.556593894958496, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.7248657941818237, + "num_tokens": 577264693.0, + "step": 22316 + }, + { + "epoch": 2.4508016692290795, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.447371244430542, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7148720026016235, + "num_tokens": 577293109.0, + "step": 22317 + }, + { + "epoch": 2.4509114869316933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3650410175323486, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7123866081237793, + "num_tokens": 577322748.0, + "step": 22318 + }, + { + "epoch": 2.451021304634307, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.926206588745117, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7069820165634155, + "num_tokens": 577346093.0, + "step": 22319 + }, + { + "epoch": 2.4511311223369208, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3691060543060303, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7134602665901184, + "num_tokens": 577376360.0, + "step": 22320 + }, + { + "epoch": 2.4512409400395345, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4358267784118652, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7035349011421204, + "num_tokens": 577406450.0, + "step": 22321 + }, + { + "epoch": 2.451350757742148, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8189494609832764, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6918913722038269, + "num_tokens": 577428068.0, + "step": 22322 + }, + { + "epoch": 2.4514605754447616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6492063999176025, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7167165279388428, + "num_tokens": 577451745.0, + "step": 22323 + }, + { + "epoch": 2.4515703931473753, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.746203660964966, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6892858743667603, + "num_tokens": 577473353.0, + "step": 22324 + }, + { + "epoch": 2.451680210849989, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.1264984607696533, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7022595405578613, + "num_tokens": 577511202.0, + "step": 22325 + }, + { + "epoch": 2.451790028552603, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5064549446105957, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7208333015441895, + "num_tokens": 577540374.0, + "step": 22326 + }, + { + "epoch": 2.451899846255216, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.42275333404541, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.6852704882621765, + "num_tokens": 577571983.0, + "step": 22327 + }, + { + "epoch": 2.45200966395783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.669797420501709, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.6988649368286133, + "num_tokens": 577596579.0, + "step": 22328 + }, + { + "epoch": 2.4521194816604437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3008506298065186, + "learning_rate": 1e-06, + "loss": 0.8931, + "mean_token_accuracy": 0.7277618646621704, + "num_tokens": 577626159.0, + "step": 22329 + }, + { + "epoch": 2.4522292993630574, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.383057117462158, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.6991323232650757, + "num_tokens": 577653838.0, + "step": 22330 + }, + { + "epoch": 2.452339117065671, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5529420375823975, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.7026851177215576, + "num_tokens": 577681468.0, + "step": 22331 + }, + { + "epoch": 2.4524489347682845, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4263129234313965, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7020875215530396, + "num_tokens": 577711263.0, + "step": 22332 + }, + { + "epoch": 2.4525587524708983, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7220587730407715, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7125029563903809, + "num_tokens": 577734729.0, + "step": 22333 + }, + { + "epoch": 2.452668570173512, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8310790061950684, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7175413370132446, + "num_tokens": 577756749.0, + "step": 22334 + }, + { + "epoch": 2.452778387876126, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.443713665008545, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7072314023971558, + "num_tokens": 577785506.0, + "step": 22335 + }, + { + "epoch": 2.452888205578739, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8194973468780518, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7084938287734985, + "num_tokens": 577808889.0, + "step": 22336 + }, + { + "epoch": 2.452998023281353, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.663952112197876, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.694023609161377, + "num_tokens": 577837719.0, + "step": 22337 + }, + { + "epoch": 2.4531078409839666, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.606995105743408, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7128563523292542, + "num_tokens": 577863447.0, + "step": 22338 + }, + { + "epoch": 2.4532176586865804, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8092164993286133, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7028006315231323, + "num_tokens": 577883916.0, + "step": 22339 + }, + { + "epoch": 2.453327476389194, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6337475776672363, + "learning_rate": 1e-06, + "loss": 1.0824, + "mean_token_accuracy": 0.688233494758606, + "num_tokens": 577910642.0, + "step": 22340 + }, + { + "epoch": 2.4534372940918074, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.545241355895996, + "learning_rate": 1e-06, + "loss": 0.8853, + "mean_token_accuracy": 0.7323118448257446, + "num_tokens": 577934895.0, + "step": 22341 + }, + { + "epoch": 2.453547111794421, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.809553861618042, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7034205198287964, + "num_tokens": 577960405.0, + "step": 22342 + }, + { + "epoch": 2.453656929497035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.9843034744262695, + "learning_rate": 1e-06, + "loss": 1.1335, + "mean_token_accuracy": 0.6755924224853516, + "num_tokens": 577991669.0, + "step": 22343 + }, + { + "epoch": 2.4537667471996487, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.72202467918396, + "learning_rate": 1e-06, + "loss": 0.9125, + "mean_token_accuracy": 0.7290042638778687, + "num_tokens": 578012901.0, + "step": 22344 + }, + { + "epoch": 2.453876564902262, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8969457149505615, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7267142534255981, + "num_tokens": 578033330.0, + "step": 22345 + }, + { + "epoch": 2.453986382604876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.937715530395508, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7213295102119446, + "num_tokens": 578053690.0, + "step": 22346 + }, + { + "epoch": 2.4540962003074895, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4545209407806396, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7081896066665649, + "num_tokens": 578084004.0, + "step": 22347 + }, + { + "epoch": 2.4542060180101033, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.512528419494629, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7198773622512817, + "num_tokens": 578109929.0, + "step": 22348 + }, + { + "epoch": 2.454315835712717, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.432138681411743, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.6997815370559692, + "num_tokens": 578143144.0, + "step": 22349 + }, + { + "epoch": 2.4544256534153304, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3906209468841553, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.7002060413360596, + "num_tokens": 578172629.0, + "step": 22350 + }, + { + "epoch": 2.454535471117944, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.538633346557617, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7146067023277283, + "num_tokens": 578196976.0, + "step": 22351 + }, + { + "epoch": 2.454645288820558, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5081305503845215, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.6967789530754089, + "num_tokens": 578226395.0, + "step": 22352 + }, + { + "epoch": 2.4547551065231716, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3873867988586426, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.7027488946914673, + "num_tokens": 578254367.0, + "step": 22353 + }, + { + "epoch": 2.4548649242257854, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.608311653137207, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7096948623657227, + "num_tokens": 578279572.0, + "step": 22354 + }, + { + "epoch": 2.4549747419283987, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6579856872558594, + "learning_rate": 1e-06, + "loss": 0.8296, + "mean_token_accuracy": 0.7533273696899414, + "num_tokens": 578302458.0, + "step": 22355 + }, + { + "epoch": 2.4550845596310125, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4848101139068604, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6936546564102173, + "num_tokens": 578333129.0, + "step": 22356 + }, + { + "epoch": 2.455194377333626, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.650975465774536, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7236459255218506, + "num_tokens": 578357560.0, + "step": 22357 + }, + { + "epoch": 2.45530419503624, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.502161741256714, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7229085564613342, + "num_tokens": 578384080.0, + "step": 22358 + }, + { + "epoch": 2.4554140127388537, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.965989589691162, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6912762522697449, + "num_tokens": 578413855.0, + "step": 22359 + }, + { + "epoch": 2.455523830441467, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6075680255889893, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.7082498073577881, + "num_tokens": 578441063.0, + "step": 22360 + }, + { + "epoch": 2.455633648144081, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.467350721359253, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.692792534828186, + "num_tokens": 578469577.0, + "step": 22361 + }, + { + "epoch": 2.4557434658466946, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6658806800842285, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7245938777923584, + "num_tokens": 578492684.0, + "step": 22362 + }, + { + "epoch": 2.4558532835493083, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.68754506111145, + "learning_rate": 1e-06, + "loss": 0.9851, + "mean_token_accuracy": 0.7091507911682129, + "num_tokens": 578516977.0, + "step": 22363 + }, + { + "epoch": 2.4559631012519216, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7289323806762695, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.6931490898132324, + "num_tokens": 578542266.0, + "step": 22364 + }, + { + "epoch": 2.4560729189545354, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.805680990219116, + "learning_rate": 1e-06, + "loss": 0.9739, + "mean_token_accuracy": 0.7147684097290039, + "num_tokens": 578564821.0, + "step": 22365 + }, + { + "epoch": 2.456182736657149, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.914668560028076, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7097439169883728, + "num_tokens": 578586215.0, + "step": 22366 + }, + { + "epoch": 2.456292554359763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6612186431884766, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7020179033279419, + "num_tokens": 578612609.0, + "step": 22367 + }, + { + "epoch": 2.456402372062376, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6670384407043457, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.686129093170166, + "num_tokens": 578637678.0, + "step": 22368 + }, + { + "epoch": 2.45651218976499, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3748550415039062, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6949954032897949, + "num_tokens": 578669710.0, + "step": 22369 + }, + { + "epoch": 2.4566220074676037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6454694271087646, + "learning_rate": 1e-06, + "loss": 0.92, + "mean_token_accuracy": 0.7269832491874695, + "num_tokens": 578693789.0, + "step": 22370 + }, + { + "epoch": 2.4567318251702175, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3257076740264893, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7050120234489441, + "num_tokens": 578724447.0, + "step": 22371 + }, + { + "epoch": 2.4568416428728312, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.468393325805664, + "learning_rate": 1e-06, + "loss": 1.08, + "mean_token_accuracy": 0.7044305205345154, + "num_tokens": 578750989.0, + "step": 22372 + }, + { + "epoch": 2.4569514605754446, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.783076286315918, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7220500707626343, + "num_tokens": 578772778.0, + "step": 22373 + }, + { + "epoch": 2.4570612782780583, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9649012088775635, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.6980540156364441, + "num_tokens": 578800895.0, + "step": 22374 + }, + { + "epoch": 2.457171095980672, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.376913070678711, + "learning_rate": 1e-06, + "loss": 0.8921, + "mean_token_accuracy": 0.7313224077224731, + "num_tokens": 578829230.0, + "step": 22375 + }, + { + "epoch": 2.457280913683286, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.526824474334717, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.694832980632782, + "num_tokens": 578860437.0, + "step": 22376 + }, + { + "epoch": 2.4573907313858996, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.552151679992676, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6848437786102295, + "num_tokens": 578887887.0, + "step": 22377 + }, + { + "epoch": 2.457500549088513, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.499911308288574, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6943305134773254, + "num_tokens": 578916142.0, + "step": 22378 + }, + { + "epoch": 2.4576103667911267, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4613776206970215, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7114806771278381, + "num_tokens": 578943239.0, + "step": 22379 + }, + { + "epoch": 2.4577201844937404, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4655613899230957, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7213863134384155, + "num_tokens": 578970595.0, + "step": 22380 + }, + { + "epoch": 2.457830002196354, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.682205200195312, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7054645419120789, + "num_tokens": 578999841.0, + "step": 22381 + }, + { + "epoch": 2.457939819898968, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5874147415161133, + "learning_rate": 1e-06, + "loss": 1.1076, + "mean_token_accuracy": 0.6741715669631958, + "num_tokens": 579028000.0, + "step": 22382 + }, + { + "epoch": 2.4580496376015812, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.773794174194336, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7044341564178467, + "num_tokens": 579051268.0, + "step": 22383 + }, + { + "epoch": 2.458159455304195, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4949915409088135, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6854866743087769, + "num_tokens": 579079968.0, + "step": 22384 + }, + { + "epoch": 2.4582692730068088, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9387474060058594, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7110384702682495, + "num_tokens": 579100722.0, + "step": 22385 + }, + { + "epoch": 2.4583790907094225, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4547882080078125, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7037156820297241, + "num_tokens": 579129843.0, + "step": 22386 + }, + { + "epoch": 2.4584889084120363, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.190112352371216, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7001729607582092, + "num_tokens": 579164875.0, + "step": 22387 + }, + { + "epoch": 2.4585987261146496, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6749799251556396, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6933357119560242, + "num_tokens": 579190306.0, + "step": 22388 + }, + { + "epoch": 2.4587085438172633, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7907357215881348, + "learning_rate": 1e-06, + "loss": 0.8862, + "mean_token_accuracy": 0.7418984770774841, + "num_tokens": 579209165.0, + "step": 22389 + }, + { + "epoch": 2.458818361519877, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.588258981704712, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6826980113983154, + "num_tokens": 579236012.0, + "step": 22390 + }, + { + "epoch": 2.458928179222491, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4988958835601807, + "learning_rate": 1e-06, + "loss": 1.0721, + "mean_token_accuracy": 0.690545916557312, + "num_tokens": 579266172.0, + "step": 22391 + }, + { + "epoch": 2.459037996925104, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5029959678649902, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7121418118476868, + "num_tokens": 579293066.0, + "step": 22392 + }, + { + "epoch": 2.459147814627718, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.575702428817749, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7004643082618713, + "num_tokens": 579318838.0, + "step": 22393 + }, + { + "epoch": 2.4592576323303317, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2898142337799072, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6934022307395935, + "num_tokens": 579352649.0, + "step": 22394 + }, + { + "epoch": 2.4593674500329454, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.568077802658081, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7162476778030396, + "num_tokens": 579376862.0, + "step": 22395 + }, + { + "epoch": 2.4594772677355587, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4658422470092773, + "learning_rate": 1e-06, + "loss": 0.9456, + "mean_token_accuracy": 0.7247723937034607, + "num_tokens": 579404419.0, + "step": 22396 + }, + { + "epoch": 2.4595870854381725, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5490829944610596, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7078893184661865, + "num_tokens": 579431207.0, + "step": 22397 + }, + { + "epoch": 2.4596969031407863, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5157623291015625, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.708160400390625, + "num_tokens": 579463351.0, + "step": 22398 + }, + { + "epoch": 2.4598067208434, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6029019355773926, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7186537384986877, + "num_tokens": 579490210.0, + "step": 22399 + }, + { + "epoch": 2.4599165385460138, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5500614643096924, + "learning_rate": 1e-06, + "loss": 0.9146, + "mean_token_accuracy": 0.7366807460784912, + "num_tokens": 579514549.0, + "step": 22400 + }, + { + "epoch": 2.460026356248627, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.680051803588867, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7098873853683472, + "num_tokens": 579538738.0, + "step": 22401 + }, + { + "epoch": 2.460136173951241, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5267434120178223, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6966399550437927, + "num_tokens": 579568753.0, + "step": 22402 + }, + { + "epoch": 2.4602459916538546, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6223607063293457, + "learning_rate": 1e-06, + "loss": 0.9399, + "mean_token_accuracy": 0.7213553190231323, + "num_tokens": 579593085.0, + "step": 22403 + }, + { + "epoch": 2.4603558093564684, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.798722505569458, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6904397010803223, + "num_tokens": 579615851.0, + "step": 22404 + }, + { + "epoch": 2.460465627059082, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4634974002838135, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.7280511856079102, + "num_tokens": 579641659.0, + "step": 22405 + }, + { + "epoch": 2.4605754447616954, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8433265686035156, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.712706446647644, + "num_tokens": 579665632.0, + "step": 22406 + }, + { + "epoch": 2.460685262464309, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.554732084274292, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7317701578140259, + "num_tokens": 579689755.0, + "step": 22407 + }, + { + "epoch": 2.460795080166923, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.5890591144561768, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7150671482086182, + "num_tokens": 579713267.0, + "step": 22408 + }, + { + "epoch": 2.4609048978695367, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6722679138183594, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6891559362411499, + "num_tokens": 579739361.0, + "step": 22409 + }, + { + "epoch": 2.4610147155721505, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.985551357269287, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6896203756332397, + "num_tokens": 579761589.0, + "step": 22410 + }, + { + "epoch": 2.4611245332747638, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5474419593811035, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6936037540435791, + "num_tokens": 579788253.0, + "step": 22411 + }, + { + "epoch": 2.4612343509773775, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4919776916503906, + "learning_rate": 1e-06, + "loss": 1.0797, + "mean_token_accuracy": 0.6851665377616882, + "num_tokens": 579817323.0, + "step": 22412 + }, + { + "epoch": 2.4613441686799913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7097251415252686, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7096453905105591, + "num_tokens": 579840291.0, + "step": 22413 + }, + { + "epoch": 2.461453986382605, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4796810150146484, + "learning_rate": 1e-06, + "loss": 1.1031, + "mean_token_accuracy": 0.6838175654411316, + "num_tokens": 579872806.0, + "step": 22414 + }, + { + "epoch": 2.4615638040852184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4786417484283447, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.7011012434959412, + "num_tokens": 579905609.0, + "step": 22415 + }, + { + "epoch": 2.461673621787832, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4563395977020264, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7060954570770264, + "num_tokens": 579934455.0, + "step": 22416 + }, + { + "epoch": 2.461783439490446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8735430240631104, + "learning_rate": 1e-06, + "loss": 0.9245, + "mean_token_accuracy": 0.723997950553894, + "num_tokens": 579956473.0, + "step": 22417 + }, + { + "epoch": 2.4618932571930596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.953814744949341, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.690830647945404, + "num_tokens": 579983569.0, + "step": 22418 + }, + { + "epoch": 2.462003074895673, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5046257972717285, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7059710025787354, + "num_tokens": 580010210.0, + "step": 22419 + }, + { + "epoch": 2.4621128925982867, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.716388463973999, + "learning_rate": 1e-06, + "loss": 1.1248, + "mean_token_accuracy": 0.6757038831710815, + "num_tokens": 580035267.0, + "step": 22420 + }, + { + "epoch": 2.4622227103009005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.588139533996582, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7146872282028198, + "num_tokens": 580059049.0, + "step": 22421 + }, + { + "epoch": 2.462332528003514, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5552914142608643, + "learning_rate": 1e-06, + "loss": 0.9359, + "mean_token_accuracy": 0.725315511226654, + "num_tokens": 580083726.0, + "step": 22422 + }, + { + "epoch": 2.462442345706128, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.747736930847168, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.694811999797821, + "num_tokens": 580109057.0, + "step": 22423 + }, + { + "epoch": 2.4625521634087413, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7881357669830322, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7076472640037537, + "num_tokens": 580131527.0, + "step": 22424 + }, + { + "epoch": 2.462661981111355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7878003120422363, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7282214164733887, + "num_tokens": 580158296.0, + "step": 22425 + }, + { + "epoch": 2.462771798813969, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.883112907409668, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6993355751037598, + "num_tokens": 580185587.0, + "step": 22426 + }, + { + "epoch": 2.4628816165165826, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520683765411377, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7102674245834351, + "num_tokens": 580211264.0, + "step": 22427 + }, + { + "epoch": 2.4629914342191963, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7708563804626465, + "learning_rate": 1e-06, + "loss": 0.9408, + "mean_token_accuracy": 0.7234729528427124, + "num_tokens": 580238128.0, + "step": 22428 + }, + { + "epoch": 2.4631012519218096, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5191402435302734, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7059314250946045, + "num_tokens": 580263380.0, + "step": 22429 + }, + { + "epoch": 2.4632110696244234, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5231058597564697, + "learning_rate": 1e-06, + "loss": 1.126, + "mean_token_accuracy": 0.6744877696037292, + "num_tokens": 580291588.0, + "step": 22430 + }, + { + "epoch": 2.463320887327037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.474456548690796, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6975979804992676, + "num_tokens": 580321155.0, + "step": 22431 + }, + { + "epoch": 2.463430705029651, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5104551315307617, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.708336353302002, + "num_tokens": 580346786.0, + "step": 22432 + }, + { + "epoch": 2.4635405227322646, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.757977247238159, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7258999347686768, + "num_tokens": 580368009.0, + "step": 22433 + }, + { + "epoch": 2.463650340434878, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.822047472000122, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7221297025680542, + "num_tokens": 580391179.0, + "step": 22434 + }, + { + "epoch": 2.4637601581374917, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.502469539642334, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7009572982788086, + "num_tokens": 580417400.0, + "step": 22435 + }, + { + "epoch": 2.4638699758401055, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7373111248016357, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.710773766040802, + "num_tokens": 580440859.0, + "step": 22436 + }, + { + "epoch": 2.4639797935427192, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7065253257751465, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7038131952285767, + "num_tokens": 580465329.0, + "step": 22437 + }, + { + "epoch": 2.464089611245333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3952739238739014, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7267634868621826, + "num_tokens": 580494802.0, + "step": 22438 + }, + { + "epoch": 2.4641994289479463, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.483299732208252, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7143125534057617, + "num_tokens": 580520357.0, + "step": 22439 + }, + { + "epoch": 2.46430924665056, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 5.091551780700684, + "learning_rate": 1e-06, + "loss": 0.8474, + "mean_token_accuracy": 0.7480201125144958, + "num_tokens": 580544172.0, + "step": 22440 + }, + { + "epoch": 2.464419064353174, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.867351531982422, + "learning_rate": 1e-06, + "loss": 0.9224, + "mean_token_accuracy": 0.7262355089187622, + "num_tokens": 580565362.0, + "step": 22441 + }, + { + "epoch": 2.4645288820557876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4925179481506348, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7117928862571716, + "num_tokens": 580592144.0, + "step": 22442 + }, + { + "epoch": 2.464638699758401, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.5051538944244385, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.7010359764099121, + "num_tokens": 580619409.0, + "step": 22443 + }, + { + "epoch": 2.4647485174610146, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0685477256774902, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7076606750488281, + "num_tokens": 580647254.0, + "step": 22444 + }, + { + "epoch": 2.4648583351636284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7334437370300293, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7085933089256287, + "num_tokens": 580669445.0, + "step": 22445 + }, + { + "epoch": 2.464968152866242, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.9091691970825195, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7173051834106445, + "num_tokens": 580698773.0, + "step": 22446 + }, + { + "epoch": 2.4650779705688555, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.406614065170288, + "learning_rate": 1e-06, + "loss": 1.0748, + "mean_token_accuracy": 0.688843309879303, + "num_tokens": 580728787.0, + "step": 22447 + }, + { + "epoch": 2.4651877882714692, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.411647319793701, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7097780704498291, + "num_tokens": 580758320.0, + "step": 22448 + }, + { + "epoch": 2.465297605974083, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3206372261047363, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.7064568400382996, + "num_tokens": 580787451.0, + "step": 22449 + }, + { + "epoch": 2.4654074236766967, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.455592632293701, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7141611576080322, + "num_tokens": 580813681.0, + "step": 22450 + }, + { + "epoch": 2.4655172413793105, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6410648822784424, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6971539258956909, + "num_tokens": 580838606.0, + "step": 22451 + }, + { + "epoch": 2.465627059081924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5094845294952393, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7113797664642334, + "num_tokens": 580866549.0, + "step": 22452 + }, + { + "epoch": 2.4657368767845376, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.63126540184021, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6935861110687256, + "num_tokens": 580893306.0, + "step": 22453 + }, + { + "epoch": 2.4658466944871513, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8989040851593018, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7069759964942932, + "num_tokens": 580916349.0, + "step": 22454 + }, + { + "epoch": 2.465956512189765, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.98654842376709, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7298705577850342, + "num_tokens": 580936238.0, + "step": 22455 + }, + { + "epoch": 2.466066329892379, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.376270294189453, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.683920681476593, + "num_tokens": 580970218.0, + "step": 22456 + }, + { + "epoch": 2.466176147594992, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7117879390716553, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7210625410079956, + "num_tokens": 580993275.0, + "step": 22457 + }, + { + "epoch": 2.466285965297606, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.296010971069336, + "learning_rate": 1e-06, + "loss": 0.9839, + "mean_token_accuracy": 0.721426248550415, + "num_tokens": 581023208.0, + "step": 22458 + }, + { + "epoch": 2.4663957830002197, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5289077758789062, + "learning_rate": 1e-06, + "loss": 0.9204, + "mean_token_accuracy": 0.73622727394104, + "num_tokens": 581047242.0, + "step": 22459 + }, + { + "epoch": 2.4665056007028334, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8710765838623047, + "learning_rate": 1e-06, + "loss": 0.9256, + "mean_token_accuracy": 0.7240854501724243, + "num_tokens": 581069604.0, + "step": 22460 + }, + { + "epoch": 2.466615418405447, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.839982509613037, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7148813605308533, + "num_tokens": 581093665.0, + "step": 22461 + }, + { + "epoch": 2.4667252361080605, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6285400390625, + "learning_rate": 1e-06, + "loss": 0.8851, + "mean_token_accuracy": 0.7361260652542114, + "num_tokens": 581117503.0, + "step": 22462 + }, + { + "epoch": 2.4668350538106742, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7305033206939697, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.719911515712738, + "num_tokens": 581141075.0, + "step": 22463 + }, + { + "epoch": 2.466944871513288, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.555415630340576, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.696933925151825, + "num_tokens": 581168850.0, + "step": 22464 + }, + { + "epoch": 2.4670546892159018, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7308385372161865, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.7105115652084351, + "num_tokens": 581194901.0, + "step": 22465 + }, + { + "epoch": 2.467164506918515, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.563831329345703, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7030913829803467, + "num_tokens": 581223253.0, + "step": 22466 + }, + { + "epoch": 2.467274324621129, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.404621124267578, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6865234375, + "num_tokens": 581257657.0, + "step": 22467 + }, + { + "epoch": 2.4673841423237426, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8493287563323975, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7079610824584961, + "num_tokens": 581280242.0, + "step": 22468 + }, + { + "epoch": 2.4674939600263563, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7117373943328857, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7047740817070007, + "num_tokens": 581309213.0, + "step": 22469 + }, + { + "epoch": 2.4676037777289697, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0721142292022705, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7099577188491821, + "num_tokens": 581328185.0, + "step": 22470 + }, + { + "epoch": 2.4677135954315834, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.063300848007202, + "learning_rate": 1e-06, + "loss": 0.8933, + "mean_token_accuracy": 0.7370282411575317, + "num_tokens": 581348910.0, + "step": 22471 + }, + { + "epoch": 2.467823413134197, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5475378036499023, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.7097654342651367, + "num_tokens": 581375189.0, + "step": 22472 + }, + { + "epoch": 2.467933230836811, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7585256099700928, + "learning_rate": 1e-06, + "loss": 0.9169, + "mean_token_accuracy": 0.7255655527114868, + "num_tokens": 581396012.0, + "step": 22473 + }, + { + "epoch": 2.4680430485394247, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.50449800491333, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7100024223327637, + "num_tokens": 581423460.0, + "step": 22474 + }, + { + "epoch": 2.468152866242038, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.523233413696289, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7235661149024963, + "num_tokens": 581449750.0, + "step": 22475 + }, + { + "epoch": 2.4682626839446518, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8956120014190674, + "learning_rate": 1e-06, + "loss": 0.9333, + "mean_token_accuracy": 0.7212694883346558, + "num_tokens": 581470226.0, + "step": 22476 + }, + { + "epoch": 2.4683725016472655, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.640782356262207, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7048019170761108, + "num_tokens": 581496730.0, + "step": 22477 + }, + { + "epoch": 2.4684823193498793, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.600341796875, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7178348302841187, + "num_tokens": 581520234.0, + "step": 22478 + }, + { + "epoch": 2.468592137052493, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.479975700378418, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7138260006904602, + "num_tokens": 581548145.0, + "step": 22479 + }, + { + "epoch": 2.4687019547551063, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7664859294891357, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7179131507873535, + "num_tokens": 581570008.0, + "step": 22480 + }, + { + "epoch": 2.46881177245772, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.661242723464966, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7225103378295898, + "num_tokens": 581594675.0, + "step": 22481 + }, + { + "epoch": 2.468921590160334, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.569549798965454, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7095506191253662, + "num_tokens": 581621903.0, + "step": 22482 + }, + { + "epoch": 2.4690314078629476, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4783525466918945, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7043956518173218, + "num_tokens": 581653424.0, + "step": 22483 + }, + { + "epoch": 2.4691412255655614, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6230757236480713, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7091113328933716, + "num_tokens": 581680659.0, + "step": 22484 + }, + { + "epoch": 2.4692510432681747, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5959012508392334, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7041000723838806, + "num_tokens": 581706034.0, + "step": 22485 + }, + { + "epoch": 2.4693608609707884, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4170548915863037, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6987178325653076, + "num_tokens": 581736537.0, + "step": 22486 + }, + { + "epoch": 2.469470678673402, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9288535118103027, + "learning_rate": 1e-06, + "loss": 0.9141, + "mean_token_accuracy": 0.732334315776825, + "num_tokens": 581758212.0, + "step": 22487 + }, + { + "epoch": 2.469580496376016, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9158737659454346, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7192661762237549, + "num_tokens": 581777728.0, + "step": 22488 + }, + { + "epoch": 2.4696903140786297, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 4.208296775817871, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.7408699989318848, + "num_tokens": 581802384.0, + "step": 22489 + }, + { + "epoch": 2.469800131781243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.682758331298828, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6962175369262695, + "num_tokens": 581831897.0, + "step": 22490 + }, + { + "epoch": 2.469909949483857, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1486494541168213, + "learning_rate": 1e-06, + "loss": 0.9367, + "mean_token_accuracy": 0.7254464626312256, + "num_tokens": 581852383.0, + "step": 22491 + }, + { + "epoch": 2.4700197671864705, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9808127880096436, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7097429037094116, + "num_tokens": 581876337.0, + "step": 22492 + }, + { + "epoch": 2.4701295848890843, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.591287851333618, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6897783875465393, + "num_tokens": 581904782.0, + "step": 22493 + }, + { + "epoch": 2.4702394025916976, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2772793769836426, + "learning_rate": 1e-06, + "loss": 1.1351, + "mean_token_accuracy": 0.6722434759140015, + "num_tokens": 581937627.0, + "step": 22494 + }, + { + "epoch": 2.4703492202943114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5928075313568115, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.7058581113815308, + "num_tokens": 581961556.0, + "step": 22495 + }, + { + "epoch": 2.470459037996925, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7269744873046875, + "learning_rate": 1e-06, + "loss": 0.8994, + "mean_token_accuracy": 0.7283419370651245, + "num_tokens": 581982982.0, + "step": 22496 + }, + { + "epoch": 2.470568855699539, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7551004886627197, + "learning_rate": 1e-06, + "loss": 0.9517, + "mean_token_accuracy": 0.7264411449432373, + "num_tokens": 582007050.0, + "step": 22497 + }, + { + "epoch": 2.470678673402152, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4566779136657715, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7153487205505371, + "num_tokens": 582037396.0, + "step": 22498 + }, + { + "epoch": 2.470788491104766, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.696399450302124, + "learning_rate": 1e-06, + "loss": 1.0797, + "mean_token_accuracy": 0.6875996589660645, + "num_tokens": 582060656.0, + "step": 22499 + }, + { + "epoch": 2.4708983088073797, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2631118297576904, + "learning_rate": 1e-06, + "loss": 1.1132, + "mean_token_accuracy": 0.6767910718917847, + "num_tokens": 582098147.0, + "step": 22500 + }, + { + "epoch": 2.4710081265099935, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8010709285736084, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6924342513084412, + "num_tokens": 582120838.0, + "step": 22501 + }, + { + "epoch": 2.471117944212607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.966677188873291, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.710822582244873, + "num_tokens": 582142587.0, + "step": 22502 + }, + { + "epoch": 2.4712277619152205, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.925745725631714, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6953125, + "num_tokens": 582164094.0, + "step": 22503 + }, + { + "epoch": 2.4713375796178343, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.444178342819214, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.700596809387207, + "num_tokens": 582191897.0, + "step": 22504 + }, + { + "epoch": 2.471447397320448, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6258296966552734, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.706376314163208, + "num_tokens": 582215152.0, + "step": 22505 + }, + { + "epoch": 2.471557215023062, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9095065593719482, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7108707427978516, + "num_tokens": 582235363.0, + "step": 22506 + }, + { + "epoch": 2.4716670327256756, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3448007106781006, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7081073522567749, + "num_tokens": 582268099.0, + "step": 22507 + }, + { + "epoch": 2.471776850428289, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.967684030532837, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7114764451980591, + "num_tokens": 582287577.0, + "step": 22508 + }, + { + "epoch": 2.4718866681309026, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5158982276916504, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7096071243286133, + "num_tokens": 582316791.0, + "step": 22509 + }, + { + "epoch": 2.4719964858335164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6092982292175293, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7140876054763794, + "num_tokens": 582343571.0, + "step": 22510 + }, + { + "epoch": 2.47210630353613, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.4129388332366943, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.7047932744026184, + "num_tokens": 582370286.0, + "step": 22511 + }, + { + "epoch": 2.472216121238744, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.420640230178833, + "learning_rate": 1e-06, + "loss": 0.9332, + "mean_token_accuracy": 0.7293359041213989, + "num_tokens": 582396706.0, + "step": 22512 + }, + { + "epoch": 2.472325938941357, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.777987480163574, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7031662464141846, + "num_tokens": 582419388.0, + "step": 22513 + }, + { + "epoch": 2.472435756643971, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4022064208984375, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6968722343444824, + "num_tokens": 582449128.0, + "step": 22514 + }, + { + "epoch": 2.4725455743465847, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2998738288879395, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7089505791664124, + "num_tokens": 582481366.0, + "step": 22515 + }, + { + "epoch": 2.4726553920491985, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.496366262435913, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.6939666867256165, + "num_tokens": 582509599.0, + "step": 22516 + }, + { + "epoch": 2.472765209751812, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5081329345703125, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.7033435106277466, + "num_tokens": 582537510.0, + "step": 22517 + }, + { + "epoch": 2.4728750274544256, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4210927486419678, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7136843204498291, + "num_tokens": 582565833.0, + "step": 22518 + }, + { + "epoch": 2.4729848451570393, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3247718811035156, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7069246172904968, + "num_tokens": 582595069.0, + "step": 22519 + }, + { + "epoch": 2.473094662859653, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4456136226654053, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.703578531742096, + "num_tokens": 582626140.0, + "step": 22520 + }, + { + "epoch": 2.473204480562267, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.333056688308716, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7175393104553223, + "num_tokens": 582656988.0, + "step": 22521 + }, + { + "epoch": 2.47331429826488, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.345740556716919, + "learning_rate": 1e-06, + "loss": 1.1423, + "mean_token_accuracy": 0.6771079301834106, + "num_tokens": 582690171.0, + "step": 22522 + }, + { + "epoch": 2.473424115967494, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.633147716522217, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7238913774490356, + "num_tokens": 582714233.0, + "step": 22523 + }, + { + "epoch": 2.4735339336701077, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.621086597442627, + "learning_rate": 1e-06, + "loss": 0.9178, + "mean_token_accuracy": 0.7306530475616455, + "num_tokens": 582738901.0, + "step": 22524 + }, + { + "epoch": 2.4736437513727214, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7051382064819336, + "learning_rate": 1e-06, + "loss": 0.9383, + "mean_token_accuracy": 0.726266622543335, + "num_tokens": 582761168.0, + "step": 22525 + }, + { + "epoch": 2.4737535690753347, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6210973262786865, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.707756757736206, + "num_tokens": 582786137.0, + "step": 22526 + }, + { + "epoch": 2.4738633867779485, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.573150157928467, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6973621845245361, + "num_tokens": 582812937.0, + "step": 22527 + }, + { + "epoch": 2.4739732044805622, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5536887645721436, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7008878588676453, + "num_tokens": 582840990.0, + "step": 22528 + }, + { + "epoch": 2.474083022183176, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.775085687637329, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7208249568939209, + "num_tokens": 582862034.0, + "step": 22529 + }, + { + "epoch": 2.4741928398857898, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.676121711730957, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7069902420043945, + "num_tokens": 582887359.0, + "step": 22530 + }, + { + "epoch": 2.474302657588403, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5755186080932617, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.7285158038139343, + "num_tokens": 582911724.0, + "step": 22531 + }, + { + "epoch": 2.474412475291017, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4671273231506348, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7104759216308594, + "num_tokens": 582940079.0, + "step": 22532 + }, + { + "epoch": 2.4745222929936306, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.616321563720703, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.7092405557632446, + "num_tokens": 582966645.0, + "step": 22533 + }, + { + "epoch": 2.4746321106962443, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.795560359954834, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7056015133857727, + "num_tokens": 582990337.0, + "step": 22534 + }, + { + "epoch": 2.474741928398858, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4537270069122314, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7038928866386414, + "num_tokens": 583019007.0, + "step": 22535 + }, + { + "epoch": 2.4748517461014714, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.613334894180298, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7098723649978638, + "num_tokens": 583046525.0, + "step": 22536 + }, + { + "epoch": 2.474961563804085, + "ewc_loss": 1.0967254638671875e-05, + "grad_norm": 80.54454803466797, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6898605823516846, + "num_tokens": 583075457.0, + "step": 22537 + }, + { + "epoch": 2.475071381506699, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9129416942596436, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7079586982727051, + "num_tokens": 583097935.0, + "step": 22538 + }, + { + "epoch": 2.4751811992093127, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0245449542999268, + "learning_rate": 1e-06, + "loss": 0.9406, + "mean_token_accuracy": 0.7196645736694336, + "num_tokens": 583116913.0, + "step": 22539 + }, + { + "epoch": 2.4752910169119264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.431169271469116, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7227120995521545, + "num_tokens": 583152270.0, + "step": 22540 + }, + { + "epoch": 2.4754008346145397, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5076842308044434, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.6985293030738831, + "num_tokens": 583181986.0, + "step": 22541 + }, + { + "epoch": 2.4755106523171535, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.257526159286499, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.6994611024856567, + "num_tokens": 583212767.0, + "step": 22542 + }, + { + "epoch": 2.4756204700197673, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.035371780395508, + "learning_rate": 1e-06, + "loss": 0.9134, + "mean_token_accuracy": 0.7315486669540405, + "num_tokens": 583230748.0, + "step": 22543 + }, + { + "epoch": 2.475730287722381, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9042017459869385, + "learning_rate": 1e-06, + "loss": 0.9286, + "mean_token_accuracy": 0.7313129305839539, + "num_tokens": 583250752.0, + "step": 22544 + }, + { + "epoch": 2.4758401054249943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.607006788253784, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.6852384805679321, + "num_tokens": 583278437.0, + "step": 22545 + }, + { + "epoch": 2.475949923127608, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.556497097015381, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6935609579086304, + "num_tokens": 583305175.0, + "step": 22546 + }, + { + "epoch": 2.476059740830222, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7660598754882812, + "learning_rate": 1e-06, + "loss": 1.0013, + "mean_token_accuracy": 0.708591103553772, + "num_tokens": 583327026.0, + "step": 22547 + }, + { + "epoch": 2.4761695585328356, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3912649154663086, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6973954439163208, + "num_tokens": 583356237.0, + "step": 22548 + }, + { + "epoch": 2.476279376235449, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.306684732437134, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7054051756858826, + "num_tokens": 583384805.0, + "step": 22549 + }, + { + "epoch": 2.4763891939380627, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.55977725982666, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7157461047172546, + "num_tokens": 583409634.0, + "step": 22550 + }, + { + "epoch": 2.4764990116406764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.42626690864563, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6867257356643677, + "num_tokens": 583439142.0, + "step": 22551 + }, + { + "epoch": 2.47660882934329, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.912036657333374, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6960164904594421, + "num_tokens": 583462208.0, + "step": 22552 + }, + { + "epoch": 2.476718647045904, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7088842391967773, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7156985998153687, + "num_tokens": 583485821.0, + "step": 22553 + }, + { + "epoch": 2.4768284647485173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.402517557144165, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6887603998184204, + "num_tokens": 583516572.0, + "step": 22554 + }, + { + "epoch": 2.476938282451131, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6097681522369385, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7057144641876221, + "num_tokens": 583542215.0, + "step": 22555 + }, + { + "epoch": 2.4770481001537448, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.641660451889038, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.692420244216919, + "num_tokens": 583568246.0, + "step": 22556 + }, + { + "epoch": 2.4771579178563585, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5686490535736084, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7175340056419373, + "num_tokens": 583593046.0, + "step": 22557 + }, + { + "epoch": 2.4772677355589723, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7144529819488525, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7203753590583801, + "num_tokens": 583617182.0, + "step": 22558 + }, + { + "epoch": 2.4773775532615856, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6767466068267822, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7069809436798096, + "num_tokens": 583640409.0, + "step": 22559 + }, + { + "epoch": 2.4774873709641994, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.532860517501831, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.707493245601654, + "num_tokens": 583667501.0, + "step": 22560 + }, + { + "epoch": 2.477597188666813, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6451072692871094, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7124667763710022, + "num_tokens": 583691981.0, + "step": 22561 + }, + { + "epoch": 2.477707006369427, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.366178274154663, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7055424451828003, + "num_tokens": 583723544.0, + "step": 22562 + }, + { + "epoch": 2.4778168240720406, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7882022857666016, + "learning_rate": 1e-06, + "loss": 0.9797, + "mean_token_accuracy": 0.7116567492485046, + "num_tokens": 583745249.0, + "step": 22563 + }, + { + "epoch": 2.477926641774654, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.682689905166626, + "learning_rate": 1e-06, + "loss": 0.935, + "mean_token_accuracy": 0.7255355715751648, + "num_tokens": 583767713.0, + "step": 22564 + }, + { + "epoch": 2.4780364594772677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.536376953125, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.710456907749176, + "num_tokens": 583796351.0, + "step": 22565 + }, + { + "epoch": 2.4781462771798815, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.769901990890503, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7193588018417358, + "num_tokens": 583818522.0, + "step": 22566 + }, + { + "epoch": 2.478256094882495, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9185400009155273, + "learning_rate": 1e-06, + "loss": 0.9331, + "mean_token_accuracy": 0.7270302772521973, + "num_tokens": 583841306.0, + "step": 22567 + }, + { + "epoch": 2.478365912585109, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.768049955368042, + "learning_rate": 1e-06, + "loss": 0.9039, + "mean_token_accuracy": 0.7352290153503418, + "num_tokens": 583862720.0, + "step": 22568 + }, + { + "epoch": 2.4784757302877223, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.530869960784912, + "learning_rate": 1e-06, + "loss": 1.1193, + "mean_token_accuracy": 0.6756387948989868, + "num_tokens": 583891557.0, + "step": 22569 + }, + { + "epoch": 2.478585547990336, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.766761302947998, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.709264874458313, + "num_tokens": 583915524.0, + "step": 22570 + }, + { + "epoch": 2.47869536569295, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.270641803741455, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7198486328125, + "num_tokens": 583947311.0, + "step": 22571 + }, + { + "epoch": 2.4788051833955635, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.558475971221924, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7043047547340393, + "num_tokens": 583971538.0, + "step": 22572 + }, + { + "epoch": 2.478915001098177, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1637110710144043, + "learning_rate": 1e-06, + "loss": 0.9576, + "mean_token_accuracy": 0.7204502820968628, + "num_tokens": 583990662.0, + "step": 22573 + }, + { + "epoch": 2.4790248188007906, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4253413677215576, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7039281725883484, + "num_tokens": 584020069.0, + "step": 22574 + }, + { + "epoch": 2.4791346365034044, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8071200847625732, + "learning_rate": 1e-06, + "loss": 0.9363, + "mean_token_accuracy": 0.7287420034408569, + "num_tokens": 584043942.0, + "step": 22575 + }, + { + "epoch": 2.479244454206018, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8543953895568848, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7101171016693115, + "num_tokens": 584064359.0, + "step": 22576 + }, + { + "epoch": 2.4793542719086314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.645972490310669, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6937755346298218, + "num_tokens": 584089446.0, + "step": 22577 + }, + { + "epoch": 2.479464089611245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6781673431396484, + "learning_rate": 1e-06, + "loss": 0.9209, + "mean_token_accuracy": 0.7307232022285461, + "num_tokens": 584112276.0, + "step": 22578 + }, + { + "epoch": 2.479573907313859, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1790337562561035, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.6942852735519409, + "num_tokens": 584131282.0, + "step": 22579 + }, + { + "epoch": 2.4796837250164727, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6958367824554443, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6990722417831421, + "num_tokens": 584157027.0, + "step": 22580 + }, + { + "epoch": 2.4797935427190865, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.551858425140381, + "learning_rate": 1e-06, + "loss": 1.1155, + "mean_token_accuracy": 0.6806313991546631, + "num_tokens": 584187271.0, + "step": 22581 + }, + { + "epoch": 2.4799033604217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6696970462799072, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7067387104034424, + "num_tokens": 584214126.0, + "step": 22582 + }, + { + "epoch": 2.4800131781243135, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6365225315093994, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7246513366699219, + "num_tokens": 584238308.0, + "step": 22583 + }, + { + "epoch": 2.4801229958269273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7583141326904297, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6828153729438782, + "num_tokens": 584260562.0, + "step": 22584 + }, + { + "epoch": 2.480232813529541, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.695772409439087, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.7002463936805725, + "num_tokens": 584285600.0, + "step": 22585 + }, + { + "epoch": 2.480342631232155, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8177967071533203, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7119308710098267, + "num_tokens": 584309999.0, + "step": 22586 + }, + { + "epoch": 2.480452448934768, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9672110080718994, + "learning_rate": 1e-06, + "loss": 0.9533, + "mean_token_accuracy": 0.7137241363525391, + "num_tokens": 584331339.0, + "step": 22587 + }, + { + "epoch": 2.480562266637382, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8505606651306152, + "learning_rate": 1e-06, + "loss": 0.8599, + "mean_token_accuracy": 0.7461434602737427, + "num_tokens": 584352018.0, + "step": 22588 + }, + { + "epoch": 2.4806720843399956, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7040328979492188, + "learning_rate": 1e-06, + "loss": 0.9128, + "mean_token_accuracy": 0.7433308362960815, + "num_tokens": 584374314.0, + "step": 22589 + }, + { + "epoch": 2.4807819020426094, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6274943351745605, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6828159093856812, + "num_tokens": 584396902.0, + "step": 22590 + }, + { + "epoch": 2.480891719745223, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6467738151550293, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7225886583328247, + "num_tokens": 584419211.0, + "step": 22591 + }, + { + "epoch": 2.4810015374478365, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.766057252883911, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7086201310157776, + "num_tokens": 584442109.0, + "step": 22592 + }, + { + "epoch": 2.4811113551504502, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.762315034866333, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7138485908508301, + "num_tokens": 584466087.0, + "step": 22593 + }, + { + "epoch": 2.481221172853064, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6036596298217773, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6979742050170898, + "num_tokens": 584490736.0, + "step": 22594 + }, + { + "epoch": 2.4813309905556777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2929108142852783, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.7047085762023926, + "num_tokens": 584525952.0, + "step": 22595 + }, + { + "epoch": 2.481440808258291, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5939197540283203, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7047293186187744, + "num_tokens": 584554175.0, + "step": 22596 + }, + { + "epoch": 2.481550625960905, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4874022006988525, + "learning_rate": 1e-06, + "loss": 0.9652, + "mean_token_accuracy": 0.7131953239440918, + "num_tokens": 584581818.0, + "step": 22597 + }, + { + "epoch": 2.4816604436635186, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5504579544067383, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7117235660552979, + "num_tokens": 584608837.0, + "step": 22598 + }, + { + "epoch": 2.4817702613661323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.487454891204834, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6915328502655029, + "num_tokens": 584637585.0, + "step": 22599 + }, + { + "epoch": 2.4818800790687456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5989274978637695, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7098301649093628, + "num_tokens": 584661031.0, + "step": 22600 + }, + { + "epoch": 2.4819898967713594, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6371631622314453, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7189585566520691, + "num_tokens": 584684536.0, + "step": 22601 + }, + { + "epoch": 2.482099714473973, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.970517873764038, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.720840573310852, + "num_tokens": 584708199.0, + "step": 22602 + }, + { + "epoch": 2.482209532176587, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.612846612930298, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.688731849193573, + "num_tokens": 584733161.0, + "step": 22603 + }, + { + "epoch": 2.4823193498792007, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.60736346244812, + "learning_rate": 1e-06, + "loss": 0.9491, + "mean_token_accuracy": 0.7240718603134155, + "num_tokens": 584757816.0, + "step": 22604 + }, + { + "epoch": 2.482429167581814, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3776848316192627, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.724826455116272, + "num_tokens": 584787031.0, + "step": 22605 + }, + { + "epoch": 2.4825389852844277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7361114025115967, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7236181497573853, + "num_tokens": 584808485.0, + "step": 22606 + }, + { + "epoch": 2.4826488029870415, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.829030990600586, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7115691900253296, + "num_tokens": 584830143.0, + "step": 22607 + }, + { + "epoch": 2.4827586206896552, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6594464778900146, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7066867351531982, + "num_tokens": 584854495.0, + "step": 22608 + }, + { + "epoch": 2.482868438392269, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7492401599884033, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.70569908618927, + "num_tokens": 584877030.0, + "step": 22609 + }, + { + "epoch": 2.4829782560948823, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.678520679473877, + "learning_rate": 1e-06, + "loss": 0.8841, + "mean_token_accuracy": 0.7369204163551331, + "num_tokens": 584899034.0, + "step": 22610 + }, + { + "epoch": 2.483088073797496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7916247844696045, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7033869028091431, + "num_tokens": 584924205.0, + "step": 22611 + }, + { + "epoch": 2.48319789150011, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6302266120910645, + "learning_rate": 1e-06, + "loss": 1.071, + "mean_token_accuracy": 0.6896166801452637, + "num_tokens": 584949529.0, + "step": 22612 + }, + { + "epoch": 2.4833077092027236, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7948880195617676, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7127399444580078, + "num_tokens": 584972765.0, + "step": 22613 + }, + { + "epoch": 2.4834175269053373, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3334615230560303, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.689673900604248, + "num_tokens": 585004051.0, + "step": 22614 + }, + { + "epoch": 2.4835273446079507, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.283613920211792, + "learning_rate": 1e-06, + "loss": 1.1136, + "mean_token_accuracy": 0.6773519515991211, + "num_tokens": 585035043.0, + "step": 22615 + }, + { + "epoch": 2.4836371623105644, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.555697441101074, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7133222818374634, + "num_tokens": 585063833.0, + "step": 22616 + }, + { + "epoch": 2.483746980013178, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.548353672027588, + "learning_rate": 1e-06, + "loss": 1.0659, + "mean_token_accuracy": 0.6923083662986755, + "num_tokens": 585089772.0, + "step": 22617 + }, + { + "epoch": 2.483856797715792, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6019287109375, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.694717288017273, + "num_tokens": 585113862.0, + "step": 22618 + }, + { + "epoch": 2.4839666154184057, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7526776790618896, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7046774625778198, + "num_tokens": 585141848.0, + "step": 22619 + }, + { + "epoch": 2.484076433121019, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6769163608551025, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7069767117500305, + "num_tokens": 585165372.0, + "step": 22620 + }, + { + "epoch": 2.4841862508236328, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.638941526412964, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6994309425354004, + "num_tokens": 585190186.0, + "step": 22621 + }, + { + "epoch": 2.4842960685262465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6386795043945312, + "learning_rate": 1e-06, + "loss": 0.9761, + "mean_token_accuracy": 0.7212806344032288, + "num_tokens": 585215818.0, + "step": 22622 + }, + { + "epoch": 2.4844058862288603, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.573552370071411, + "learning_rate": 1e-06, + "loss": 0.9175, + "mean_token_accuracy": 0.728789210319519, + "num_tokens": 585238265.0, + "step": 22623 + }, + { + "epoch": 2.4845157039314736, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5894687175750732, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7240914106369019, + "num_tokens": 585266467.0, + "step": 22624 + }, + { + "epoch": 2.4846255216340873, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.345240354537964, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.7206901907920837, + "num_tokens": 585295652.0, + "step": 22625 + }, + { + "epoch": 2.484735339336701, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.761204481124878, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7214375734329224, + "num_tokens": 585323200.0, + "step": 22626 + }, + { + "epoch": 2.484845157039315, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.61509108543396, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7134360074996948, + "num_tokens": 585347754.0, + "step": 22627 + }, + { + "epoch": 2.484954974741928, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4736733436584473, + "learning_rate": 1e-06, + "loss": 0.9814, + "mean_token_accuracy": 0.7110900282859802, + "num_tokens": 585373003.0, + "step": 22628 + }, + { + "epoch": 2.485064792444542, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9216928482055664, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.7037447690963745, + "num_tokens": 585399580.0, + "step": 22629 + }, + { + "epoch": 2.4851746101471557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.613828659057617, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7190330028533936, + "num_tokens": 585423217.0, + "step": 22630 + }, + { + "epoch": 2.4852844278497694, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5671660900115967, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.7045571804046631, + "num_tokens": 585451117.0, + "step": 22631 + }, + { + "epoch": 2.485394245552383, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3811428546905518, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.708508312702179, + "num_tokens": 585480424.0, + "step": 22632 + }, + { + "epoch": 2.4855040632549965, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3788678646087646, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7204800248146057, + "num_tokens": 585508401.0, + "step": 22633 + }, + { + "epoch": 2.4856138809576103, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.381254196166992, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6948288679122925, + "num_tokens": 585537066.0, + "step": 22634 + }, + { + "epoch": 2.485723698660224, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.426226854324341, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.700128972530365, + "num_tokens": 585564235.0, + "step": 22635 + }, + { + "epoch": 2.485833516362838, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8805017471313477, + "learning_rate": 1e-06, + "loss": 0.9603, + "mean_token_accuracy": 0.7155330181121826, + "num_tokens": 585585897.0, + "step": 22636 + }, + { + "epoch": 2.4859433340654515, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6376209259033203, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.7072139978408813, + "num_tokens": 585613902.0, + "step": 22637 + }, + { + "epoch": 2.486053151768065, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7519373893737793, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7060441970825195, + "num_tokens": 585640191.0, + "step": 22638 + }, + { + "epoch": 2.4861629694706786, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5273914337158203, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7057082056999207, + "num_tokens": 585668203.0, + "step": 22639 + }, + { + "epoch": 2.4862727871732924, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.667738914489746, + "learning_rate": 1e-06, + "loss": 0.98, + "mean_token_accuracy": 0.7132232785224915, + "num_tokens": 585692082.0, + "step": 22640 + }, + { + "epoch": 2.486382604875906, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.996555805206299, + "learning_rate": 1e-06, + "loss": 0.8918, + "mean_token_accuracy": 0.7298682332038879, + "num_tokens": 585713766.0, + "step": 22641 + }, + { + "epoch": 2.48649242257852, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.536734104156494, + "learning_rate": 1e-06, + "loss": 0.9148, + "mean_token_accuracy": 0.7277457118034363, + "num_tokens": 585739024.0, + "step": 22642 + }, + { + "epoch": 2.486602240281133, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.368441104888916, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7026872634887695, + "num_tokens": 585768397.0, + "step": 22643 + }, + { + "epoch": 2.486712057983747, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.817809820175171, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7093198299407959, + "num_tokens": 585790832.0, + "step": 22644 + }, + { + "epoch": 2.4868218756863607, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.650299310684204, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6974650621414185, + "num_tokens": 585816244.0, + "step": 22645 + }, + { + "epoch": 2.4869316933889745, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5680155754089355, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7253544330596924, + "num_tokens": 585841796.0, + "step": 22646 + }, + { + "epoch": 2.4870415110915878, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.566789388656616, + "learning_rate": 1e-06, + "loss": 1.0902, + "mean_token_accuracy": 0.6864476203918457, + "num_tokens": 585868691.0, + "step": 22647 + }, + { + "epoch": 2.4871513287942015, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6045594215393066, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.7003448009490967, + "num_tokens": 585896525.0, + "step": 22648 + }, + { + "epoch": 2.4872611464968153, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.513218402862549, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7054336071014404, + "num_tokens": 585922581.0, + "step": 22649 + }, + { + "epoch": 2.487370964199429, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5834133625030518, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7183557152748108, + "num_tokens": 585952214.0, + "step": 22650 + }, + { + "epoch": 2.487480781902043, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4405510425567627, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7095680236816406, + "num_tokens": 585981483.0, + "step": 22651 + }, + { + "epoch": 2.487590599604656, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.519352674484253, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7099606990814209, + "num_tokens": 586006401.0, + "step": 22652 + }, + { + "epoch": 2.48770041730727, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.462207078933716, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6942600011825562, + "num_tokens": 586035868.0, + "step": 22653 + }, + { + "epoch": 2.4878102350098836, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.50163197517395, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6989191770553589, + "num_tokens": 586063210.0, + "step": 22654 + }, + { + "epoch": 2.4879200527124974, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5849649906158447, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7263751029968262, + "num_tokens": 586088749.0, + "step": 22655 + }, + { + "epoch": 2.4880298704151107, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5877845287323, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7156891226768494, + "num_tokens": 586114721.0, + "step": 22656 + }, + { + "epoch": 2.4881396881177245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6576778888702393, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7209267616271973, + "num_tokens": 586138740.0, + "step": 22657 + }, + { + "epoch": 2.488249505820338, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.541689395904541, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6995830535888672, + "num_tokens": 586163806.0, + "step": 22658 + }, + { + "epoch": 2.488359323522952, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.557750940322876, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6945514678955078, + "num_tokens": 586189894.0, + "step": 22659 + }, + { + "epoch": 2.4884691412255657, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.450773239135742, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7012490630149841, + "num_tokens": 586218294.0, + "step": 22660 + }, + { + "epoch": 2.488578958928179, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8245372772216797, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7098946571350098, + "num_tokens": 586241428.0, + "step": 22661 + }, + { + "epoch": 2.488688776630793, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9260647296905518, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7219964265823364, + "num_tokens": 586259610.0, + "step": 22662 + }, + { + "epoch": 2.4887985943334066, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.594508409500122, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.694683313369751, + "num_tokens": 586287264.0, + "step": 22663 + }, + { + "epoch": 2.4889084120360203, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.495474338531494, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7203500270843506, + "num_tokens": 586315879.0, + "step": 22664 + }, + { + "epoch": 2.489018229738634, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8698620796203613, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6931830048561096, + "num_tokens": 586336976.0, + "step": 22665 + }, + { + "epoch": 2.4891280474412474, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2766709327697754, + "learning_rate": 1e-06, + "loss": 1.0905, + "mean_token_accuracy": 0.6917614936828613, + "num_tokens": 586369199.0, + "step": 22666 + }, + { + "epoch": 2.489237865143861, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.02280330657959, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7115669846534729, + "num_tokens": 586387568.0, + "step": 22667 + }, + { + "epoch": 2.489347682846475, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.583327531814575, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7012150883674622, + "num_tokens": 586415242.0, + "step": 22668 + }, + { + "epoch": 2.4894575005490887, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596116542816162, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7126336693763733, + "num_tokens": 586441756.0, + "step": 22669 + }, + { + "epoch": 2.4895673182517024, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4364986419677734, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7185858488082886, + "num_tokens": 586468933.0, + "step": 22670 + }, + { + "epoch": 2.4896771359543157, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6757912635803223, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6963748335838318, + "num_tokens": 586495660.0, + "step": 22671 + }, + { + "epoch": 2.4897869536569295, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5602903366088867, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.699612557888031, + "num_tokens": 586522449.0, + "step": 22672 + }, + { + "epoch": 2.4898967713595432, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3811681270599365, + "learning_rate": 1e-06, + "loss": 0.9048, + "mean_token_accuracy": 0.7417176961898804, + "num_tokens": 586549179.0, + "step": 22673 + }, + { + "epoch": 2.490006589062157, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.564918279647827, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7119874954223633, + "num_tokens": 586574623.0, + "step": 22674 + }, + { + "epoch": 2.4901164067647703, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.595221757888794, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.698660135269165, + "num_tokens": 586600188.0, + "step": 22675 + }, + { + "epoch": 2.490226224467384, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.57922101020813, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7075146436691284, + "num_tokens": 586623978.0, + "step": 22676 + }, + { + "epoch": 2.490336042169998, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7693991661071777, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6968660354614258, + "num_tokens": 586647401.0, + "step": 22677 + }, + { + "epoch": 2.4904458598726116, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.945403814315796, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.6974533796310425, + "num_tokens": 586668775.0, + "step": 22678 + }, + { + "epoch": 2.490555677575225, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.609445095062256, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7143617868423462, + "num_tokens": 586694421.0, + "step": 22679 + }, + { + "epoch": 2.4906654952778386, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5334084033966064, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7037622928619385, + "num_tokens": 586721979.0, + "step": 22680 + }, + { + "epoch": 2.4907753129804524, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.74516224861145, + "learning_rate": 1e-06, + "loss": 1.1546, + "mean_token_accuracy": 0.6706698536872864, + "num_tokens": 586752442.0, + "step": 22681 + }, + { + "epoch": 2.490885130683066, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.506021022796631, + "learning_rate": 1e-06, + "loss": 1.0856, + "mean_token_accuracy": 0.6879526972770691, + "num_tokens": 586780628.0, + "step": 22682 + }, + { + "epoch": 2.49099494838568, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2577483654022217, + "learning_rate": 1e-06, + "loss": 0.9291, + "mean_token_accuracy": 0.726812481880188, + "num_tokens": 586809136.0, + "step": 22683 + }, + { + "epoch": 2.4911047660882932, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.417052745819092, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7061915397644043, + "num_tokens": 586838550.0, + "step": 22684 + }, + { + "epoch": 2.491214583790907, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4218294620513916, + "learning_rate": 1e-06, + "loss": 1.1313, + "mean_token_accuracy": 0.6742727160453796, + "num_tokens": 586872032.0, + "step": 22685 + }, + { + "epoch": 2.4913244014935207, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.325265645980835, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.6989165544509888, + "num_tokens": 586906818.0, + "step": 22686 + }, + { + "epoch": 2.4914342191961345, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9252090454101562, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7161202430725098, + "num_tokens": 586928153.0, + "step": 22687 + }, + { + "epoch": 2.4915440368987483, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.581261396408081, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.6982631683349609, + "num_tokens": 586954338.0, + "step": 22688 + }, + { + "epoch": 2.4916538546013616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.618910312652588, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7006803750991821, + "num_tokens": 586981709.0, + "step": 22689 + }, + { + "epoch": 2.4917636723039753, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5620124340057373, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7252643704414368, + "num_tokens": 587009605.0, + "step": 22690 + }, + { + "epoch": 2.491873490006589, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6854751110076904, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.720683217048645, + "num_tokens": 587032841.0, + "step": 22691 + }, + { + "epoch": 2.491983307709203, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6282386779785156, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7077633142471313, + "num_tokens": 587056958.0, + "step": 22692 + }, + { + "epoch": 2.4920931254118166, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5745067596435547, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.7129791378974915, + "num_tokens": 587083782.0, + "step": 22693 + }, + { + "epoch": 2.49220294311443, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.613896131515503, + "learning_rate": 1e-06, + "loss": 1.086, + "mean_token_accuracy": 0.6873092651367188, + "num_tokens": 587110823.0, + "step": 22694 + }, + { + "epoch": 2.4923127608170437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.475857734680176, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6916313767433167, + "num_tokens": 587140248.0, + "step": 22695 + }, + { + "epoch": 2.4924225785196574, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5164551734924316, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.6832987070083618, + "num_tokens": 587169688.0, + "step": 22696 + }, + { + "epoch": 2.492532396222271, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.755173683166504, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.718677818775177, + "num_tokens": 587192285.0, + "step": 22697 + }, + { + "epoch": 2.4926422139248845, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6789591312408447, + "learning_rate": 1e-06, + "loss": 0.9403, + "mean_token_accuracy": 0.7226055860519409, + "num_tokens": 587213982.0, + "step": 22698 + }, + { + "epoch": 2.4927520316274983, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.2911088466644287, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7117760181427002, + "num_tokens": 587241666.0, + "step": 22699 + }, + { + "epoch": 2.492861849330112, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.5650269985198975, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7165881395339966, + "num_tokens": 587265561.0, + "step": 22700 + }, + { + "epoch": 2.4929716670327258, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4617111682891846, + "learning_rate": 1e-06, + "loss": 1.0839, + "mean_token_accuracy": 0.6883525848388672, + "num_tokens": 587295396.0, + "step": 22701 + }, + { + "epoch": 2.4930814847353395, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561917543411255, + "learning_rate": 1e-06, + "loss": 0.9129, + "mean_token_accuracy": 0.7317836284637451, + "num_tokens": 587321031.0, + "step": 22702 + }, + { + "epoch": 2.493191302437953, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.975565195083618, + "learning_rate": 1e-06, + "loss": 0.9465, + "mean_token_accuracy": 0.718921959400177, + "num_tokens": 587339850.0, + "step": 22703 + }, + { + "epoch": 2.4933011201405666, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.442754030227661, + "learning_rate": 1e-06, + "loss": 1.1224, + "mean_token_accuracy": 0.6789408326148987, + "num_tokens": 587369105.0, + "step": 22704 + }, + { + "epoch": 2.4934109378431804, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.574371576309204, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6892444491386414, + "num_tokens": 587397579.0, + "step": 22705 + }, + { + "epoch": 2.493520755545794, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2210607528686523, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.710591197013855, + "num_tokens": 587423378.0, + "step": 22706 + }, + { + "epoch": 2.4936305732484074, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.959812641143799, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7093883752822876, + "num_tokens": 587446342.0, + "step": 22707 + }, + { + "epoch": 2.493740390951021, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4838075637817383, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.700263500213623, + "num_tokens": 587476983.0, + "step": 22708 + }, + { + "epoch": 2.493850208653635, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0395283699035645, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.7093697786331177, + "num_tokens": 587501236.0, + "step": 22709 + }, + { + "epoch": 2.4939600263562487, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.941650867462158, + "learning_rate": 1e-06, + "loss": 0.9957, + "mean_token_accuracy": 0.7082301378250122, + "num_tokens": 587521020.0, + "step": 22710 + }, + { + "epoch": 2.4940698440588625, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.536797523498535, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7029016017913818, + "num_tokens": 587547275.0, + "step": 22711 + }, + { + "epoch": 2.4941796617614758, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6345903873443604, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7201911211013794, + "num_tokens": 587573084.0, + "step": 22712 + }, + { + "epoch": 2.4942894794640895, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4548416137695312, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7004121541976929, + "num_tokens": 587603257.0, + "step": 22713 + }, + { + "epoch": 2.4943992971667033, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.53719162940979, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7020687460899353, + "num_tokens": 587630187.0, + "step": 22714 + }, + { + "epoch": 2.494509114869317, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.914201021194458, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7074668407440186, + "num_tokens": 587649842.0, + "step": 22715 + }, + { + "epoch": 2.494618932571931, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5127623081207275, + "learning_rate": 1e-06, + "loss": 1.0538, + "mean_token_accuracy": 0.7033450603485107, + "num_tokens": 587675550.0, + "step": 22716 + }, + { + "epoch": 2.494728750274544, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9669454097747803, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7135182619094849, + "num_tokens": 587698137.0, + "step": 22717 + }, + { + "epoch": 2.494838567977158, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.784640073776245, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7137181758880615, + "num_tokens": 587726673.0, + "step": 22718 + }, + { + "epoch": 2.4949483856797716, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.2070510387420654, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.7105722427368164, + "num_tokens": 587756002.0, + "step": 22719 + }, + { + "epoch": 2.4950582033823854, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6193933486938477, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6961987018585205, + "num_tokens": 587784302.0, + "step": 22720 + }, + { + "epoch": 2.495168021084999, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4247612953186035, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7108258605003357, + "num_tokens": 587813162.0, + "step": 22721 + }, + { + "epoch": 2.4952778387876124, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2893550395965576, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.721526026725769, + "num_tokens": 587844242.0, + "step": 22722 + }, + { + "epoch": 2.495387656490226, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6297478675842285, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6963647603988647, + "num_tokens": 587872383.0, + "step": 22723 + }, + { + "epoch": 2.49549747419284, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6965229511260986, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7073241472244263, + "num_tokens": 587895424.0, + "step": 22724 + }, + { + "epoch": 2.4956072918954537, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.289522647857666, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7072225213050842, + "num_tokens": 587929377.0, + "step": 22725 + }, + { + "epoch": 2.495717109598067, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7975008487701416, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7159384489059448, + "num_tokens": 587952057.0, + "step": 22726 + }, + { + "epoch": 2.495826927300681, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6959002017974854, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.7329916954040527, + "num_tokens": 587973851.0, + "step": 22727 + }, + { + "epoch": 2.4959367450032945, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.555084466934204, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7123171091079712, + "num_tokens": 588000156.0, + "step": 22728 + }, + { + "epoch": 2.4960465627059083, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5526926517486572, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7098062038421631, + "num_tokens": 588027864.0, + "step": 22729 + }, + { + "epoch": 2.4961563804085216, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.660728693008423, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6850781440734863, + "num_tokens": 588053574.0, + "step": 22730 + }, + { + "epoch": 2.4962661981111354, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2990784645080566, + "learning_rate": 1e-06, + "loss": 0.937, + "mean_token_accuracy": 0.720659613609314, + "num_tokens": 588083717.0, + "step": 22731 + }, + { + "epoch": 2.496376015813749, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8147313594818115, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7090178728103638, + "num_tokens": 588106376.0, + "step": 22732 + }, + { + "epoch": 2.496485833516363, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6470274925231934, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.707751989364624, + "num_tokens": 588134677.0, + "step": 22733 + }, + { + "epoch": 2.4965956512189766, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.397444725036621, + "learning_rate": 1e-06, + "loss": 1.0534, + "mean_token_accuracy": 0.6977647542953491, + "num_tokens": 588163054.0, + "step": 22734 + }, + { + "epoch": 2.49670546892159, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6128122806549072, + "learning_rate": 1e-06, + "loss": 1.0058, + "mean_token_accuracy": 0.7044365406036377, + "num_tokens": 588192624.0, + "step": 22735 + }, + { + "epoch": 2.4968152866242037, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.545517921447754, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7003433704376221, + "num_tokens": 588217520.0, + "step": 22736 + }, + { + "epoch": 2.4969251043268175, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.585556745529175, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.6945397853851318, + "num_tokens": 588246847.0, + "step": 22737 + }, + { + "epoch": 2.4970349220294312, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.76210355758667, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7031643390655518, + "num_tokens": 588271089.0, + "step": 22738 + }, + { + "epoch": 2.497144739732045, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5275087356567383, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7043216228485107, + "num_tokens": 588296509.0, + "step": 22739 + }, + { + "epoch": 2.4972545574346583, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7977588176727295, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7029625177383423, + "num_tokens": 588320644.0, + "step": 22740 + }, + { + "epoch": 2.497364375137272, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.321392774581909, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7136387825012207, + "num_tokens": 588351365.0, + "step": 22741 + }, + { + "epoch": 2.497474192839886, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6182303428649902, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7140377759933472, + "num_tokens": 588374301.0, + "step": 22742 + }, + { + "epoch": 2.4975840105424996, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7854833602905273, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7139341235160828, + "num_tokens": 588394979.0, + "step": 22743 + }, + { + "epoch": 2.4976938282451133, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6879451274871826, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7022572159767151, + "num_tokens": 588418678.0, + "step": 22744 + }, + { + "epoch": 2.4978036459477266, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6469199657440186, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6992911696434021, + "num_tokens": 588442045.0, + "step": 22745 + }, + { + "epoch": 2.4979134636503404, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8114564418792725, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7083353400230408, + "num_tokens": 588464154.0, + "step": 22746 + }, + { + "epoch": 2.498023281352954, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6340532302856445, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7248339056968689, + "num_tokens": 588490177.0, + "step": 22747 + }, + { + "epoch": 2.498133099055568, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.581679582595825, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.6988879442214966, + "num_tokens": 588514830.0, + "step": 22748 + }, + { + "epoch": 2.4982429167581817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.031829357147217, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7219347953796387, + "num_tokens": 588533952.0, + "step": 22749 + }, + { + "epoch": 2.498352734460795, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9346506595611572, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7074881792068481, + "num_tokens": 588553512.0, + "step": 22750 + }, + { + "epoch": 2.4984625521634087, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.050502300262451, + "learning_rate": 1e-06, + "loss": 0.8903, + "mean_token_accuracy": 0.7347228527069092, + "num_tokens": 588570616.0, + "step": 22751 + }, + { + "epoch": 2.4985723698660225, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7026662826538086, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.6992038488388062, + "num_tokens": 588595651.0, + "step": 22752 + }, + { + "epoch": 2.4986821875686362, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7732105255126953, + "learning_rate": 1e-06, + "loss": 0.9219, + "mean_token_accuracy": 0.7247397899627686, + "num_tokens": 588617392.0, + "step": 22753 + }, + { + "epoch": 2.4987920052712496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.336801528930664, + "learning_rate": 1e-06, + "loss": 1.1227, + "mean_token_accuracy": 0.6796547770500183, + "num_tokens": 588649690.0, + "step": 22754 + }, + { + "epoch": 2.4989018229738633, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4547548294067383, + "learning_rate": 1e-06, + "loss": 1.0474, + "mean_token_accuracy": 0.7021225094795227, + "num_tokens": 588680395.0, + "step": 22755 + }, + { + "epoch": 2.499011640676477, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.1799116134643555, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6950457692146301, + "num_tokens": 588715486.0, + "step": 22756 + }, + { + "epoch": 2.499121458379091, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5968523025512695, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6970885396003723, + "num_tokens": 588741555.0, + "step": 22757 + }, + { + "epoch": 2.499231276081704, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.285186529159546, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7098580598831177, + "num_tokens": 588771413.0, + "step": 22758 + }, + { + "epoch": 2.499341093784318, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8904309272766113, + "learning_rate": 1e-06, + "loss": 0.9044, + "mean_token_accuracy": 0.7343879342079163, + "num_tokens": 588791492.0, + "step": 22759 + }, + { + "epoch": 2.4994509114869317, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.044074296951294, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7171297073364258, + "num_tokens": 588811461.0, + "step": 22760 + }, + { + "epoch": 2.4995607291895454, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.964540481567383, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6977779865264893, + "num_tokens": 588840279.0, + "step": 22761 + }, + { + "epoch": 2.499670546892159, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.643195390701294, + "learning_rate": 1e-06, + "loss": 1.0211, + "mean_token_accuracy": 0.7088471055030823, + "num_tokens": 588866667.0, + "step": 22762 + }, + { + "epoch": 2.4997803645947725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.416811466217041, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7028868198394775, + "num_tokens": 588894255.0, + "step": 22763 + }, + { + "epoch": 2.4998901822973862, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5371463298797607, + "learning_rate": 1e-06, + "loss": 1.0217, + "mean_token_accuracy": 0.7031106948852539, + "num_tokens": 588921334.0, + "step": 22764 + }, + { + "epoch": 2.5, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8291211128234863, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7024581432342529, + "num_tokens": 588942704.0, + "step": 22765 + }, + { + "epoch": 2.5001098177026138, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3893866539001465, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.698789656162262, + "num_tokens": 588971823.0, + "step": 22766 + }, + { + "epoch": 2.5002196354052275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.728001117706299, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7218023538589478, + "num_tokens": 588993280.0, + "step": 22767 + }, + { + "epoch": 2.500329453107841, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.881870985031128, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6963862180709839, + "num_tokens": 589016288.0, + "step": 22768 + }, + { + "epoch": 2.5004392708104546, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.728513240814209, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7095811367034912, + "num_tokens": 589042750.0, + "step": 22769 + }, + { + "epoch": 2.5005490885130683, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.643242835998535, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6922802329063416, + "num_tokens": 589068047.0, + "step": 22770 + }, + { + "epoch": 2.500658906215682, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.651463508605957, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.718275785446167, + "num_tokens": 589090217.0, + "step": 22771 + }, + { + "epoch": 2.500768723918296, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1115221977233887, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7049301862716675, + "num_tokens": 589120073.0, + "step": 22772 + }, + { + "epoch": 2.500878541620909, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5896263122558594, + "learning_rate": 1e-06, + "loss": 0.9033, + "mean_token_accuracy": 0.7311810255050659, + "num_tokens": 589144251.0, + "step": 22773 + }, + { + "epoch": 2.500988359323523, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6367032527923584, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7071771621704102, + "num_tokens": 589168790.0, + "step": 22774 + }, + { + "epoch": 2.5010981770261367, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.570638418197632, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7208085060119629, + "num_tokens": 589193331.0, + "step": 22775 + }, + { + "epoch": 2.5012079947287504, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3950977325439453, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7047203183174133, + "num_tokens": 589221114.0, + "step": 22776 + }, + { + "epoch": 2.501317812431364, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5026066303253174, + "learning_rate": 1e-06, + "loss": 0.8797, + "mean_token_accuracy": 0.737135648727417, + "num_tokens": 589246004.0, + "step": 22777 + }, + { + "epoch": 2.5014276301339775, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.081303119659424, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7069534659385681, + "num_tokens": 589265631.0, + "step": 22778 + }, + { + "epoch": 2.5015374478365913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7521867752075195, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7330204248428345, + "num_tokens": 589287728.0, + "step": 22779 + }, + { + "epoch": 2.501647265539205, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4907214641571045, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7068434953689575, + "num_tokens": 589313655.0, + "step": 22780 + }, + { + "epoch": 2.5017570832418183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.697298765182495, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7110500931739807, + "num_tokens": 589337141.0, + "step": 22781 + }, + { + "epoch": 2.501866900944432, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.700220823287964, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7083146572113037, + "num_tokens": 589359967.0, + "step": 22782 + }, + { + "epoch": 2.501976718647046, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.50895357131958, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.7340123057365417, + "num_tokens": 589387337.0, + "step": 22783 + }, + { + "epoch": 2.5020865363496596, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4932327270507812, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7061149477958679, + "num_tokens": 589414927.0, + "step": 22784 + }, + { + "epoch": 2.5021963540522734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.866156816482544, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7153172492980957, + "num_tokens": 589436169.0, + "step": 22785 + }, + { + "epoch": 2.5023061717548867, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559504508972168, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.7042657136917114, + "num_tokens": 589464166.0, + "step": 22786 + }, + { + "epoch": 2.5024159894575004, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.353095293045044, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.6914758682250977, + "num_tokens": 589495499.0, + "step": 22787 + }, + { + "epoch": 2.502525807160114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565051555633545, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6997079849243164, + "num_tokens": 589522298.0, + "step": 22788 + }, + { + "epoch": 2.502635624862728, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.828606605529785, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7193435430526733, + "num_tokens": 589542211.0, + "step": 22789 + }, + { + "epoch": 2.5027454425653417, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8336844444274902, + "learning_rate": 1e-06, + "loss": 1.0493, + "mean_token_accuracy": 0.6903002262115479, + "num_tokens": 589563880.0, + "step": 22790 + }, + { + "epoch": 2.502855260267955, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4955997467041016, + "learning_rate": 1e-06, + "loss": 0.8447, + "mean_token_accuracy": 0.7487470507621765, + "num_tokens": 589590574.0, + "step": 22791 + }, + { + "epoch": 2.5029650779705688, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561091899871826, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7177445888519287, + "num_tokens": 589619617.0, + "step": 22792 + }, + { + "epoch": 2.5030748956731825, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9152591228485107, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.6990805864334106, + "num_tokens": 589640033.0, + "step": 22793 + }, + { + "epoch": 2.5031847133757963, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4665560722351074, + "learning_rate": 1e-06, + "loss": 0.9455, + "mean_token_accuracy": 0.7244545221328735, + "num_tokens": 589665609.0, + "step": 22794 + }, + { + "epoch": 2.50329453107841, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.779496908187866, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7192333340644836, + "num_tokens": 589687734.0, + "step": 22795 + }, + { + "epoch": 2.5034043487810234, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.001906156539917, + "learning_rate": 1e-06, + "loss": 0.9397, + "mean_token_accuracy": 0.7359427213668823, + "num_tokens": 589707580.0, + "step": 22796 + }, + { + "epoch": 2.503514166483637, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3756771087646484, + "learning_rate": 1e-06, + "loss": 0.8918, + "mean_token_accuracy": 0.7398964166641235, + "num_tokens": 589734307.0, + "step": 22797 + }, + { + "epoch": 2.503623984186251, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.691258430480957, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7131839990615845, + "num_tokens": 589757196.0, + "step": 22798 + }, + { + "epoch": 2.5037338018888646, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6478817462921143, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7206624746322632, + "num_tokens": 589780928.0, + "step": 22799 + }, + { + "epoch": 2.5038436195914784, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4927117824554443, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7176249623298645, + "num_tokens": 589810612.0, + "step": 22800 + }, + { + "epoch": 2.5039534372940917, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.918527841567993, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7076032161712646, + "num_tokens": 589831931.0, + "step": 22801 + }, + { + "epoch": 2.5040632549967055, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2215194702148438, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7000249028205872, + "num_tokens": 589861413.0, + "step": 22802 + }, + { + "epoch": 2.504173072699319, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5219056606292725, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.7023075819015503, + "num_tokens": 589891007.0, + "step": 22803 + }, + { + "epoch": 2.5042828904019325, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5705950260162354, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7026404142379761, + "num_tokens": 589918620.0, + "step": 22804 + }, + { + "epoch": 2.5043927081045467, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6232831478118896, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.709047257900238, + "num_tokens": 589946586.0, + "step": 22805 + }, + { + "epoch": 2.50450252580716, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.066917896270752, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7175874710083008, + "num_tokens": 589974246.0, + "step": 22806 + }, + { + "epoch": 2.504612343509774, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3602278232574463, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.7044593095779419, + "num_tokens": 590002770.0, + "step": 22807 + }, + { + "epoch": 2.5047221612123876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5747954845428467, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.7085771560668945, + "num_tokens": 590027131.0, + "step": 22808 + }, + { + "epoch": 2.504831978915001, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9763667583465576, + "learning_rate": 1e-06, + "loss": 0.979, + "mean_token_accuracy": 0.7069516181945801, + "num_tokens": 590046283.0, + "step": 22809 + }, + { + "epoch": 2.5049417966176146, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4071907997131348, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.6864150166511536, + "num_tokens": 590076302.0, + "step": 22810 + }, + { + "epoch": 2.5050516143202284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5316367149353027, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6944313049316406, + "num_tokens": 590102808.0, + "step": 22811 + }, + { + "epoch": 2.505161432022842, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7089169025421143, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7063584327697754, + "num_tokens": 590128825.0, + "step": 22812 + }, + { + "epoch": 2.505271249725456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.501342296600342, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7183919548988342, + "num_tokens": 590153176.0, + "step": 22813 + }, + { + "epoch": 2.505381067428069, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4885120391845703, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7109131217002869, + "num_tokens": 590179251.0, + "step": 22814 + }, + { + "epoch": 2.505490885130683, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9585108757019043, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7159584760665894, + "num_tokens": 590199334.0, + "step": 22815 + }, + { + "epoch": 2.5056007028332967, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4432168006896973, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.7075585126876831, + "num_tokens": 590230521.0, + "step": 22816 + }, + { + "epoch": 2.5057105205359105, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6785457134246826, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6951565742492676, + "num_tokens": 590256396.0, + "step": 22817 + }, + { + "epoch": 2.5058203382385242, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.675309419631958, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7072159647941589, + "num_tokens": 590280641.0, + "step": 22818 + }, + { + "epoch": 2.5059301559411375, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6504344940185547, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.694554328918457, + "num_tokens": 590307933.0, + "step": 22819 + }, + { + "epoch": 2.5060399736437513, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7356832027435303, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.6976266503334045, + "num_tokens": 590328960.0, + "step": 22820 + }, + { + "epoch": 2.506149791346365, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.62625789642334, + "learning_rate": 1e-06, + "loss": 0.9185, + "mean_token_accuracy": 0.7274233102798462, + "num_tokens": 590350891.0, + "step": 22821 + }, + { + "epoch": 2.506259609048979, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.834113121032715, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7227098941802979, + "num_tokens": 590371449.0, + "step": 22822 + }, + { + "epoch": 2.5063694267515926, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7909257411956787, + "learning_rate": 1e-06, + "loss": 0.8674, + "mean_token_accuracy": 0.743171215057373, + "num_tokens": 590393247.0, + "step": 22823 + }, + { + "epoch": 2.506479244454206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.908451795578003, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7054693698883057, + "num_tokens": 590416572.0, + "step": 22824 + }, + { + "epoch": 2.5065890621568196, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.758789539337158, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7229061126708984, + "num_tokens": 590442330.0, + "step": 22825 + }, + { + "epoch": 2.5066988798594334, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.555842161178589, + "learning_rate": 1e-06, + "loss": 1.0808, + "mean_token_accuracy": 0.6888123154640198, + "num_tokens": 590474719.0, + "step": 22826 + }, + { + "epoch": 2.506808697562047, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.547989845275879, + "learning_rate": 1e-06, + "loss": 0.8629, + "mean_token_accuracy": 0.7470418214797974, + "num_tokens": 590502613.0, + "step": 22827 + }, + { + "epoch": 2.506918515264661, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6642003059387207, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7076135873794556, + "num_tokens": 590528490.0, + "step": 22828 + }, + { + "epoch": 2.5070283329672742, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5824131965637207, + "learning_rate": 1e-06, + "loss": 1.1162, + "mean_token_accuracy": 0.6907734274864197, + "num_tokens": 590556084.0, + "step": 22829 + }, + { + "epoch": 2.507138150669888, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.662606954574585, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7058383226394653, + "num_tokens": 590579798.0, + "step": 22830 + }, + { + "epoch": 2.5072479683725017, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8523030281066895, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7089064121246338, + "num_tokens": 590601789.0, + "step": 22831 + }, + { + "epoch": 2.507357786075115, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.064626932144165, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7084434032440186, + "num_tokens": 590626173.0, + "step": 22832 + }, + { + "epoch": 2.507467603777729, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6450910568237305, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.6974161863327026, + "num_tokens": 590653382.0, + "step": 22833 + }, + { + "epoch": 2.5075774214803426, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6299967765808105, + "learning_rate": 1e-06, + "loss": 0.9091, + "mean_token_accuracy": 0.7308320999145508, + "num_tokens": 590681835.0, + "step": 22834 + }, + { + "epoch": 2.5076872391829563, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.733119249343872, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7207391858100891, + "num_tokens": 590703115.0, + "step": 22835 + }, + { + "epoch": 2.50779705688557, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4430553913116455, + "learning_rate": 1e-06, + "loss": 0.9356, + "mean_token_accuracy": 0.7223811149597168, + "num_tokens": 590729143.0, + "step": 22836 + }, + { + "epoch": 2.5079068745881834, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5556116104125977, + "learning_rate": 1e-06, + "loss": 1.0626, + "mean_token_accuracy": 0.6941949129104614, + "num_tokens": 590757784.0, + "step": 22837 + }, + { + "epoch": 2.508016692290797, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.086167335510254, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7192413210868835, + "num_tokens": 590777166.0, + "step": 22838 + }, + { + "epoch": 2.508126509993411, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7092745304107666, + "learning_rate": 1e-06, + "loss": 0.9019, + "mean_token_accuracy": 0.7377767562866211, + "num_tokens": 590801084.0, + "step": 22839 + }, + { + "epoch": 2.5082363276960247, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.847407102584839, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7116779088973999, + "num_tokens": 590823667.0, + "step": 22840 + }, + { + "epoch": 2.5083461453986384, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4602036476135254, + "learning_rate": 1e-06, + "loss": 1.0781, + "mean_token_accuracy": 0.686520516872406, + "num_tokens": 590853423.0, + "step": 22841 + }, + { + "epoch": 2.5084559631012517, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.727928638458252, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.7091086506843567, + "num_tokens": 590876373.0, + "step": 22842 + }, + { + "epoch": 2.5085657808038655, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8050098419189453, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.7067074775695801, + "num_tokens": 590900751.0, + "step": 22843 + }, + { + "epoch": 2.5086755985064793, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.472801446914673, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7000833749771118, + "num_tokens": 590929624.0, + "step": 22844 + }, + { + "epoch": 2.508785416209093, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.476586103439331, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7038263082504272, + "num_tokens": 590957559.0, + "step": 22845 + }, + { + "epoch": 2.5088952339117068, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.971923828125, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7160359025001526, + "num_tokens": 590977189.0, + "step": 22846 + }, + { + "epoch": 2.50900505161432, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4460904598236084, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7161941528320312, + "num_tokens": 591006782.0, + "step": 22847 + }, + { + "epoch": 2.509114869316934, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1715965270996094, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7205665111541748, + "num_tokens": 591032821.0, + "step": 22848 + }, + { + "epoch": 2.5092246870195476, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5455446243286133, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.716431736946106, + "num_tokens": 591063078.0, + "step": 22849 + }, + { + "epoch": 2.5093345047221614, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545114278793335, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.7058805227279663, + "num_tokens": 591093057.0, + "step": 22850 + }, + { + "epoch": 2.509444322424775, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7067387104034424, + "learning_rate": 1e-06, + "loss": 0.9257, + "mean_token_accuracy": 0.7268422842025757, + "num_tokens": 591117707.0, + "step": 22851 + }, + { + "epoch": 2.5095541401273884, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.909860849380493, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7150741815567017, + "num_tokens": 591139261.0, + "step": 22852 + }, + { + "epoch": 2.509663957830002, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7422430515289307, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7086864709854126, + "num_tokens": 591163001.0, + "step": 22853 + }, + { + "epoch": 2.509773775532616, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4360191822052, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7168377637863159, + "num_tokens": 591191088.0, + "step": 22854 + }, + { + "epoch": 2.5098835932352292, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4688098430633545, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6840759515762329, + "num_tokens": 591217394.0, + "step": 22855 + }, + { + "epoch": 2.5099934109378434, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.844327211380005, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7179592251777649, + "num_tokens": 591238614.0, + "step": 22856 + }, + { + "epoch": 2.5101032286404568, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5896012783050537, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7053382396697998, + "num_tokens": 591264604.0, + "step": 22857 + }, + { + "epoch": 2.5102130463430705, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8448069095611572, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7270654439926147, + "num_tokens": 591287228.0, + "step": 22858 + }, + { + "epoch": 2.5103228640456843, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.557509183883667, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.6998946070671082, + "num_tokens": 591312753.0, + "step": 22859 + }, + { + "epoch": 2.5104326817482976, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5849902629852295, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7061541080474854, + "num_tokens": 591338562.0, + "step": 22860 + }, + { + "epoch": 2.5105424994509113, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.498441219329834, + "learning_rate": 1e-06, + "loss": 1.0956, + "mean_token_accuracy": 0.6838338971138, + "num_tokens": 591366615.0, + "step": 22861 + }, + { + "epoch": 2.510652317153525, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7357733249664307, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7041287422180176, + "num_tokens": 591390433.0, + "step": 22862 + }, + { + "epoch": 2.510762134856139, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.674337863922119, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6976777911186218, + "num_tokens": 591412696.0, + "step": 22863 + }, + { + "epoch": 2.5108719525587526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5339949131011963, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7117847800254822, + "num_tokens": 591439559.0, + "step": 22864 + }, + { + "epoch": 2.510981770261366, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6953518390655518, + "learning_rate": 1e-06, + "loss": 0.8676, + "mean_token_accuracy": 0.7390839457511902, + "num_tokens": 591459784.0, + "step": 22865 + }, + { + "epoch": 2.5110915879639797, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.568777084350586, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7123804688453674, + "num_tokens": 591483730.0, + "step": 22866 + }, + { + "epoch": 2.5112014056665934, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6262362003326416, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7143366932868958, + "num_tokens": 591509318.0, + "step": 22867 + }, + { + "epoch": 2.511311223369207, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.679105043411255, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7125657796859741, + "num_tokens": 591532774.0, + "step": 22868 + }, + { + "epoch": 2.511421041071821, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.883500099182129, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.6983094215393066, + "num_tokens": 591553811.0, + "step": 22869 + }, + { + "epoch": 2.5115308587744343, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3529038429260254, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6966850757598877, + "num_tokens": 591587246.0, + "step": 22870 + }, + { + "epoch": 2.511640676477048, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.419313907623291, + "learning_rate": 1e-06, + "loss": 1.082, + "mean_token_accuracy": 0.6827915906906128, + "num_tokens": 591616083.0, + "step": 22871 + }, + { + "epoch": 2.511750494179662, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.478168249130249, + "learning_rate": 1e-06, + "loss": 0.9063, + "mean_token_accuracy": 0.7347182631492615, + "num_tokens": 591642943.0, + "step": 22872 + }, + { + "epoch": 2.5118603118822755, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.608666181564331, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7270528078079224, + "num_tokens": 591668235.0, + "step": 22873 + }, + { + "epoch": 2.5119701295848893, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5264780521392822, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.703736424446106, + "num_tokens": 591696854.0, + "step": 22874 + }, + { + "epoch": 2.5120799472875026, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4853756427764893, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.7019927501678467, + "num_tokens": 591722220.0, + "step": 22875 + }, + { + "epoch": 2.5121897649901164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5533530712127686, + "learning_rate": 1e-06, + "loss": 1.0807, + "mean_token_accuracy": 0.6941894888877869, + "num_tokens": 591749674.0, + "step": 22876 + }, + { + "epoch": 2.51229958269273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4120006561279297, + "learning_rate": 1e-06, + "loss": 0.924, + "mean_token_accuracy": 0.7346069812774658, + "num_tokens": 591777655.0, + "step": 22877 + }, + { + "epoch": 2.512409400395344, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3094289302825928, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7085120677947998, + "num_tokens": 591810654.0, + "step": 22878 + }, + { + "epoch": 2.5125192180979576, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.515699625015259, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7101864218711853, + "num_tokens": 591838576.0, + "step": 22879 + }, + { + "epoch": 2.512629035800571, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.630009412765503, + "learning_rate": 1e-06, + "loss": 0.8881, + "mean_token_accuracy": 0.7370243668556213, + "num_tokens": 591860586.0, + "step": 22880 + }, + { + "epoch": 2.5127388535031847, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7278804779052734, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7166544198989868, + "num_tokens": 591885342.0, + "step": 22881 + }, + { + "epoch": 2.5128486712057985, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7015302181243896, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.719755232334137, + "num_tokens": 591907658.0, + "step": 22882 + }, + { + "epoch": 2.512958488908412, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6850247383117676, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7052065134048462, + "num_tokens": 591933944.0, + "step": 22883 + }, + { + "epoch": 2.5130683066110255, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4696409702301025, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.701385498046875, + "num_tokens": 591965979.0, + "step": 22884 + }, + { + "epoch": 2.5131781243136393, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8896658420562744, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.7272468209266663, + "num_tokens": 591985122.0, + "step": 22885 + }, + { + "epoch": 2.513287942016253, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.772808313369751, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7116525173187256, + "num_tokens": 592009719.0, + "step": 22886 + }, + { + "epoch": 2.513397759718867, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.484739065170288, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.706199049949646, + "num_tokens": 592036278.0, + "step": 22887 + }, + { + "epoch": 2.51350757742148, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.468658924102783, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7038233280181885, + "num_tokens": 592062284.0, + "step": 22888 + }, + { + "epoch": 2.513617395124094, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.1653711795806885, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6967815160751343, + "num_tokens": 592098411.0, + "step": 22889 + }, + { + "epoch": 2.5137272128267076, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.648711681365967, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7179500460624695, + "num_tokens": 592124820.0, + "step": 22890 + }, + { + "epoch": 2.5138370305293214, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8567657470703125, + "learning_rate": 1e-06, + "loss": 0.927, + "mean_token_accuracy": 0.7205874919891357, + "num_tokens": 592143372.0, + "step": 22891 + }, + { + "epoch": 2.513946848231935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4802181720733643, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7040838003158569, + "num_tokens": 592171269.0, + "step": 22892 + }, + { + "epoch": 2.5140566659345485, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6471569538116455, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7049372792243958, + "num_tokens": 592196793.0, + "step": 22893 + }, + { + "epoch": 2.514166483637162, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6336171627044678, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7159421443939209, + "num_tokens": 592221733.0, + "step": 22894 + }, + { + "epoch": 2.514276301339776, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5953469276428223, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6972516179084778, + "num_tokens": 592247370.0, + "step": 22895 + }, + { + "epoch": 2.5143861190423897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4376702308654785, + "learning_rate": 1e-06, + "loss": 1.0917, + "mean_token_accuracy": 0.6836802959442139, + "num_tokens": 592276498.0, + "step": 22896 + }, + { + "epoch": 2.5144959367450035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6052584648132324, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7105973958969116, + "num_tokens": 592300544.0, + "step": 22897 + }, + { + "epoch": 2.514605754447617, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.679447889328003, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6927044987678528, + "num_tokens": 592330526.0, + "step": 22898 + }, + { + "epoch": 2.5147155721502306, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4700675010681152, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7195572853088379, + "num_tokens": 592358300.0, + "step": 22899 + }, + { + "epoch": 2.5148253898528443, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6155848503112793, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7119250893592834, + "num_tokens": 592385102.0, + "step": 22900 + }, + { + "epoch": 2.514935207555458, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4252912998199463, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7169455289840698, + "num_tokens": 592415240.0, + "step": 22901 + }, + { + "epoch": 2.515045025258072, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.97341251373291, + "learning_rate": 1e-06, + "loss": 0.9027, + "mean_token_accuracy": 0.7273805141448975, + "num_tokens": 592432834.0, + "step": 22902 + }, + { + "epoch": 2.515154842960685, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.615103006362915, + "learning_rate": 1e-06, + "loss": 0.9688, + "mean_token_accuracy": 0.726603627204895, + "num_tokens": 592457682.0, + "step": 22903 + }, + { + "epoch": 2.515264660663299, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6555116176605225, + "learning_rate": 1e-06, + "loss": 0.8748, + "mean_token_accuracy": 0.734602689743042, + "num_tokens": 592481379.0, + "step": 22904 + }, + { + "epoch": 2.5153744783659127, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3993489742279053, + "learning_rate": 1e-06, + "loss": 0.9121, + "mean_token_accuracy": 0.7331807613372803, + "num_tokens": 592511498.0, + "step": 22905 + }, + { + "epoch": 2.515484296068526, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8601016998291016, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.7044017314910889, + "num_tokens": 592535201.0, + "step": 22906 + }, + { + "epoch": 2.51559411377114, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2963709831237793, + "learning_rate": 1e-06, + "loss": 1.0086, + "mean_token_accuracy": 0.7070713639259338, + "num_tokens": 592569210.0, + "step": 22907 + }, + { + "epoch": 2.5157039314737535, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.307891368865967, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.6917383670806885, + "num_tokens": 592606097.0, + "step": 22908 + }, + { + "epoch": 2.5158137491763672, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.018204927444458, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7117503881454468, + "num_tokens": 592625761.0, + "step": 22909 + }, + { + "epoch": 2.515923566878981, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6193454265594482, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.731292188167572, + "num_tokens": 592649864.0, + "step": 22910 + }, + { + "epoch": 2.5160333845815943, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.667778730392456, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.704890787601471, + "num_tokens": 592674280.0, + "step": 22911 + }, + { + "epoch": 2.516143202284208, + "ewc_loss": 1.0848045349121094e-05, + "grad_norm": 4.36874532699585, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6923216581344604, + "num_tokens": 592696943.0, + "step": 22912 + }, + { + "epoch": 2.516253019986822, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6965630054473877, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7116053104400635, + "num_tokens": 592722978.0, + "step": 22913 + }, + { + "epoch": 2.5163628376894356, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7621943950653076, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.7271031141281128, + "num_tokens": 592746688.0, + "step": 22914 + }, + { + "epoch": 2.5164726553920493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.538440704345703, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.708258330821991, + "num_tokens": 592776387.0, + "step": 22915 + }, + { + "epoch": 2.5165824730946627, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7165746688842773, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6940629482269287, + "num_tokens": 592800075.0, + "step": 22916 + }, + { + "epoch": 2.5166922907972764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4261415004730225, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7241135835647583, + "num_tokens": 592826613.0, + "step": 22917 + }, + { + "epoch": 2.51680210849989, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.696256160736084, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7112442255020142, + "num_tokens": 592851609.0, + "step": 22918 + }, + { + "epoch": 2.516911926202504, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4347872734069824, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7072641849517822, + "num_tokens": 592879143.0, + "step": 22919 + }, + { + "epoch": 2.5170217439051177, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5928690433502197, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7096941471099854, + "num_tokens": 592905359.0, + "step": 22920 + }, + { + "epoch": 2.517131561607731, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4753332138061523, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6996314525604248, + "num_tokens": 592934634.0, + "step": 22921 + }, + { + "epoch": 2.5172413793103448, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5725901126861572, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.7226194143295288, + "num_tokens": 592961039.0, + "step": 22922 + }, + { + "epoch": 2.5173511970129585, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6014013290405273, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7010303735733032, + "num_tokens": 592988378.0, + "step": 22923 + }, + { + "epoch": 2.5174610147155723, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7660529613494873, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7004532217979431, + "num_tokens": 593017157.0, + "step": 22924 + }, + { + "epoch": 2.517570832418186, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3982512950897217, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.6822347640991211, + "num_tokens": 593052808.0, + "step": 22925 + }, + { + "epoch": 2.5176806501207993, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6639978885650635, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7290741205215454, + "num_tokens": 593077480.0, + "step": 22926 + }, + { + "epoch": 2.517790467823413, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3441648483276367, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7125287055969238, + "num_tokens": 593111996.0, + "step": 22927 + }, + { + "epoch": 2.517900285526027, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4839389324188232, + "learning_rate": 1e-06, + "loss": 1.0932, + "mean_token_accuracy": 0.6871429681777954, + "num_tokens": 593139114.0, + "step": 22928 + }, + { + "epoch": 2.5180101032286406, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.521524667739868, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7060569524765015, + "num_tokens": 593167042.0, + "step": 22929 + }, + { + "epoch": 2.5181199209312544, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.703813552856445, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.700846791267395, + "num_tokens": 593191849.0, + "step": 22930 + }, + { + "epoch": 2.5182297386338677, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6084415912628174, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7123541831970215, + "num_tokens": 593218714.0, + "step": 22931 + }, + { + "epoch": 2.5183395563364814, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7001612186431885, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7009358406066895, + "num_tokens": 593247208.0, + "step": 22932 + }, + { + "epoch": 2.518449374039095, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6699483394622803, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6980913877487183, + "num_tokens": 593272635.0, + "step": 22933 + }, + { + "epoch": 2.5185591917417085, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6622116565704346, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7073891758918762, + "num_tokens": 593297594.0, + "step": 22934 + }, + { + "epoch": 2.5186690094443223, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4636921882629395, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.7005413770675659, + "num_tokens": 593329123.0, + "step": 22935 + }, + { + "epoch": 2.518778827146936, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.526716470718384, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7077100276947021, + "num_tokens": 593358256.0, + "step": 22936 + }, + { + "epoch": 2.5188886448495498, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.818488121032715, + "learning_rate": 1e-06, + "loss": 0.8966, + "mean_token_accuracy": 0.7328130006790161, + "num_tokens": 593379323.0, + "step": 22937 + }, + { + "epoch": 2.5189984625521635, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3981540203094482, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7177711725234985, + "num_tokens": 593412384.0, + "step": 22938 + }, + { + "epoch": 2.519108280254777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.556938648223877, + "learning_rate": 1e-06, + "loss": 1.0705, + "mean_token_accuracy": 0.6965993642807007, + "num_tokens": 593439200.0, + "step": 22939 + }, + { + "epoch": 2.5192180979573906, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4341323375701904, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6881279945373535, + "num_tokens": 593467494.0, + "step": 22940 + }, + { + "epoch": 2.5193279156600044, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.363067626953125, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.712303102016449, + "num_tokens": 593497676.0, + "step": 22941 + }, + { + "epoch": 2.519437733362618, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8304860591888428, + "learning_rate": 1e-06, + "loss": 0.9287, + "mean_token_accuracy": 0.7282944917678833, + "num_tokens": 593519999.0, + "step": 22942 + }, + { + "epoch": 2.519547551065232, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.536963701248169, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.688495397567749, + "num_tokens": 593546497.0, + "step": 22943 + }, + { + "epoch": 2.519657368767845, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4164950847625732, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.712517261505127, + "num_tokens": 593571869.0, + "step": 22944 + }, + { + "epoch": 2.519767186470459, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7052159309387207, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7142823934555054, + "num_tokens": 593595902.0, + "step": 22945 + }, + { + "epoch": 2.5198770041730727, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0054383277893066, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7313275337219238, + "num_tokens": 593616021.0, + "step": 22946 + }, + { + "epoch": 2.5199868218756865, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7793664932250977, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6952674388885498, + "num_tokens": 593640677.0, + "step": 22947 + }, + { + "epoch": 2.5200966395783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.598381519317627, + "learning_rate": 1e-06, + "loss": 1.0912, + "mean_token_accuracy": 0.6905150413513184, + "num_tokens": 593675273.0, + "step": 22948 + }, + { + "epoch": 2.5202064572809135, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7487940788269043, + "learning_rate": 1e-06, + "loss": 1.029, + "mean_token_accuracy": 0.7014720439910889, + "num_tokens": 593698371.0, + "step": 22949 + }, + { + "epoch": 2.5203162749835273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.86981201171875, + "learning_rate": 1e-06, + "loss": 1.0736, + "mean_token_accuracy": 0.6855688095092773, + "num_tokens": 593719487.0, + "step": 22950 + }, + { + "epoch": 2.520426092686141, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.804672956466675, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7115535736083984, + "num_tokens": 593748209.0, + "step": 22951 + }, + { + "epoch": 2.520535910388755, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3849174976348877, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7036222219467163, + "num_tokens": 593779992.0, + "step": 22952 + }, + { + "epoch": 2.5206457280913686, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6080291271209717, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7159708142280579, + "num_tokens": 593805162.0, + "step": 22953 + }, + { + "epoch": 2.520755545793982, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7308509349823, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.7289948463439941, + "num_tokens": 593826809.0, + "step": 22954 + }, + { + "epoch": 2.5208653634965956, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.573918342590332, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7068381309509277, + "num_tokens": 593854502.0, + "step": 22955 + }, + { + "epoch": 2.5209751811992094, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.497178792953491, + "learning_rate": 1e-06, + "loss": 1.0831, + "mean_token_accuracy": 0.6864780783653259, + "num_tokens": 593882364.0, + "step": 22956 + }, + { + "epoch": 2.521084998901823, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0914947986602783, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7024260759353638, + "num_tokens": 593914002.0, + "step": 22957 + }, + { + "epoch": 2.521194816604437, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3928096294403076, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7286389470100403, + "num_tokens": 593942545.0, + "step": 22958 + }, + { + "epoch": 2.52130463430705, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.537492275238037, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7049623727798462, + "num_tokens": 593969663.0, + "step": 22959 + }, + { + "epoch": 2.521414452009664, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.596914768218994, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6965391635894775, + "num_tokens": 593993735.0, + "step": 22960 + }, + { + "epoch": 2.5215242697122777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3309831619262695, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7113302946090698, + "num_tokens": 594022149.0, + "step": 22961 + }, + { + "epoch": 2.521634087414891, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5309677124023438, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7150005102157593, + "num_tokens": 594048254.0, + "step": 22962 + }, + { + "epoch": 2.521743905117505, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9112231731414795, + "learning_rate": 1e-06, + "loss": 0.9448, + "mean_token_accuracy": 0.718586266040802, + "num_tokens": 594069964.0, + "step": 22963 + }, + { + "epoch": 2.5218537228201185, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 6.642421722412109, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7223479747772217, + "num_tokens": 594094639.0, + "step": 22964 + }, + { + "epoch": 2.5219635405227323, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.0264170169830322, + "learning_rate": 1e-06, + "loss": 0.9401, + "mean_token_accuracy": 0.7256782054901123, + "num_tokens": 594115174.0, + "step": 22965 + }, + { + "epoch": 2.522073358225346, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.311053514480591, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7150982022285461, + "num_tokens": 594140095.0, + "step": 22966 + }, + { + "epoch": 2.5221831759279594, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7738096714019775, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.701621413230896, + "num_tokens": 594162787.0, + "step": 22967 + }, + { + "epoch": 2.522292993630573, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.746507406234741, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7069340944290161, + "num_tokens": 594188094.0, + "step": 22968 + }, + { + "epoch": 2.522402811333187, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7327916622161865, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7063581347465515, + "num_tokens": 594213783.0, + "step": 22969 + }, + { + "epoch": 2.5225126290358006, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5517423152923584, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6959706544876099, + "num_tokens": 594240306.0, + "step": 22970 + }, + { + "epoch": 2.5226224467384144, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3559770584106445, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7125710844993591, + "num_tokens": 594269709.0, + "step": 22971 + }, + { + "epoch": 2.5227322644410277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9284138679504395, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7248386144638062, + "num_tokens": 594289564.0, + "step": 22972 + }, + { + "epoch": 2.5228420821436415, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4951443672180176, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7045976519584656, + "num_tokens": 594317962.0, + "step": 22973 + }, + { + "epoch": 2.5229518998462552, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6515214443206787, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7248817682266235, + "num_tokens": 594340334.0, + "step": 22974 + }, + { + "epoch": 2.523061717548869, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7038705348968506, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6937770247459412, + "num_tokens": 594364305.0, + "step": 22975 + }, + { + "epoch": 2.5231715352514827, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8596744537353516, + "learning_rate": 1e-06, + "loss": 0.8997, + "mean_token_accuracy": 0.7348901629447937, + "num_tokens": 594383972.0, + "step": 22976 + }, + { + "epoch": 2.523281352954096, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4396679401397705, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7143124341964722, + "num_tokens": 594413752.0, + "step": 22977 + }, + { + "epoch": 2.52339117065671, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5021419525146484, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6926247477531433, + "num_tokens": 594442079.0, + "step": 22978 + }, + { + "epoch": 2.5235009883593236, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.70833158493042, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6875682473182678, + "num_tokens": 594465901.0, + "step": 22979 + }, + { + "epoch": 2.5236108060619373, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.479158401489258, + "learning_rate": 1e-06, + "loss": 1.0575, + "mean_token_accuracy": 0.6954600811004639, + "num_tokens": 594496278.0, + "step": 22980 + }, + { + "epoch": 2.523720623764551, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.880605459213257, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7149246335029602, + "num_tokens": 594520948.0, + "step": 22981 + }, + { + "epoch": 2.5238304414671644, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5254127979278564, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7183089256286621, + "num_tokens": 594548810.0, + "step": 22982 + }, + { + "epoch": 2.523940259169778, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.887603759765625, + "learning_rate": 1e-06, + "loss": 1.1061, + "mean_token_accuracy": 0.6818534135818481, + "num_tokens": 594574861.0, + "step": 22983 + }, + { + "epoch": 2.524050076872392, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.441774368286133, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6925352215766907, + "num_tokens": 594604307.0, + "step": 22984 + }, + { + "epoch": 2.5241598945750052, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5710856914520264, + "learning_rate": 1e-06, + "loss": 1.0693, + "mean_token_accuracy": 0.6967941522598267, + "num_tokens": 594629850.0, + "step": 22985 + }, + { + "epoch": 2.5242697122776194, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.018167734146118, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7127745747566223, + "num_tokens": 594649519.0, + "step": 22986 + }, + { + "epoch": 2.5243795299802327, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.978559732437134, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.708803653717041, + "num_tokens": 594669664.0, + "step": 22987 + }, + { + "epoch": 2.5244893476828465, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5379343032836914, + "learning_rate": 1e-06, + "loss": 1.0785, + "mean_token_accuracy": 0.6991307735443115, + "num_tokens": 594698318.0, + "step": 22988 + }, + { + "epoch": 2.5245991653854603, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.597489833831787, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.6979683637619019, + "num_tokens": 594724229.0, + "step": 22989 + }, + { + "epoch": 2.5247089830880736, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.627955436706543, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7021892070770264, + "num_tokens": 594750082.0, + "step": 22990 + }, + { + "epoch": 2.5248188007906873, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.892923355102539, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7188169956207275, + "num_tokens": 594769454.0, + "step": 22991 + }, + { + "epoch": 2.524928618493301, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.862551689147949, + "learning_rate": 1e-06, + "loss": 0.9509, + "mean_token_accuracy": 0.7225239276885986, + "num_tokens": 594790967.0, + "step": 22992 + }, + { + "epoch": 2.525038436195915, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7422547340393066, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6955627799034119, + "num_tokens": 594813218.0, + "step": 22993 + }, + { + "epoch": 2.5251482538985286, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8297135829925537, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6986554265022278, + "num_tokens": 594844603.0, + "step": 22994 + }, + { + "epoch": 2.525258071601142, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4488964080810547, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7029761075973511, + "num_tokens": 594871719.0, + "step": 22995 + }, + { + "epoch": 2.5253678893037557, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6220057010650635, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7043347358703613, + "num_tokens": 594895488.0, + "step": 22996 + }, + { + "epoch": 2.5254777070063694, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.6269609928131104, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7151162624359131, + "num_tokens": 594919364.0, + "step": 22997 + }, + { + "epoch": 2.525587524708983, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.509573459625244, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7085616588592529, + "num_tokens": 594946118.0, + "step": 22998 + }, + { + "epoch": 2.525697342411597, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.82417893409729, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6963956356048584, + "num_tokens": 594969450.0, + "step": 22999 + }, + { + "epoch": 2.5258071601142102, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9925718307495117, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.714309811592102, + "num_tokens": 594988795.0, + "step": 23000 + }, + { + "epoch": 2.525916977816824, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6769192218780518, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7226005792617798, + "num_tokens": 595016239.0, + "step": 23001 + }, + { + "epoch": 2.5260267955194378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6380417346954346, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7230427265167236, + "num_tokens": 595041564.0, + "step": 23002 + }, + { + "epoch": 2.5261366132220515, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.595463752746582, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7175293564796448, + "num_tokens": 595064933.0, + "step": 23003 + }, + { + "epoch": 2.5262464309246653, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5073723793029785, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6955606341362, + "num_tokens": 595094820.0, + "step": 23004 + }, + { + "epoch": 2.5263562486272786, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5286107063293457, + "learning_rate": 1e-06, + "loss": 0.921, + "mean_token_accuracy": 0.7320297956466675, + "num_tokens": 595120478.0, + "step": 23005 + }, + { + "epoch": 2.5264660663298923, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0722384452819824, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.7231642603874207, + "num_tokens": 595144689.0, + "step": 23006 + }, + { + "epoch": 2.526575884032506, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4499218463897705, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7094129323959351, + "num_tokens": 595171859.0, + "step": 23007 + }, + { + "epoch": 2.52668570173512, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4952211380004883, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7161461114883423, + "num_tokens": 595196206.0, + "step": 23008 + }, + { + "epoch": 2.5267955194377336, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.673750877380371, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.692427933216095, + "num_tokens": 595219237.0, + "step": 23009 + }, + { + "epoch": 2.526905337140347, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.399383544921875, + "learning_rate": 1e-06, + "loss": 1.1517, + "mean_token_accuracy": 0.6729286909103394, + "num_tokens": 595249726.0, + "step": 23010 + }, + { + "epoch": 2.5270151548429607, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5486068725585938, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7074781060218811, + "num_tokens": 595274027.0, + "step": 23011 + }, + { + "epoch": 2.5271249725455744, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7587177753448486, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.6947699785232544, + "num_tokens": 595296661.0, + "step": 23012 + }, + { + "epoch": 2.5272347902481878, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 5.231664657592773, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7201651930809021, + "num_tokens": 595319437.0, + "step": 23013 + }, + { + "epoch": 2.5273446079508015, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.500206470489502, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6917870044708252, + "num_tokens": 595351096.0, + "step": 23014 + }, + { + "epoch": 2.5274544256534153, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6348342895507812, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.703890323638916, + "num_tokens": 595376735.0, + "step": 23015 + }, + { + "epoch": 2.527564243356029, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5014851093292236, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.7027004361152649, + "num_tokens": 595404045.0, + "step": 23016 + }, + { + "epoch": 2.527674061058643, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6799914836883545, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7017878293991089, + "num_tokens": 595429491.0, + "step": 23017 + }, + { + "epoch": 2.527783878761256, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.548914909362793, + "learning_rate": 1e-06, + "loss": 1.0714, + "mean_token_accuracy": 0.6915771961212158, + "num_tokens": 595457524.0, + "step": 23018 + }, + { + "epoch": 2.52789369646387, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.351144552230835, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7138275504112244, + "num_tokens": 595490527.0, + "step": 23019 + }, + { + "epoch": 2.5280035141664836, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9011242389678955, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.7038275599479675, + "num_tokens": 595515326.0, + "step": 23020 + }, + { + "epoch": 2.5281133318690974, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.211292266845703, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.6999455690383911, + "num_tokens": 595547889.0, + "step": 23021 + }, + { + "epoch": 2.528223149571711, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.498372793197632, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6973943114280701, + "num_tokens": 595576016.0, + "step": 23022 + }, + { + "epoch": 2.5283329672743244, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7319352626800537, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.6991156935691833, + "num_tokens": 595599972.0, + "step": 23023 + }, + { + "epoch": 2.528442784976938, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3801658153533936, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7138832211494446, + "num_tokens": 595629268.0, + "step": 23024 + }, + { + "epoch": 2.528552602679552, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7087297439575195, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.711448073387146, + "num_tokens": 595653835.0, + "step": 23025 + }, + { + "epoch": 2.5286624203821657, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5999155044555664, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7026239633560181, + "num_tokens": 595679525.0, + "step": 23026 + }, + { + "epoch": 2.5287722380847795, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6142077445983887, + "learning_rate": 1e-06, + "loss": 1.05, + "mean_token_accuracy": 0.6920384168624878, + "num_tokens": 595704461.0, + "step": 23027 + }, + { + "epoch": 2.528882055787393, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.774569511413574, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7087059020996094, + "num_tokens": 595726467.0, + "step": 23028 + }, + { + "epoch": 2.5289918734900065, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7327823638916016, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.7180370092391968, + "num_tokens": 595749917.0, + "step": 23029 + }, + { + "epoch": 2.5291016911926203, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6084187030792236, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6941400766372681, + "num_tokens": 595778799.0, + "step": 23030 + }, + { + "epoch": 2.529211508895234, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.919538736343384, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7091645002365112, + "num_tokens": 595798731.0, + "step": 23031 + }, + { + "epoch": 2.529321326597848, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.730682373046875, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.7003280520439148, + "num_tokens": 595824073.0, + "step": 23032 + }, + { + "epoch": 2.529431144300461, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5913102626800537, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.699102520942688, + "num_tokens": 595849564.0, + "step": 23033 + }, + { + "epoch": 2.529540962003075, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5006983280181885, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7056716680526733, + "num_tokens": 595878108.0, + "step": 23034 + }, + { + "epoch": 2.5296507797056886, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7859456539154053, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6953539848327637, + "num_tokens": 595906890.0, + "step": 23035 + }, + { + "epoch": 2.529760597408302, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9257092475891113, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7146109342575073, + "num_tokens": 595926522.0, + "step": 23036 + }, + { + "epoch": 2.529870415110916, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6992805004119873, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.701047420501709, + "num_tokens": 595949645.0, + "step": 23037 + }, + { + "epoch": 2.5299802328135295, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5120978355407715, + "learning_rate": 1e-06, + "loss": 1.0387, + "mean_token_accuracy": 0.7007570862770081, + "num_tokens": 595977614.0, + "step": 23038 + }, + { + "epoch": 2.530090050516143, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.341306209564209, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.698906660079956, + "num_tokens": 596007884.0, + "step": 23039 + }, + { + "epoch": 2.530199868218757, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6016979217529297, + "learning_rate": 1e-06, + "loss": 0.9453, + "mean_token_accuracy": 0.7282296419143677, + "num_tokens": 596031739.0, + "step": 23040 + }, + { + "epoch": 2.5303096859213703, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9929916858673096, + "learning_rate": 1e-06, + "loss": 0.9688, + "mean_token_accuracy": 0.717705249786377, + "num_tokens": 596050106.0, + "step": 23041 + }, + { + "epoch": 2.530419503623984, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9004311561584473, + "learning_rate": 1e-06, + "loss": 0.8435, + "mean_token_accuracy": 0.7473359107971191, + "num_tokens": 596069132.0, + "step": 23042 + }, + { + "epoch": 2.530529321326598, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.388549566268921, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7153081893920898, + "num_tokens": 596097303.0, + "step": 23043 + }, + { + "epoch": 2.5306391390292116, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6990966796875, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7230044007301331, + "num_tokens": 596121083.0, + "step": 23044 + }, + { + "epoch": 2.5307489567318253, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.484433174133301, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6990063190460205, + "num_tokens": 596148564.0, + "step": 23045 + }, + { + "epoch": 2.5308587744344386, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6520462036132812, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7215523719787598, + "num_tokens": 596174298.0, + "step": 23046 + }, + { + "epoch": 2.5309685921370524, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7825562953948975, + "learning_rate": 1e-06, + "loss": 1.0827, + "mean_token_accuracy": 0.6827272176742554, + "num_tokens": 596198003.0, + "step": 23047 + }, + { + "epoch": 2.531078409839666, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8510470390319824, + "learning_rate": 1e-06, + "loss": 0.913, + "mean_token_accuracy": 0.7291867733001709, + "num_tokens": 596220404.0, + "step": 23048 + }, + { + "epoch": 2.53118822754228, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.444279670715332, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6925541162490845, + "num_tokens": 596253105.0, + "step": 23049 + }, + { + "epoch": 2.5312980452448937, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.596768856048584, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7077661752700806, + "num_tokens": 596277084.0, + "step": 23050 + }, + { + "epoch": 2.531407862947507, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6416702270507812, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.6938968896865845, + "num_tokens": 596300341.0, + "step": 23051 + }, + { + "epoch": 2.5315176806501207, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.902454376220703, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7232896089553833, + "num_tokens": 596322781.0, + "step": 23052 + }, + { + "epoch": 2.5316274983527345, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7141175270080566, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.709304928779602, + "num_tokens": 596345388.0, + "step": 23053 + }, + { + "epoch": 2.5317373160553482, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.488696336746216, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7013682126998901, + "num_tokens": 596373827.0, + "step": 23054 + }, + { + "epoch": 2.531847133757962, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5483791828155518, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7225388288497925, + "num_tokens": 596396903.0, + "step": 23055 + }, + { + "epoch": 2.5319569514605753, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5434160232543945, + "learning_rate": 1e-06, + "loss": 1.0897, + "mean_token_accuracy": 0.69439697265625, + "num_tokens": 596424384.0, + "step": 23056 + }, + { + "epoch": 2.532066769163189, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.598679780960083, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7078368663787842, + "num_tokens": 596448517.0, + "step": 23057 + }, + { + "epoch": 2.532176586865803, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9535164833068848, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6875358819961548, + "num_tokens": 596476470.0, + "step": 23058 + }, + { + "epoch": 2.5322864045684166, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.9463155269622803, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7130314111709595, + "num_tokens": 596504634.0, + "step": 23059 + }, + { + "epoch": 2.5323962222710303, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4589972496032715, + "learning_rate": 1e-06, + "loss": 1.1033, + "mean_token_accuracy": 0.6943053007125854, + "num_tokens": 596533937.0, + "step": 23060 + }, + { + "epoch": 2.5325060399736437, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.472276210784912, + "learning_rate": 1e-06, + "loss": 1.1005, + "mean_token_accuracy": 0.6797863245010376, + "num_tokens": 596566767.0, + "step": 23061 + }, + { + "epoch": 2.5326158576762574, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6057608127593994, + "learning_rate": 1e-06, + "loss": 1.0563, + "mean_token_accuracy": 0.6952048540115356, + "num_tokens": 596592004.0, + "step": 23062 + }, + { + "epoch": 2.532725675378871, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.538205623626709, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7133739590644836, + "num_tokens": 596617623.0, + "step": 23063 + }, + { + "epoch": 2.5328354930814845, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4169230461120605, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7128582000732422, + "num_tokens": 596647337.0, + "step": 23064 + }, + { + "epoch": 2.5329453107840982, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.753842830657959, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7095484733581543, + "num_tokens": 596668598.0, + "step": 23065 + }, + { + "epoch": 2.533055128486712, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.395904779434204, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7044339776039124, + "num_tokens": 596699776.0, + "step": 23066 + }, + { + "epoch": 2.5331649461893258, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6237969398498535, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6928714513778687, + "num_tokens": 596725281.0, + "step": 23067 + }, + { + "epoch": 2.5332747638919395, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.267056941986084, + "learning_rate": 1e-06, + "loss": 0.932, + "mean_token_accuracy": 0.7247262001037598, + "num_tokens": 596747043.0, + "step": 23068 + }, + { + "epoch": 2.533384581594553, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9697141647338867, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6809166669845581, + "num_tokens": 596771535.0, + "step": 23069 + }, + { + "epoch": 2.5334943992971666, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6369471549987793, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6952431201934814, + "num_tokens": 596798443.0, + "step": 23070 + }, + { + "epoch": 2.5336042169997803, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9200551509857178, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7229217886924744, + "num_tokens": 596819292.0, + "step": 23071 + }, + { + "epoch": 2.533714034702394, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7317440509796143, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7249220013618469, + "num_tokens": 596848840.0, + "step": 23072 + }, + { + "epoch": 2.533823852405008, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.449861764907837, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7116712331771851, + "num_tokens": 596876828.0, + "step": 23073 + }, + { + "epoch": 2.533933670107621, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6102051734924316, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6976242065429688, + "num_tokens": 596903668.0, + "step": 23074 + }, + { + "epoch": 2.534043487810235, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.344057083129883, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6907778978347778, + "num_tokens": 596939752.0, + "step": 23075 + }, + { + "epoch": 2.5341533055128487, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8082847595214844, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.695071280002594, + "num_tokens": 596965149.0, + "step": 23076 + }, + { + "epoch": 2.5342631232154624, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5988526344299316, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7165698409080505, + "num_tokens": 596988327.0, + "step": 23077 + }, + { + "epoch": 2.534372940918076, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6785402297973633, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7114925980567932, + "num_tokens": 597013645.0, + "step": 23078 + }, + { + "epoch": 2.5344827586206895, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.693138599395752, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7168921232223511, + "num_tokens": 597036964.0, + "step": 23079 + }, + { + "epoch": 2.5345925763233033, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8936867713928223, + "learning_rate": 1e-06, + "loss": 0.9036, + "mean_token_accuracy": 0.7334967851638794, + "num_tokens": 597055558.0, + "step": 23080 + }, + { + "epoch": 2.534702394025917, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6195855140686035, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.696687638759613, + "num_tokens": 597081249.0, + "step": 23081 + }, + { + "epoch": 2.5348122117285308, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.592859983444214, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.718376874923706, + "num_tokens": 597104520.0, + "step": 23082 + }, + { + "epoch": 2.5349220294311445, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4840195178985596, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7131991982460022, + "num_tokens": 597136510.0, + "step": 23083 + }, + { + "epoch": 2.535031847133758, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7818796634674072, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.7018014788627625, + "num_tokens": 597157658.0, + "step": 23084 + }, + { + "epoch": 2.5351416648363716, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3666012287139893, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7082765102386475, + "num_tokens": 597189398.0, + "step": 23085 + }, + { + "epoch": 2.5352514825389854, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.925875663757324, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7262501120567322, + "num_tokens": 597210921.0, + "step": 23086 + }, + { + "epoch": 2.5353613002415987, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.396605968475342, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7233558893203735, + "num_tokens": 597237123.0, + "step": 23087 + }, + { + "epoch": 2.535471117944213, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.652285575866699, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7203539609909058, + "num_tokens": 597260578.0, + "step": 23088 + }, + { + "epoch": 2.535580935646826, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4334781169891357, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7187435626983643, + "num_tokens": 597289215.0, + "step": 23089 + }, + { + "epoch": 2.53569075334944, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.535825729370117, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7045804262161255, + "num_tokens": 597316786.0, + "step": 23090 + }, + { + "epoch": 2.5358005710520537, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.604562759399414, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.6981781125068665, + "num_tokens": 597340930.0, + "step": 23091 + }, + { + "epoch": 2.535910388754667, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8077101707458496, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7038808465003967, + "num_tokens": 597364706.0, + "step": 23092 + }, + { + "epoch": 2.5360202064572808, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.994208574295044, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7024611830711365, + "num_tokens": 597386630.0, + "step": 23093 + }, + { + "epoch": 2.5361300241598945, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6267600059509277, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7118624448776245, + "num_tokens": 597410659.0, + "step": 23094 + }, + { + "epoch": 2.5362398418625083, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5345847606658936, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7074435353279114, + "num_tokens": 597437572.0, + "step": 23095 + }, + { + "epoch": 2.536349659565122, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5754239559173584, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7116312980651855, + "num_tokens": 597460736.0, + "step": 23096 + }, + { + "epoch": 2.5364594772677354, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6784629821777344, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7075499296188354, + "num_tokens": 597485885.0, + "step": 23097 + }, + { + "epoch": 2.536569294970349, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5145933628082275, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7106847763061523, + "num_tokens": 597511911.0, + "step": 23098 + }, + { + "epoch": 2.536679112672963, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.404294490814209, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.713303804397583, + "num_tokens": 597540549.0, + "step": 23099 + }, + { + "epoch": 2.5367889303755766, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8158936500549316, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7055284976959229, + "num_tokens": 597568988.0, + "step": 23100 + }, + { + "epoch": 2.5368987480781904, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5629055500030518, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.6994374990463257, + "num_tokens": 597599355.0, + "step": 23101 + }, + { + "epoch": 2.5370085657808037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5102460384368896, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.710685670375824, + "num_tokens": 597626892.0, + "step": 23102 + }, + { + "epoch": 2.5371183834834174, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.449239492416382, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6998343467712402, + "num_tokens": 597656696.0, + "step": 23103 + }, + { + "epoch": 2.537228201186031, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5531022548675537, + "learning_rate": 1e-06, + "loss": 1.1696, + "mean_token_accuracy": 0.6713374853134155, + "num_tokens": 597686492.0, + "step": 23104 + }, + { + "epoch": 2.537338018888645, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4679648876190186, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6861375570297241, + "num_tokens": 597720272.0, + "step": 23105 + }, + { + "epoch": 2.5374478365912587, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.80643892288208, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7067537903785706, + "num_tokens": 597745245.0, + "step": 23106 + }, + { + "epoch": 2.537557654293872, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.780442237854004, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7072290182113647, + "num_tokens": 597769483.0, + "step": 23107 + }, + { + "epoch": 2.537667471996486, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584096670150757, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.7196310758590698, + "num_tokens": 597794789.0, + "step": 23108 + }, + { + "epoch": 2.5377772896990995, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.515630006790161, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.7084618806838989, + "num_tokens": 597823774.0, + "step": 23109 + }, + { + "epoch": 2.5378871074017133, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.399780750274658, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.701596736907959, + "num_tokens": 597854226.0, + "step": 23110 + }, + { + "epoch": 2.537996925104327, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.77740740776062, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7190369367599487, + "num_tokens": 597879666.0, + "step": 23111 + }, + { + "epoch": 2.5381067428069404, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.696991205215454, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7215136289596558, + "num_tokens": 597903409.0, + "step": 23112 + }, + { + "epoch": 2.538216560509554, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5737626552581787, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7161449790000916, + "num_tokens": 597932638.0, + "step": 23113 + }, + { + "epoch": 2.538326378212168, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6849262714385986, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7222565412521362, + "num_tokens": 597954380.0, + "step": 23114 + }, + { + "epoch": 2.538436195914781, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6313576698303223, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.7037312984466553, + "num_tokens": 597980225.0, + "step": 23115 + }, + { + "epoch": 2.538546013617395, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.809540271759033, + "learning_rate": 1e-06, + "loss": 0.997, + "mean_token_accuracy": 0.7141497135162354, + "num_tokens": 598001818.0, + "step": 23116 + }, + { + "epoch": 2.5386558313200087, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5107791423797607, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7165005803108215, + "num_tokens": 598031330.0, + "step": 23117 + }, + { + "epoch": 2.5387656490226225, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8327560424804688, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7192806005477905, + "num_tokens": 598052679.0, + "step": 23118 + }, + { + "epoch": 2.5388754667252362, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.200648784637451, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7204921245574951, + "num_tokens": 598073395.0, + "step": 23119 + }, + { + "epoch": 2.5389852844278495, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6035826206207275, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7103555202484131, + "num_tokens": 598099590.0, + "step": 23120 + }, + { + "epoch": 2.5390951021304633, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.210611581802368, + "learning_rate": 1e-06, + "loss": 0.8889, + "mean_token_accuracy": 0.7353589534759521, + "num_tokens": 598117357.0, + "step": 23121 + }, + { + "epoch": 2.539204919833077, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.327953577041626, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7277765870094299, + "num_tokens": 598146407.0, + "step": 23122 + }, + { + "epoch": 2.539314737535691, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8727643489837646, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7039207220077515, + "num_tokens": 598168807.0, + "step": 23123 + }, + { + "epoch": 2.5394245552383046, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7022361755371094, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7125372886657715, + "num_tokens": 598193948.0, + "step": 23124 + }, + { + "epoch": 2.539534372940918, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7575302124023438, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7110507488250732, + "num_tokens": 598220707.0, + "step": 23125 + }, + { + "epoch": 2.5396441906435316, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6021742820739746, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.719067394733429, + "num_tokens": 598246641.0, + "step": 23126 + }, + { + "epoch": 2.5397540083461454, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5221304893493652, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7131354212760925, + "num_tokens": 598275816.0, + "step": 23127 + }, + { + "epoch": 2.539863826048759, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.606916666030884, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7081758975982666, + "num_tokens": 598304609.0, + "step": 23128 + }, + { + "epoch": 2.539973643751373, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4396283626556396, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6956320405006409, + "num_tokens": 598332952.0, + "step": 23129 + }, + { + "epoch": 2.5400834614539862, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4417810440063477, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.6911861896514893, + "num_tokens": 598361065.0, + "step": 23130 + }, + { + "epoch": 2.5401932791566, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5500869750976562, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.6861321926116943, + "num_tokens": 598394400.0, + "step": 23131 + }, + { + "epoch": 2.5403030968592137, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5185420513153076, + "learning_rate": 1e-06, + "loss": 1.0057, + "mean_token_accuracy": 0.7140326499938965, + "num_tokens": 598421433.0, + "step": 23132 + }, + { + "epoch": 2.5404129145618275, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9550893306732178, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7090810537338257, + "num_tokens": 598441169.0, + "step": 23133 + }, + { + "epoch": 2.5405227322644413, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7740848064422607, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.7086493968963623, + "num_tokens": 598463886.0, + "step": 23134 + }, + { + "epoch": 2.5406325499670546, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.930711269378662, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7016793489456177, + "num_tokens": 598485379.0, + "step": 23135 + }, + { + "epoch": 2.5407423676696683, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6476571559906006, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7087089419364929, + "num_tokens": 598511676.0, + "step": 23136 + }, + { + "epoch": 2.540852185372282, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.959242343902588, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6935786008834839, + "num_tokens": 598533470.0, + "step": 23137 + }, + { + "epoch": 2.540962003074896, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7287187576293945, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.6972323656082153, + "num_tokens": 598560612.0, + "step": 23138 + }, + { + "epoch": 2.5410718207775096, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7267682552337646, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7200504541397095, + "num_tokens": 598583473.0, + "step": 23139 + }, + { + "epoch": 2.541181638480123, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 4.665312767028809, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7079567909240723, + "num_tokens": 598611489.0, + "step": 23140 + }, + { + "epoch": 2.5412914561827367, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.012141704559326, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7084085941314697, + "num_tokens": 598633506.0, + "step": 23141 + }, + { + "epoch": 2.5414012738853504, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8824005126953125, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.72020423412323, + "num_tokens": 598655695.0, + "step": 23142 + }, + { + "epoch": 2.5415110915879637, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6341350078582764, + "learning_rate": 1e-06, + "loss": 0.9411, + "mean_token_accuracy": 0.7230679392814636, + "num_tokens": 598678323.0, + "step": 23143 + }, + { + "epoch": 2.5416209092905775, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.756114959716797, + "learning_rate": 1e-06, + "loss": 0.936, + "mean_token_accuracy": 0.7189614176750183, + "num_tokens": 598702544.0, + "step": 23144 + }, + { + "epoch": 2.5417307269931912, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6266696453094482, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6902490854263306, + "num_tokens": 598728721.0, + "step": 23145 + }, + { + "epoch": 2.541840544695805, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9250242710113525, + "learning_rate": 1e-06, + "loss": 0.9127, + "mean_token_accuracy": 0.7294172048568726, + "num_tokens": 598748373.0, + "step": 23146 + }, + { + "epoch": 2.5419503623984188, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6513454914093018, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7094401717185974, + "num_tokens": 598772238.0, + "step": 23147 + }, + { + "epoch": 2.542060180101032, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5655007362365723, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.711963415145874, + "num_tokens": 598798213.0, + "step": 23148 + }, + { + "epoch": 2.542169997803646, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5795891284942627, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6859341263771057, + "num_tokens": 598825979.0, + "step": 23149 + }, + { + "epoch": 2.5422798155062596, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.794616222381592, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.7215850949287415, + "num_tokens": 598848689.0, + "step": 23150 + }, + { + "epoch": 2.5423896332088733, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3688364028930664, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7116557955741882, + "num_tokens": 598879821.0, + "step": 23151 + }, + { + "epoch": 2.542499450911487, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4065699577331543, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6927833557128906, + "num_tokens": 598912920.0, + "step": 23152 + }, + { + "epoch": 2.5426092686141004, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4216370582580566, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7137719392776489, + "num_tokens": 598941236.0, + "step": 23153 + }, + { + "epoch": 2.542719086316714, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2717537879943848, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7166764736175537, + "num_tokens": 598974363.0, + "step": 23154 + }, + { + "epoch": 2.542828904019328, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7908971309661865, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6986348628997803, + "num_tokens": 598998809.0, + "step": 23155 + }, + { + "epoch": 2.5429387217219417, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.286301612854004, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7065714001655579, + "num_tokens": 599030383.0, + "step": 23156 + }, + { + "epoch": 2.5430485394245554, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4456770420074463, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7044668793678284, + "num_tokens": 599061889.0, + "step": 23157 + }, + { + "epoch": 2.5431583571271688, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.647094488143921, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.694689154624939, + "num_tokens": 599087343.0, + "step": 23158 + }, + { + "epoch": 2.5432681748297825, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5074398517608643, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6972764730453491, + "num_tokens": 599116440.0, + "step": 23159 + }, + { + "epoch": 2.5433779925323963, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8123435974121094, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7260634303092957, + "num_tokens": 599137004.0, + "step": 23160 + }, + { + "epoch": 2.54348781023501, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 16.81546974182129, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.6978094577789307, + "num_tokens": 599164010.0, + "step": 23161 + }, + { + "epoch": 2.543597627937624, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.720994234085083, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7260230183601379, + "num_tokens": 599186751.0, + "step": 23162 + }, + { + "epoch": 2.543707445640237, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4926280975341797, + "learning_rate": 1e-06, + "loss": 1.0708, + "mean_token_accuracy": 0.6922343969345093, + "num_tokens": 599216747.0, + "step": 23163 + }, + { + "epoch": 2.543817263342851, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6886179447174072, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7112408876419067, + "num_tokens": 599242956.0, + "step": 23164 + }, + { + "epoch": 2.5439270810454646, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7515511512756348, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7108588218688965, + "num_tokens": 599268162.0, + "step": 23165 + }, + { + "epoch": 2.544036898748078, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7097220420837402, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7113074064254761, + "num_tokens": 599292277.0, + "step": 23166 + }, + { + "epoch": 2.544146716450692, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9663803577423096, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7225435376167297, + "num_tokens": 599314496.0, + "step": 23167 + }, + { + "epoch": 2.5442565341533054, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.602879762649536, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6886469721794128, + "num_tokens": 599341428.0, + "step": 23168 + }, + { + "epoch": 2.544366351855919, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.004124879837036, + "learning_rate": 1e-06, + "loss": 0.9089, + "mean_token_accuracy": 0.7376280426979065, + "num_tokens": 599362376.0, + "step": 23169 + }, + { + "epoch": 2.544476169558533, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.835728883743286, + "learning_rate": 1e-06, + "loss": 0.9353, + "mean_token_accuracy": 0.7332716584205627, + "num_tokens": 599382189.0, + "step": 23170 + }, + { + "epoch": 2.5445859872611463, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8180642127990723, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7169625163078308, + "num_tokens": 599404613.0, + "step": 23171 + }, + { + "epoch": 2.54469580496376, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.1042699813842773, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7117570638656616, + "num_tokens": 599422510.0, + "step": 23172 + }, + { + "epoch": 2.544805622666374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5777132511138916, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7023466229438782, + "num_tokens": 599448293.0, + "step": 23173 + }, + { + "epoch": 2.5449154403689875, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.441762685775757, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.689894437789917, + "num_tokens": 599478848.0, + "step": 23174 + }, + { + "epoch": 2.5450252580716013, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6553053855895996, + "learning_rate": 1e-06, + "loss": 0.9364, + "mean_token_accuracy": 0.7249060869216919, + "num_tokens": 599502001.0, + "step": 23175 + }, + { + "epoch": 2.5451350757742146, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5163228511810303, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7075694799423218, + "num_tokens": 599528639.0, + "step": 23176 + }, + { + "epoch": 2.5452448934768284, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7245800495147705, + "learning_rate": 1e-06, + "loss": 1.134, + "mean_token_accuracy": 0.675193190574646, + "num_tokens": 599556815.0, + "step": 23177 + }, + { + "epoch": 2.545354711179442, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7043538093566895, + "learning_rate": 1e-06, + "loss": 0.9555, + "mean_token_accuracy": 0.7266886234283447, + "num_tokens": 599580546.0, + "step": 23178 + }, + { + "epoch": 2.545464528882056, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5594496726989746, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6971864700317383, + "num_tokens": 599610573.0, + "step": 23179 + }, + { + "epoch": 2.5455743465846696, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.812666654586792, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.6944654583930969, + "num_tokens": 599632809.0, + "step": 23180 + }, + { + "epoch": 2.545684164287283, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6407508850097656, + "learning_rate": 1e-06, + "loss": 1.0851, + "mean_token_accuracy": 0.6900507211685181, + "num_tokens": 599660754.0, + "step": 23181 + }, + { + "epoch": 2.5457939819898967, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3777804374694824, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.6994998455047607, + "num_tokens": 599692272.0, + "step": 23182 + }, + { + "epoch": 2.5459037996925105, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4321072101593018, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6951704621315002, + "num_tokens": 599722418.0, + "step": 23183 + }, + { + "epoch": 2.546013617395124, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.468484401702881, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7114112377166748, + "num_tokens": 599750615.0, + "step": 23184 + }, + { + "epoch": 2.546123435097738, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.562685489654541, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7208701372146606, + "num_tokens": 599779424.0, + "step": 23185 + }, + { + "epoch": 2.5462332528003513, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718982458114624, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7181270718574524, + "num_tokens": 599804042.0, + "step": 23186 + }, + { + "epoch": 2.546343070502965, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5591228008270264, + "learning_rate": 1e-06, + "loss": 1.1527, + "mean_token_accuracy": 0.669773519039154, + "num_tokens": 599831383.0, + "step": 23187 + }, + { + "epoch": 2.546452888205579, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5979011058807373, + "learning_rate": 1e-06, + "loss": 0.9638, + "mean_token_accuracy": 0.7218910455703735, + "num_tokens": 599855286.0, + "step": 23188 + }, + { + "epoch": 2.5465627059081926, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7785568237304688, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.71673583984375, + "num_tokens": 599878271.0, + "step": 23189 + }, + { + "epoch": 2.5466725236108063, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5464863777160645, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7093609571456909, + "num_tokens": 599905729.0, + "step": 23190 + }, + { + "epoch": 2.5467823413134196, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.845984935760498, + "learning_rate": 1e-06, + "loss": 0.9369, + "mean_token_accuracy": 0.7323360443115234, + "num_tokens": 599928077.0, + "step": 23191 + }, + { + "epoch": 2.5468921590160334, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.782109260559082, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7255772352218628, + "num_tokens": 599951375.0, + "step": 23192 + }, + { + "epoch": 2.547001976718647, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.625249147415161, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7058742046356201, + "num_tokens": 599976493.0, + "step": 23193 + }, + { + "epoch": 2.5471117944212605, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7227797508239746, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7181424498558044, + "num_tokens": 600000689.0, + "step": 23194 + }, + { + "epoch": 2.547221612123874, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4371864795684814, + "learning_rate": 1e-06, + "loss": 1.1453, + "mean_token_accuracy": 0.6735514402389526, + "num_tokens": 600031697.0, + "step": 23195 + }, + { + "epoch": 2.547331429826488, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.508162260055542, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.7271928787231445, + "num_tokens": 600056997.0, + "step": 23196 + }, + { + "epoch": 2.5474412475291017, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0873234272003174, + "learning_rate": 1e-06, + "loss": 0.9443, + "mean_token_accuracy": 0.7226722240447998, + "num_tokens": 600081186.0, + "step": 23197 + }, + { + "epoch": 2.5475510652317155, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.65775465965271, + "learning_rate": 1e-06, + "loss": 1.0495, + "mean_token_accuracy": 0.6915130019187927, + "num_tokens": 600106506.0, + "step": 23198 + }, + { + "epoch": 2.547660882934329, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4690027236938477, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.690333902835846, + "num_tokens": 600137370.0, + "step": 23199 + }, + { + "epoch": 2.5477707006369426, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5186378955841064, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7174610495567322, + "num_tokens": 600164594.0, + "step": 23200 + }, + { + "epoch": 2.5478805183395563, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.537672281265259, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.70598304271698, + "num_tokens": 600189231.0, + "step": 23201 + }, + { + "epoch": 2.54799033604217, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.612464666366577, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7132168412208557, + "num_tokens": 600214379.0, + "step": 23202 + }, + { + "epoch": 2.548100153744784, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.470395803451538, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.7247704863548279, + "num_tokens": 600241873.0, + "step": 23203 + }, + { + "epoch": 2.548209971447397, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.585308790206909, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7159197330474854, + "num_tokens": 600265780.0, + "step": 23204 + }, + { + "epoch": 2.548319789150011, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.598590135574341, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7128680944442749, + "num_tokens": 600294184.0, + "step": 23205 + }, + { + "epoch": 2.5484296068526247, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5832056999206543, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7242239117622375, + "num_tokens": 600322499.0, + "step": 23206 + }, + { + "epoch": 2.5485394245552384, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.537844657897949, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7028670310974121, + "num_tokens": 600351212.0, + "step": 23207 + }, + { + "epoch": 2.548649242257852, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.360614538192749, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6987552046775818, + "num_tokens": 600376114.0, + "step": 23208 + }, + { + "epoch": 2.5487590599604655, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.968341827392578, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7043773531913757, + "num_tokens": 600397781.0, + "step": 23209 + }, + { + "epoch": 2.5488688776630792, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.503706455230713, + "learning_rate": 1e-06, + "loss": 0.9232, + "mean_token_accuracy": 0.7293266654014587, + "num_tokens": 600425370.0, + "step": 23210 + }, + { + "epoch": 2.548978695365693, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.60461163520813, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6917043924331665, + "num_tokens": 600451785.0, + "step": 23211 + }, + { + "epoch": 2.5490885130683067, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.435154914855957, + "learning_rate": 1e-06, + "loss": 1.003, + "mean_token_accuracy": 0.7100632786750793, + "num_tokens": 600485089.0, + "step": 23212 + }, + { + "epoch": 2.5491983307709205, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.826091766357422, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7229413986206055, + "num_tokens": 600505764.0, + "step": 23213 + }, + { + "epoch": 2.549308148473534, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.652200698852539, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7161498069763184, + "num_tokens": 600529607.0, + "step": 23214 + }, + { + "epoch": 2.5494179661761476, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.05769681930542, + "learning_rate": 1e-06, + "loss": 0.8987, + "mean_token_accuracy": 0.7398719191551208, + "num_tokens": 600549122.0, + "step": 23215 + }, + { + "epoch": 2.5495277838787613, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4514925479888916, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.712500810623169, + "num_tokens": 600578980.0, + "step": 23216 + }, + { + "epoch": 2.5496376015813746, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.705090284347534, + "learning_rate": 1e-06, + "loss": 0.9021, + "mean_token_accuracy": 0.7407959699630737, + "num_tokens": 600602436.0, + "step": 23217 + }, + { + "epoch": 2.549747419283989, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5021841526031494, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.723207950592041, + "num_tokens": 600631297.0, + "step": 23218 + }, + { + "epoch": 2.549857236986602, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7011544704437256, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7065267562866211, + "num_tokens": 600656923.0, + "step": 23219 + }, + { + "epoch": 2.549967054689216, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.439302444458008, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7063188552856445, + "num_tokens": 600683939.0, + "step": 23220 + }, + { + "epoch": 2.5500768723918297, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4191277027130127, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7046427130699158, + "num_tokens": 600712750.0, + "step": 23221 + }, + { + "epoch": 2.550186690094443, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.695040702819824, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6957100629806519, + "num_tokens": 600736385.0, + "step": 23222 + }, + { + "epoch": 2.5502965077970567, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.123222827911377, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7227269411087036, + "num_tokens": 600756943.0, + "step": 23223 + }, + { + "epoch": 2.5504063254996705, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.489652156829834, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7027750611305237, + "num_tokens": 600784891.0, + "step": 23224 + }, + { + "epoch": 2.5505161432022843, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2928130626678467, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7044168710708618, + "num_tokens": 600815435.0, + "step": 23225 + }, + { + "epoch": 2.550625960904898, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6546924114227295, + "learning_rate": 1e-06, + "loss": 1.066, + "mean_token_accuracy": 0.6863728761672974, + "num_tokens": 600842585.0, + "step": 23226 + }, + { + "epoch": 2.5507357786075113, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.451300621032715, + "learning_rate": 1e-06, + "loss": 0.9171, + "mean_token_accuracy": 0.7299401164054871, + "num_tokens": 600870637.0, + "step": 23227 + }, + { + "epoch": 2.550845596310125, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7062337398529053, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7149955034255981, + "num_tokens": 600894183.0, + "step": 23228 + }, + { + "epoch": 2.550955414012739, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6463663578033447, + "learning_rate": 1e-06, + "loss": 0.9833, + "mean_token_accuracy": 0.7140260934829712, + "num_tokens": 600917597.0, + "step": 23229 + }, + { + "epoch": 2.5510652317153526, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3081398010253906, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7046138048171997, + "num_tokens": 600950127.0, + "step": 23230 + }, + { + "epoch": 2.5511750494179664, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.589536190032959, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7077632546424866, + "num_tokens": 600976942.0, + "step": 23231 + }, + { + "epoch": 2.5512848671205797, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.653703451156616, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7091699242591858, + "num_tokens": 601000981.0, + "step": 23232 + }, + { + "epoch": 2.5513946848231934, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6158370971679688, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7216336131095886, + "num_tokens": 601027115.0, + "step": 23233 + }, + { + "epoch": 2.551504502525807, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.790498733520508, + "learning_rate": 1e-06, + "loss": 0.8844, + "mean_token_accuracy": 0.7383174300193787, + "num_tokens": 601050975.0, + "step": 23234 + }, + { + "epoch": 2.551614320228421, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.503960609436035, + "learning_rate": 1e-06, + "loss": 0.957, + "mean_token_accuracy": 0.7188900113105774, + "num_tokens": 601079541.0, + "step": 23235 + }, + { + "epoch": 2.5517241379310347, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4966442584991455, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6990635991096497, + "num_tokens": 601109464.0, + "step": 23236 + }, + { + "epoch": 2.551833955633648, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4083261489868164, + "learning_rate": 1e-06, + "loss": 0.9297, + "mean_token_accuracy": 0.7273133397102356, + "num_tokens": 601137656.0, + "step": 23237 + }, + { + "epoch": 2.5519437733362618, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7393462657928467, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7133141756057739, + "num_tokens": 601160857.0, + "step": 23238 + }, + { + "epoch": 2.5520535910388755, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.493469476699829, + "learning_rate": 1e-06, + "loss": 0.9506, + "mean_token_accuracy": 0.7225293517112732, + "num_tokens": 601187777.0, + "step": 23239 + }, + { + "epoch": 2.5521634087414893, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.526916265487671, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.7259492874145508, + "num_tokens": 601211846.0, + "step": 23240 + }, + { + "epoch": 2.552273226444103, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.535834789276123, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7005941867828369, + "num_tokens": 601236545.0, + "step": 23241 + }, + { + "epoch": 2.5523830441467164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.417724609375, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7121862769126892, + "num_tokens": 601264423.0, + "step": 23242 + }, + { + "epoch": 2.55249286184933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5029358863830566, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.702560544013977, + "num_tokens": 601290790.0, + "step": 23243 + }, + { + "epoch": 2.552602679551944, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8495032787323, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7180924415588379, + "num_tokens": 601311762.0, + "step": 23244 + }, + { + "epoch": 2.552712497254557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.659902334213257, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7100296020507812, + "num_tokens": 601338352.0, + "step": 23245 + }, + { + "epoch": 2.552822314957171, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.269211769104004, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.7306990623474121, + "num_tokens": 601360971.0, + "step": 23246 + }, + { + "epoch": 2.5529321326597847, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.907433271408081, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.7158198356628418, + "num_tokens": 601382171.0, + "step": 23247 + }, + { + "epoch": 2.5530419503623984, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7808942794799805, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7150702476501465, + "num_tokens": 601405621.0, + "step": 23248 + }, + { + "epoch": 2.553151768065012, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 20.539875030517578, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7043800950050354, + "num_tokens": 601433299.0, + "step": 23249 + }, + { + "epoch": 2.5532615857676255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.56052303314209, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7136244177818298, + "num_tokens": 601461285.0, + "step": 23250 + }, + { + "epoch": 2.5533714034702393, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6825361251831055, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6884473562240601, + "num_tokens": 601491882.0, + "step": 23251 + }, + { + "epoch": 2.553481221172853, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2511496543884277, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7097781896591187, + "num_tokens": 601513290.0, + "step": 23252 + }, + { + "epoch": 2.553591038875467, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6168525218963623, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6838020086288452, + "num_tokens": 601538138.0, + "step": 23253 + }, + { + "epoch": 2.5537008565780805, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4205782413482666, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.688838541507721, + "num_tokens": 601568430.0, + "step": 23254 + }, + { + "epoch": 2.553810674280694, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5494322776794434, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.7107349634170532, + "num_tokens": 601594573.0, + "step": 23255 + }, + { + "epoch": 2.5539204919833076, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4854416847229004, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7222180366516113, + "num_tokens": 601622798.0, + "step": 23256 + }, + { + "epoch": 2.5540303096859214, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7549617290496826, + "learning_rate": 1e-06, + "loss": 1.0122, + "mean_token_accuracy": 0.7046955227851868, + "num_tokens": 601645628.0, + "step": 23257 + }, + { + "epoch": 2.554140127388535, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.016160726547241, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.7012736797332764, + "num_tokens": 601666359.0, + "step": 23258 + }, + { + "epoch": 2.554249945091149, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.732952833175659, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7039729356765747, + "num_tokens": 601687415.0, + "step": 23259 + }, + { + "epoch": 2.554359762793762, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6204447746276855, + "learning_rate": 1e-06, + "loss": 1.097, + "mean_token_accuracy": 0.6859196424484253, + "num_tokens": 601712944.0, + "step": 23260 + }, + { + "epoch": 2.554469580496376, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7015748023986816, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7031145691871643, + "num_tokens": 601740554.0, + "step": 23261 + }, + { + "epoch": 2.5545793981989897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7407302856445312, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7063881158828735, + "num_tokens": 601762863.0, + "step": 23262 + }, + { + "epoch": 2.5546892159016035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.511821746826172, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7037994861602783, + "num_tokens": 601790171.0, + "step": 23263 + }, + { + "epoch": 2.5547990336042172, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5029678344726562, + "learning_rate": 1e-06, + "loss": 0.9332, + "mean_token_accuracy": 0.7306964993476868, + "num_tokens": 601817728.0, + "step": 23264 + }, + { + "epoch": 2.5549088513068305, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.744098424911499, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7129815816879272, + "num_tokens": 601840399.0, + "step": 23265 + }, + { + "epoch": 2.5550186690094443, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.567436695098877, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7252436876296997, + "num_tokens": 601864120.0, + "step": 23266 + }, + { + "epoch": 2.555128486712058, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.782201051712036, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7067353129386902, + "num_tokens": 601887840.0, + "step": 23267 + }, + { + "epoch": 2.5552383044146714, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7685744762420654, + "learning_rate": 1e-06, + "loss": 1.0772, + "mean_token_accuracy": 0.6929572224617004, + "num_tokens": 601911940.0, + "step": 23268 + }, + { + "epoch": 2.5553481221172856, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9553847312927246, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7022694945335388, + "num_tokens": 601932390.0, + "step": 23269 + }, + { + "epoch": 2.555457939819899, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7160086631774902, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7064661979675293, + "num_tokens": 601961872.0, + "step": 23270 + }, + { + "epoch": 2.5555677575225126, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5229010581970215, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.6978317499160767, + "num_tokens": 601990854.0, + "step": 23271 + }, + { + "epoch": 2.5556775752251264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.986443519592285, + "learning_rate": 1e-06, + "loss": 0.9697, + "mean_token_accuracy": 0.7141974568367004, + "num_tokens": 602010988.0, + "step": 23272 + }, + { + "epoch": 2.5557873929277397, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5634777545928955, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7200213670730591, + "num_tokens": 602035175.0, + "step": 23273 + }, + { + "epoch": 2.5558972106303535, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8664259910583496, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.6998772025108337, + "num_tokens": 602057633.0, + "step": 23274 + }, + { + "epoch": 2.5560070283329672, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.697519540786743, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.7179535031318665, + "num_tokens": 602082040.0, + "step": 23275 + }, + { + "epoch": 2.556116846035581, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.759615182876587, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7048735618591309, + "num_tokens": 602104266.0, + "step": 23276 + }, + { + "epoch": 2.5562266637381947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4425902366638184, + "learning_rate": 1e-06, + "loss": 1.0614, + "mean_token_accuracy": 0.6917558908462524, + "num_tokens": 602131165.0, + "step": 23277 + }, + { + "epoch": 2.556336481440808, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.553618907928467, + "learning_rate": 1e-06, + "loss": 0.8766, + "mean_token_accuracy": 0.7385338544845581, + "num_tokens": 602155657.0, + "step": 23278 + }, + { + "epoch": 2.556446299143422, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5067126750946045, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7107304930686951, + "num_tokens": 602181965.0, + "step": 23279 + }, + { + "epoch": 2.5565561168460356, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6122970581054688, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6977576017379761, + "num_tokens": 602211976.0, + "step": 23280 + }, + { + "epoch": 2.5566659345486493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.49489426612854, + "learning_rate": 1e-06, + "loss": 0.9296, + "mean_token_accuracy": 0.7336945533752441, + "num_tokens": 602239196.0, + "step": 23281 + }, + { + "epoch": 2.556775752251263, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3766324520111084, + "learning_rate": 1e-06, + "loss": 0.9708, + "mean_token_accuracy": 0.7145225405693054, + "num_tokens": 602267444.0, + "step": 23282 + }, + { + "epoch": 2.5568855699538764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.386587619781494, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7056005001068115, + "num_tokens": 602298803.0, + "step": 23283 + }, + { + "epoch": 2.55699538765649, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.439370632171631, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7116786241531372, + "num_tokens": 602328423.0, + "step": 23284 + }, + { + "epoch": 2.557105205359104, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8020806312561035, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6982304453849792, + "num_tokens": 602350943.0, + "step": 23285 + }, + { + "epoch": 2.5572150230617177, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4960105419158936, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7137190103530884, + "num_tokens": 602378886.0, + "step": 23286 + }, + { + "epoch": 2.5573248407643314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.715169668197632, + "learning_rate": 1e-06, + "loss": 0.9059, + "mean_token_accuracy": 0.734430193901062, + "num_tokens": 602400804.0, + "step": 23287 + }, + { + "epoch": 2.5574346584669447, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7951407432556152, + "learning_rate": 1e-06, + "loss": 1.0367, + "mean_token_accuracy": 0.7007322311401367, + "num_tokens": 602426240.0, + "step": 23288 + }, + { + "epoch": 2.5575444761695585, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9216713905334473, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.709978461265564, + "num_tokens": 602447651.0, + "step": 23289 + }, + { + "epoch": 2.5576542938721722, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.041954517364502, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7038004994392395, + "num_tokens": 602469324.0, + "step": 23290 + }, + { + "epoch": 2.557764111574786, + "ewc_loss": 1.0967254638671875e-05, + "grad_norm": 80.5480728149414, + "learning_rate": 1e-06, + "loss": 0.9221, + "mean_token_accuracy": 0.7279317378997803, + "num_tokens": 602493436.0, + "step": 23291 + }, + { + "epoch": 2.5578739292773998, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.781599760055542, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.698609471321106, + "num_tokens": 602524451.0, + "step": 23292 + }, + { + "epoch": 2.557983746980013, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6422548294067383, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7163524031639099, + "num_tokens": 602552139.0, + "step": 23293 + }, + { + "epoch": 2.558093564682627, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.320552110671997, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7125820517539978, + "num_tokens": 602584609.0, + "step": 23294 + }, + { + "epoch": 2.5582033823852406, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3612704277038574, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6880548000335693, + "num_tokens": 602613567.0, + "step": 23295 + }, + { + "epoch": 2.558313200087854, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1363754272460938, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7208647727966309, + "num_tokens": 602630565.0, + "step": 23296 + }, + { + "epoch": 2.5584230177904677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5873634815216064, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.6960114240646362, + "num_tokens": 602655516.0, + "step": 23297 + }, + { + "epoch": 2.5585328354930814, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.465574026107788, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7137166261672974, + "num_tokens": 602681620.0, + "step": 23298 + }, + { + "epoch": 2.558642653195695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6022756099700928, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7246720194816589, + "num_tokens": 602707308.0, + "step": 23299 + }, + { + "epoch": 2.558752470898309, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5137906074523926, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7124347686767578, + "num_tokens": 602731336.0, + "step": 23300 + }, + { + "epoch": 2.5588622886009222, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8929171562194824, + "learning_rate": 1e-06, + "loss": 0.9119, + "mean_token_accuracy": 0.7315762639045715, + "num_tokens": 602751478.0, + "step": 23301 + }, + { + "epoch": 2.558972106303536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6294174194335938, + "learning_rate": 1e-06, + "loss": 1.0842, + "mean_token_accuracy": 0.6903186440467834, + "num_tokens": 602776797.0, + "step": 23302 + }, + { + "epoch": 2.5590819240061498, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7163689136505127, + "learning_rate": 1e-06, + "loss": 0.9457, + "mean_token_accuracy": 0.7216243743896484, + "num_tokens": 602799985.0, + "step": 23303 + }, + { + "epoch": 2.5591917417087635, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6888816356658936, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7054952383041382, + "num_tokens": 602825496.0, + "step": 23304 + }, + { + "epoch": 2.5593015594113773, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.762665033340454, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.6983693838119507, + "num_tokens": 602849205.0, + "step": 23305 + }, + { + "epoch": 2.5594113771139906, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.477109432220459, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6917757987976074, + "num_tokens": 602879621.0, + "step": 23306 + }, + { + "epoch": 2.5595211948166043, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.514493465423584, + "learning_rate": 1e-06, + "loss": 0.9669, + "mean_token_accuracy": 0.7190425992012024, + "num_tokens": 602908426.0, + "step": 23307 + }, + { + "epoch": 2.559631012519218, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5219619274139404, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7184218168258667, + "num_tokens": 602934702.0, + "step": 23308 + }, + { + "epoch": 2.559740830221832, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7461118698120117, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7075737714767456, + "num_tokens": 602956640.0, + "step": 23309 + }, + { + "epoch": 2.5598506479244456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.618601083755493, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7228776216506958, + "num_tokens": 602980767.0, + "step": 23310 + }, + { + "epoch": 2.559960465627059, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6800408363342285, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7108281850814819, + "num_tokens": 603004907.0, + "step": 23311 + }, + { + "epoch": 2.5600702833296727, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.474961757659912, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.716566801071167, + "num_tokens": 603031709.0, + "step": 23312 + }, + { + "epoch": 2.5601801010322864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.55472469329834, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6929287910461426, + "num_tokens": 603061858.0, + "step": 23313 + }, + { + "epoch": 2.5602899187349, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.765429735183716, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7160255312919617, + "num_tokens": 603085881.0, + "step": 23314 + }, + { + "epoch": 2.560399736437514, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0339417457580566, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7145867943763733, + "num_tokens": 603103768.0, + "step": 23315 + }, + { + "epoch": 2.5605095541401273, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5191166400909424, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.7124157547950745, + "num_tokens": 603132650.0, + "step": 23316 + }, + { + "epoch": 2.560619371842741, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7118029594421387, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7168701887130737, + "num_tokens": 603157782.0, + "step": 23317 + }, + { + "epoch": 2.560729189545355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9391160011291504, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7006796598434448, + "num_tokens": 603177854.0, + "step": 23318 + }, + { + "epoch": 2.5608390072479685, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.863898992538452, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7237169146537781, + "num_tokens": 603199065.0, + "step": 23319 + }, + { + "epoch": 2.5609488249505823, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.979214668273926, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7107788324356079, + "num_tokens": 603225984.0, + "step": 23320 + }, + { + "epoch": 2.5610586426531956, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6895973682403564, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7128145694732666, + "num_tokens": 603253866.0, + "step": 23321 + }, + { + "epoch": 2.5611684603558094, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7060582637786865, + "learning_rate": 1e-06, + "loss": 0.95, + "mean_token_accuracy": 0.7227790355682373, + "num_tokens": 603276823.0, + "step": 23322 + }, + { + "epoch": 2.561278278058423, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.748175859451294, + "learning_rate": 1e-06, + "loss": 0.936, + "mean_token_accuracy": 0.7292359471321106, + "num_tokens": 603298508.0, + "step": 23323 + }, + { + "epoch": 2.5613880957610364, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5443520545959473, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.6997910737991333, + "num_tokens": 603329742.0, + "step": 23324 + }, + { + "epoch": 2.56149791346365, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.506655693054199, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.68635094165802, + "num_tokens": 603358198.0, + "step": 23325 + }, + { + "epoch": 2.561607731166264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6147990226745605, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7158998847007751, + "num_tokens": 603383642.0, + "step": 23326 + }, + { + "epoch": 2.5617175488688777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.645512819290161, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7244990468025208, + "num_tokens": 603409275.0, + "step": 23327 + }, + { + "epoch": 2.5618273665714915, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3511803150177, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.691192626953125, + "num_tokens": 603440126.0, + "step": 23328 + }, + { + "epoch": 2.5619371842741048, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.900402784347534, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7134889364242554, + "num_tokens": 603459754.0, + "step": 23329 + }, + { + "epoch": 2.5620470019767185, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4428138732910156, + "learning_rate": 1e-06, + "loss": 1.0531, + "mean_token_accuracy": 0.6922556161880493, + "num_tokens": 603491305.0, + "step": 23330 + }, + { + "epoch": 2.5621568196793323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.5628416538238525, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7200237512588501, + "num_tokens": 603516967.0, + "step": 23331 + }, + { + "epoch": 2.562266637381946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5740814208984375, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.71503746509552, + "num_tokens": 603543598.0, + "step": 23332 + }, + { + "epoch": 2.56237645508456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.541740655899048, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7087223529815674, + "num_tokens": 603572869.0, + "step": 23333 + }, + { + "epoch": 2.562486272787173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.878657579421997, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.699762761592865, + "num_tokens": 603595639.0, + "step": 23334 + }, + { + "epoch": 2.562596090489787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9396138191223145, + "learning_rate": 1e-06, + "loss": 1.0816, + "mean_token_accuracy": 0.6882745027542114, + "num_tokens": 603619042.0, + "step": 23335 + }, + { + "epoch": 2.5627059081924006, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3898215293884277, + "learning_rate": 1e-06, + "loss": 1.0666, + "mean_token_accuracy": 0.6943380832672119, + "num_tokens": 603654904.0, + "step": 23336 + }, + { + "epoch": 2.5628157258950144, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7653284072875977, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.7057176828384399, + "num_tokens": 603679865.0, + "step": 23337 + }, + { + "epoch": 2.562925543597628, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6231021881103516, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7081465125083923, + "num_tokens": 603705678.0, + "step": 23338 + }, + { + "epoch": 2.5630353613002415, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8408284187316895, + "learning_rate": 1e-06, + "loss": 1.1168, + "mean_token_accuracy": 0.6876528263092041, + "num_tokens": 603729771.0, + "step": 23339 + }, + { + "epoch": 2.563145179002855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.592092275619507, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.6988434791564941, + "num_tokens": 603755805.0, + "step": 23340 + }, + { + "epoch": 2.563254996705469, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5628654956817627, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6945077776908875, + "num_tokens": 603784069.0, + "step": 23341 + }, + { + "epoch": 2.5633648144080827, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5110108852386475, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7176864147186279, + "num_tokens": 603809966.0, + "step": 23342 + }, + { + "epoch": 2.5634746321106965, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7970449924468994, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.69118332862854, + "num_tokens": 603833997.0, + "step": 23343 + }, + { + "epoch": 2.56358444981331, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.686668872833252, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.6913585066795349, + "num_tokens": 603862599.0, + "step": 23344 + }, + { + "epoch": 2.5636942675159236, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7040929794311523, + "learning_rate": 1e-06, + "loss": 1.1082, + "mean_token_accuracy": 0.6893339157104492, + "num_tokens": 603889002.0, + "step": 23345 + }, + { + "epoch": 2.5638040852185373, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5937411785125732, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6968616247177124, + "num_tokens": 603915766.0, + "step": 23346 + }, + { + "epoch": 2.5639139029211506, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2978994846343994, + "learning_rate": 1e-06, + "loss": 1.0861, + "mean_token_accuracy": 0.6843253970146179, + "num_tokens": 603949935.0, + "step": 23347 + }, + { + "epoch": 2.564023720623765, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.545842409133911, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7281320095062256, + "num_tokens": 603976908.0, + "step": 23348 + }, + { + "epoch": 2.564133538326378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584867477416992, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.702286958694458, + "num_tokens": 604002638.0, + "step": 23349 + }, + { + "epoch": 2.564243356028992, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5021004676818848, + "learning_rate": 1e-06, + "loss": 1.0727, + "mean_token_accuracy": 0.6911864280700684, + "num_tokens": 604032425.0, + "step": 23350 + }, + { + "epoch": 2.5643531737316057, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3758397102355957, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7071642279624939, + "num_tokens": 604062291.0, + "step": 23351 + }, + { + "epoch": 2.564462991434219, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7469866275787354, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7046234607696533, + "num_tokens": 604085436.0, + "step": 23352 + }, + { + "epoch": 2.5645728091368327, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8030786514282227, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6971081495285034, + "num_tokens": 604108637.0, + "step": 23353 + }, + { + "epoch": 2.5646826268394465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6446921825408936, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6996833682060242, + "num_tokens": 604133226.0, + "step": 23354 + }, + { + "epoch": 2.5647924445420602, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6057066917419434, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7042620778083801, + "num_tokens": 604159178.0, + "step": 23355 + }, + { + "epoch": 2.564902262244674, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7274773120880127, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7149434685707092, + "num_tokens": 604181713.0, + "step": 23356 + }, + { + "epoch": 2.5650120799472873, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4095475673675537, + "learning_rate": 1e-06, + "loss": 1.0542, + "mean_token_accuracy": 0.7033237218856812, + "num_tokens": 604211696.0, + "step": 23357 + }, + { + "epoch": 2.565121897649901, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4769043922424316, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7186989784240723, + "num_tokens": 604238132.0, + "step": 23358 + }, + { + "epoch": 2.565231715352515, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.527937412261963, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7098429799079895, + "num_tokens": 604267759.0, + "step": 23359 + }, + { + "epoch": 2.5653415330551286, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4733834266662598, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7038986682891846, + "num_tokens": 604295286.0, + "step": 23360 + }, + { + "epoch": 2.5654513507577423, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6558313369750977, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7139633893966675, + "num_tokens": 604318751.0, + "step": 23361 + }, + { + "epoch": 2.5655611684603556, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5341336727142334, + "learning_rate": 1e-06, + "loss": 0.8968, + "mean_token_accuracy": 0.7342320680618286, + "num_tokens": 604343713.0, + "step": 23362 + }, + { + "epoch": 2.5656709861629694, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.395869255065918, + "learning_rate": 1e-06, + "loss": 1.0665, + "mean_token_accuracy": 0.6955603957176208, + "num_tokens": 604376454.0, + "step": 23363 + }, + { + "epoch": 2.565780803865583, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9492859840393066, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7224684953689575, + "num_tokens": 604397121.0, + "step": 23364 + }, + { + "epoch": 2.565890621568197, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8046486377716064, + "learning_rate": 1e-06, + "loss": 0.8823, + "mean_token_accuracy": 0.740007221698761, + "num_tokens": 604417671.0, + "step": 23365 + }, + { + "epoch": 2.5660004392708107, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.631340503692627, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.698884904384613, + "num_tokens": 604441561.0, + "step": 23366 + }, + { + "epoch": 2.566110256973424, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.539677381515503, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7037794589996338, + "num_tokens": 604467657.0, + "step": 23367 + }, + { + "epoch": 2.5662200746760377, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.461906671524048, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6954097747802734, + "num_tokens": 604497964.0, + "step": 23368 + }, + { + "epoch": 2.5663298923786515, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6835756301879883, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7031030654907227, + "num_tokens": 604523342.0, + "step": 23369 + }, + { + "epoch": 2.5664397100812653, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.881326198577881, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6993407011032104, + "num_tokens": 604548761.0, + "step": 23370 + }, + { + "epoch": 2.566549527783879, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6968131065368652, + "learning_rate": 1e-06, + "loss": 1.0643, + "mean_token_accuracy": 0.692326545715332, + "num_tokens": 604574531.0, + "step": 23371 + }, + { + "epoch": 2.5666593454864923, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.542210817337036, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.692253053188324, + "num_tokens": 604605288.0, + "step": 23372 + }, + { + "epoch": 2.566769163189106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9103314876556396, + "learning_rate": 1e-06, + "loss": 0.9254, + "mean_token_accuracy": 0.7334331274032593, + "num_tokens": 604624112.0, + "step": 23373 + }, + { + "epoch": 2.56687898089172, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6241414546966553, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7175682783126831, + "num_tokens": 604647904.0, + "step": 23374 + }, + { + "epoch": 2.566988798594333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6228060722351074, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7061209678649902, + "num_tokens": 604674123.0, + "step": 23375 + }, + { + "epoch": 2.567098616296947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8035428524017334, + "learning_rate": 1e-06, + "loss": 0.976, + "mean_token_accuracy": 0.7198294401168823, + "num_tokens": 604698545.0, + "step": 23376 + }, + { + "epoch": 2.5672084339995607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.384488821029663, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7073763608932495, + "num_tokens": 604728830.0, + "step": 23377 + }, + { + "epoch": 2.5673182517021744, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8804593086242676, + "learning_rate": 1e-06, + "loss": 0.947, + "mean_token_accuracy": 0.7299550771713257, + "num_tokens": 604753387.0, + "step": 23378 + }, + { + "epoch": 2.567428069404788, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6282873153686523, + "learning_rate": 1e-06, + "loss": 0.9271, + "mean_token_accuracy": 0.727101743221283, + "num_tokens": 604778137.0, + "step": 23379 + }, + { + "epoch": 2.5675378871074015, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.236743450164795, + "learning_rate": 1e-06, + "loss": 1.0085, + "mean_token_accuracy": 0.7139042019844055, + "num_tokens": 604814087.0, + "step": 23380 + }, + { + "epoch": 2.5676477048100153, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6890921592712402, + "learning_rate": 1e-06, + "loss": 0.8973, + "mean_token_accuracy": 0.733879566192627, + "num_tokens": 604836600.0, + "step": 23381 + }, + { + "epoch": 2.567757522512629, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.3095145225524902, + "learning_rate": 1e-06, + "loss": 0.9092, + "mean_token_accuracy": 0.7355600595474243, + "num_tokens": 604855732.0, + "step": 23382 + }, + { + "epoch": 2.5678673402152428, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4375486373901367, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7306667566299438, + "num_tokens": 604883780.0, + "step": 23383 + }, + { + "epoch": 2.5679771579178565, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6892142295837402, + "learning_rate": 1e-06, + "loss": 0.9264, + "mean_token_accuracy": 0.7295122742652893, + "num_tokens": 604906772.0, + "step": 23384 + }, + { + "epoch": 2.56808697562047, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.46509051322937, + "learning_rate": 1e-06, + "loss": 1.1, + "mean_token_accuracy": 0.6774219274520874, + "num_tokens": 604937195.0, + "step": 23385 + }, + { + "epoch": 2.5681967933230836, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5714809894561768, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.7027945518493652, + "num_tokens": 604964163.0, + "step": 23386 + }, + { + "epoch": 2.5683066110256974, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4327595233917236, + "learning_rate": 1e-06, + "loss": 1.0408, + "mean_token_accuracy": 0.6941601634025574, + "num_tokens": 604992272.0, + "step": 23387 + }, + { + "epoch": 2.568416428728311, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.569763660430908, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7077391147613525, + "num_tokens": 605016750.0, + "step": 23388 + }, + { + "epoch": 2.568526246430925, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5681424140930176, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7304734587669373, + "num_tokens": 605041357.0, + "step": 23389 + }, + { + "epoch": 2.568636064133538, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4727554321289062, + "learning_rate": 1e-06, + "loss": 0.934, + "mean_token_accuracy": 0.7241193056106567, + "num_tokens": 605067758.0, + "step": 23390 + }, + { + "epoch": 2.568745881836152, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.623246431350708, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.714111864566803, + "num_tokens": 605092976.0, + "step": 23391 + }, + { + "epoch": 2.5688556995387657, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.44856595993042, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.704502284526825, + "num_tokens": 605121236.0, + "step": 23392 + }, + { + "epoch": 2.5689655172413794, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3606607913970947, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7041910886764526, + "num_tokens": 605152048.0, + "step": 23393 + }, + { + "epoch": 2.569075334943993, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8967723846435547, + "learning_rate": 1e-06, + "loss": 0.9478, + "mean_token_accuracy": 0.7254093289375305, + "num_tokens": 605172717.0, + "step": 23394 + }, + { + "epoch": 2.5691851526466065, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.747894287109375, + "learning_rate": 1e-06, + "loss": 0.9138, + "mean_token_accuracy": 0.7292066812515259, + "num_tokens": 605193782.0, + "step": 23395 + }, + { + "epoch": 2.5692949703492203, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.819018840789795, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7074688673019409, + "num_tokens": 605218931.0, + "step": 23396 + }, + { + "epoch": 2.569404788051834, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5763111114501953, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7003660202026367, + "num_tokens": 605246891.0, + "step": 23397 + }, + { + "epoch": 2.5695146057544473, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6021616458892822, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7173805236816406, + "num_tokens": 605272368.0, + "step": 23398 + }, + { + "epoch": 2.5696244234570615, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.873790979385376, + "learning_rate": 1e-06, + "loss": 0.9221, + "mean_token_accuracy": 0.7247508764266968, + "num_tokens": 605292360.0, + "step": 23399 + }, + { + "epoch": 2.569734241159675, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6125123500823975, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6902081966400146, + "num_tokens": 605322031.0, + "step": 23400 + }, + { + "epoch": 2.5698440588622886, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7424628734588623, + "learning_rate": 1e-06, + "loss": 0.9745, + "mean_token_accuracy": 0.7121495008468628, + "num_tokens": 605347994.0, + "step": 23401 + }, + { + "epoch": 2.5699538765649024, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8154029846191406, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.7033257484436035, + "num_tokens": 605372584.0, + "step": 23402 + }, + { + "epoch": 2.5700636942675157, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5479326248168945, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7153496146202087, + "num_tokens": 605398031.0, + "step": 23403 + }, + { + "epoch": 2.5701735119701294, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.812628984451294, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.7030555605888367, + "num_tokens": 605420220.0, + "step": 23404 + }, + { + "epoch": 2.570283329672743, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.46134614944458, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6912631988525391, + "num_tokens": 605447632.0, + "step": 23405 + }, + { + "epoch": 2.570393147375357, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.999507188796997, + "learning_rate": 1e-06, + "loss": 0.9588, + "mean_token_accuracy": 0.7145212292671204, + "num_tokens": 605469430.0, + "step": 23406 + }, + { + "epoch": 2.5705029650779707, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.659909248352051, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6861493587493896, + "num_tokens": 605499226.0, + "step": 23407 + }, + { + "epoch": 2.570612782780584, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559688091278076, + "learning_rate": 1e-06, + "loss": 1.0545, + "mean_token_accuracy": 0.6969926953315735, + "num_tokens": 605527397.0, + "step": 23408 + }, + { + "epoch": 2.570722600483198, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4320807456970215, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.6950687766075134, + "num_tokens": 605553009.0, + "step": 23409 + }, + { + "epoch": 2.5708324181858115, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.862748146057129, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.7288151979446411, + "num_tokens": 605572554.0, + "step": 23410 + }, + { + "epoch": 2.5709422358884253, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.579679012298584, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.7013148069381714, + "num_tokens": 605600291.0, + "step": 23411 + }, + { + "epoch": 2.571052053591039, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5641887187957764, + "learning_rate": 1e-06, + "loss": 0.8458, + "mean_token_accuracy": 0.7523261308670044, + "num_tokens": 605624462.0, + "step": 23412 + }, + { + "epoch": 2.5711618712936524, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.568608283996582, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.715720534324646, + "num_tokens": 605654859.0, + "step": 23413 + }, + { + "epoch": 2.571271688996266, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1203315258026123, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.726014256477356, + "num_tokens": 605677494.0, + "step": 23414 + }, + { + "epoch": 2.57138150669888, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.557791233062744, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7031742930412292, + "num_tokens": 605703942.0, + "step": 23415 + }, + { + "epoch": 2.5714913244014936, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.690433979034424, + "learning_rate": 1e-06, + "loss": 1.0699, + "mean_token_accuracy": 0.6854464411735535, + "num_tokens": 605728956.0, + "step": 23416 + }, + { + "epoch": 2.5716011421041074, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.640493631362915, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7296982407569885, + "num_tokens": 605752376.0, + "step": 23417 + }, + { + "epoch": 2.5717109598067207, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7231462001800537, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7021958827972412, + "num_tokens": 605776295.0, + "step": 23418 + }, + { + "epoch": 2.5718207775093345, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.806438684463501, + "learning_rate": 1e-06, + "loss": 0.9185, + "mean_token_accuracy": 0.7330501675605774, + "num_tokens": 605798727.0, + "step": 23419 + }, + { + "epoch": 2.5719305952119482, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.667543649673462, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6842280030250549, + "num_tokens": 605830369.0, + "step": 23420 + }, + { + "epoch": 2.572040412914562, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.835191011428833, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7069806456565857, + "num_tokens": 605853381.0, + "step": 23421 + }, + { + "epoch": 2.5721502306171757, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5097529888153076, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7013812065124512, + "num_tokens": 605881397.0, + "step": 23422 + }, + { + "epoch": 2.572260048319789, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.612809181213379, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7037765383720398, + "num_tokens": 605905674.0, + "step": 23423 + }, + { + "epoch": 2.572369866022403, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.712716579437256, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.7060517072677612, + "num_tokens": 605931293.0, + "step": 23424 + }, + { + "epoch": 2.5724796837250166, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0813262462615967, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7080473899841309, + "num_tokens": 605949840.0, + "step": 23425 + }, + { + "epoch": 2.57258950142763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.889242172241211, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.7306605577468872, + "num_tokens": 605969918.0, + "step": 23426 + }, + { + "epoch": 2.5726993191302436, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6044137477874756, + "learning_rate": 1e-06, + "loss": 1.1146, + "mean_token_accuracy": 0.6806025505065918, + "num_tokens": 605997400.0, + "step": 23427 + }, + { + "epoch": 2.5728091368328574, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5876617431640625, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6924515962600708, + "num_tokens": 606022978.0, + "step": 23428 + }, + { + "epoch": 2.572918954535471, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7754485607147217, + "learning_rate": 1e-06, + "loss": 0.8878, + "mean_token_accuracy": 0.7396237850189209, + "num_tokens": 606043132.0, + "step": 23429 + }, + { + "epoch": 2.573028772238085, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.605135202407837, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6896525621414185, + "num_tokens": 606070287.0, + "step": 23430 + }, + { + "epoch": 2.573138589940698, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.480494499206543, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7029850482940674, + "num_tokens": 606100701.0, + "step": 23431 + }, + { + "epoch": 2.573248407643312, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565656900405884, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7119247913360596, + "num_tokens": 606128149.0, + "step": 23432 + }, + { + "epoch": 2.5733582253459257, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9221296310424805, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.700903058052063, + "num_tokens": 606150533.0, + "step": 23433 + }, + { + "epoch": 2.5734680430485395, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0258796215057373, + "learning_rate": 1e-06, + "loss": 0.8642, + "mean_token_accuracy": 0.7459526658058167, + "num_tokens": 606169482.0, + "step": 23434 + }, + { + "epoch": 2.5735778607511532, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5568649768829346, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7159351706504822, + "num_tokens": 606195832.0, + "step": 23435 + }, + { + "epoch": 2.5736876784537666, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.493652582168579, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7078380584716797, + "num_tokens": 606225471.0, + "step": 23436 + }, + { + "epoch": 2.5737974961563803, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.522733449935913, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6939365863800049, + "num_tokens": 606251218.0, + "step": 23437 + }, + { + "epoch": 2.573907313858994, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3423008918762207, + "learning_rate": 1e-06, + "loss": 1.1022, + "mean_token_accuracy": 0.6872124671936035, + "num_tokens": 606287133.0, + "step": 23438 + }, + { + "epoch": 2.574017131561608, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3091158866882324, + "learning_rate": 1e-06, + "loss": 1.0558, + "mean_token_accuracy": 0.6936337351799011, + "num_tokens": 606319512.0, + "step": 23439 + }, + { + "epoch": 2.5741269492642216, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.326699733734131, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7102327346801758, + "num_tokens": 606351132.0, + "step": 23440 + }, + { + "epoch": 2.574236766966835, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4682862758636475, + "learning_rate": 1e-06, + "loss": 0.8933, + "mean_token_accuracy": 0.7415099143981934, + "num_tokens": 606377231.0, + "step": 23441 + }, + { + "epoch": 2.5743465846694487, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.703296422958374, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7031862735748291, + "num_tokens": 606401545.0, + "step": 23442 + }, + { + "epoch": 2.5744564023720624, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5829124450683594, + "learning_rate": 1e-06, + "loss": 0.9518, + "mean_token_accuracy": 0.7196743488311768, + "num_tokens": 606426396.0, + "step": 23443 + }, + { + "epoch": 2.574566220074676, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6592531204223633, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7157158851623535, + "num_tokens": 606455352.0, + "step": 23444 + }, + { + "epoch": 2.57467603777729, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.674715280532837, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7189722061157227, + "num_tokens": 606478255.0, + "step": 23445 + }, + { + "epoch": 2.5747858554799032, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.728156328201294, + "learning_rate": 1e-06, + "loss": 1.0779, + "mean_token_accuracy": 0.6946603059768677, + "num_tokens": 606502966.0, + "step": 23446 + }, + { + "epoch": 2.574895673182517, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.224714994430542, + "learning_rate": 1e-06, + "loss": 0.9378, + "mean_token_accuracy": 0.7204715013504028, + "num_tokens": 606520752.0, + "step": 23447 + }, + { + "epoch": 2.5750054908851308, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.419919967651367, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.7153192758560181, + "num_tokens": 606549079.0, + "step": 23448 + }, + { + "epoch": 2.575115308587744, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6575164794921875, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7144066095352173, + "num_tokens": 606571157.0, + "step": 23449 + }, + { + "epoch": 2.5752251262903583, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.530410051345825, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.6995823979377747, + "num_tokens": 606596777.0, + "step": 23450 + }, + { + "epoch": 2.5753349439929716, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6821179389953613, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7115171551704407, + "num_tokens": 606619862.0, + "step": 23451 + }, + { + "epoch": 2.5754447616955853, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8410329818725586, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7102961540222168, + "num_tokens": 606643526.0, + "step": 23452 + }, + { + "epoch": 2.575554579398199, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.884960889816284, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.7201370000839233, + "num_tokens": 606665192.0, + "step": 23453 + }, + { + "epoch": 2.5756643971008124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5111429691314697, + "learning_rate": 1e-06, + "loss": 1.1104, + "mean_token_accuracy": 0.6897758841514587, + "num_tokens": 606694666.0, + "step": 23454 + }, + { + "epoch": 2.575774214803426, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.573474168777466, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6959303617477417, + "num_tokens": 606721178.0, + "step": 23455 + }, + { + "epoch": 2.57588403250604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4708847999572754, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7093201279640198, + "num_tokens": 606751290.0, + "step": 23456 + }, + { + "epoch": 2.5759938502086537, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6850595474243164, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6979044079780579, + "num_tokens": 606777418.0, + "step": 23457 + }, + { + "epoch": 2.5761036679112674, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.586458206176758, + "learning_rate": 1e-06, + "loss": 1.0219, + "mean_token_accuracy": 0.7095808982849121, + "num_tokens": 606802705.0, + "step": 23458 + }, + { + "epoch": 2.5762134856138807, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6841583251953125, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7199842929840088, + "num_tokens": 606826176.0, + "step": 23459 + }, + { + "epoch": 2.5763233033164945, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8272793292999268, + "learning_rate": 1e-06, + "loss": 0.9123, + "mean_token_accuracy": 0.7347708940505981, + "num_tokens": 606851494.0, + "step": 23460 + }, + { + "epoch": 2.5764331210191083, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4681835174560547, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.6961140632629395, + "num_tokens": 606880711.0, + "step": 23461 + }, + { + "epoch": 2.576542938721722, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.445171356201172, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6933614015579224, + "num_tokens": 606909157.0, + "step": 23462 + }, + { + "epoch": 2.5766527564243358, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.609377861022949, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7010531425476074, + "num_tokens": 606934181.0, + "step": 23463 + }, + { + "epoch": 2.576762574126949, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.388829231262207, + "learning_rate": 1e-06, + "loss": 1.1196, + "mean_token_accuracy": 0.6820173263549805, + "num_tokens": 606967102.0, + "step": 23464 + }, + { + "epoch": 2.576872391829563, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.437588930130005, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7220957279205322, + "num_tokens": 606994772.0, + "step": 23465 + }, + { + "epoch": 2.5769822095321766, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.64093017578125, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6849633455276489, + "num_tokens": 607024344.0, + "step": 23466 + }, + { + "epoch": 2.5770920272347904, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8448445796966553, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7062749862670898, + "num_tokens": 607045444.0, + "step": 23467 + }, + { + "epoch": 2.577201844937404, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.419471502304077, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7233728170394897, + "num_tokens": 607072488.0, + "step": 23468 + }, + { + "epoch": 2.5773116626400174, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.872016429901123, + "learning_rate": 1e-06, + "loss": 1.096, + "mean_token_accuracy": 0.6921731233596802, + "num_tokens": 607095180.0, + "step": 23469 + }, + { + "epoch": 2.577421480342631, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.840198278427124, + "learning_rate": 1e-06, + "loss": 0.9636, + "mean_token_accuracy": 0.7222059965133667, + "num_tokens": 607118068.0, + "step": 23470 + }, + { + "epoch": 2.577531298045245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3472065925598145, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7041891813278198, + "num_tokens": 607151100.0, + "step": 23471 + }, + { + "epoch": 2.5776411157478587, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5880322456359863, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.7027648687362671, + "num_tokens": 607176109.0, + "step": 23472 + }, + { + "epoch": 2.5777509334504725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6944925785064697, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7221879959106445, + "num_tokens": 607200271.0, + "step": 23473 + }, + { + "epoch": 2.5778607511530858, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.625905752182007, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6899963021278381, + "num_tokens": 607231627.0, + "step": 23474 + }, + { + "epoch": 2.5779705688556995, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.398340940475464, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7099118828773499, + "num_tokens": 607261567.0, + "step": 23475 + }, + { + "epoch": 2.5780803865583133, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7912585735321045, + "learning_rate": 1e-06, + "loss": 0.9419, + "mean_token_accuracy": 0.7163617610931396, + "num_tokens": 607283953.0, + "step": 23476 + }, + { + "epoch": 2.5781902042609266, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.413090467453003, + "learning_rate": 1e-06, + "loss": 1.1126, + "mean_token_accuracy": 0.6774882078170776, + "num_tokens": 607314055.0, + "step": 23477 + }, + { + "epoch": 2.578300021963541, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.447925090789795, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7154295444488525, + "num_tokens": 607341826.0, + "step": 23478 + }, + { + "epoch": 2.578409839666154, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.688662528991699, + "learning_rate": 1e-06, + "loss": 0.9295, + "mean_token_accuracy": 0.7248794436454773, + "num_tokens": 607362328.0, + "step": 23479 + }, + { + "epoch": 2.578519657368768, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8712289333343506, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7075219750404358, + "num_tokens": 607385401.0, + "step": 23480 + }, + { + "epoch": 2.5786294750713816, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.479388952255249, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7110769152641296, + "num_tokens": 607411585.0, + "step": 23481 + }, + { + "epoch": 2.578739292773995, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6214184761047363, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.69781494140625, + "num_tokens": 607438405.0, + "step": 23482 + }, + { + "epoch": 2.5788491104766087, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5650250911712646, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7194923758506775, + "num_tokens": 607464513.0, + "step": 23483 + }, + { + "epoch": 2.5789589281792225, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.607264518737793, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7170450687408447, + "num_tokens": 607489036.0, + "step": 23484 + }, + { + "epoch": 2.579068745881836, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3820607662200928, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7040811777114868, + "num_tokens": 607518788.0, + "step": 23485 + }, + { + "epoch": 2.57917856358445, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.515395164489746, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7196986675262451, + "num_tokens": 607545199.0, + "step": 23486 + }, + { + "epoch": 2.5792883812870633, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.780280113220215, + "learning_rate": 1e-06, + "loss": 1.0976, + "mean_token_accuracy": 0.6937543749809265, + "num_tokens": 607571705.0, + "step": 23487 + }, + { + "epoch": 2.579398198989677, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4452850818634033, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7224708795547485, + "num_tokens": 607600658.0, + "step": 23488 + }, + { + "epoch": 2.579508016692291, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8599531650543213, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7221875190734863, + "num_tokens": 607622523.0, + "step": 23489 + }, + { + "epoch": 2.5796178343949046, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.533461093902588, + "learning_rate": 1e-06, + "loss": 0.9247, + "mean_token_accuracy": 0.7302821278572083, + "num_tokens": 607649235.0, + "step": 23490 + }, + { + "epoch": 2.5797276520975183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4826602935791016, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7132518291473389, + "num_tokens": 607674183.0, + "step": 23491 + }, + { + "epoch": 2.5798374698001316, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.894495725631714, + "learning_rate": 1e-06, + "loss": 1.0754, + "mean_token_accuracy": 0.6943967938423157, + "num_tokens": 607697653.0, + "step": 23492 + }, + { + "epoch": 2.5799472875027454, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.886199474334717, + "learning_rate": 1e-06, + "loss": 0.9886, + "mean_token_accuracy": 0.7096149921417236, + "num_tokens": 607720883.0, + "step": 23493 + }, + { + "epoch": 2.580057105205359, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8190009593963623, + "learning_rate": 1e-06, + "loss": 0.9478, + "mean_token_accuracy": 0.7179709672927856, + "num_tokens": 607742152.0, + "step": 23494 + }, + { + "epoch": 2.580166922907973, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6318819522857666, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7123507857322693, + "num_tokens": 607768335.0, + "step": 23495 + }, + { + "epoch": 2.5802767406105866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.622954845428467, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7030223608016968, + "num_tokens": 607795566.0, + "step": 23496 + }, + { + "epoch": 2.5803865583132, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.489880084991455, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.692409873008728, + "num_tokens": 607825518.0, + "step": 23497 + }, + { + "epoch": 2.5804963760158137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.554443597793579, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.716181755065918, + "num_tokens": 607850631.0, + "step": 23498 + }, + { + "epoch": 2.5806061937184275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3220207691192627, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6785770654678345, + "num_tokens": 607880161.0, + "step": 23499 + }, + { + "epoch": 2.5807160114210412, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.347730875015259, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6872402429580688, + "num_tokens": 607912535.0, + "step": 23500 + }, + { + "epoch": 2.580825829123655, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3807616233825684, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7059661746025085, + "num_tokens": 607949154.0, + "step": 23501 + }, + { + "epoch": 2.5809356468262683, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4768571853637695, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.694372296333313, + "num_tokens": 607978259.0, + "step": 23502 + }, + { + "epoch": 2.581045464528882, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9163661003112793, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7028788924217224, + "num_tokens": 607999828.0, + "step": 23503 + }, + { + "epoch": 2.581155282231496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6432790756225586, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7006123065948486, + "num_tokens": 608026370.0, + "step": 23504 + }, + { + "epoch": 2.581265099934109, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7518465518951416, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.701459527015686, + "num_tokens": 608052186.0, + "step": 23505 + }, + { + "epoch": 2.581374917636723, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8936171531677246, + "learning_rate": 1e-06, + "loss": 0.9241, + "mean_token_accuracy": 0.7305004596710205, + "num_tokens": 608071921.0, + "step": 23506 + }, + { + "epoch": 2.5814847353393366, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9531450271606445, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7115423083305359, + "num_tokens": 608094605.0, + "step": 23507 + }, + { + "epoch": 2.5815945530419504, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.521554946899414, + "learning_rate": 1e-06, + "loss": 1.1197, + "mean_token_accuracy": 0.6797943115234375, + "num_tokens": 608121527.0, + "step": 23508 + }, + { + "epoch": 2.581704370744564, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5515170097351074, + "learning_rate": 1e-06, + "loss": 1.0124, + "mean_token_accuracy": 0.7002644538879395, + "num_tokens": 608148738.0, + "step": 23509 + }, + { + "epoch": 2.5818141884471775, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.469783306121826, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6956296563148499, + "num_tokens": 608178076.0, + "step": 23510 + }, + { + "epoch": 2.5819240061497912, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.800901174545288, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7312111854553223, + "num_tokens": 608200723.0, + "step": 23511 + }, + { + "epoch": 2.582033823852405, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8531551361083984, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7169709205627441, + "num_tokens": 608223552.0, + "step": 23512 + }, + { + "epoch": 2.5821436415550187, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6644928455352783, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.6978029012680054, + "num_tokens": 608248956.0, + "step": 23513 + }, + { + "epoch": 2.5822534592576325, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8436176776885986, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7090336084365845, + "num_tokens": 608281144.0, + "step": 23514 + }, + { + "epoch": 2.582363276960246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.081078052520752, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7144071459770203, + "num_tokens": 608298919.0, + "step": 23515 + }, + { + "epoch": 2.5824730946628596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.936121940612793, + "learning_rate": 1e-06, + "loss": 0.8922, + "mean_token_accuracy": 0.7408397197723389, + "num_tokens": 608317932.0, + "step": 23516 + }, + { + "epoch": 2.5825829123654733, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9194304943084717, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7076166868209839, + "num_tokens": 608343042.0, + "step": 23517 + }, + { + "epoch": 2.582692730068087, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6839983463287354, + "learning_rate": 1e-06, + "loss": 0.8702, + "mean_token_accuracy": 0.7402989864349365, + "num_tokens": 608366394.0, + "step": 23518 + }, + { + "epoch": 2.582802547770701, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6654505729675293, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6978476047515869, + "num_tokens": 608391877.0, + "step": 23519 + }, + { + "epoch": 2.582912365473314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4615330696105957, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6928443908691406, + "num_tokens": 608422930.0, + "step": 23520 + }, + { + "epoch": 2.583022183175928, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.391353130340576, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6782335042953491, + "num_tokens": 608452170.0, + "step": 23521 + }, + { + "epoch": 2.5831320008785417, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7808837890625, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.7235386371612549, + "num_tokens": 608474138.0, + "step": 23522 + }, + { + "epoch": 2.5832418185811554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565711736679077, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7226930856704712, + "num_tokens": 608500710.0, + "step": 23523 + }, + { + "epoch": 2.583351636283769, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.7604751586914062, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7230805158615112, + "num_tokens": 608526318.0, + "step": 23524 + }, + { + "epoch": 2.5834614539863825, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9797701835632324, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7137036323547363, + "num_tokens": 608549646.0, + "step": 23525 + }, + { + "epoch": 2.5835712716889963, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.650226354598999, + "learning_rate": 1e-06, + "loss": 1.0766, + "mean_token_accuracy": 0.6919074058532715, + "num_tokens": 608576690.0, + "step": 23526 + }, + { + "epoch": 2.58368108939161, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4197726249694824, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.6959099173545837, + "num_tokens": 608605824.0, + "step": 23527 + }, + { + "epoch": 2.5837909070942233, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5565035343170166, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7079952359199524, + "num_tokens": 608631410.0, + "step": 23528 + }, + { + "epoch": 2.5839007247968375, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5009942054748535, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.7046235799789429, + "num_tokens": 608659695.0, + "step": 23529 + }, + { + "epoch": 2.584010542499451, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4807610511779785, + "learning_rate": 1e-06, + "loss": 1.0533, + "mean_token_accuracy": 0.7003006339073181, + "num_tokens": 608684659.0, + "step": 23530 + }, + { + "epoch": 2.5841203602020646, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5889031887054443, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7072945833206177, + "num_tokens": 608712998.0, + "step": 23531 + }, + { + "epoch": 2.5842301779046783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3498191833496094, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7008512020111084, + "num_tokens": 608742565.0, + "step": 23532 + }, + { + "epoch": 2.5843399956072917, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.39269757270813, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7163288593292236, + "num_tokens": 608772595.0, + "step": 23533 + }, + { + "epoch": 2.5844498133099054, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.701781988143921, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6909493803977966, + "num_tokens": 608800380.0, + "step": 23534 + }, + { + "epoch": 2.584559631012519, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7601609230041504, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6878677010536194, + "num_tokens": 608826483.0, + "step": 23535 + }, + { + "epoch": 2.584669448715133, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.61531400680542, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6968324184417725, + "num_tokens": 608851677.0, + "step": 23536 + }, + { + "epoch": 2.5847792664177467, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.770348072052002, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7044761776924133, + "num_tokens": 608877458.0, + "step": 23537 + }, + { + "epoch": 2.58488908412036, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.547161817550659, + "learning_rate": 1e-06, + "loss": 0.93, + "mean_token_accuracy": 0.7234717607498169, + "num_tokens": 608903584.0, + "step": 23538 + }, + { + "epoch": 2.5849989018229738, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.71034574508667, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7068815231323242, + "num_tokens": 608931746.0, + "step": 23539 + }, + { + "epoch": 2.5851087195255875, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6437528133392334, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6821072697639465, + "num_tokens": 608960365.0, + "step": 23540 + }, + { + "epoch": 2.5852185372282013, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.475098133087158, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.7014548778533936, + "num_tokens": 608990885.0, + "step": 23541 + }, + { + "epoch": 2.585328354930815, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4183688163757324, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7074146866798401, + "num_tokens": 609018899.0, + "step": 23542 + }, + { + "epoch": 2.5854381726334283, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7557032108306885, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.7011079788208008, + "num_tokens": 609042820.0, + "step": 23543 + }, + { + "epoch": 2.585547990336042, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4942715167999268, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.703362762928009, + "num_tokens": 609071626.0, + "step": 23544 + }, + { + "epoch": 2.585657808038656, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.799401044845581, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7006074786186218, + "num_tokens": 609094227.0, + "step": 23545 + }, + { + "epoch": 2.5857676257412696, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.451050043106079, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7018001079559326, + "num_tokens": 609125138.0, + "step": 23546 + }, + { + "epoch": 2.5858774434438834, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.499239683151245, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7250786423683167, + "num_tokens": 609153763.0, + "step": 23547 + }, + { + "epoch": 2.5859872611464967, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4947266578674316, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7040517926216125, + "num_tokens": 609181468.0, + "step": 23548 + }, + { + "epoch": 2.5860970788491104, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8240153789520264, + "learning_rate": 1e-06, + "loss": 0.902, + "mean_token_accuracy": 0.7277134656906128, + "num_tokens": 609203204.0, + "step": 23549 + }, + { + "epoch": 2.586206896551724, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6830992698669434, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.719623327255249, + "num_tokens": 609226339.0, + "step": 23550 + }, + { + "epoch": 2.586316714254338, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.741196393966675, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7005975246429443, + "num_tokens": 609251273.0, + "step": 23551 + }, + { + "epoch": 2.5864265319569517, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.704702615737915, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7054679989814758, + "num_tokens": 609275377.0, + "step": 23552 + }, + { + "epoch": 2.586536349659565, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6052756309509277, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7053623199462891, + "num_tokens": 609302709.0, + "step": 23553 + }, + { + "epoch": 2.586646167362179, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.988064765930176, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.6918447613716125, + "num_tokens": 609328684.0, + "step": 23554 + }, + { + "epoch": 2.5867559850647925, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3950576782226562, + "learning_rate": 1e-06, + "loss": 1.1273, + "mean_token_accuracy": 0.67532879114151, + "num_tokens": 609358995.0, + "step": 23555 + }, + { + "epoch": 2.586865802767406, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6887869834899902, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.6977095603942871, + "num_tokens": 609383638.0, + "step": 23556 + }, + { + "epoch": 2.5869756204700196, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8216588497161865, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.696454644203186, + "num_tokens": 609406471.0, + "step": 23557 + }, + { + "epoch": 2.5870854381726334, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5631814002990723, + "learning_rate": 1e-06, + "loss": 1.0573, + "mean_token_accuracy": 0.693881094455719, + "num_tokens": 609432105.0, + "step": 23558 + }, + { + "epoch": 2.587195255875247, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5707039833068848, + "learning_rate": 1e-06, + "loss": 1.0633, + "mean_token_accuracy": 0.6921142339706421, + "num_tokens": 609460303.0, + "step": 23559 + }, + { + "epoch": 2.587305073577861, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7844808101654053, + "learning_rate": 1e-06, + "loss": 1.0874, + "mean_token_accuracy": 0.6875431537628174, + "num_tokens": 609482805.0, + "step": 23560 + }, + { + "epoch": 2.587414891280474, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6135387420654297, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.717071533203125, + "num_tokens": 609508378.0, + "step": 23561 + }, + { + "epoch": 2.587524708983088, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7994918823242188, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7054575681686401, + "num_tokens": 609532991.0, + "step": 23562 + }, + { + "epoch": 2.5876345266857017, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6884984970092773, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.6865329742431641, + "num_tokens": 609562786.0, + "step": 23563 + }, + { + "epoch": 2.5877443443883155, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.691155195236206, + "learning_rate": 1e-06, + "loss": 0.8759, + "mean_token_accuracy": 0.7373332381248474, + "num_tokens": 609584798.0, + "step": 23564 + }, + { + "epoch": 2.587854162090929, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4862635135650635, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7100731134414673, + "num_tokens": 609613866.0, + "step": 23565 + }, + { + "epoch": 2.5879639797935425, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.712508201599121, + "learning_rate": 1e-06, + "loss": 0.9159, + "mean_token_accuracy": 0.7392982244491577, + "num_tokens": 609637725.0, + "step": 23566 + }, + { + "epoch": 2.5880737974961563, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3826489448547363, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7110490798950195, + "num_tokens": 609666176.0, + "step": 23567 + }, + { + "epoch": 2.58818361519877, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7440409660339355, + "learning_rate": 1e-06, + "loss": 0.9475, + "mean_token_accuracy": 0.7150859236717224, + "num_tokens": 609688773.0, + "step": 23568 + }, + { + "epoch": 2.588293432901384, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7479472160339355, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7108088731765747, + "num_tokens": 609713472.0, + "step": 23569 + }, + { + "epoch": 2.5884032506039976, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6980233192443848, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7229000329971313, + "num_tokens": 609740360.0, + "step": 23570 + }, + { + "epoch": 2.588513068306611, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.730870485305786, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7159491181373596, + "num_tokens": 609767501.0, + "step": 23571 + }, + { + "epoch": 2.5886228860092246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6697466373443604, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7162468433380127, + "num_tokens": 609791624.0, + "step": 23572 + }, + { + "epoch": 2.5887327037118384, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.924288272857666, + "learning_rate": 1e-06, + "loss": 0.9217, + "mean_token_accuracy": 0.7322492599487305, + "num_tokens": 609811024.0, + "step": 23573 + }, + { + "epoch": 2.588842521414452, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.558175802230835, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7200300693511963, + "num_tokens": 609835938.0, + "step": 23574 + }, + { + "epoch": 2.588952339117066, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3875937461853027, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7064815163612366, + "num_tokens": 609865852.0, + "step": 23575 + }, + { + "epoch": 2.589062156819679, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5664641857147217, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.6992776393890381, + "num_tokens": 609892294.0, + "step": 23576 + }, + { + "epoch": 2.589171974522293, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.659270763397217, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7054110765457153, + "num_tokens": 609918345.0, + "step": 23577 + }, + { + "epoch": 2.5892817922249067, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6859681606292725, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7039872407913208, + "num_tokens": 609943155.0, + "step": 23578 + }, + { + "epoch": 2.58939160992752, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.539389133453369, + "learning_rate": 1e-06, + "loss": 0.8823, + "mean_token_accuracy": 0.734087347984314, + "num_tokens": 609967669.0, + "step": 23579 + }, + { + "epoch": 2.5895014276301342, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3253674507141113, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.7004222273826599, + "num_tokens": 609999713.0, + "step": 23580 + }, + { + "epoch": 2.5896112453327476, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5307350158691406, + "learning_rate": 1e-06, + "loss": 1.0946, + "mean_token_accuracy": 0.6916018724441528, + "num_tokens": 610029806.0, + "step": 23581 + }, + { + "epoch": 2.5897210630353613, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.575342893600464, + "learning_rate": 1e-06, + "loss": 1.0941, + "mean_token_accuracy": 0.6850488185882568, + "num_tokens": 610060118.0, + "step": 23582 + }, + { + "epoch": 2.589830880737975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7546815872192383, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6987163424491882, + "num_tokens": 610086735.0, + "step": 23583 + }, + { + "epoch": 2.5899406984405884, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4183859825134277, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6877045631408691, + "num_tokens": 610115401.0, + "step": 23584 + }, + { + "epoch": 2.590050516143202, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.709735155105591, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7209355235099792, + "num_tokens": 610138102.0, + "step": 23585 + }, + { + "epoch": 2.590160333845816, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.569782257080078, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.6980716586112976, + "num_tokens": 610169734.0, + "step": 23586 + }, + { + "epoch": 2.5902701515484297, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.590003490447998, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.7025456428527832, + "num_tokens": 610194219.0, + "step": 23587 + }, + { + "epoch": 2.5903799692510434, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.991666555404663, + "learning_rate": 1e-06, + "loss": 1.083, + "mean_token_accuracy": 0.6830809712409973, + "num_tokens": 610218464.0, + "step": 23588 + }, + { + "epoch": 2.5904897869536567, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.669163942337036, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7126137018203735, + "num_tokens": 610244318.0, + "step": 23589 + }, + { + "epoch": 2.5905996046562705, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9120750427246094, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.7044214010238647, + "num_tokens": 610265950.0, + "step": 23590 + }, + { + "epoch": 2.5907094223588842, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.685009002685547, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7060286998748779, + "num_tokens": 610289450.0, + "step": 23591 + }, + { + "epoch": 2.590819240061498, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5718705654144287, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7069928646087646, + "num_tokens": 610318546.0, + "step": 23592 + }, + { + "epoch": 2.5909290577641118, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4965529441833496, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7077229619026184, + "num_tokens": 610344578.0, + "step": 23593 + }, + { + "epoch": 2.591038875466725, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7153966426849365, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7236002087593079, + "num_tokens": 610368562.0, + "step": 23594 + }, + { + "epoch": 2.591148693169339, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.529533863067627, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7085385918617249, + "num_tokens": 610393880.0, + "step": 23595 + }, + { + "epoch": 2.5912585108719526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4859142303466797, + "learning_rate": 1e-06, + "loss": 1.1312, + "mean_token_accuracy": 0.6748698353767395, + "num_tokens": 610427393.0, + "step": 23596 + }, + { + "epoch": 2.5913683285745663, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9259159564971924, + "learning_rate": 1e-06, + "loss": 0.9491, + "mean_token_accuracy": 0.7256483435630798, + "num_tokens": 610449326.0, + "step": 23597 + }, + { + "epoch": 2.59147814627718, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5439391136169434, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6955177187919617, + "num_tokens": 610477099.0, + "step": 23598 + }, + { + "epoch": 2.5915879639797934, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8042569160461426, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6948717832565308, + "num_tokens": 610502104.0, + "step": 23599 + }, + { + "epoch": 2.591697781682407, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.600388288497925, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.7020936608314514, + "num_tokens": 610526704.0, + "step": 23600 + }, + { + "epoch": 2.591807599385021, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.720447301864624, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.706475019454956, + "num_tokens": 610552280.0, + "step": 23601 + }, + { + "epoch": 2.5919174170876347, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.092738389968872, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7255643606185913, + "num_tokens": 610570278.0, + "step": 23602 + }, + { + "epoch": 2.5920272347902484, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6732075214385986, + "learning_rate": 1e-06, + "loss": 1.0231, + "mean_token_accuracy": 0.7020063400268555, + "num_tokens": 610594914.0, + "step": 23603 + }, + { + "epoch": 2.5921370524928617, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7543861865997314, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7252882719039917, + "num_tokens": 610618667.0, + "step": 23604 + }, + { + "epoch": 2.5922468701954755, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6807117462158203, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.725669264793396, + "num_tokens": 610641748.0, + "step": 23605 + }, + { + "epoch": 2.5923566878980893, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2469100952148438, + "learning_rate": 1e-06, + "loss": 1.1573, + "mean_token_accuracy": 0.6712120771408081, + "num_tokens": 610678830.0, + "step": 23606 + }, + { + "epoch": 2.5924665056007026, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3061416149139404, + "learning_rate": 1e-06, + "loss": 0.9031, + "mean_token_accuracy": 0.7301080226898193, + "num_tokens": 610706433.0, + "step": 23607 + }, + { + "epoch": 2.5925763233033163, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5632996559143066, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7072992324829102, + "num_tokens": 610732306.0, + "step": 23608 + }, + { + "epoch": 2.59268614100593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8584628105163574, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7150835990905762, + "num_tokens": 610755381.0, + "step": 23609 + }, + { + "epoch": 2.592795958708544, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.813441038131714, + "learning_rate": 1e-06, + "loss": 1.0634, + "mean_token_accuracy": 0.6896924376487732, + "num_tokens": 610779631.0, + "step": 23610 + }, + { + "epoch": 2.5929057764111576, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5619609355926514, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.6798800230026245, + "num_tokens": 610807265.0, + "step": 23611 + }, + { + "epoch": 2.593015594113771, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.291794776916504, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7122272253036499, + "num_tokens": 610837654.0, + "step": 23612 + }, + { + "epoch": 2.5931254118163847, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.437500476837158, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7210015058517456, + "num_tokens": 610863735.0, + "step": 23613 + }, + { + "epoch": 2.5932352295189984, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.380472183227539, + "learning_rate": 1e-06, + "loss": 1.0391, + "mean_token_accuracy": 0.7092755436897278, + "num_tokens": 610894086.0, + "step": 23614 + }, + { + "epoch": 2.593345047221612, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7199971675872803, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7093638181686401, + "num_tokens": 610916455.0, + "step": 23615 + }, + { + "epoch": 2.593454864924226, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8901264667510986, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.7016030550003052, + "num_tokens": 610939189.0, + "step": 23616 + }, + { + "epoch": 2.5935646826268393, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.765618085861206, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7212804555892944, + "num_tokens": 610962545.0, + "step": 23617 + }, + { + "epoch": 2.593674500329453, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.286067247390747, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7102984189987183, + "num_tokens": 610979802.0, + "step": 23618 + }, + { + "epoch": 2.5937843180320668, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5343706607818604, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7057182192802429, + "num_tokens": 611007790.0, + "step": 23619 + }, + { + "epoch": 2.5938941357346805, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.855025291442871, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.702567458152771, + "num_tokens": 611028766.0, + "step": 23620 + }, + { + "epoch": 2.5940039534372943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4185338020324707, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.690820574760437, + "num_tokens": 611060140.0, + "step": 23621 + }, + { + "epoch": 2.5941137711399076, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6091911792755127, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7119671106338501, + "num_tokens": 611084267.0, + "step": 23622 + }, + { + "epoch": 2.5942235888425214, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.556818962097168, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6954876780509949, + "num_tokens": 611110761.0, + "step": 23623 + }, + { + "epoch": 2.594333406545135, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.568321943283081, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6804687976837158, + "num_tokens": 611136743.0, + "step": 23624 + }, + { + "epoch": 2.594443224247749, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.613431215286255, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7095049619674683, + "num_tokens": 611162697.0, + "step": 23625 + }, + { + "epoch": 2.5945530419503626, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.55008864402771, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7000712752342224, + "num_tokens": 611190809.0, + "step": 23626 + }, + { + "epoch": 2.594662859652976, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5221729278564453, + "learning_rate": 1e-06, + "loss": 1.0871, + "mean_token_accuracy": 0.6790444850921631, + "num_tokens": 611219321.0, + "step": 23627 + }, + { + "epoch": 2.5947726773555897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.894326686859131, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7094362378120422, + "num_tokens": 611240407.0, + "step": 23628 + }, + { + "epoch": 2.5948824950582035, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6098737716674805, + "learning_rate": 1e-06, + "loss": 1.0416, + "mean_token_accuracy": 0.7020262479782104, + "num_tokens": 611269693.0, + "step": 23629 + }, + { + "epoch": 2.594992312760817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6319077014923096, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6848185062408447, + "num_tokens": 611294073.0, + "step": 23630 + }, + { + "epoch": 2.595102130463431, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4647374153137207, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.7016709446907043, + "num_tokens": 611322370.0, + "step": 23631 + }, + { + "epoch": 2.5952119481660443, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.671633720397949, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6948016881942749, + "num_tokens": 611346265.0, + "step": 23632 + }, + { + "epoch": 2.595321765868658, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5802438259124756, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7094109058380127, + "num_tokens": 611371607.0, + "step": 23633 + }, + { + "epoch": 2.595431583571272, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4158315658569336, + "learning_rate": 1e-06, + "loss": 0.9375, + "mean_token_accuracy": 0.7266766428947449, + "num_tokens": 611399678.0, + "step": 23634 + }, + { + "epoch": 2.595541401273885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.487872838973999, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7050296068191528, + "num_tokens": 611428886.0, + "step": 23635 + }, + { + "epoch": 2.595651218976499, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.721954822540283, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7030500173568726, + "num_tokens": 611451301.0, + "step": 23636 + }, + { + "epoch": 2.5957610366791126, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4281957149505615, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7059462666511536, + "num_tokens": 611478880.0, + "step": 23637 + }, + { + "epoch": 2.5958708543817264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.524240016937256, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7041329741477966, + "num_tokens": 611506458.0, + "step": 23638 + }, + { + "epoch": 2.59598067208434, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2710394859313965, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.72404545545578, + "num_tokens": 611532584.0, + "step": 23639 + }, + { + "epoch": 2.5960904897869534, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2665576934814453, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7140189409255981, + "num_tokens": 611568513.0, + "step": 23640 + }, + { + "epoch": 2.596200307489567, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6311962604522705, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6906677484512329, + "num_tokens": 611595791.0, + "step": 23641 + }, + { + "epoch": 2.596310125192181, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6331427097320557, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.694675087928772, + "num_tokens": 611623092.0, + "step": 23642 + }, + { + "epoch": 2.5964199428947947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.639461040496826, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7128828167915344, + "num_tokens": 611646677.0, + "step": 23643 + }, + { + "epoch": 2.5965297605974085, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5612905025482178, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7136317491531372, + "num_tokens": 611672403.0, + "step": 23644 + }, + { + "epoch": 2.596639578300022, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4937520027160645, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7100232839584351, + "num_tokens": 611702719.0, + "step": 23645 + }, + { + "epoch": 2.5967493960026355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.910557508468628, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6845113039016724, + "num_tokens": 611740960.0, + "step": 23646 + }, + { + "epoch": 2.5968592137052493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4983577728271484, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7103396654129028, + "num_tokens": 611770401.0, + "step": 23647 + }, + { + "epoch": 2.596969031407863, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6392388343811035, + "learning_rate": 1e-06, + "loss": 0.9536, + "mean_token_accuracy": 0.7244600057601929, + "num_tokens": 611791834.0, + "step": 23648 + }, + { + "epoch": 2.597078849110477, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5874645709991455, + "learning_rate": 1e-06, + "loss": 1.0288, + "mean_token_accuracy": 0.7039400339126587, + "num_tokens": 611818006.0, + "step": 23649 + }, + { + "epoch": 2.59718866681309, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0782272815704346, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7109785676002502, + "num_tokens": 611836610.0, + "step": 23650 + }, + { + "epoch": 2.597298484515704, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4751789569854736, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6918046474456787, + "num_tokens": 611868349.0, + "step": 23651 + }, + { + "epoch": 2.5974083022183176, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7715911865234375, + "learning_rate": 1e-06, + "loss": 0.9369, + "mean_token_accuracy": 0.7296991348266602, + "num_tokens": 611888622.0, + "step": 23652 + }, + { + "epoch": 2.5975181199209314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.40181303024292, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7245998978614807, + "num_tokens": 611919472.0, + "step": 23653 + }, + { + "epoch": 2.597627937623545, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3510043621063232, + "learning_rate": 1e-06, + "loss": 1.0092, + "mean_token_accuracy": 0.7072328329086304, + "num_tokens": 611949302.0, + "step": 23654 + }, + { + "epoch": 2.5977377553261585, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5134670734405518, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7118034362792969, + "num_tokens": 611977106.0, + "step": 23655 + }, + { + "epoch": 2.5978475730287722, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6522977352142334, + "learning_rate": 1e-06, + "loss": 0.9964, + "mean_token_accuracy": 0.7095469236373901, + "num_tokens": 612002776.0, + "step": 23656 + }, + { + "epoch": 2.597957390731386, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.745993137359619, + "learning_rate": 1e-06, + "loss": 1.0556, + "mean_token_accuracy": 0.6916025876998901, + "num_tokens": 612028747.0, + "step": 23657 + }, + { + "epoch": 2.5980672084339993, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.846329689025879, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7220621705055237, + "num_tokens": 612051398.0, + "step": 23658 + }, + { + "epoch": 2.5981770261366135, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.625899076461792, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7125110626220703, + "num_tokens": 612076770.0, + "step": 23659 + }, + { + "epoch": 2.598286843839227, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596735954284668, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7182551622390747, + "num_tokens": 612100720.0, + "step": 23660 + }, + { + "epoch": 2.5983966615418406, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.706563949584961, + "learning_rate": 1e-06, + "loss": 0.918, + "mean_token_accuracy": 0.7291457653045654, + "num_tokens": 612124654.0, + "step": 23661 + }, + { + "epoch": 2.5985064792444543, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8262555599212646, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7374310493469238, + "num_tokens": 612144913.0, + "step": 23662 + }, + { + "epoch": 2.5986162969470676, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5946457386016846, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7309065461158752, + "num_tokens": 612171849.0, + "step": 23663 + }, + { + "epoch": 2.5987261146496814, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6241164207458496, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6875647306442261, + "num_tokens": 612203713.0, + "step": 23664 + }, + { + "epoch": 2.598835932352295, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4437196254730225, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7082815170288086, + "num_tokens": 612232019.0, + "step": 23665 + }, + { + "epoch": 2.598945750054909, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.740257263183594, + "learning_rate": 1e-06, + "loss": 1.0081, + "mean_token_accuracy": 0.7061378955841064, + "num_tokens": 612254444.0, + "step": 23666 + }, + { + "epoch": 2.5990555677575227, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8254804611206055, + "learning_rate": 1e-06, + "loss": 0.9149, + "mean_token_accuracy": 0.7303038835525513, + "num_tokens": 612280856.0, + "step": 23667 + }, + { + "epoch": 2.599165385460136, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7209255695343018, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7043991088867188, + "num_tokens": 612304793.0, + "step": 23668 + }, + { + "epoch": 2.5992752031627497, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8052539825439453, + "learning_rate": 1e-06, + "loss": 1.1011, + "mean_token_accuracy": 0.6892138123512268, + "num_tokens": 612330615.0, + "step": 23669 + }, + { + "epoch": 2.5993850208653635, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7009050846099854, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.719367504119873, + "num_tokens": 612356540.0, + "step": 23670 + }, + { + "epoch": 2.5994948385679773, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8236491680145264, + "learning_rate": 1e-06, + "loss": 0.9738, + "mean_token_accuracy": 0.716507077217102, + "num_tokens": 612378341.0, + "step": 23671 + }, + { + "epoch": 2.599604656270591, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4223239421844482, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6983531713485718, + "num_tokens": 612409321.0, + "step": 23672 + }, + { + "epoch": 2.5997144739732043, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5528171062469482, + "learning_rate": 1e-06, + "loss": 1.0869, + "mean_token_accuracy": 0.6845579147338867, + "num_tokens": 612436828.0, + "step": 23673 + }, + { + "epoch": 2.599824291675818, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6366093158721924, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.7030189037322998, + "num_tokens": 612461194.0, + "step": 23674 + }, + { + "epoch": 2.599934109378432, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6746163368225098, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7046055197715759, + "num_tokens": 612484439.0, + "step": 23675 + }, + { + "epoch": 2.6000439270810456, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5316388607025146, + "learning_rate": 1e-06, + "loss": 0.984, + "mean_token_accuracy": 0.7156450152397156, + "num_tokens": 612510719.0, + "step": 23676 + }, + { + "epoch": 2.6001537447836593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9247140884399414, + "learning_rate": 1e-06, + "loss": 0.9615, + "mean_token_accuracy": 0.718464195728302, + "num_tokens": 612531870.0, + "step": 23677 + }, + { + "epoch": 2.6002635624862727, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.76385760307312, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7180415391921997, + "num_tokens": 612559693.0, + "step": 23678 + }, + { + "epoch": 2.6003733801888864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7401604652404785, + "learning_rate": 1e-06, + "loss": 1.0442, + "mean_token_accuracy": 0.7096748352050781, + "num_tokens": 612582916.0, + "step": 23679 + }, + { + "epoch": 2.6004831978915, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.696842670440674, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.7091420888900757, + "num_tokens": 612607430.0, + "step": 23680 + }, + { + "epoch": 2.600593015594114, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7287046909332275, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7096077799797058, + "num_tokens": 612632203.0, + "step": 23681 + }, + { + "epoch": 2.6007028332967277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.498504877090454, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.6809654235839844, + "num_tokens": 612663563.0, + "step": 23682 + }, + { + "epoch": 2.600812650999341, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.742889642715454, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7123144865036011, + "num_tokens": 612687162.0, + "step": 23683 + }, + { + "epoch": 2.6009224687019548, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.884098529815674, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.6993843913078308, + "num_tokens": 612708150.0, + "step": 23684 + }, + { + "epoch": 2.6010322864045685, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.959378957748413, + "learning_rate": 1e-06, + "loss": 1.0723, + "mean_token_accuracy": 0.6952203512191772, + "num_tokens": 612732273.0, + "step": 23685 + }, + { + "epoch": 2.601142104107182, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3946332931518555, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6976377964019775, + "num_tokens": 612762367.0, + "step": 23686 + }, + { + "epoch": 2.6012519218097956, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.139975070953369, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7183316349983215, + "num_tokens": 612784479.0, + "step": 23687 + }, + { + "epoch": 2.6013617395124093, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7149689197540283, + "learning_rate": 1e-06, + "loss": 0.917, + "mean_token_accuracy": 0.7325133085250854, + "num_tokens": 612807784.0, + "step": 23688 + }, + { + "epoch": 2.601471557215023, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.769077777862549, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7139570713043213, + "num_tokens": 612830441.0, + "step": 23689 + }, + { + "epoch": 2.601581374917637, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.435330629348755, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.7278368473052979, + "num_tokens": 612860287.0, + "step": 23690 + }, + { + "epoch": 2.60169119262025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7471203804016113, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6982598304748535, + "num_tokens": 612885853.0, + "step": 23691 + }, + { + "epoch": 2.601801010322864, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.551316022872925, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7003401517868042, + "num_tokens": 612913970.0, + "step": 23692 + }, + { + "epoch": 2.6019108280254777, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.701495409011841, + "learning_rate": 1e-06, + "loss": 0.9499, + "mean_token_accuracy": 0.7185283899307251, + "num_tokens": 612936848.0, + "step": 23693 + }, + { + "epoch": 2.6020206457280914, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5220582485198975, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7034602165222168, + "num_tokens": 612962683.0, + "step": 23694 + }, + { + "epoch": 2.602130463430705, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 5.339782238006592, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.729252815246582, + "num_tokens": 612990744.0, + "step": 23695 + }, + { + "epoch": 2.6022402811333185, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6191060543060303, + "learning_rate": 1e-06, + "loss": 0.9597, + "mean_token_accuracy": 0.7162743806838989, + "num_tokens": 613017058.0, + "step": 23696 + }, + { + "epoch": 2.6023500988359323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.791256904602051, + "learning_rate": 1e-06, + "loss": 0.9394, + "mean_token_accuracy": 0.7241114377975464, + "num_tokens": 613038577.0, + "step": 23697 + }, + { + "epoch": 2.602459916538546, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.759963989257812, + "learning_rate": 1e-06, + "loss": 1.0444, + "mean_token_accuracy": 0.6953852772712708, + "num_tokens": 613067846.0, + "step": 23698 + }, + { + "epoch": 2.60256973424116, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9005773067474365, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7004512548446655, + "num_tokens": 613089819.0, + "step": 23699 + }, + { + "epoch": 2.6026795519437735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.057166576385498, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7106709480285645, + "num_tokens": 613111268.0, + "step": 23700 + }, + { + "epoch": 2.602789369646387, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.396690845489502, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7116521000862122, + "num_tokens": 613138421.0, + "step": 23701 + }, + { + "epoch": 2.6028991873490006, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7239277362823486, + "learning_rate": 1e-06, + "loss": 1.1046, + "mean_token_accuracy": 0.6863487958908081, + "num_tokens": 613163254.0, + "step": 23702 + }, + { + "epoch": 2.6030090050516144, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4036011695861816, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.7000889182090759, + "num_tokens": 613196287.0, + "step": 23703 + }, + { + "epoch": 2.603118822754228, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7524261474609375, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6979905366897583, + "num_tokens": 613220044.0, + "step": 23704 + }, + { + "epoch": 2.603228640456842, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.738755464553833, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6954636573791504, + "num_tokens": 613241739.0, + "step": 23705 + }, + { + "epoch": 2.603338458159455, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3848659992218018, + "learning_rate": 1e-06, + "loss": 1.1154, + "mean_token_accuracy": 0.6815729737281799, + "num_tokens": 613269674.0, + "step": 23706 + }, + { + "epoch": 2.603448275862069, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.633878469467163, + "learning_rate": 1e-06, + "loss": 1.0148, + "mean_token_accuracy": 0.7072137594223022, + "num_tokens": 613299200.0, + "step": 23707 + }, + { + "epoch": 2.6035580935646827, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.397465229034424, + "learning_rate": 1e-06, + "loss": 0.9359, + "mean_token_accuracy": 0.7261548638343811, + "num_tokens": 613326806.0, + "step": 23708 + }, + { + "epoch": 2.603667911267296, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2292275428771973, + "learning_rate": 1e-06, + "loss": 1.0899, + "mean_token_accuracy": 0.6836287379264832, + "num_tokens": 613363218.0, + "step": 23709 + }, + { + "epoch": 2.60377772896991, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.391864538192749, + "learning_rate": 1e-06, + "loss": 0.9064, + "mean_token_accuracy": 0.72971111536026, + "num_tokens": 613392091.0, + "step": 23710 + }, + { + "epoch": 2.6038875466725235, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5075607299804688, + "learning_rate": 1e-06, + "loss": 0.9244, + "mean_token_accuracy": 0.727706253528595, + "num_tokens": 613415949.0, + "step": 23711 + }, + { + "epoch": 2.6039973643751373, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5916383266448975, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.6984374523162842, + "num_tokens": 613442712.0, + "step": 23712 + }, + { + "epoch": 2.604107182077751, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.749537229537964, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.6972419619560242, + "num_tokens": 613468117.0, + "step": 23713 + }, + { + "epoch": 2.6042169997803644, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.452259063720703, + "learning_rate": 1e-06, + "loss": 1.0907, + "mean_token_accuracy": 0.6854872703552246, + "num_tokens": 613497539.0, + "step": 23714 + }, + { + "epoch": 2.604326817482978, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.564699649810791, + "learning_rate": 1e-06, + "loss": 0.9875, + "mean_token_accuracy": 0.7152151465415955, + "num_tokens": 613522873.0, + "step": 23715 + }, + { + "epoch": 2.604436635185592, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.605997085571289, + "learning_rate": 1e-06, + "loss": 1.098, + "mean_token_accuracy": 0.6814948320388794, + "num_tokens": 613549805.0, + "step": 23716 + }, + { + "epoch": 2.6045464528882056, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.552778482437134, + "learning_rate": 1e-06, + "loss": 1.0917, + "mean_token_accuracy": 0.6923234462738037, + "num_tokens": 613576634.0, + "step": 23717 + }, + { + "epoch": 2.6046562705908194, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7553112506866455, + "learning_rate": 1e-06, + "loss": 1.0767, + "mean_token_accuracy": 0.7024099230766296, + "num_tokens": 613600814.0, + "step": 23718 + }, + { + "epoch": 2.6047660882934327, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.510476589202881, + "learning_rate": 1e-06, + "loss": 1.009, + "mean_token_accuracy": 0.7063407897949219, + "num_tokens": 613626305.0, + "step": 23719 + }, + { + "epoch": 2.6048759059960465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4456264972686768, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7146105766296387, + "num_tokens": 613654566.0, + "step": 23720 + }, + { + "epoch": 2.60498572369866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.874077796936035, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7220650911331177, + "num_tokens": 613678702.0, + "step": 23721 + }, + { + "epoch": 2.605095541401274, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.617699146270752, + "learning_rate": 1e-06, + "loss": 1.1165, + "mean_token_accuracy": 0.677670419216156, + "num_tokens": 613705791.0, + "step": 23722 + }, + { + "epoch": 2.6052053591038877, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.47965145111084, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6975059509277344, + "num_tokens": 613734905.0, + "step": 23723 + }, + { + "epoch": 2.605315176806501, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6741559505462646, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7269899845123291, + "num_tokens": 613758211.0, + "step": 23724 + }, + { + "epoch": 2.605424994509115, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.951509714126587, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.6980087161064148, + "num_tokens": 613778836.0, + "step": 23725 + }, + { + "epoch": 2.6055348122117286, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7808759212493896, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7002356648445129, + "num_tokens": 613801624.0, + "step": 23726 + }, + { + "epoch": 2.6056446299143423, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.540759563446045, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7054377198219299, + "num_tokens": 613829644.0, + "step": 23727 + }, + { + "epoch": 2.605754447616956, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6421995162963867, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7138480544090271, + "num_tokens": 613853936.0, + "step": 23728 + }, + { + "epoch": 2.6058642653195694, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.983344793319702, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7246745228767395, + "num_tokens": 613873765.0, + "step": 23729 + }, + { + "epoch": 2.605974083022183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.408061981201172, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7051019668579102, + "num_tokens": 613901446.0, + "step": 23730 + }, + { + "epoch": 2.606083900724797, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.92592191696167, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7169556021690369, + "num_tokens": 613926484.0, + "step": 23731 + }, + { + "epoch": 2.6061937184274107, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.532485008239746, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7061402201652527, + "num_tokens": 613955178.0, + "step": 23732 + }, + { + "epoch": 2.6063035361300244, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.902061939239502, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6968719363212585, + "num_tokens": 613986262.0, + "step": 23733 + }, + { + "epoch": 2.6064133538326377, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0056543350219727, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.7340788841247559, + "num_tokens": 614004905.0, + "step": 23734 + }, + { + "epoch": 2.6065231715352515, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7230451107025146, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7137049436569214, + "num_tokens": 614028283.0, + "step": 23735 + }, + { + "epoch": 2.6066329892378652, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9877045154571533, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.718646228313446, + "num_tokens": 614046571.0, + "step": 23736 + }, + { + "epoch": 2.6067428069404786, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.830760955810547, + "learning_rate": 1e-06, + "loss": 0.9417, + "mean_token_accuracy": 0.7249478697776794, + "num_tokens": 614067079.0, + "step": 23737 + }, + { + "epoch": 2.6068526246430923, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.548020839691162, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7016240358352661, + "num_tokens": 614092766.0, + "step": 23738 + }, + { + "epoch": 2.606962442345706, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9587039947509766, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7130587100982666, + "num_tokens": 614115213.0, + "step": 23739 + }, + { + "epoch": 2.60707226004832, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7554800510406494, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7129936814308167, + "num_tokens": 614137069.0, + "step": 23740 + }, + { + "epoch": 2.6071820777509336, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.927521228790283, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7315707206726074, + "num_tokens": 614160859.0, + "step": 23741 + }, + { + "epoch": 2.607291895453547, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3894593715667725, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7031122446060181, + "num_tokens": 614188900.0, + "step": 23742 + }, + { + "epoch": 2.6074017131561606, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5128989219665527, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7066786289215088, + "num_tokens": 614217035.0, + "step": 23743 + }, + { + "epoch": 2.6075115308587744, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8000476360321045, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.6978431344032288, + "num_tokens": 614239744.0, + "step": 23744 + }, + { + "epoch": 2.607621348561388, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9370529651641846, + "learning_rate": 1e-06, + "loss": 0.9664, + "mean_token_accuracy": 0.7153294682502747, + "num_tokens": 614266888.0, + "step": 23745 + }, + { + "epoch": 2.607731166264002, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4133918285369873, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6918430924415588, + "num_tokens": 614295459.0, + "step": 23746 + }, + { + "epoch": 2.6078409839666152, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.078019142150879, + "learning_rate": 1e-06, + "loss": 0.956, + "mean_token_accuracy": 0.7176949381828308, + "num_tokens": 614319657.0, + "step": 23747 + }, + { + "epoch": 2.607950801669229, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7666842937469482, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7086607217788696, + "num_tokens": 614343494.0, + "step": 23748 + }, + { + "epoch": 2.6080606193718427, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5454351902008057, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7104630470275879, + "num_tokens": 614371395.0, + "step": 23749 + }, + { + "epoch": 2.6081704370744565, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7110519409179688, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7130296230316162, + "num_tokens": 614398280.0, + "step": 23750 + }, + { + "epoch": 2.6082802547770703, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8020286560058594, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7095474004745483, + "num_tokens": 614423501.0, + "step": 23751 + }, + { + "epoch": 2.6083900724796836, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.975637197494507, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7224342226982117, + "num_tokens": 614443447.0, + "step": 23752 + }, + { + "epoch": 2.6084998901822973, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.528998851776123, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6868582963943481, + "num_tokens": 614470168.0, + "step": 23753 + }, + { + "epoch": 2.608609707884911, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.470330238342285, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.709335446357727, + "num_tokens": 614496569.0, + "step": 23754 + }, + { + "epoch": 2.608719525587525, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.557379722595215, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.6987663507461548, + "num_tokens": 614522567.0, + "step": 23755 + }, + { + "epoch": 2.6088293432901386, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.644254207611084, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7015313506126404, + "num_tokens": 614547755.0, + "step": 23756 + }, + { + "epoch": 2.608939160992752, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5976836681365967, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.6985551714897156, + "num_tokens": 614574353.0, + "step": 23757 + }, + { + "epoch": 2.6090489786953657, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4782989025115967, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7101782560348511, + "num_tokens": 614600494.0, + "step": 23758 + }, + { + "epoch": 2.6091587963979794, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8904929161071777, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7016786932945251, + "num_tokens": 614624797.0, + "step": 23759 + }, + { + "epoch": 2.6092686141005927, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.394324541091919, + "learning_rate": 1e-06, + "loss": 1.1184, + "mean_token_accuracy": 0.6766283512115479, + "num_tokens": 614655913.0, + "step": 23760 + }, + { + "epoch": 2.609378431803207, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.78580641746521, + "learning_rate": 1e-06, + "loss": 0.9102, + "mean_token_accuracy": 0.7296858429908752, + "num_tokens": 614679388.0, + "step": 23761 + }, + { + "epoch": 2.6094882495058203, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4451324939727783, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7192444801330566, + "num_tokens": 614709117.0, + "step": 23762 + }, + { + "epoch": 2.609598067208434, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.193756580352783, + "learning_rate": 1e-06, + "loss": 0.9191, + "mean_token_accuracy": 0.726107120513916, + "num_tokens": 614727989.0, + "step": 23763 + }, + { + "epoch": 2.6097078849110478, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.874091863632202, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7214794158935547, + "num_tokens": 614750503.0, + "step": 23764 + }, + { + "epoch": 2.609817702613661, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7539820671081543, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7063217759132385, + "num_tokens": 614772555.0, + "step": 23765 + }, + { + "epoch": 2.609927520316275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.925858974456787, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.6958764791488647, + "num_tokens": 614798477.0, + "step": 23766 + }, + { + "epoch": 2.6100373380188886, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6182031631469727, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.7239915132522583, + "num_tokens": 614821899.0, + "step": 23767 + }, + { + "epoch": 2.6101471557215024, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4454801082611084, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7152307033538818, + "num_tokens": 614853929.0, + "step": 23768 + }, + { + "epoch": 2.610256973424116, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.480595588684082, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7281298637390137, + "num_tokens": 614880210.0, + "step": 23769 + }, + { + "epoch": 2.6103667911267294, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.694429636001587, + "learning_rate": 1e-06, + "loss": 0.9203, + "mean_token_accuracy": 0.7308746576309204, + "num_tokens": 614903738.0, + "step": 23770 + }, + { + "epoch": 2.610476608829343, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4400041103363037, + "learning_rate": 1e-06, + "loss": 0.927, + "mean_token_accuracy": 0.7332975268363953, + "num_tokens": 614929478.0, + "step": 23771 + }, + { + "epoch": 2.610586426531957, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5494649410247803, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.6995042562484741, + "num_tokens": 614955346.0, + "step": 23772 + }, + { + "epoch": 2.6106962442345707, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4426677227020264, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7181867957115173, + "num_tokens": 614982587.0, + "step": 23773 + }, + { + "epoch": 2.6108060619371845, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.683652400970459, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6943339109420776, + "num_tokens": 615006632.0, + "step": 23774 + }, + { + "epoch": 2.6109158796397978, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6184349060058594, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7097219228744507, + "num_tokens": 615035959.0, + "step": 23775 + }, + { + "epoch": 2.6110256973424115, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.430431604385376, + "learning_rate": 1e-06, + "loss": 1.0968, + "mean_token_accuracy": 0.691657543182373, + "num_tokens": 615064681.0, + "step": 23776 + }, + { + "epoch": 2.6111355150450253, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.680103302001953, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.707062840461731, + "num_tokens": 615090058.0, + "step": 23777 + }, + { + "epoch": 2.611245332747639, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.573199510574341, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7231022119522095, + "num_tokens": 615115131.0, + "step": 23778 + }, + { + "epoch": 2.611355150450253, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7390332221984863, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7332700490951538, + "num_tokens": 615137660.0, + "step": 23779 + }, + { + "epoch": 2.611464968152866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4773998260498047, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7162858843803406, + "num_tokens": 615164764.0, + "step": 23780 + }, + { + "epoch": 2.61157478585548, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.416290044784546, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7146151065826416, + "num_tokens": 615193518.0, + "step": 23781 + }, + { + "epoch": 2.6116846035580936, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1034886837005615, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.719963788986206, + "num_tokens": 615214833.0, + "step": 23782 + }, + { + "epoch": 2.6117944212607074, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.508216619491577, + "learning_rate": 1e-06, + "loss": 0.9998, + "mean_token_accuracy": 0.7116662263870239, + "num_tokens": 615240146.0, + "step": 23783 + }, + { + "epoch": 2.611904238963321, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5064592361450195, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7143984436988831, + "num_tokens": 615266486.0, + "step": 23784 + }, + { + "epoch": 2.6120140566659344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5729966163635254, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7115423679351807, + "num_tokens": 615295792.0, + "step": 23785 + }, + { + "epoch": 2.612123874368548, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.492675542831421, + "learning_rate": 1e-06, + "loss": 0.8529, + "mean_token_accuracy": 0.7421200275421143, + "num_tokens": 615322940.0, + "step": 23786 + }, + { + "epoch": 2.612233692071162, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.2281103134155273, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7133846879005432, + "num_tokens": 615344774.0, + "step": 23787 + }, + { + "epoch": 2.6123435097737753, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8777008056640625, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7207639813423157, + "num_tokens": 615365924.0, + "step": 23788 + }, + { + "epoch": 2.612453327476389, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8940751552581787, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.708713173866272, + "num_tokens": 615390288.0, + "step": 23789 + }, + { + "epoch": 2.612563145179003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6917221546173096, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7337559461593628, + "num_tokens": 615413048.0, + "step": 23790 + }, + { + "epoch": 2.6126729628816165, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7322001457214355, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7225849032402039, + "num_tokens": 615436740.0, + "step": 23791 + }, + { + "epoch": 2.6127827805842303, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536659002304077, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7142044305801392, + "num_tokens": 615462034.0, + "step": 23792 + }, + { + "epoch": 2.6128925982868436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.694150447845459, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.7051633596420288, + "num_tokens": 615488700.0, + "step": 23793 + }, + { + "epoch": 2.6130024159894574, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.817082166671753, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7097269892692566, + "num_tokens": 615513606.0, + "step": 23794 + }, + { + "epoch": 2.613112233692071, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7636771202087402, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7060550451278687, + "num_tokens": 615535227.0, + "step": 23795 + }, + { + "epoch": 2.613222051394685, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3955533504486084, + "learning_rate": 1e-06, + "loss": 1.0429, + "mean_token_accuracy": 0.6964163780212402, + "num_tokens": 615562831.0, + "step": 23796 + }, + { + "epoch": 2.6133318690972986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4405040740966797, + "learning_rate": 1e-06, + "loss": 1.1089, + "mean_token_accuracy": 0.6841857433319092, + "num_tokens": 615592648.0, + "step": 23797 + }, + { + "epoch": 2.613441686799912, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.004279851913452, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7088881731033325, + "num_tokens": 615614363.0, + "step": 23798 + }, + { + "epoch": 2.6135515045025257, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.493182420730591, + "learning_rate": 1e-06, + "loss": 1.0996, + "mean_token_accuracy": 0.6800011396408081, + "num_tokens": 615647265.0, + "step": 23799 + }, + { + "epoch": 2.6136613222051395, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6312592029571533, + "learning_rate": 1e-06, + "loss": 0.9149, + "mean_token_accuracy": 0.726227879524231, + "num_tokens": 615671285.0, + "step": 23800 + }, + { + "epoch": 2.6137711399077532, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7358312606811523, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6918105483055115, + "num_tokens": 615704263.0, + "step": 23801 + }, + { + "epoch": 2.613880957610367, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5906102657318115, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7076455354690552, + "num_tokens": 615728820.0, + "step": 23802 + }, + { + "epoch": 2.6139907753129803, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3480968475341797, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6932893991470337, + "num_tokens": 615758594.0, + "step": 23803 + }, + { + "epoch": 2.614100593015594, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3899853229522705, + "learning_rate": 1e-06, + "loss": 1.1079, + "mean_token_accuracy": 0.684739351272583, + "num_tokens": 615789841.0, + "step": 23804 + }, + { + "epoch": 2.614210410718208, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7470059394836426, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7063782215118408, + "num_tokens": 615811739.0, + "step": 23805 + }, + { + "epoch": 2.6143202284208216, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3123769760131836, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7016080617904663, + "num_tokens": 615843106.0, + "step": 23806 + }, + { + "epoch": 2.6144300461234353, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4041454792022705, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7178237438201904, + "num_tokens": 615870852.0, + "step": 23807 + }, + { + "epoch": 2.6145398638260486, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5850162506103516, + "learning_rate": 1e-06, + "loss": 0.8962, + "mean_token_accuracy": 0.7312673330307007, + "num_tokens": 615893427.0, + "step": 23808 + }, + { + "epoch": 2.6146496815286624, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.423595666885376, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6846483945846558, + "num_tokens": 615922853.0, + "step": 23809 + }, + { + "epoch": 2.614759499231276, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8501245975494385, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7044199705123901, + "num_tokens": 615944594.0, + "step": 23810 + }, + { + "epoch": 2.61486931693389, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8465664386749268, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7133867740631104, + "num_tokens": 615968025.0, + "step": 23811 + }, + { + "epoch": 2.6149791346365037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7404584884643555, + "learning_rate": 1e-06, + "loss": 1.1121, + "mean_token_accuracy": 0.6871803998947144, + "num_tokens": 615996177.0, + "step": 23812 + }, + { + "epoch": 2.615088952339117, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.592952013015747, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7087181210517883, + "num_tokens": 616023342.0, + "step": 23813 + }, + { + "epoch": 2.6151987700417307, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.329678773880005, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7047603130340576, + "num_tokens": 616052964.0, + "step": 23814 + }, + { + "epoch": 2.6153085877443445, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0758864879608154, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6965552568435669, + "num_tokens": 616078639.0, + "step": 23815 + }, + { + "epoch": 2.615418405446958, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4655706882476807, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7155957818031311, + "num_tokens": 616108430.0, + "step": 23816 + }, + { + "epoch": 2.6155282231495716, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6378228664398193, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7205639481544495, + "num_tokens": 616132651.0, + "step": 23817 + }, + { + "epoch": 2.6156380408521853, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.495818614959717, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6904999017715454, + "num_tokens": 616160477.0, + "step": 23818 + }, + { + "epoch": 2.615747858554799, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.892765998840332, + "learning_rate": 1e-06, + "loss": 0.9584, + "mean_token_accuracy": 0.7109482288360596, + "num_tokens": 616181366.0, + "step": 23819 + }, + { + "epoch": 2.615857676257413, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.863438367843628, + "learning_rate": 1e-06, + "loss": 1.0602, + "mean_token_accuracy": 0.6997066140174866, + "num_tokens": 616203339.0, + "step": 23820 + }, + { + "epoch": 2.615967493960026, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7932181358337402, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.719458818435669, + "num_tokens": 616226918.0, + "step": 23821 + }, + { + "epoch": 2.61607731166264, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7234585285186768, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6885101795196533, + "num_tokens": 616253500.0, + "step": 23822 + }, + { + "epoch": 2.6161871293652537, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.695845603942871, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7275552749633789, + "num_tokens": 616275443.0, + "step": 23823 + }, + { + "epoch": 2.6162969470678674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8977770805358887, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7091819047927856, + "num_tokens": 616298483.0, + "step": 23824 + }, + { + "epoch": 2.616406764770481, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8024280071258545, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7067399024963379, + "num_tokens": 616321924.0, + "step": 23825 + }, + { + "epoch": 2.6165165824730945, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.58290433883667, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7048836946487427, + "num_tokens": 616349063.0, + "step": 23826 + }, + { + "epoch": 2.6166264001757082, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.46278715133667, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7015931606292725, + "num_tokens": 616375379.0, + "step": 23827 + }, + { + "epoch": 2.616736217878322, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.745071411132812, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.6999204754829407, + "num_tokens": 616401438.0, + "step": 23828 + }, + { + "epoch": 2.6168460355809358, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5829923152923584, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7084705829620361, + "num_tokens": 616430323.0, + "step": 23829 + }, + { + "epoch": 2.6169558532835495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5383529663085938, + "learning_rate": 1e-06, + "loss": 0.9636, + "mean_token_accuracy": 0.7223724126815796, + "num_tokens": 616457365.0, + "step": 23830 + }, + { + "epoch": 2.617065670986163, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.544619560241699, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7135676145553589, + "num_tokens": 616484086.0, + "step": 23831 + }, + { + "epoch": 2.6171754886887766, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.322205066680908, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6949477195739746, + "num_tokens": 616515260.0, + "step": 23832 + }, + { + "epoch": 2.6172853063913903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.302581548690796, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7017050385475159, + "num_tokens": 616545524.0, + "step": 23833 + }, + { + "epoch": 2.617395124094004, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6333611011505127, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7141174077987671, + "num_tokens": 616570629.0, + "step": 23834 + }, + { + "epoch": 2.617504941796618, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.393467903137207, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7072688937187195, + "num_tokens": 616600138.0, + "step": 23835 + }, + { + "epoch": 2.617614759499231, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9037258625030518, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.6902909874916077, + "num_tokens": 616621489.0, + "step": 23836 + }, + { + "epoch": 2.617724577201845, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9240164756774902, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7181918025016785, + "num_tokens": 616642458.0, + "step": 23837 + }, + { + "epoch": 2.6178343949044587, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5967183113098145, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7198943495750427, + "num_tokens": 616668748.0, + "step": 23838 + }, + { + "epoch": 2.617944212607072, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3414418697357178, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.708014965057373, + "num_tokens": 616700767.0, + "step": 23839 + }, + { + "epoch": 2.618054030309686, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5845510959625244, + "learning_rate": 1e-06, + "loss": 0.9496, + "mean_token_accuracy": 0.7255605459213257, + "num_tokens": 616725784.0, + "step": 23840 + }, + { + "epoch": 2.6181638480122995, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.709155321121216, + "learning_rate": 1e-06, + "loss": 0.936, + "mean_token_accuracy": 0.7274994850158691, + "num_tokens": 616749963.0, + "step": 23841 + }, + { + "epoch": 2.6182736657149133, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.730703115463257, + "learning_rate": 1e-06, + "loss": 1.0, + "mean_token_accuracy": 0.7122502326965332, + "num_tokens": 616773116.0, + "step": 23842 + }, + { + "epoch": 2.618383483417527, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.580148696899414, + "learning_rate": 1e-06, + "loss": 1.1138, + "mean_token_accuracy": 0.6793386936187744, + "num_tokens": 616799662.0, + "step": 23843 + }, + { + "epoch": 2.6184933011201403, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.366689443588257, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7031136751174927, + "num_tokens": 616832181.0, + "step": 23844 + }, + { + "epoch": 2.618603118822754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.98179817199707, + "learning_rate": 1e-06, + "loss": 0.8642, + "mean_token_accuracy": 0.741878867149353, + "num_tokens": 616856032.0, + "step": 23845 + }, + { + "epoch": 2.618712936525368, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7011878490448, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6970903277397156, + "num_tokens": 616882922.0, + "step": 23846 + }, + { + "epoch": 2.6188227542279816, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.817949056625366, + "learning_rate": 1e-06, + "loss": 0.9871, + "mean_token_accuracy": 0.7130715847015381, + "num_tokens": 616905003.0, + "step": 23847 + }, + { + "epoch": 2.6189325719305954, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.837344169616699, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7066224217414856, + "num_tokens": 616934182.0, + "step": 23848 + }, + { + "epoch": 2.6190423896332087, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.550741195678711, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6914963722229004, + "num_tokens": 616962792.0, + "step": 23849 + }, + { + "epoch": 2.6191522073358224, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.39892840385437, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6988049745559692, + "num_tokens": 616993715.0, + "step": 23850 + }, + { + "epoch": 2.619262025038436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8898189067840576, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.714788019657135, + "num_tokens": 617016131.0, + "step": 23851 + }, + { + "epoch": 2.61937184274105, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0447328090667725, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.69535893201828, + "num_tokens": 617047681.0, + "step": 23852 + }, + { + "epoch": 2.6194816604436637, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.485429048538208, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7262518405914307, + "num_tokens": 617074011.0, + "step": 23853 + }, + { + "epoch": 2.619591478146277, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7287912368774414, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7177631855010986, + "num_tokens": 617098524.0, + "step": 23854 + }, + { + "epoch": 2.6197012958488908, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6043057441711426, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.70115065574646, + "num_tokens": 617125190.0, + "step": 23855 + }, + { + "epoch": 2.6198111135515045, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6695871353149414, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7202869653701782, + "num_tokens": 617149480.0, + "step": 23856 + }, + { + "epoch": 2.6199209312541183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520900011062622, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7053930759429932, + "num_tokens": 617175751.0, + "step": 23857 + }, + { + "epoch": 2.620030748956732, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 7.915458679199219, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7077561020851135, + "num_tokens": 617207187.0, + "step": 23858 + }, + { + "epoch": 2.6201405666593454, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7443771362304688, + "learning_rate": 1e-06, + "loss": 0.9836, + "mean_token_accuracy": 0.7144933938980103, + "num_tokens": 617230627.0, + "step": 23859 + }, + { + "epoch": 2.620250384361959, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6458241939544678, + "learning_rate": 1e-06, + "loss": 1.1094, + "mean_token_accuracy": 0.6754902601242065, + "num_tokens": 617255474.0, + "step": 23860 + }, + { + "epoch": 2.620360202064573, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.552000045776367, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7133451700210571, + "num_tokens": 617283950.0, + "step": 23861 + }, + { + "epoch": 2.6204700197671866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4756999015808105, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.7018935084342957, + "num_tokens": 617311662.0, + "step": 23862 + }, + { + "epoch": 2.6205798374698004, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6806108951568604, + "learning_rate": 1e-06, + "loss": 0.9106, + "mean_token_accuracy": 0.7241992354393005, + "num_tokens": 617337524.0, + "step": 23863 + }, + { + "epoch": 2.6206896551724137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6102375984191895, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7104358673095703, + "num_tokens": 617362717.0, + "step": 23864 + }, + { + "epoch": 2.6207994728750275, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.691312551498413, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.704667329788208, + "num_tokens": 617390882.0, + "step": 23865 + }, + { + "epoch": 2.620909290577641, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.479346990585327, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6911581754684448, + "num_tokens": 617419244.0, + "step": 23866 + }, + { + "epoch": 2.6210191082802545, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7870888710021973, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7187803983688354, + "num_tokens": 617440822.0, + "step": 23867 + }, + { + "epoch": 2.6211289259828683, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4251389503479004, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6995391249656677, + "num_tokens": 617472444.0, + "step": 23868 + }, + { + "epoch": 2.621238743685482, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7644002437591553, + "learning_rate": 1e-06, + "loss": 0.9288, + "mean_token_accuracy": 0.7181529998779297, + "num_tokens": 617498360.0, + "step": 23869 + }, + { + "epoch": 2.621348561388096, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5573112964630127, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7118973135948181, + "num_tokens": 617524776.0, + "step": 23870 + }, + { + "epoch": 2.6214583790907096, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3423428535461426, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6931030750274658, + "num_tokens": 617559133.0, + "step": 23871 + }, + { + "epoch": 2.621568196793323, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.249075412750244, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.6994978189468384, + "num_tokens": 617591498.0, + "step": 23872 + }, + { + "epoch": 2.6216780144959366, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4664723873138428, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7062292098999023, + "num_tokens": 617622204.0, + "step": 23873 + }, + { + "epoch": 2.6217878321985504, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0278453826904297, + "learning_rate": 1e-06, + "loss": 1.0427, + "mean_token_accuracy": 0.6972275972366333, + "num_tokens": 617645255.0, + "step": 23874 + }, + { + "epoch": 2.621897649901164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.676140308380127, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7204891443252563, + "num_tokens": 617668651.0, + "step": 23875 + }, + { + "epoch": 2.622007467603778, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.682119846343994, + "learning_rate": 1e-06, + "loss": 1.1026, + "mean_token_accuracy": 0.6823321580886841, + "num_tokens": 617693808.0, + "step": 23876 + }, + { + "epoch": 2.622117285306391, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7643232345581055, + "learning_rate": 1e-06, + "loss": 0.9129, + "mean_token_accuracy": 0.7313662767410278, + "num_tokens": 617714789.0, + "step": 23877 + }, + { + "epoch": 2.622227103009005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6833388805389404, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.7050753235816956, + "num_tokens": 617740029.0, + "step": 23878 + }, + { + "epoch": 2.6223369207116187, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.705880880355835, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.7176570296287537, + "num_tokens": 617763305.0, + "step": 23879 + }, + { + "epoch": 2.6224467384142325, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1856160163879395, + "learning_rate": 1e-06, + "loss": 0.9216, + "mean_token_accuracy": 0.7306304574012756, + "num_tokens": 617788975.0, + "step": 23880 + }, + { + "epoch": 2.6225565561168462, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3773345947265625, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7064382433891296, + "num_tokens": 617821155.0, + "step": 23881 + }, + { + "epoch": 2.6226663738194596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7885582447052, + "learning_rate": 1e-06, + "loss": 0.9538, + "mean_token_accuracy": 0.7291873097419739, + "num_tokens": 617843279.0, + "step": 23882 + }, + { + "epoch": 2.6227761915220733, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4535725116729736, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.698715090751648, + "num_tokens": 617871053.0, + "step": 23883 + }, + { + "epoch": 2.622886009224687, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.389540672302246, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7193222045898438, + "num_tokens": 617901241.0, + "step": 23884 + }, + { + "epoch": 2.622995826927301, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7524497509002686, + "learning_rate": 1e-06, + "loss": 1.0838, + "mean_token_accuracy": 0.6897014379501343, + "num_tokens": 617925776.0, + "step": 23885 + }, + { + "epoch": 2.6231056446299146, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9201526641845703, + "learning_rate": 1e-06, + "loss": 0.8872, + "mean_token_accuracy": 0.7455652356147766, + "num_tokens": 617947516.0, + "step": 23886 + }, + { + "epoch": 2.623215462332528, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0887205600738525, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7182014584541321, + "num_tokens": 617966419.0, + "step": 23887 + }, + { + "epoch": 2.6233252800351416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.622654438018799, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7112642526626587, + "num_tokens": 617991109.0, + "step": 23888 + }, + { + "epoch": 2.6234350977377554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.544799566268921, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.6982242465019226, + "num_tokens": 618020680.0, + "step": 23889 + }, + { + "epoch": 2.6235449154403687, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.70788311958313, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.726933479309082, + "num_tokens": 618043607.0, + "step": 23890 + }, + { + "epoch": 2.623654733142983, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.632169246673584, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7028825879096985, + "num_tokens": 618072111.0, + "step": 23891 + }, + { + "epoch": 2.6237645508455962, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6098315715789795, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7073429822921753, + "num_tokens": 618095676.0, + "step": 23892 + }, + { + "epoch": 2.62387436854821, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4375529289245605, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.7051448225975037, + "num_tokens": 618122311.0, + "step": 23893 + }, + { + "epoch": 2.6239841862508237, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.678149938583374, + "learning_rate": 1e-06, + "loss": 1.0018, + "mean_token_accuracy": 0.7155663967132568, + "num_tokens": 618146648.0, + "step": 23894 + }, + { + "epoch": 2.624094003953437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8975400924682617, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7285054326057434, + "num_tokens": 618167716.0, + "step": 23895 + }, + { + "epoch": 2.624203821656051, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5001184940338135, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7121627330780029, + "num_tokens": 618194484.0, + "step": 23896 + }, + { + "epoch": 2.6243136393586646, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9012653827667236, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7161026000976562, + "num_tokens": 618215614.0, + "step": 23897 + }, + { + "epoch": 2.6244234570612783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7596209049224854, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6995065212249756, + "num_tokens": 618241722.0, + "step": 23898 + }, + { + "epoch": 2.624533274763892, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4361724853515625, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.7014137506484985, + "num_tokens": 618270986.0, + "step": 23899 + }, + { + "epoch": 2.6246430924665054, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.529609203338623, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7094194293022156, + "num_tokens": 618298025.0, + "step": 23900 + }, + { + "epoch": 2.624752910169119, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9077396392822266, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7059314250946045, + "num_tokens": 618320589.0, + "step": 23901 + }, + { + "epoch": 2.624862727871733, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.688859462738037, + "learning_rate": 1e-06, + "loss": 0.9404, + "mean_token_accuracy": 0.7362287044525146, + "num_tokens": 618343548.0, + "step": 23902 + }, + { + "epoch": 2.6249725455743467, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.645857572555542, + "learning_rate": 1e-06, + "loss": 0.8851, + "mean_token_accuracy": 0.7342188358306885, + "num_tokens": 618366871.0, + "step": 23903 + }, + { + "epoch": 2.6250823632769604, + "ewc_loss": 1.1026859283447266e-05, + "grad_norm": 80.53548431396484, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7083035707473755, + "num_tokens": 618397442.0, + "step": 23904 + }, + { + "epoch": 2.6251921809795737, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5949673652648926, + "learning_rate": 1e-06, + "loss": 1.0108, + "mean_token_accuracy": 0.7047249674797058, + "num_tokens": 618425474.0, + "step": 23905 + }, + { + "epoch": 2.6253019986821875, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.500190496444702, + "learning_rate": 1e-06, + "loss": 0.8729, + "mean_token_accuracy": 0.7380226850509644, + "num_tokens": 618451059.0, + "step": 23906 + }, + { + "epoch": 2.6254118163848013, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4834115505218506, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.6931936740875244, + "num_tokens": 618481342.0, + "step": 23907 + }, + { + "epoch": 2.625521634087415, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5392653942108154, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7103955745697021, + "num_tokens": 618512365.0, + "step": 23908 + }, + { + "epoch": 2.6256314517900288, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.808434247970581, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7132395505905151, + "num_tokens": 618540595.0, + "step": 23909 + }, + { + "epoch": 2.625741269492642, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9262607097625732, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.706295907497406, + "num_tokens": 618562278.0, + "step": 23910 + }, + { + "epoch": 2.625851087195256, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.640143394470215, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7158648371696472, + "num_tokens": 618587828.0, + "step": 23911 + }, + { + "epoch": 2.6259609048978696, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.50421404838562, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.6912686824798584, + "num_tokens": 618615196.0, + "step": 23912 + }, + { + "epoch": 2.6260707226004834, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4776811599731445, + "learning_rate": 1e-06, + "loss": 0.9254, + "mean_token_accuracy": 0.7295087575912476, + "num_tokens": 618643927.0, + "step": 23913 + }, + { + "epoch": 2.626180540303097, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3268110752105713, + "learning_rate": 1e-06, + "loss": 1.0928, + "mean_token_accuracy": 0.6906776428222656, + "num_tokens": 618676520.0, + "step": 23914 + }, + { + "epoch": 2.6262903580057104, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.58697247505188, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7029780745506287, + "num_tokens": 618701042.0, + "step": 23915 + }, + { + "epoch": 2.626400175708324, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9789555072784424, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7164806127548218, + "num_tokens": 618727378.0, + "step": 23916 + }, + { + "epoch": 2.626509993410938, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.433664560317993, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7023313045501709, + "num_tokens": 618754896.0, + "step": 23917 + }, + { + "epoch": 2.6266198111135513, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5863983631134033, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.715928316116333, + "num_tokens": 618778671.0, + "step": 23918 + }, + { + "epoch": 2.626729628816165, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.577601671218872, + "learning_rate": 1e-06, + "loss": 1.0722, + "mean_token_accuracy": 0.702182948589325, + "num_tokens": 618804593.0, + "step": 23919 + }, + { + "epoch": 2.6268394465187788, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.577564239501953, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7119587063789368, + "num_tokens": 618829558.0, + "step": 23920 + }, + { + "epoch": 2.6269492642213925, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.029782295227051, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7079099416732788, + "num_tokens": 618856096.0, + "step": 23921 + }, + { + "epoch": 2.6270590819240063, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5247304439544678, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.703147292137146, + "num_tokens": 618887366.0, + "step": 23922 + }, + { + "epoch": 2.6271688996266196, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3551933765411377, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.700210452079773, + "num_tokens": 618918351.0, + "step": 23923 + }, + { + "epoch": 2.6272787173292333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6288554668426514, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7009299993515015, + "num_tokens": 618944233.0, + "step": 23924 + }, + { + "epoch": 2.627388535031847, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0927841663360596, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7096630334854126, + "num_tokens": 618968428.0, + "step": 23925 + }, + { + "epoch": 2.627498352734461, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9990198612213135, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7220156788825989, + "num_tokens": 618989435.0, + "step": 23926 + }, + { + "epoch": 2.6276081704370746, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8022713661193848, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7022174596786499, + "num_tokens": 619011294.0, + "step": 23927 + }, + { + "epoch": 2.627717988139688, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.896287679672241, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7180893421173096, + "num_tokens": 619031707.0, + "step": 23928 + }, + { + "epoch": 2.6278278058423017, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6850647926330566, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7230043411254883, + "num_tokens": 619056051.0, + "step": 23929 + }, + { + "epoch": 2.6279376235449154, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.670531988143921, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7060822248458862, + "num_tokens": 619078179.0, + "step": 23930 + }, + { + "epoch": 2.628047441247529, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.38393497467041, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7102773785591125, + "num_tokens": 619107033.0, + "step": 23931 + }, + { + "epoch": 2.628157258950143, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4095072746276855, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6987115144729614, + "num_tokens": 619134208.0, + "step": 23932 + }, + { + "epoch": 2.6282670766527563, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4640252590179443, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.705805778503418, + "num_tokens": 619160739.0, + "step": 23933 + }, + { + "epoch": 2.62837689435537, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.3465781211853027, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7055346369743347, + "num_tokens": 619182600.0, + "step": 23934 + }, + { + "epoch": 2.628486712057984, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.681034803390503, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.6978693008422852, + "num_tokens": 619207172.0, + "step": 23935 + }, + { + "epoch": 2.6285965297605975, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4858896732330322, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.717417299747467, + "num_tokens": 619234152.0, + "step": 23936 + }, + { + "epoch": 2.6287063474632113, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4198062419891357, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.7035312652587891, + "num_tokens": 619264240.0, + "step": 23937 + }, + { + "epoch": 2.6288161651658246, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4837539196014404, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7037081718444824, + "num_tokens": 619292047.0, + "step": 23938 + }, + { + "epoch": 2.6289259828684384, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.728067398071289, + "learning_rate": 1e-06, + "loss": 0.9417, + "mean_token_accuracy": 0.7199957370758057, + "num_tokens": 619315084.0, + "step": 23939 + }, + { + "epoch": 2.629035800571052, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.085855484008789, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7019962668418884, + "num_tokens": 619335129.0, + "step": 23940 + }, + { + "epoch": 2.6291456182736654, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.539321184158325, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.7026892304420471, + "num_tokens": 619362365.0, + "step": 23941 + }, + { + "epoch": 2.6292554359762796, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.757772207260132, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.701898455619812, + "num_tokens": 619385774.0, + "step": 23942 + }, + { + "epoch": 2.629365253678893, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3967368602752686, + "learning_rate": 1e-06, + "loss": 0.8712, + "mean_token_accuracy": 0.7418368458747864, + "num_tokens": 619411958.0, + "step": 23943 + }, + { + "epoch": 2.6294750713815067, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.303997278213501, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6888569593429565, + "num_tokens": 619446957.0, + "step": 23944 + }, + { + "epoch": 2.6295848890841205, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.509970188140869, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.709094226360321, + "num_tokens": 619477566.0, + "step": 23945 + }, + { + "epoch": 2.629694706786734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6853842735290527, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7163070440292358, + "num_tokens": 619501131.0, + "step": 23946 + }, + { + "epoch": 2.6298045244893475, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.334120035171509, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7052018642425537, + "num_tokens": 619535315.0, + "step": 23947 + }, + { + "epoch": 2.6299143421919613, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.889570474624634, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.711968183517456, + "num_tokens": 619556423.0, + "step": 23948 + }, + { + "epoch": 2.630024159894575, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4973063468933105, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6903733015060425, + "num_tokens": 619587941.0, + "step": 23949 + }, + { + "epoch": 2.630133977597189, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.633674383163452, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.715126633644104, + "num_tokens": 619612379.0, + "step": 23950 + }, + { + "epoch": 2.630243795299802, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6969149112701416, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7200835943222046, + "num_tokens": 619634800.0, + "step": 23951 + }, + { + "epoch": 2.630353613002416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1630303859710693, + "learning_rate": 1e-06, + "loss": 0.9774, + "mean_token_accuracy": 0.7110888957977295, + "num_tokens": 619661653.0, + "step": 23952 + }, + { + "epoch": 2.6304634307050296, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.440605401992798, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7150118350982666, + "num_tokens": 619688426.0, + "step": 23953 + }, + { + "epoch": 2.6305732484076434, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7136495113372803, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7066152095794678, + "num_tokens": 619711954.0, + "step": 23954 + }, + { + "epoch": 2.630683066110257, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8097314834594727, + "learning_rate": 1e-06, + "loss": 1.0939, + "mean_token_accuracy": 0.6981626152992249, + "num_tokens": 619740104.0, + "step": 23955 + }, + { + "epoch": 2.6307928838128705, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8745875358581543, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.7030973434448242, + "num_tokens": 619763446.0, + "step": 23956 + }, + { + "epoch": 2.630902701515484, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.628894329071045, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6905755996704102, + "num_tokens": 619789381.0, + "step": 23957 + }, + { + "epoch": 2.631012519218098, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7517123222351074, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7114130854606628, + "num_tokens": 619815071.0, + "step": 23958 + }, + { + "epoch": 2.6311223369207117, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.929940700531006, + "learning_rate": 1e-06, + "loss": 0.935, + "mean_token_accuracy": 0.7230003476142883, + "num_tokens": 619836300.0, + "step": 23959 + }, + { + "epoch": 2.6312321546233255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5197110176086426, + "learning_rate": 1e-06, + "loss": 1.0574, + "mean_token_accuracy": 0.6963793635368347, + "num_tokens": 619864827.0, + "step": 23960 + }, + { + "epoch": 2.631341972325939, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 4.4779438972473145, + "learning_rate": 1e-06, + "loss": 0.9429, + "mean_token_accuracy": 0.7226594090461731, + "num_tokens": 619886514.0, + "step": 23961 + }, + { + "epoch": 2.6314517900285526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6513452529907227, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7105048894882202, + "num_tokens": 619911601.0, + "step": 23962 + }, + { + "epoch": 2.6315616077311663, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8785367012023926, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7125349640846252, + "num_tokens": 619933542.0, + "step": 23963 + }, + { + "epoch": 2.63167142543378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4185099601745605, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7115899324417114, + "num_tokens": 619962953.0, + "step": 23964 + }, + { + "epoch": 2.631781243136394, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.488952159881592, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.7087900638580322, + "num_tokens": 619990685.0, + "step": 23965 + }, + { + "epoch": 2.631891060839007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6468183994293213, + "learning_rate": 1e-06, + "loss": 1.0306, + "mean_token_accuracy": 0.7050138115882874, + "num_tokens": 620016539.0, + "step": 23966 + }, + { + "epoch": 2.632000878541621, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6721978187561035, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6965475678443909, + "num_tokens": 620048256.0, + "step": 23967 + }, + { + "epoch": 2.6321106962442347, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8763179779052734, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7122243642807007, + "num_tokens": 620071741.0, + "step": 23968 + }, + { + "epoch": 2.632220513946848, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6574926376342773, + "learning_rate": 1e-06, + "loss": 0.9954, + "mean_token_accuracy": 0.7032085657119751, + "num_tokens": 620096596.0, + "step": 23969 + }, + { + "epoch": 2.6323303316494617, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.962634563446045, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.7089805603027344, + "num_tokens": 620118757.0, + "step": 23970 + }, + { + "epoch": 2.6324401493520755, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4324216842651367, + "learning_rate": 1e-06, + "loss": 1.1025, + "mean_token_accuracy": 0.6764404773712158, + "num_tokens": 620152284.0, + "step": 23971 + }, + { + "epoch": 2.6325499670546892, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8242337703704834, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7042639851570129, + "num_tokens": 620174262.0, + "step": 23972 + }, + { + "epoch": 2.632659784757303, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4636950492858887, + "learning_rate": 1e-06, + "loss": 1.1533, + "mean_token_accuracy": 0.6666547060012817, + "num_tokens": 620205473.0, + "step": 23973 + }, + { + "epoch": 2.6327696024599163, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8583450317382812, + "learning_rate": 1e-06, + "loss": 1.0473, + "mean_token_accuracy": 0.6923626661300659, + "num_tokens": 620231736.0, + "step": 23974 + }, + { + "epoch": 2.63287942016253, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4516773223876953, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7293182611465454, + "num_tokens": 620260991.0, + "step": 23975 + }, + { + "epoch": 2.632989237865144, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.523963689804077, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.7097007632255554, + "num_tokens": 620288854.0, + "step": 23976 + }, + { + "epoch": 2.6330990555677576, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.769137382507324, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6860113143920898, + "num_tokens": 620317596.0, + "step": 23977 + }, + { + "epoch": 2.6332088732703713, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.656639814376831, + "learning_rate": 1e-06, + "loss": 0.9077, + "mean_token_accuracy": 0.732207179069519, + "num_tokens": 620341641.0, + "step": 23978 + }, + { + "epoch": 2.6333186909729847, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4975860118865967, + "learning_rate": 1e-06, + "loss": 1.0151, + "mean_token_accuracy": 0.708153247833252, + "num_tokens": 620371222.0, + "step": 23979 + }, + { + "epoch": 2.6334285086755984, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.676321268081665, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.718990683555603, + "num_tokens": 620395234.0, + "step": 23980 + }, + { + "epoch": 2.633538326378212, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.429912567138672, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7111029624938965, + "num_tokens": 620423716.0, + "step": 23981 + }, + { + "epoch": 2.633648144080826, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.907064914703369, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7077008485794067, + "num_tokens": 620445412.0, + "step": 23982 + }, + { + "epoch": 2.6337579617834397, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.797147512435913, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7273358106613159, + "num_tokens": 620468398.0, + "step": 23983 + }, + { + "epoch": 2.633867779486053, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6110663414001465, + "learning_rate": 1e-06, + "loss": 1.0134, + "mean_token_accuracy": 0.7102015018463135, + "num_tokens": 620492287.0, + "step": 23984 + }, + { + "epoch": 2.6339775971886668, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.758848190307617, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7148244976997375, + "num_tokens": 620514595.0, + "step": 23985 + }, + { + "epoch": 2.6340874148912805, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5545754432678223, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7082154750823975, + "num_tokens": 620543280.0, + "step": 23986 + }, + { + "epoch": 2.6341972325938943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.692681312561035, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7066137790679932, + "num_tokens": 620564893.0, + "step": 23987 + }, + { + "epoch": 2.634307050296508, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8101909160614014, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6945968866348267, + "num_tokens": 620592556.0, + "step": 23988 + }, + { + "epoch": 2.6344168679991213, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.996997833251953, + "learning_rate": 1e-06, + "loss": 0.9321, + "mean_token_accuracy": 0.7287219166755676, + "num_tokens": 620612564.0, + "step": 23989 + }, + { + "epoch": 2.634526685701735, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.609672784805298, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7235617637634277, + "num_tokens": 620637624.0, + "step": 23990 + }, + { + "epoch": 2.634636503404349, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7282705307006836, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7017775774002075, + "num_tokens": 620662855.0, + "step": 23991 + }, + { + "epoch": 2.6347463211069626, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7996931076049805, + "learning_rate": 1e-06, + "loss": 0.9399, + "mean_token_accuracy": 0.7286897301673889, + "num_tokens": 620684595.0, + "step": 23992 + }, + { + "epoch": 2.6348561388095764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.693598747253418, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.7066845893859863, + "num_tokens": 620710732.0, + "step": 23993 + }, + { + "epoch": 2.6349659565121897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.64378023147583, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6903835535049438, + "num_tokens": 620740712.0, + "step": 23994 + }, + { + "epoch": 2.6350757742148034, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.695213794708252, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7130295634269714, + "num_tokens": 620764298.0, + "step": 23995 + }, + { + "epoch": 2.635185591917417, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5406320095062256, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.6999037265777588, + "num_tokens": 620790394.0, + "step": 23996 + }, + { + "epoch": 2.6352954096200305, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4856128692626953, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6894346475601196, + "num_tokens": 620817893.0, + "step": 23997 + }, + { + "epoch": 2.6354052273226443, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.455817461013794, + "learning_rate": 1e-06, + "loss": 0.9291, + "mean_token_accuracy": 0.7247822880744934, + "num_tokens": 620844533.0, + "step": 23998 + }, + { + "epoch": 2.635515045025258, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.869335174560547, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.7002151012420654, + "num_tokens": 620867060.0, + "step": 23999 + }, + { + "epoch": 2.6356248627278718, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6653177738189697, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7083186507225037, + "num_tokens": 620893365.0, + "step": 24000 + }, + { + "epoch": 2.6357346804304855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5996346473693848, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7222386598587036, + "num_tokens": 620920208.0, + "step": 24001 + }, + { + "epoch": 2.635844498133099, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7782962322235107, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.7009286880493164, + "num_tokens": 620942974.0, + "step": 24002 + }, + { + "epoch": 2.6359543158357126, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5256946086883545, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6930878162384033, + "num_tokens": 620970769.0, + "step": 24003 + }, + { + "epoch": 2.6360641335383264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.370063066482544, + "learning_rate": 1e-06, + "loss": 1.1084, + "mean_token_accuracy": 0.6829439997673035, + "num_tokens": 621003008.0, + "step": 24004 + }, + { + "epoch": 2.63617395124094, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4715471267700195, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7032573223114014, + "num_tokens": 621033307.0, + "step": 24005 + }, + { + "epoch": 2.636283768943554, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3527581691741943, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7054897546768188, + "num_tokens": 621063311.0, + "step": 24006 + }, + { + "epoch": 2.636393586646167, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.521806478500366, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7047163248062134, + "num_tokens": 621091081.0, + "step": 24007 + }, + { + "epoch": 2.636503404348781, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9144845008850098, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6905403137207031, + "num_tokens": 621111113.0, + "step": 24008 + }, + { + "epoch": 2.6366132220513947, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6867709159851074, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7053655982017517, + "num_tokens": 621137657.0, + "step": 24009 + }, + { + "epoch": 2.6367230397540085, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.929173707962036, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7279233932495117, + "num_tokens": 621157945.0, + "step": 24010 + }, + { + "epoch": 2.636832857456622, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.87174129486084, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.689501941204071, + "num_tokens": 621178900.0, + "step": 24011 + }, + { + "epoch": 2.6369426751592355, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5367860794067383, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7100839614868164, + "num_tokens": 621204561.0, + "step": 24012 + }, + { + "epoch": 2.6370524928618493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.790966272354126, + "learning_rate": 1e-06, + "loss": 1.0269, + "mean_token_accuracy": 0.7085294723510742, + "num_tokens": 621226032.0, + "step": 24013 + }, + { + "epoch": 2.637162310564463, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7765355110168457, + "learning_rate": 1e-06, + "loss": 1.0123, + "mean_token_accuracy": 0.7031672596931458, + "num_tokens": 621245472.0, + "step": 24014 + }, + { + "epoch": 2.637272128267077, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6706905364990234, + "learning_rate": 1e-06, + "loss": 1.037, + "mean_token_accuracy": 0.6993381977081299, + "num_tokens": 621266922.0, + "step": 24015 + }, + { + "epoch": 2.6373819459696906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4698405265808105, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.7005022764205933, + "num_tokens": 621296209.0, + "step": 24016 + }, + { + "epoch": 2.637491763672304, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8447206020355225, + "learning_rate": 1e-06, + "loss": 1.0691, + "mean_token_accuracy": 0.6988812685012817, + "num_tokens": 621318693.0, + "step": 24017 + }, + { + "epoch": 2.6376015813749176, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4487063884735107, + "learning_rate": 1e-06, + "loss": 1.1208, + "mean_token_accuracy": 0.6772818565368652, + "num_tokens": 621347752.0, + "step": 24018 + }, + { + "epoch": 2.6377113990775314, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.518059015274048, + "learning_rate": 1e-06, + "loss": 0.9089, + "mean_token_accuracy": 0.737292468547821, + "num_tokens": 621373631.0, + "step": 24019 + }, + { + "epoch": 2.6378212167801447, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.618793249130249, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7188091278076172, + "num_tokens": 621398262.0, + "step": 24020 + }, + { + "epoch": 2.637931034482759, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.751279592514038, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7353298664093018, + "num_tokens": 621423360.0, + "step": 24021 + }, + { + "epoch": 2.638040852185372, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.503439426422119, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7113732695579529, + "num_tokens": 621451716.0, + "step": 24022 + }, + { + "epoch": 2.638150669887986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.558685302734375, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.6974679827690125, + "num_tokens": 621475198.0, + "step": 24023 + }, + { + "epoch": 2.6382604875905997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.666935682296753, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7035158276557922, + "num_tokens": 621499649.0, + "step": 24024 + }, + { + "epoch": 2.638370305293213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.658050298690796, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7191635966300964, + "num_tokens": 621524844.0, + "step": 24025 + }, + { + "epoch": 2.638480122995827, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3903558254241943, + "learning_rate": 1e-06, + "loss": 1.1089, + "mean_token_accuracy": 0.6823450326919556, + "num_tokens": 621556794.0, + "step": 24026 + }, + { + "epoch": 2.6385899406984406, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.774977922439575, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7023991346359253, + "num_tokens": 621576931.0, + "step": 24027 + }, + { + "epoch": 2.6386997584010543, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4742114543914795, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6992119550704956, + "num_tokens": 621603944.0, + "step": 24028 + }, + { + "epoch": 2.638809576103668, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.442502498626709, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6908222436904907, + "num_tokens": 621635189.0, + "step": 24029 + }, + { + "epoch": 2.6389193938062814, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.867945432662964, + "learning_rate": 1e-06, + "loss": 0.9125, + "mean_token_accuracy": 0.7269349098205566, + "num_tokens": 621656516.0, + "step": 24030 + }, + { + "epoch": 2.639029211508895, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7197115421295166, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.7025131583213806, + "num_tokens": 621681421.0, + "step": 24031 + }, + { + "epoch": 2.639139029211509, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7290217876434326, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6761929988861084, + "num_tokens": 621705867.0, + "step": 24032 + }, + { + "epoch": 2.6392488469141226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6159560680389404, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7116355895996094, + "num_tokens": 621731823.0, + "step": 24033 + }, + { + "epoch": 2.6393586646167364, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.022524118423462, + "learning_rate": 1e-06, + "loss": 0.9079, + "mean_token_accuracy": 0.7268072366714478, + "num_tokens": 621751871.0, + "step": 24034 + }, + { + "epoch": 2.6394684823193497, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.365821361541748, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7106295824050903, + "num_tokens": 621778959.0, + "step": 24035 + }, + { + "epoch": 2.6395783000219635, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5363268852233887, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6970067024230957, + "num_tokens": 621806512.0, + "step": 24036 + }, + { + "epoch": 2.6396881177245772, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 8.106322288513184, + "learning_rate": 1e-06, + "loss": 0.9514, + "mean_token_accuracy": 0.7193576097488403, + "num_tokens": 621823949.0, + "step": 24037 + }, + { + "epoch": 2.639797935427191, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4606423377990723, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7222967147827148, + "num_tokens": 621851867.0, + "step": 24038 + }, + { + "epoch": 2.6399077531298047, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.704761266708374, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7101019620895386, + "num_tokens": 621877830.0, + "step": 24039 + }, + { + "epoch": 2.640017570832418, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.62817645072937, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6912264227867126, + "num_tokens": 621904075.0, + "step": 24040 + }, + { + "epoch": 2.640127388535032, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8449037075042725, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7227464914321899, + "num_tokens": 621923382.0, + "step": 24041 + }, + { + "epoch": 2.6402372062376456, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5357296466827393, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7070215940475464, + "num_tokens": 621953238.0, + "step": 24042 + }, + { + "epoch": 2.6403470239402593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9963204860687256, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7200126647949219, + "num_tokens": 621972582.0, + "step": 24043 + }, + { + "epoch": 2.640456841642873, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.634197950363159, + "learning_rate": 1e-06, + "loss": 1.0213, + "mean_token_accuracy": 0.7053049802780151, + "num_tokens": 621997752.0, + "step": 24044 + }, + { + "epoch": 2.6405666593454864, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.784976005554199, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7191600203514099, + "num_tokens": 622020119.0, + "step": 24045 + }, + { + "epoch": 2.6406764770481, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8589928150177, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7141256928443909, + "num_tokens": 622043496.0, + "step": 24046 + }, + { + "epoch": 2.640786294750714, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536208391189575, + "learning_rate": 1e-06, + "loss": 1.0351, + "mean_token_accuracy": 0.6943941116333008, + "num_tokens": 622070300.0, + "step": 24047 + }, + { + "epoch": 2.6408961124533272, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.773344039916992, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7069582939147949, + "num_tokens": 622094411.0, + "step": 24048 + }, + { + "epoch": 2.641005930155941, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.435948610305786, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6860232353210449, + "num_tokens": 622128698.0, + "step": 24049 + }, + { + "epoch": 2.6411157478585547, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.870443344116211, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7069035172462463, + "num_tokens": 622148936.0, + "step": 24050 + }, + { + "epoch": 2.6412255655611685, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3583929538726807, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7119200229644775, + "num_tokens": 622176301.0, + "step": 24051 + }, + { + "epoch": 2.6413353832637823, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.262270212173462, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6911396980285645, + "num_tokens": 622208911.0, + "step": 24052 + }, + { + "epoch": 2.6414452009663956, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.551523208618164, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6930081844329834, + "num_tokens": 622236356.0, + "step": 24053 + }, + { + "epoch": 2.6415550186690093, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4880623817443848, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7053778767585754, + "num_tokens": 622262816.0, + "step": 24054 + }, + { + "epoch": 2.641664836371623, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7847635746002197, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.6950004696846008, + "num_tokens": 622284963.0, + "step": 24055 + }, + { + "epoch": 2.641774654074237, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0265676975250244, + "learning_rate": 1e-06, + "loss": 0.8981, + "mean_token_accuracy": 0.7326127290725708, + "num_tokens": 622305749.0, + "step": 24056 + }, + { + "epoch": 2.6418844717768506, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.543588876724243, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.7270940542221069, + "num_tokens": 622331647.0, + "step": 24057 + }, + { + "epoch": 2.641994289479464, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.711148738861084, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.7010868787765503, + "num_tokens": 622354461.0, + "step": 24058 + }, + { + "epoch": 2.6421041071820777, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.755155324935913, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.6998335123062134, + "num_tokens": 622378606.0, + "step": 24059 + }, + { + "epoch": 2.6422139248846914, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.557455539703369, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.7123579382896423, + "num_tokens": 622404972.0, + "step": 24060 + }, + { + "epoch": 2.642323742587305, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.441265106201172, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6948726177215576, + "num_tokens": 622437224.0, + "step": 24061 + }, + { + "epoch": 2.642433560289919, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.392638683319092, + "learning_rate": 1e-06, + "loss": 1.1287, + "mean_token_accuracy": 0.6714036464691162, + "num_tokens": 622470711.0, + "step": 24062 + }, + { + "epoch": 2.6425433779925322, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7734270095825195, + "learning_rate": 1e-06, + "loss": 0.9532, + "mean_token_accuracy": 0.719679594039917, + "num_tokens": 622495174.0, + "step": 24063 + }, + { + "epoch": 2.642653195695146, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 8.012908935546875, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.7258092761039734, + "num_tokens": 622518425.0, + "step": 24064 + }, + { + "epoch": 2.6427630133977598, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.417917013168335, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7002753019332886, + "num_tokens": 622549915.0, + "step": 24065 + }, + { + "epoch": 2.6428728311003735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8437225818634033, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6828271150588989, + "num_tokens": 622572444.0, + "step": 24066 + }, + { + "epoch": 2.6429826488029873, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8525617122650146, + "learning_rate": 1e-06, + "loss": 0.9197, + "mean_token_accuracy": 0.727763295173645, + "num_tokens": 622593190.0, + "step": 24067 + }, + { + "epoch": 2.6430924665056006, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7670645713806152, + "learning_rate": 1e-06, + "loss": 1.0516, + "mean_token_accuracy": 0.7064735889434814, + "num_tokens": 622620392.0, + "step": 24068 + }, + { + "epoch": 2.6432022842082143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.512601852416992, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.70445716381073, + "num_tokens": 622649585.0, + "step": 24069 + }, + { + "epoch": 2.643312101910828, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.43257212638855, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7188071012496948, + "num_tokens": 622676984.0, + "step": 24070 + }, + { + "epoch": 2.6434219196134414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4498088359832764, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.6987555623054504, + "num_tokens": 622705084.0, + "step": 24071 + }, + { + "epoch": 2.6435317373160556, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6053473949432373, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7036228179931641, + "num_tokens": 622732563.0, + "step": 24072 + }, + { + "epoch": 2.643641555018669, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.725780725479126, + "learning_rate": 1e-06, + "loss": 0.9454, + "mean_token_accuracy": 0.7243382930755615, + "num_tokens": 622757536.0, + "step": 24073 + }, + { + "epoch": 2.6437513727212827, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.576080560684204, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7049837112426758, + "num_tokens": 622787075.0, + "step": 24074 + }, + { + "epoch": 2.6438611904238964, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8359251022338867, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7008408904075623, + "num_tokens": 622811946.0, + "step": 24075 + }, + { + "epoch": 2.6439710081265098, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4998373985290527, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7123298645019531, + "num_tokens": 622839809.0, + "step": 24076 + }, + { + "epoch": 2.6440808258291235, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8332507610321045, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7209768295288086, + "num_tokens": 622858369.0, + "step": 24077 + }, + { + "epoch": 2.6441906435317373, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9033203125, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7162668704986572, + "num_tokens": 622879353.0, + "step": 24078 + }, + { + "epoch": 2.644300461234351, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8895576000213623, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.6923830509185791, + "num_tokens": 622906053.0, + "step": 24079 + }, + { + "epoch": 2.644410278936965, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.549710988998413, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.6977638602256775, + "num_tokens": 622935393.0, + "step": 24080 + }, + { + "epoch": 2.644520096639578, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.503460645675659, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.6987544298171997, + "num_tokens": 622961271.0, + "step": 24081 + }, + { + "epoch": 2.644629914342192, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.517193555831909, + "learning_rate": 1e-06, + "loss": 0.9389, + "mean_token_accuracy": 0.7223268747329712, + "num_tokens": 622988400.0, + "step": 24082 + }, + { + "epoch": 2.6447397320448056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6542768478393555, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7079310417175293, + "num_tokens": 623011022.0, + "step": 24083 + }, + { + "epoch": 2.6448495497474194, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8085336685180664, + "learning_rate": 1e-06, + "loss": 0.8943, + "mean_token_accuracy": 0.7459350228309631, + "num_tokens": 623030796.0, + "step": 24084 + }, + { + "epoch": 2.644959367450033, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.519331455230713, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6947864890098572, + "num_tokens": 623060043.0, + "step": 24085 + }, + { + "epoch": 2.6450691851526464, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.693297863006592, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7261824607849121, + "num_tokens": 623082033.0, + "step": 24086 + }, + { + "epoch": 2.64517900285526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520573377609253, + "learning_rate": 1e-06, + "loss": 1.0529, + "mean_token_accuracy": 0.6975575685501099, + "num_tokens": 623109445.0, + "step": 24087 + }, + { + "epoch": 2.645288820557874, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.582402229309082, + "learning_rate": 1e-06, + "loss": 1.0744, + "mean_token_accuracy": 0.6944009065628052, + "num_tokens": 623138438.0, + "step": 24088 + }, + { + "epoch": 2.6453986382604877, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.568319797515869, + "learning_rate": 1e-06, + "loss": 0.9527, + "mean_token_accuracy": 0.7203229665756226, + "num_tokens": 623162892.0, + "step": 24089 + }, + { + "epoch": 2.6455084559631015, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.223536491394043, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7167407274246216, + "num_tokens": 623181200.0, + "step": 24090 + }, + { + "epoch": 2.645618273665715, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4567506313323975, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7070569396018982, + "num_tokens": 623209572.0, + "step": 24091 + }, + { + "epoch": 2.6457280913683285, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7578234672546387, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.6962082386016846, + "num_tokens": 623233462.0, + "step": 24092 + }, + { + "epoch": 2.6458379090709423, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5311264991760254, + "learning_rate": 1e-06, + "loss": 1.0146, + "mean_token_accuracy": 0.7062442302703857, + "num_tokens": 623260152.0, + "step": 24093 + }, + { + "epoch": 2.645947726773556, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4678754806518555, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.700283408164978, + "num_tokens": 623290976.0, + "step": 24094 + }, + { + "epoch": 2.64605754447617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.698183536529541, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6911169290542603, + "num_tokens": 623314911.0, + "step": 24095 + }, + { + "epoch": 2.646167362178783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5578579902648926, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7149957418441772, + "num_tokens": 623342413.0, + "step": 24096 + }, + { + "epoch": 2.646277179881397, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.399106979370117, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.690720796585083, + "num_tokens": 623371308.0, + "step": 24097 + }, + { + "epoch": 2.6463869975840106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.316790819168091, + "learning_rate": 1e-06, + "loss": 0.8858, + "mean_token_accuracy": 0.7393044233322144, + "num_tokens": 623398913.0, + "step": 24098 + }, + { + "epoch": 2.646496815286624, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3971686363220215, + "learning_rate": 1e-06, + "loss": 1.101, + "mean_token_accuracy": 0.6892368197441101, + "num_tokens": 623430023.0, + "step": 24099 + }, + { + "epoch": 2.6466066329892377, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.326855182647705, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.6971833109855652, + "num_tokens": 623461192.0, + "step": 24100 + }, + { + "epoch": 2.6467164506918515, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5022125244140625, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.720758318901062, + "num_tokens": 623487347.0, + "step": 24101 + }, + { + "epoch": 2.646826268394465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8004558086395264, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6990127563476562, + "num_tokens": 623511378.0, + "step": 24102 + }, + { + "epoch": 2.646936086097079, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6333746910095215, + "learning_rate": 1e-06, + "loss": 1.0457, + "mean_token_accuracy": 0.6987258791923523, + "num_tokens": 623535263.0, + "step": 24103 + }, + { + "epoch": 2.6470459037996923, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.680065155029297, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.6934218406677246, + "num_tokens": 623562398.0, + "step": 24104 + }, + { + "epoch": 2.647155721502306, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.688809633255005, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.712259829044342, + "num_tokens": 623586222.0, + "step": 24105 + }, + { + "epoch": 2.64726553920492, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.896873950958252, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.703524649143219, + "num_tokens": 623608154.0, + "step": 24106 + }, + { + "epoch": 2.6473753569075336, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.717406988143921, + "learning_rate": 1e-06, + "loss": 1.1055, + "mean_token_accuracy": 0.6826041340827942, + "num_tokens": 623634219.0, + "step": 24107 + }, + { + "epoch": 2.6474851746101473, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.70975923538208, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.6999093294143677, + "num_tokens": 623661239.0, + "step": 24108 + }, + { + "epoch": 2.6475949923127606, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.509686231613159, + "learning_rate": 1e-06, + "loss": 0.9184, + "mean_token_accuracy": 0.7293931245803833, + "num_tokens": 623687698.0, + "step": 24109 + }, + { + "epoch": 2.6477048100153744, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.657384157180786, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7100023031234741, + "num_tokens": 623712239.0, + "step": 24110 + }, + { + "epoch": 2.647814627717988, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9005417823791504, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.723381519317627, + "num_tokens": 623734006.0, + "step": 24111 + }, + { + "epoch": 2.647924445420602, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6357102394104004, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7020347118377686, + "num_tokens": 623761280.0, + "step": 24112 + }, + { + "epoch": 2.6480342631232157, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.680774211883545, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6964339017868042, + "num_tokens": 623788872.0, + "step": 24113 + }, + { + "epoch": 2.648144080825829, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.649365186691284, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6852721571922302, + "num_tokens": 623815823.0, + "step": 24114 + }, + { + "epoch": 2.6482538985284427, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520789861679077, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7043176889419556, + "num_tokens": 623841142.0, + "step": 24115 + }, + { + "epoch": 2.6483637162310565, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.508647918701172, + "learning_rate": 1e-06, + "loss": 1.1248, + "mean_token_accuracy": 0.6729438304901123, + "num_tokens": 623868966.0, + "step": 24116 + }, + { + "epoch": 2.6484735339336702, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.730415105819702, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7122313380241394, + "num_tokens": 623889833.0, + "step": 24117 + }, + { + "epoch": 2.648583351636284, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.594780445098877, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6905807256698608, + "num_tokens": 623916392.0, + "step": 24118 + }, + { + "epoch": 2.6486931693388973, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5576891899108887, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.7034890651702881, + "num_tokens": 623941110.0, + "step": 24119 + }, + { + "epoch": 2.648802987041511, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6652534008026123, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.713931679725647, + "num_tokens": 623965608.0, + "step": 24120 + }, + { + "epoch": 2.648912804744125, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6657042503356934, + "learning_rate": 1e-06, + "loss": 1.021, + "mean_token_accuracy": 0.7079100608825684, + "num_tokens": 623992636.0, + "step": 24121 + }, + { + "epoch": 2.649022622446738, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5562472343444824, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7201160192489624, + "num_tokens": 624016363.0, + "step": 24122 + }, + { + "epoch": 2.6491324401493523, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.484004020690918, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7235012054443359, + "num_tokens": 624043771.0, + "step": 24123 + }, + { + "epoch": 2.6492422578519657, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.442150831222534, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.7010999917984009, + "num_tokens": 624072386.0, + "step": 24124 + }, + { + "epoch": 2.6493520755545794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6014440059661865, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.708136260509491, + "num_tokens": 624097076.0, + "step": 24125 + }, + { + "epoch": 2.649461893257193, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7976267337799072, + "learning_rate": 1e-06, + "loss": 0.919, + "mean_token_accuracy": 0.7259261608123779, + "num_tokens": 624120131.0, + "step": 24126 + }, + { + "epoch": 2.6495717109598065, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.529555320739746, + "learning_rate": 1e-06, + "loss": 0.967, + "mean_token_accuracy": 0.7171174883842468, + "num_tokens": 624148254.0, + "step": 24127 + }, + { + "epoch": 2.6496815286624202, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6051442623138428, + "learning_rate": 1e-06, + "loss": 1.0881, + "mean_token_accuracy": 0.6935639977455139, + "num_tokens": 624176985.0, + "step": 24128 + }, + { + "epoch": 2.649791346365034, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6662535667419434, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.710081160068512, + "num_tokens": 624202640.0, + "step": 24129 + }, + { + "epoch": 2.6499011640676478, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.745319366455078, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7231512069702148, + "num_tokens": 624225318.0, + "step": 24130 + }, + { + "epoch": 2.6500109817702615, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.738067150115967, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7114039063453674, + "num_tokens": 624249215.0, + "step": 24131 + }, + { + "epoch": 2.650120799472875, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7091057300567627, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7115700244903564, + "num_tokens": 624271947.0, + "step": 24132 + }, + { + "epoch": 2.6502306171754886, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7008001804351807, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7166205644607544, + "num_tokens": 624294819.0, + "step": 24133 + }, + { + "epoch": 2.6503404348781023, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7814388275146484, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7089484333992004, + "num_tokens": 624319693.0, + "step": 24134 + }, + { + "epoch": 2.650450252580716, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3302114009857178, + "learning_rate": 1e-06, + "loss": 1.012, + "mean_token_accuracy": 0.7068415880203247, + "num_tokens": 624350989.0, + "step": 24135 + }, + { + "epoch": 2.65056007028333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.377760171890259, + "learning_rate": 1e-06, + "loss": 0.913, + "mean_token_accuracy": 0.733931839466095, + "num_tokens": 624380414.0, + "step": 24136 + }, + { + "epoch": 2.650669887985943, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4716732501983643, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.6969060897827148, + "num_tokens": 624406399.0, + "step": 24137 + }, + { + "epoch": 2.650779705688557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5462493896484375, + "learning_rate": 1e-06, + "loss": 1.1036, + "mean_token_accuracy": 0.6874305009841919, + "num_tokens": 624433901.0, + "step": 24138 + }, + { + "epoch": 2.6508895233911707, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.61405086517334, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7149705290794373, + "num_tokens": 624458660.0, + "step": 24139 + }, + { + "epoch": 2.6509993410937844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7085964679718018, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7121795415878296, + "num_tokens": 624481456.0, + "step": 24140 + }, + { + "epoch": 2.651109158796398, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8330421447753906, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7048584818840027, + "num_tokens": 624504676.0, + "step": 24141 + }, + { + "epoch": 2.6512189764990115, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.44992733001709, + "learning_rate": 1e-06, + "loss": 1.1633, + "mean_token_accuracy": 0.6715428829193115, + "num_tokens": 624535033.0, + "step": 24142 + }, + { + "epoch": 2.6513287942016253, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8033459186553955, + "learning_rate": 1e-06, + "loss": 0.9421, + "mean_token_accuracy": 0.7172524929046631, + "num_tokens": 624557535.0, + "step": 24143 + }, + { + "epoch": 2.651438611904239, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6553659439086914, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7064671516418457, + "num_tokens": 624583930.0, + "step": 24144 + }, + { + "epoch": 2.6515484296068528, + "ewc_loss": 1.1086463928222656e-05, + "grad_norm": 12.91801643371582, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.707787036895752, + "num_tokens": 624607216.0, + "step": 24145 + }, + { + "epoch": 2.6516582473094665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.786633014678955, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7222697138786316, + "num_tokens": 624629725.0, + "step": 24146 + }, + { + "epoch": 2.65176806501208, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 8.065690040588379, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7043205499649048, + "num_tokens": 624654995.0, + "step": 24147 + }, + { + "epoch": 2.6518778827146936, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.729351758956909, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.72968590259552, + "num_tokens": 624677595.0, + "step": 24148 + }, + { + "epoch": 2.6519877004173074, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.633661985397339, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.696183979511261, + "num_tokens": 624705954.0, + "step": 24149 + }, + { + "epoch": 2.6520975181199207, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4571192264556885, + "learning_rate": 1e-06, + "loss": 1.0619, + "mean_token_accuracy": 0.6925347447395325, + "num_tokens": 624735000.0, + "step": 24150 + }, + { + "epoch": 2.6522073358225344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8756322860717773, + "learning_rate": 1e-06, + "loss": 1.0256, + "mean_token_accuracy": 0.6977440118789673, + "num_tokens": 624756511.0, + "step": 24151 + }, + { + "epoch": 2.652317153525148, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.430013418197632, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6882171034812927, + "num_tokens": 624784308.0, + "step": 24152 + }, + { + "epoch": 2.652426971227762, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.669503927230835, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7057812809944153, + "num_tokens": 624809534.0, + "step": 24153 + }, + { + "epoch": 2.6525367889303757, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617908239364624, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.7227110862731934, + "num_tokens": 624834936.0, + "step": 24154 + }, + { + "epoch": 2.652646606632989, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.882972240447998, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.6901615262031555, + "num_tokens": 624864487.0, + "step": 24155 + }, + { + "epoch": 2.6527564243356028, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9940106868743896, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6916267275810242, + "num_tokens": 624889573.0, + "step": 24156 + }, + { + "epoch": 2.6528662420382165, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4584033489227295, + "learning_rate": 1e-06, + "loss": 1.1098, + "mean_token_accuracy": 0.6862614154815674, + "num_tokens": 624918369.0, + "step": 24157 + }, + { + "epoch": 2.6529760597408303, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4847824573516846, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7145874500274658, + "num_tokens": 624944122.0, + "step": 24158 + }, + { + "epoch": 2.653085877443444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7252607345581055, + "learning_rate": 1e-06, + "loss": 0.8865, + "mean_token_accuracy": 0.7339611649513245, + "num_tokens": 624964762.0, + "step": 24159 + }, + { + "epoch": 2.6531956951460574, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.855947971343994, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6955240368843079, + "num_tokens": 624989315.0, + "step": 24160 + }, + { + "epoch": 2.653305512848671, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6661458015441895, + "learning_rate": 1e-06, + "loss": 1.1002, + "mean_token_accuracy": 0.6888779997825623, + "num_tokens": 625016020.0, + "step": 24161 + }, + { + "epoch": 2.653415330551285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.296628713607788, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.7110942006111145, + "num_tokens": 625046600.0, + "step": 24162 + }, + { + "epoch": 2.6535251482538986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.544557809829712, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7052726745605469, + "num_tokens": 625073179.0, + "step": 24163 + }, + { + "epoch": 2.6536349659565124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.670123338699341, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.7224222421646118, + "num_tokens": 625100817.0, + "step": 24164 + }, + { + "epoch": 2.6537447836591257, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4807534217834473, + "learning_rate": 1e-06, + "loss": 1.1088, + "mean_token_accuracy": 0.6831867694854736, + "num_tokens": 625132091.0, + "step": 24165 + }, + { + "epoch": 2.6538546013617395, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.706517457962036, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7075194120407104, + "num_tokens": 625156293.0, + "step": 24166 + }, + { + "epoch": 2.653964419064353, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5082924365997314, + "learning_rate": 1e-06, + "loss": 1.0761, + "mean_token_accuracy": 0.6856608986854553, + "num_tokens": 625185119.0, + "step": 24167 + }, + { + "epoch": 2.654074236766967, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4932875633239746, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6997333765029907, + "num_tokens": 625216115.0, + "step": 24168 + }, + { + "epoch": 2.6541840544695807, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6093268394470215, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7050946950912476, + "num_tokens": 625243558.0, + "step": 24169 + }, + { + "epoch": 2.654293872172194, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9783692359924316, + "learning_rate": 1e-06, + "loss": 0.9917, + "mean_token_accuracy": 0.713451623916626, + "num_tokens": 625266309.0, + "step": 24170 + }, + { + "epoch": 2.654403689874808, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5246613025665283, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6855430006980896, + "num_tokens": 625295082.0, + "step": 24171 + }, + { + "epoch": 2.6545135075774215, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6696958541870117, + "learning_rate": 1e-06, + "loss": 0.9193, + "mean_token_accuracy": 0.7285563349723816, + "num_tokens": 625319200.0, + "step": 24172 + }, + { + "epoch": 2.6546233252800353, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5502657890319824, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7330849170684814, + "num_tokens": 625344812.0, + "step": 24173 + }, + { + "epoch": 2.654733142982649, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.433046817779541, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7099108099937439, + "num_tokens": 625371736.0, + "step": 24174 + }, + { + "epoch": 2.6548429606852624, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5967724323272705, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7168819308280945, + "num_tokens": 625400843.0, + "step": 24175 + }, + { + "epoch": 2.654952778387876, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1266069412231445, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.6966047286987305, + "num_tokens": 625421941.0, + "step": 24176 + }, + { + "epoch": 2.65506259609049, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.756976842880249, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7096700072288513, + "num_tokens": 625443964.0, + "step": 24177 + }, + { + "epoch": 2.655172413793103, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.91725754737854, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.722810685634613, + "num_tokens": 625463927.0, + "step": 24178 + }, + { + "epoch": 2.655282231495717, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.673389196395874, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7152611613273621, + "num_tokens": 625492348.0, + "step": 24179 + }, + { + "epoch": 2.6553920491983307, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9396731853485107, + "learning_rate": 1e-06, + "loss": 0.9768, + "mean_token_accuracy": 0.7148860692977905, + "num_tokens": 625513604.0, + "step": 24180 + }, + { + "epoch": 2.6555018669009445, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9374289512634277, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.7109260559082031, + "num_tokens": 625534927.0, + "step": 24181 + }, + { + "epoch": 2.6556116846035582, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7225704193115234, + "learning_rate": 1e-06, + "loss": 0.941, + "mean_token_accuracy": 0.724211573600769, + "num_tokens": 625558477.0, + "step": 24182 + }, + { + "epoch": 2.6557215023061715, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4430673122406006, + "learning_rate": 1e-06, + "loss": 0.981, + "mean_token_accuracy": 0.71394944190979, + "num_tokens": 625585865.0, + "step": 24183 + }, + { + "epoch": 2.6558313200087853, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.751204252243042, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7064954042434692, + "num_tokens": 625610298.0, + "step": 24184 + }, + { + "epoch": 2.655941137711399, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.651838779449463, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7147924304008484, + "num_tokens": 625635271.0, + "step": 24185 + }, + { + "epoch": 2.656050955414013, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.462719440460205, + "learning_rate": 1e-06, + "loss": 1.1093, + "mean_token_accuracy": 0.6790594458580017, + "num_tokens": 625667778.0, + "step": 24186 + }, + { + "epoch": 2.6561607731166266, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5405097007751465, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6976145505905151, + "num_tokens": 625693015.0, + "step": 24187 + }, + { + "epoch": 2.65627059081924, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3588738441467285, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6965197324752808, + "num_tokens": 625726287.0, + "step": 24188 + }, + { + "epoch": 2.6563804085218536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7206931114196777, + "learning_rate": 1e-06, + "loss": 0.995, + "mean_token_accuracy": 0.7081084251403809, + "num_tokens": 625747179.0, + "step": 24189 + }, + { + "epoch": 2.6564902262244674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.751248836517334, + "learning_rate": 1e-06, + "loss": 1.0694, + "mean_token_accuracy": 0.6987816095352173, + "num_tokens": 625772030.0, + "step": 24190 + }, + { + "epoch": 2.656600043927081, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4756460189819336, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.7013592720031738, + "num_tokens": 625800036.0, + "step": 24191 + }, + { + "epoch": 2.656709861629695, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4623219966888428, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7019484639167786, + "num_tokens": 625828947.0, + "step": 24192 + }, + { + "epoch": 2.6568196793323082, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5876433849334717, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.6995482444763184, + "num_tokens": 625855482.0, + "step": 24193 + }, + { + "epoch": 2.656929497034922, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.614016056060791, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7250726222991943, + "num_tokens": 625881288.0, + "step": 24194 + }, + { + "epoch": 2.6570393147375357, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8146321773529053, + "learning_rate": 1e-06, + "loss": 1.0623, + "mean_token_accuracy": 0.6926223635673523, + "num_tokens": 625905574.0, + "step": 24195 + }, + { + "epoch": 2.6571491324401495, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.350020408630371, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.7082476615905762, + "num_tokens": 625934805.0, + "step": 24196 + }, + { + "epoch": 2.6572589501427633, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9391720294952393, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7200107574462891, + "num_tokens": 625957195.0, + "step": 24197 + }, + { + "epoch": 2.6573687678453766, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8832285404205322, + "learning_rate": 1e-06, + "loss": 0.8272, + "mean_token_accuracy": 0.7510353326797485, + "num_tokens": 625976035.0, + "step": 24198 + }, + { + "epoch": 2.6574785855479903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6284682750701904, + "learning_rate": 1e-06, + "loss": 1.0403, + "mean_token_accuracy": 0.6997002363204956, + "num_tokens": 625999073.0, + "step": 24199 + }, + { + "epoch": 2.657588403250604, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6999127864837646, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7039929032325745, + "num_tokens": 626026380.0, + "step": 24200 + }, + { + "epoch": 2.6576982209532174, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6612229347229004, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.6992532014846802, + "num_tokens": 626052800.0, + "step": 24201 + }, + { + "epoch": 2.6578080386558316, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7098183631896973, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.70156329870224, + "num_tokens": 626076458.0, + "step": 24202 + }, + { + "epoch": 2.657917856358445, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.845222234725952, + "learning_rate": 1e-06, + "loss": 0.9512, + "mean_token_accuracy": 0.7233105897903442, + "num_tokens": 626097548.0, + "step": 24203 + }, + { + "epoch": 2.6580276740610587, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5255374908447266, + "learning_rate": 1e-06, + "loss": 0.9566, + "mean_token_accuracy": 0.716429591178894, + "num_tokens": 626121697.0, + "step": 24204 + }, + { + "epoch": 2.6581374917636724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.690554618835449, + "learning_rate": 1e-06, + "loss": 0.9777, + "mean_token_accuracy": 0.7122783064842224, + "num_tokens": 626143533.0, + "step": 24205 + }, + { + "epoch": 2.6582473094662857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4506423473358154, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.6948046684265137, + "num_tokens": 626169981.0, + "step": 24206 + }, + { + "epoch": 2.6583571271688995, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.538668394088745, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.6999874114990234, + "num_tokens": 626197217.0, + "step": 24207 + }, + { + "epoch": 2.6584669448715132, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4644250869750977, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7113175988197327, + "num_tokens": 626229495.0, + "step": 24208 + }, + { + "epoch": 2.658576762574127, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4450714588165283, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.6996886134147644, + "num_tokens": 626257022.0, + "step": 24209 + }, + { + "epoch": 2.6586865802767408, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7114131450653076, + "learning_rate": 1e-06, + "loss": 1.0417, + "mean_token_accuracy": 0.7000323534011841, + "num_tokens": 626280771.0, + "step": 24210 + }, + { + "epoch": 2.658796397979354, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6695005893707275, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6997888684272766, + "num_tokens": 626305088.0, + "step": 24211 + }, + { + "epoch": 2.658906215681968, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8873069286346436, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6923935413360596, + "num_tokens": 626327391.0, + "step": 24212 + }, + { + "epoch": 2.6590160333845816, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6780331134796143, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7177826762199402, + "num_tokens": 626354456.0, + "step": 24213 + }, + { + "epoch": 2.6591258510871953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5666344165802, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7172709703445435, + "num_tokens": 626380975.0, + "step": 24214 + }, + { + "epoch": 2.659235668789809, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4649744033813477, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.724294900894165, + "num_tokens": 626408294.0, + "step": 24215 + }, + { + "epoch": 2.6593454864924224, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6338586807250977, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7120338678359985, + "num_tokens": 626436301.0, + "step": 24216 + }, + { + "epoch": 2.659455304195036, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6595101356506348, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.6980098485946655, + "num_tokens": 626461838.0, + "step": 24217 + }, + { + "epoch": 2.65956512189765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.393925428390503, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6944824457168579, + "num_tokens": 626491759.0, + "step": 24218 + }, + { + "epoch": 2.6596749396002637, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.62966251373291, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7107788920402527, + "num_tokens": 626516598.0, + "step": 24219 + }, + { + "epoch": 2.6597847573028774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.57665753364563, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.6884986758232117, + "num_tokens": 626542760.0, + "step": 24220 + }, + { + "epoch": 2.6598945750054908, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6097912788391113, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7138349413871765, + "num_tokens": 626568063.0, + "step": 24221 + }, + { + "epoch": 2.6600043927081045, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4721171855926514, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7141438722610474, + "num_tokens": 626602420.0, + "step": 24222 + }, + { + "epoch": 2.6601142104107183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.224807024002075, + "learning_rate": 1e-06, + "loss": 0.8982, + "mean_token_accuracy": 0.7335600256919861, + "num_tokens": 626622888.0, + "step": 24223 + }, + { + "epoch": 2.660224028113332, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3773868083953857, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7218250036239624, + "num_tokens": 626655392.0, + "step": 24224 + }, + { + "epoch": 2.660333845815946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3482630252838135, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.6937220096588135, + "num_tokens": 626687978.0, + "step": 24225 + }, + { + "epoch": 2.660443663518559, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.57746958732605, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7107710242271423, + "num_tokens": 626714453.0, + "step": 24226 + }, + { + "epoch": 2.660553481221173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.77705454826355, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7106241583824158, + "num_tokens": 626738129.0, + "step": 24227 + }, + { + "epoch": 2.6606632989237866, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7320783138275146, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.7114360928535461, + "num_tokens": 626762084.0, + "step": 24228 + }, + { + "epoch": 2.6607731166264, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8497588634490967, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.708330512046814, + "num_tokens": 626786533.0, + "step": 24229 + }, + { + "epoch": 2.6608829343290137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.0255074501037598, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7125639915466309, + "num_tokens": 626811012.0, + "step": 24230 + }, + { + "epoch": 2.6609927520316274, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8347573280334473, + "learning_rate": 1e-06, + "loss": 0.9169, + "mean_token_accuracy": 0.7311786413192749, + "num_tokens": 626835039.0, + "step": 24231 + }, + { + "epoch": 2.661102569734241, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.605778932571411, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7070937156677246, + "num_tokens": 626859162.0, + "step": 24232 + }, + { + "epoch": 2.661212387436855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4463887214660645, + "learning_rate": 1e-06, + "loss": 1.0876, + "mean_token_accuracy": 0.6853845715522766, + "num_tokens": 626890271.0, + "step": 24233 + }, + { + "epoch": 2.6613222051394683, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.902442216873169, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7008752226829529, + "num_tokens": 626916816.0, + "step": 24234 + }, + { + "epoch": 2.661432022842082, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4712603092193604, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6912685632705688, + "num_tokens": 626948544.0, + "step": 24235 + }, + { + "epoch": 2.661541840544696, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8554937839508057, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.695790708065033, + "num_tokens": 626969038.0, + "step": 24236 + }, + { + "epoch": 2.6616516582473095, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4423434734344482, + "learning_rate": 1e-06, + "loss": 1.0564, + "mean_token_accuracy": 0.7014281749725342, + "num_tokens": 627001751.0, + "step": 24237 + }, + { + "epoch": 2.6617614759499233, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6559667587280273, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7027784585952759, + "num_tokens": 627025947.0, + "step": 24238 + }, + { + "epoch": 2.6618712936525366, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9543657302856445, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7129799127578735, + "num_tokens": 627045642.0, + "step": 24239 + }, + { + "epoch": 2.6619811113551504, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6335346698760986, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.7018954157829285, + "num_tokens": 627072345.0, + "step": 24240 + }, + { + "epoch": 2.662090929057764, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4268057346343994, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7127599716186523, + "num_tokens": 627100585.0, + "step": 24241 + }, + { + "epoch": 2.662200746760378, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6764867305755615, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6931976079940796, + "num_tokens": 627126318.0, + "step": 24242 + }, + { + "epoch": 2.6623105644629916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4710516929626465, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.6943106651306152, + "num_tokens": 627153684.0, + "step": 24243 + }, + { + "epoch": 2.662420382165605, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4538984298706055, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7048202753067017, + "num_tokens": 627182507.0, + "step": 24244 + }, + { + "epoch": 2.6625301998682187, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.474379539489746, + "learning_rate": 1e-06, + "loss": 0.8589, + "mean_token_accuracy": 0.7465861439704895, + "num_tokens": 627210042.0, + "step": 24245 + }, + { + "epoch": 2.6626400175708325, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4261138439178467, + "learning_rate": 1e-06, + "loss": 0.8887, + "mean_token_accuracy": 0.7387231588363647, + "num_tokens": 627235725.0, + "step": 24246 + }, + { + "epoch": 2.662749835273446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4031577110290527, + "learning_rate": 1e-06, + "loss": 1.04, + "mean_token_accuracy": 0.6986658573150635, + "num_tokens": 627264374.0, + "step": 24247 + }, + { + "epoch": 2.66285965297606, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4517290592193604, + "learning_rate": 1e-06, + "loss": 1.0923, + "mean_token_accuracy": 0.6912797093391418, + "num_tokens": 627292718.0, + "step": 24248 + }, + { + "epoch": 2.6629694706786733, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.397679567337036, + "learning_rate": 1e-06, + "loss": 1.0888, + "mean_token_accuracy": 0.682399332523346, + "num_tokens": 627323275.0, + "step": 24249 + }, + { + "epoch": 2.663079288381287, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7688021659851074, + "learning_rate": 1e-06, + "loss": 0.9456, + "mean_token_accuracy": 0.7230484485626221, + "num_tokens": 627347055.0, + "step": 24250 + }, + { + "epoch": 2.663189106083901, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4333951473236084, + "learning_rate": 1e-06, + "loss": 1.0366, + "mean_token_accuracy": 0.6953125, + "num_tokens": 627375292.0, + "step": 24251 + }, + { + "epoch": 2.663298923786514, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.559399127960205, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6888917684555054, + "num_tokens": 627404543.0, + "step": 24252 + }, + { + "epoch": 2.6634087414891283, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4354782104492188, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.7015566229820251, + "num_tokens": 627435946.0, + "step": 24253 + }, + { + "epoch": 2.6635185591917416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.549862861633301, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6936884522438049, + "num_tokens": 627463374.0, + "step": 24254 + }, + { + "epoch": 2.6636283768943554, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.798784017562866, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7128574848175049, + "num_tokens": 627486138.0, + "step": 24255 + }, + { + "epoch": 2.663738194596969, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5168004035949707, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7080155611038208, + "num_tokens": 627512833.0, + "step": 24256 + }, + { + "epoch": 2.6638480122995825, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4383363723754883, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.6970934867858887, + "num_tokens": 627545399.0, + "step": 24257 + }, + { + "epoch": 2.663957830002196, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4472389221191406, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.6985710859298706, + "num_tokens": 627571910.0, + "step": 24258 + }, + { + "epoch": 2.66406764770481, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.527885913848877, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6996601223945618, + "num_tokens": 627600017.0, + "step": 24259 + }, + { + "epoch": 2.6641774654074237, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.152332305908203, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.705602765083313, + "num_tokens": 627630488.0, + "step": 24260 + }, + { + "epoch": 2.6642872831100375, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.604851722717285, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.7140018939971924, + "num_tokens": 627654978.0, + "step": 24261 + }, + { + "epoch": 2.664397100812651, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7863991260528564, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6964997053146362, + "num_tokens": 627676512.0, + "step": 24262 + }, + { + "epoch": 2.6645069185152646, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5444698333740234, + "learning_rate": 1e-06, + "loss": 1.0879, + "mean_token_accuracy": 0.6851035952568054, + "num_tokens": 627702865.0, + "step": 24263 + }, + { + "epoch": 2.6646167362178783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6915013790130615, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7137379050254822, + "num_tokens": 627729759.0, + "step": 24264 + }, + { + "epoch": 2.664726553920492, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3899543285369873, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.695000410079956, + "num_tokens": 627761297.0, + "step": 24265 + }, + { + "epoch": 2.664836371623106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8190009593963623, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7128080129623413, + "num_tokens": 627783809.0, + "step": 24266 + }, + { + "epoch": 2.664946189325719, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.705516815185547, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7167745232582092, + "num_tokens": 627808469.0, + "step": 24267 + }, + { + "epoch": 2.665056007028333, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5697786808013916, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7246687412261963, + "num_tokens": 627832696.0, + "step": 24268 + }, + { + "epoch": 2.6651658247309467, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.695675849914551, + "learning_rate": 1e-06, + "loss": 0.9298, + "mean_token_accuracy": 0.7226666212081909, + "num_tokens": 627857509.0, + "step": 24269 + }, + { + "epoch": 2.6652756424335604, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5022060871124268, + "learning_rate": 1e-06, + "loss": 0.9378, + "mean_token_accuracy": 0.7286564111709595, + "num_tokens": 627883020.0, + "step": 24270 + }, + { + "epoch": 2.665385460136174, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.664844512939453, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7076327800750732, + "num_tokens": 627907355.0, + "step": 24271 + }, + { + "epoch": 2.6654952778387875, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7468533515930176, + "learning_rate": 1e-06, + "loss": 0.9786, + "mean_token_accuracy": 0.7080544829368591, + "num_tokens": 627934318.0, + "step": 24272 + }, + { + "epoch": 2.6656050955414012, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5910160541534424, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7145829796791077, + "num_tokens": 627962342.0, + "step": 24273 + }, + { + "epoch": 2.665714913244015, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.604107141494751, + "learning_rate": 1e-06, + "loss": 1.0788, + "mean_token_accuracy": 0.6850484609603882, + "num_tokens": 627989770.0, + "step": 24274 + }, + { + "epoch": 2.6658247309466288, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.871938705444336, + "learning_rate": 1e-06, + "loss": 0.8682, + "mean_token_accuracy": 0.7397104501724243, + "num_tokens": 628009019.0, + "step": 24275 + }, + { + "epoch": 2.6659345486492425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2940683364868164, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6807547211647034, + "num_tokens": 628044962.0, + "step": 24276 + }, + { + "epoch": 2.666044366351856, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.646230936050415, + "learning_rate": 1e-06, + "loss": 0.9335, + "mean_token_accuracy": 0.7171200513839722, + "num_tokens": 628069155.0, + "step": 24277 + }, + { + "epoch": 2.6661541840544696, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596202850341797, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7038811445236206, + "num_tokens": 628094640.0, + "step": 24278 + }, + { + "epoch": 2.6662640017570833, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6188395023345947, + "learning_rate": 1e-06, + "loss": 1.0504, + "mean_token_accuracy": 0.6997935771942139, + "num_tokens": 628118746.0, + "step": 24279 + }, + { + "epoch": 2.6663738194596966, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.498530626296997, + "learning_rate": 1e-06, + "loss": 0.8853, + "mean_token_accuracy": 0.7375816106796265, + "num_tokens": 628143374.0, + "step": 24280 + }, + { + "epoch": 2.6664836371623104, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.340158700942993, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.723275899887085, + "num_tokens": 628175194.0, + "step": 24281 + }, + { + "epoch": 2.666593454864924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5274658203125, + "learning_rate": 1e-06, + "loss": 0.9796, + "mean_token_accuracy": 0.7210856676101685, + "num_tokens": 628199375.0, + "step": 24282 + }, + { + "epoch": 2.666703272567538, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5601110458374023, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7176688313484192, + "num_tokens": 628224584.0, + "step": 24283 + }, + { + "epoch": 2.6668130902701517, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.3892178535461426, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.6979537606239319, + "num_tokens": 628246848.0, + "step": 24284 + }, + { + "epoch": 2.666922907972765, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.626073122024536, + "learning_rate": 1e-06, + "loss": 1.1096, + "mean_token_accuracy": 0.681161642074585, + "num_tokens": 628274470.0, + "step": 24285 + }, + { + "epoch": 2.6670327256753787, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.608380079269409, + "learning_rate": 1e-06, + "loss": 0.9323, + "mean_token_accuracy": 0.7245206832885742, + "num_tokens": 628299899.0, + "step": 24286 + }, + { + "epoch": 2.6671425433779925, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.796576499938965, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6920529007911682, + "num_tokens": 628323819.0, + "step": 24287 + }, + { + "epoch": 2.6672523610806063, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.686678409576416, + "learning_rate": 1e-06, + "loss": 1.0053, + "mean_token_accuracy": 0.7135351300239563, + "num_tokens": 628346219.0, + "step": 24288 + }, + { + "epoch": 2.66736217878322, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6343212127685547, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7076864242553711, + "num_tokens": 628370534.0, + "step": 24289 + }, + { + "epoch": 2.6674719964858333, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.555123805999756, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.6972768306732178, + "num_tokens": 628397247.0, + "step": 24290 + }, + { + "epoch": 2.667581814188447, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4731874465942383, + "learning_rate": 1e-06, + "loss": 0.97, + "mean_token_accuracy": 0.7113774418830872, + "num_tokens": 628424134.0, + "step": 24291 + }, + { + "epoch": 2.667691631891061, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.402498483657837, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.729020357131958, + "num_tokens": 628451886.0, + "step": 24292 + }, + { + "epoch": 2.6678014495936746, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.495490074157715, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7095128297805786, + "num_tokens": 628475401.0, + "step": 24293 + }, + { + "epoch": 2.6679112672962884, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.762624502182007, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7056925892829895, + "num_tokens": 628496463.0, + "step": 24294 + }, + { + "epoch": 2.6680210849989017, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.765376329421997, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7203218936920166, + "num_tokens": 628518543.0, + "step": 24295 + }, + { + "epoch": 2.6681309027015154, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.537963628768921, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.6998864412307739, + "num_tokens": 628546183.0, + "step": 24296 + }, + { + "epoch": 2.668240720404129, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.709320306777954, + "learning_rate": 1e-06, + "loss": 1.0012, + "mean_token_accuracy": 0.7082560658454895, + "num_tokens": 628569383.0, + "step": 24297 + }, + { + "epoch": 2.668350538106743, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6489336490631104, + "learning_rate": 1e-06, + "loss": 0.9446, + "mean_token_accuracy": 0.7213104963302612, + "num_tokens": 628592940.0, + "step": 24298 + }, + { + "epoch": 2.6684603558093567, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5587666034698486, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7025585770606995, + "num_tokens": 628620055.0, + "step": 24299 + }, + { + "epoch": 2.66857017351197, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.482598066329956, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.7102266550064087, + "num_tokens": 628645968.0, + "step": 24300 + }, + { + "epoch": 2.6686799912145838, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.637690782546997, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7068655490875244, + "num_tokens": 628667921.0, + "step": 24301 + }, + { + "epoch": 2.6687898089171975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.616591453552246, + "learning_rate": 1e-06, + "loss": 0.9597, + "mean_token_accuracy": 0.7167900800704956, + "num_tokens": 628693110.0, + "step": 24302 + }, + { + "epoch": 2.668899626619811, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7209651470184326, + "learning_rate": 1e-06, + "loss": 1.0568, + "mean_token_accuracy": 0.6955446600914001, + "num_tokens": 628716271.0, + "step": 24303 + }, + { + "epoch": 2.669009444322425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5655746459960938, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.699475884437561, + "num_tokens": 628741671.0, + "step": 24304 + }, + { + "epoch": 2.6691192620250384, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536046028137207, + "learning_rate": 1e-06, + "loss": 1.0882, + "mean_token_accuracy": 0.6912723779678345, + "num_tokens": 628768746.0, + "step": 24305 + }, + { + "epoch": 2.669229079727652, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6808083057403564, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7263625860214233, + "num_tokens": 628792673.0, + "step": 24306 + }, + { + "epoch": 2.669338897430266, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5155625343322754, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.6942747235298157, + "num_tokens": 628817965.0, + "step": 24307 + }, + { + "epoch": 2.669448715132879, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3099544048309326, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6823751330375671, + "num_tokens": 628852368.0, + "step": 24308 + }, + { + "epoch": 2.669558532835493, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7705612182617188, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.7085411548614502, + "num_tokens": 628876487.0, + "step": 24309 + }, + { + "epoch": 2.6696683505381067, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.46490216255188, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7040635943412781, + "num_tokens": 628904968.0, + "step": 24310 + }, + { + "epoch": 2.6697781682407205, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4616827964782715, + "learning_rate": 1e-06, + "loss": 1.1087, + "mean_token_accuracy": 0.6847628355026245, + "num_tokens": 628936609.0, + "step": 24311 + }, + { + "epoch": 2.669887985943334, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7739179134368896, + "learning_rate": 1e-06, + "loss": 0.8723, + "mean_token_accuracy": 0.7388702034950256, + "num_tokens": 628956272.0, + "step": 24312 + }, + { + "epoch": 2.6699978036459475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.653010606765747, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7089576721191406, + "num_tokens": 628978794.0, + "step": 24313 + }, + { + "epoch": 2.6701076213485613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.399557590484619, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7021782398223877, + "num_tokens": 629004138.0, + "step": 24314 + }, + { + "epoch": 2.670217439051175, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.638916492462158, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7056490778923035, + "num_tokens": 629030168.0, + "step": 24315 + }, + { + "epoch": 2.670327256753789, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8262133598327637, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7067030072212219, + "num_tokens": 629057981.0, + "step": 24316 + }, + { + "epoch": 2.6704370744564025, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.451549768447876, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7118929624557495, + "num_tokens": 629086282.0, + "step": 24317 + }, + { + "epoch": 2.670546892159016, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5531699657440186, + "learning_rate": 1e-06, + "loss": 0.9831, + "mean_token_accuracy": 0.7170621156692505, + "num_tokens": 629111573.0, + "step": 24318 + }, + { + "epoch": 2.6706567098616296, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.844312906265259, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7068431377410889, + "num_tokens": 629134667.0, + "step": 24319 + }, + { + "epoch": 2.6707665275642434, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3718583583831787, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6956443786621094, + "num_tokens": 629169421.0, + "step": 24320 + }, + { + "epoch": 2.670876345266857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.4899418354034424, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7217941284179688, + "num_tokens": 629190802.0, + "step": 24321 + }, + { + "epoch": 2.670986162969471, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.695519208908081, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7033427357673645, + "num_tokens": 629214455.0, + "step": 24322 + }, + { + "epoch": 2.671095980672084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.512110471725464, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7182374596595764, + "num_tokens": 629239574.0, + "step": 24323 + }, + { + "epoch": 2.671205798374698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4828970432281494, + "learning_rate": 1e-06, + "loss": 1.0644, + "mean_token_accuracy": 0.6899205446243286, + "num_tokens": 629266362.0, + "step": 24324 + }, + { + "epoch": 2.6713156160773117, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4860198497772217, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7138989567756653, + "num_tokens": 629290907.0, + "step": 24325 + }, + { + "epoch": 2.6714254337799255, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5814738273620605, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7064160108566284, + "num_tokens": 629317945.0, + "step": 24326 + }, + { + "epoch": 2.6715352514825392, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.724468469619751, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7110397815704346, + "num_tokens": 629340176.0, + "step": 24327 + }, + { + "epoch": 2.6716450691851525, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565680503845215, + "learning_rate": 1e-06, + "loss": 1.1092, + "mean_token_accuracy": 0.680281400680542, + "num_tokens": 629367833.0, + "step": 24328 + }, + { + "epoch": 2.6717548868877663, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.5128750801086426, + "learning_rate": 1e-06, + "loss": 1.0648, + "mean_token_accuracy": 0.6968333125114441, + "num_tokens": 629395065.0, + "step": 24329 + }, + { + "epoch": 2.67186470459038, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.589919328689575, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7116196155548096, + "num_tokens": 629420042.0, + "step": 24330 + }, + { + "epoch": 2.6719745222929934, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.076169490814209, + "learning_rate": 1e-06, + "loss": 0.9031, + "mean_token_accuracy": 0.7337294816970825, + "num_tokens": 629438858.0, + "step": 24331 + }, + { + "epoch": 2.672084339995607, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5273330211639404, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7135217189788818, + "num_tokens": 629463482.0, + "step": 24332 + }, + { + "epoch": 2.672194157698221, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565718650817871, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.6977638006210327, + "num_tokens": 629489477.0, + "step": 24333 + }, + { + "epoch": 2.6723039754008346, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.377609968185425, + "learning_rate": 1e-06, + "loss": 0.9181, + "mean_token_accuracy": 0.7254033088684082, + "num_tokens": 629516528.0, + "step": 24334 + }, + { + "epoch": 2.6724137931034484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8018240928649902, + "learning_rate": 1e-06, + "loss": 0.9067, + "mean_token_accuracy": 0.7265498638153076, + "num_tokens": 629538327.0, + "step": 24335 + }, + { + "epoch": 2.6725236108060617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9200551509857178, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7147998809814453, + "num_tokens": 629558109.0, + "step": 24336 + }, + { + "epoch": 2.6726334285086755, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5871658325195312, + "learning_rate": 1e-06, + "loss": 0.9372, + "mean_token_accuracy": 0.7273266315460205, + "num_tokens": 629581221.0, + "step": 24337 + }, + { + "epoch": 2.6727432462112892, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.49579119682312, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6971354484558105, + "num_tokens": 629610989.0, + "step": 24338 + }, + { + "epoch": 2.672853063913903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9929659366607666, + "learning_rate": 1e-06, + "loss": 1.1008, + "mean_token_accuracy": 0.6749088764190674, + "num_tokens": 629639140.0, + "step": 24339 + }, + { + "epoch": 2.6729628816165167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9028329849243164, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7240154147148132, + "num_tokens": 629658917.0, + "step": 24340 + }, + { + "epoch": 2.67307269931913, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8135056495666504, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.716078519821167, + "num_tokens": 629683062.0, + "step": 24341 + }, + { + "epoch": 2.673182517021744, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.288444757461548, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6832908391952515, + "num_tokens": 629717786.0, + "step": 24342 + }, + { + "epoch": 2.6732923347243576, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7770049571990967, + "learning_rate": 1e-06, + "loss": 0.9349, + "mean_token_accuracy": 0.7179719805717468, + "num_tokens": 629738750.0, + "step": 24343 + }, + { + "epoch": 2.6734021524269713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8175318241119385, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.694705069065094, + "num_tokens": 629759552.0, + "step": 24344 + }, + { + "epoch": 2.673511970129585, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4749815464019775, + "learning_rate": 1e-06, + "loss": 0.9858, + "mean_token_accuracy": 0.7194706797599792, + "num_tokens": 629787628.0, + "step": 24345 + }, + { + "epoch": 2.6736217878321984, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 8.791534423828125, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7154374122619629, + "num_tokens": 629812867.0, + "step": 24346 + }, + { + "epoch": 2.673731605534812, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.592169761657715, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7034952640533447, + "num_tokens": 629839150.0, + "step": 24347 + }, + { + "epoch": 2.673841423237426, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6807467937469482, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7036099433898926, + "num_tokens": 629863973.0, + "step": 24348 + }, + { + "epoch": 2.6739512409400397, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6189980506896973, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7085019946098328, + "num_tokens": 629888021.0, + "step": 24349 + }, + { + "epoch": 2.6740610586426534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3928799629211426, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6911990642547607, + "num_tokens": 629920231.0, + "step": 24350 + }, + { + "epoch": 2.6741708763452667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5800318717956543, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7017781734466553, + "num_tokens": 629947745.0, + "step": 24351 + }, + { + "epoch": 2.6742806940478805, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.901381015777588, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.707341194152832, + "num_tokens": 629969679.0, + "step": 24352 + }, + { + "epoch": 2.6743905117504942, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4388012886047363, + "learning_rate": 1e-06, + "loss": 1.079, + "mean_token_accuracy": 0.6879798769950867, + "num_tokens": 630001784.0, + "step": 24353 + }, + { + "epoch": 2.674500329453108, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.424837112426758, + "learning_rate": 1e-06, + "loss": 1.0928, + "mean_token_accuracy": 0.6932056546211243, + "num_tokens": 630033023.0, + "step": 24354 + }, + { + "epoch": 2.6746101471557218, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6946229934692383, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.700693666934967, + "num_tokens": 630059274.0, + "step": 24355 + }, + { + "epoch": 2.674719964858335, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.526268243789673, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6992858648300171, + "num_tokens": 630085439.0, + "step": 24356 + }, + { + "epoch": 2.674829782560949, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8002688884735107, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6899994015693665, + "num_tokens": 630111706.0, + "step": 24357 + }, + { + "epoch": 2.6749396002635626, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5791406631469727, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.703890323638916, + "num_tokens": 630140800.0, + "step": 24358 + }, + { + "epoch": 2.675049417966176, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.464935064315796, + "learning_rate": 1e-06, + "loss": 1.0292, + "mean_token_accuracy": 0.7002861499786377, + "num_tokens": 630170430.0, + "step": 24359 + }, + { + "epoch": 2.6751592356687897, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.704223871231079, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.718173086643219, + "num_tokens": 630192434.0, + "step": 24360 + }, + { + "epoch": 2.6752690533714034, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6419544219970703, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7123923897743225, + "num_tokens": 630216439.0, + "step": 24361 + }, + { + "epoch": 2.675378871074017, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.570401191711426, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6943882703781128, + "num_tokens": 630244907.0, + "step": 24362 + }, + { + "epoch": 2.675488688776631, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4240329265594482, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7089970707893372, + "num_tokens": 630273273.0, + "step": 24363 + }, + { + "epoch": 2.6755985064792442, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6336841583251953, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7029457688331604, + "num_tokens": 630299721.0, + "step": 24364 + }, + { + "epoch": 2.675708324181858, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.564899206161499, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7018500566482544, + "num_tokens": 630325033.0, + "step": 24365 + }, + { + "epoch": 2.6758181418844718, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.574551582336426, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.695614755153656, + "num_tokens": 630351318.0, + "step": 24366 + }, + { + "epoch": 2.6759279595870855, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.533318042755127, + "learning_rate": 1e-06, + "loss": 1.0281, + "mean_token_accuracy": 0.7002535462379456, + "num_tokens": 630378922.0, + "step": 24367 + }, + { + "epoch": 2.6760377772896993, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.610661029815674, + "learning_rate": 1e-06, + "loss": 1.0506, + "mean_token_accuracy": 0.6923025846481323, + "num_tokens": 630408757.0, + "step": 24368 + }, + { + "epoch": 2.6761475949923126, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4583215713500977, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7224319577217102, + "num_tokens": 630435441.0, + "step": 24369 + }, + { + "epoch": 2.6762574126949263, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.460190534591675, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.6924433708190918, + "num_tokens": 630462981.0, + "step": 24370 + }, + { + "epoch": 2.67636723039754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9923174381256104, + "learning_rate": 1e-06, + "loss": 0.9471, + "mean_token_accuracy": 0.7265711426734924, + "num_tokens": 630484003.0, + "step": 24371 + }, + { + "epoch": 2.676477048100154, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7365846633911133, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.7024857997894287, + "num_tokens": 630510191.0, + "step": 24372 + }, + { + "epoch": 2.6765868658027676, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.783285617828369, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7140709161758423, + "num_tokens": 630532773.0, + "step": 24373 + }, + { + "epoch": 2.676696683505381, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6197078227996826, + "learning_rate": 1e-06, + "loss": 0.9464, + "mean_token_accuracy": 0.7193477749824524, + "num_tokens": 630557174.0, + "step": 24374 + }, + { + "epoch": 2.6768065012079947, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7382988929748535, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7193299531936646, + "num_tokens": 630580489.0, + "step": 24375 + }, + { + "epoch": 2.6769163189106084, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4331655502319336, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7040910720825195, + "num_tokens": 630609870.0, + "step": 24376 + }, + { + "epoch": 2.677026136613222, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.867274761199951, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7064576745033264, + "num_tokens": 630630976.0, + "step": 24377 + }, + { + "epoch": 2.677135954315836, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9819250106811523, + "learning_rate": 1e-06, + "loss": 0.9254, + "mean_token_accuracy": 0.7230851650238037, + "num_tokens": 630651152.0, + "step": 24378 + }, + { + "epoch": 2.6772457720184493, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7351057529449463, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6898049712181091, + "num_tokens": 630679285.0, + "step": 24379 + }, + { + "epoch": 2.677355589721063, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.548943042755127, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6963809728622437, + "num_tokens": 630706399.0, + "step": 24380 + }, + { + "epoch": 2.677465407423677, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.551893949508667, + "learning_rate": 1e-06, + "loss": 0.8566, + "mean_token_accuracy": 0.7447494268417358, + "num_tokens": 630732261.0, + "step": 24381 + }, + { + "epoch": 2.67757522512629, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.566291332244873, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7172491550445557, + "num_tokens": 630755814.0, + "step": 24382 + }, + { + "epoch": 2.6776850428289043, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.482698678970337, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7010562419891357, + "num_tokens": 630782783.0, + "step": 24383 + }, + { + "epoch": 2.6777948605315176, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3059816360473633, + "learning_rate": 1e-06, + "loss": 1.0706, + "mean_token_accuracy": 0.6875157356262207, + "num_tokens": 630815627.0, + "step": 24384 + }, + { + "epoch": 2.6779046782341314, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2464187145233154, + "learning_rate": 1e-06, + "loss": 0.9303, + "mean_token_accuracy": 0.7273834347724915, + "num_tokens": 630832948.0, + "step": 24385 + }, + { + "epoch": 2.678014495936745, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5892045497894287, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7133516073226929, + "num_tokens": 630860164.0, + "step": 24386 + }, + { + "epoch": 2.6781243136393584, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5971717834472656, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7034915685653687, + "num_tokens": 630889955.0, + "step": 24387 + }, + { + "epoch": 2.678234131341972, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4788129329681396, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6917909979820251, + "num_tokens": 630917747.0, + "step": 24388 + }, + { + "epoch": 2.678343949044586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5260863304138184, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.697810709476471, + "num_tokens": 630945446.0, + "step": 24389 + }, + { + "epoch": 2.6784537667471997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8985273838043213, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.6823545098304749, + "num_tokens": 630973759.0, + "step": 24390 + }, + { + "epoch": 2.6785635844498135, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.625753402709961, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6965137720108032, + "num_tokens": 631001578.0, + "step": 24391 + }, + { + "epoch": 2.6786734021524268, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.262871503829956, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7054730653762817, + "num_tokens": 631032517.0, + "step": 24392 + }, + { + "epoch": 2.6787832198550405, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8274037837982178, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7056318521499634, + "num_tokens": 631054209.0, + "step": 24393 + }, + { + "epoch": 2.6788930375576543, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.008307933807373, + "learning_rate": 1e-06, + "loss": 0.9281, + "mean_token_accuracy": 0.7264073491096497, + "num_tokens": 631073210.0, + "step": 24394 + }, + { + "epoch": 2.679002855260268, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6270899772644043, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7122735977172852, + "num_tokens": 631099917.0, + "step": 24395 + }, + { + "epoch": 2.679112672962882, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.565835952758789, + "learning_rate": 1e-06, + "loss": 1.1147, + "mean_token_accuracy": 0.677973747253418, + "num_tokens": 631128902.0, + "step": 24396 + }, + { + "epoch": 2.679222490665495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5738463401794434, + "learning_rate": 1e-06, + "loss": 1.0883, + "mean_token_accuracy": 0.685890257358551, + "num_tokens": 631157803.0, + "step": 24397 + }, + { + "epoch": 2.679332308368109, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8323793411254883, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7135071158409119, + "num_tokens": 631177743.0, + "step": 24398 + }, + { + "epoch": 2.6794421260707226, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.694434642791748, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7182525396347046, + "num_tokens": 631201585.0, + "step": 24399 + }, + { + "epoch": 2.6795519437733364, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.235309600830078, + "learning_rate": 1e-06, + "loss": 0.947, + "mean_token_accuracy": 0.7180919647216797, + "num_tokens": 631235262.0, + "step": 24400 + }, + { + "epoch": 2.67966176147595, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.814939022064209, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.697882890701294, + "num_tokens": 631257057.0, + "step": 24401 + }, + { + "epoch": 2.6797715791785635, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3313281536102295, + "learning_rate": 1e-06, + "loss": 1.1513, + "mean_token_accuracy": 0.6915227770805359, + "num_tokens": 631287863.0, + "step": 24402 + }, + { + "epoch": 2.679881396881177, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.59373140335083, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7008440494537354, + "num_tokens": 631312985.0, + "step": 24403 + }, + { + "epoch": 2.679991214583791, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.726423978805542, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7072539329528809, + "num_tokens": 631337562.0, + "step": 24404 + }, + { + "epoch": 2.6801010322864047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.862287759780884, + "learning_rate": 1e-06, + "loss": 0.9275, + "mean_token_accuracy": 0.7221996188163757, + "num_tokens": 631357719.0, + "step": 24405 + }, + { + "epoch": 2.6802108499890185, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.404651641845703, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7088751196861267, + "num_tokens": 631389130.0, + "step": 24406 + }, + { + "epoch": 2.680320667691632, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8200671672821045, + "learning_rate": 1e-06, + "loss": 0.9665, + "mean_token_accuracy": 0.7164744734764099, + "num_tokens": 631410423.0, + "step": 24407 + }, + { + "epoch": 2.6804304853942456, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4302196502685547, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.7265807390213013, + "num_tokens": 631437968.0, + "step": 24408 + }, + { + "epoch": 2.6805403030968593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.498786449432373, + "learning_rate": 1e-06, + "loss": 0.8748, + "mean_token_accuracy": 0.7388979196548462, + "num_tokens": 631463678.0, + "step": 24409 + }, + { + "epoch": 2.6806501207994726, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.290675401687622, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7148460745811462, + "num_tokens": 631494062.0, + "step": 24410 + }, + { + "epoch": 2.6807599385020864, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.713495969772339, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7114819884300232, + "num_tokens": 631519307.0, + "step": 24411 + }, + { + "epoch": 2.6808697562047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7209177017211914, + "learning_rate": 1e-06, + "loss": 0.9705, + "mean_token_accuracy": 0.7187907099723816, + "num_tokens": 631542890.0, + "step": 24412 + }, + { + "epoch": 2.680979573907314, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3364522457122803, + "learning_rate": 1e-06, + "loss": 1.109, + "mean_token_accuracy": 0.6792179346084595, + "num_tokens": 631577584.0, + "step": 24413 + }, + { + "epoch": 2.6810893916099277, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5720577239990234, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.6985945701599121, + "num_tokens": 631602628.0, + "step": 24414 + }, + { + "epoch": 2.681199209312541, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7632758617401123, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7155324220657349, + "num_tokens": 631625565.0, + "step": 24415 + }, + { + "epoch": 2.6813090270151547, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.762024402618408, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7118288278579712, + "num_tokens": 631648466.0, + "step": 24416 + }, + { + "epoch": 2.6814188447177685, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6705923080444336, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7005963325500488, + "num_tokens": 631672970.0, + "step": 24417 + }, + { + "epoch": 2.6815286624203822, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.688657522201538, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.6944146752357483, + "num_tokens": 631700133.0, + "step": 24418 + }, + { + "epoch": 2.681638480122996, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.548766613006592, + "learning_rate": 1e-06, + "loss": 1.0855, + "mean_token_accuracy": 0.6823798418045044, + "num_tokens": 631726860.0, + "step": 24419 + }, + { + "epoch": 2.6817482978256093, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.443741798400879, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7016364336013794, + "num_tokens": 631754956.0, + "step": 24420 + }, + { + "epoch": 2.681858115528223, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6877126693725586, + "learning_rate": 1e-06, + "loss": 0.8914, + "mean_token_accuracy": 0.7334581613540649, + "num_tokens": 631775773.0, + "step": 24421 + }, + { + "epoch": 2.681967933230837, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8633241653442383, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7177499532699585, + "num_tokens": 631798551.0, + "step": 24422 + }, + { + "epoch": 2.6820777509334506, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.65555477142334, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7067855596542358, + "num_tokens": 631826646.0, + "step": 24423 + }, + { + "epoch": 2.6821875686360643, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3408257961273193, + "learning_rate": 1e-06, + "loss": 0.9131, + "mean_token_accuracy": 0.7371320724487305, + "num_tokens": 631858987.0, + "step": 24424 + }, + { + "epoch": 2.6822973863386776, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5910229682922363, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7079349160194397, + "num_tokens": 631882069.0, + "step": 24425 + }, + { + "epoch": 2.6824072040412914, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3906683921813965, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7064732909202576, + "num_tokens": 631911029.0, + "step": 24426 + }, + { + "epoch": 2.682517021743905, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.590571165084839, + "learning_rate": 1e-06, + "loss": 1.1236, + "mean_token_accuracy": 0.6772612929344177, + "num_tokens": 631937513.0, + "step": 24427 + }, + { + "epoch": 2.682626839446519, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5398454666137695, + "learning_rate": 1e-06, + "loss": 0.9102, + "mean_token_accuracy": 0.7251145243644714, + "num_tokens": 631964628.0, + "step": 24428 + }, + { + "epoch": 2.6827366571491327, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0840883255004883, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7028571367263794, + "num_tokens": 631987694.0, + "step": 24429 + }, + { + "epoch": 2.682846474851746, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5756139755249023, + "learning_rate": 1e-06, + "loss": 1.0799, + "mean_token_accuracy": 0.6897114515304565, + "num_tokens": 632018211.0, + "step": 24430 + }, + { + "epoch": 2.6829562925543597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.555850028991699, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6987038850784302, + "num_tokens": 632047901.0, + "step": 24431 + }, + { + "epoch": 2.6830661102569735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.70466947555542, + "learning_rate": 1e-06, + "loss": 0.8834, + "mean_token_accuracy": 0.734004020690918, + "num_tokens": 632068329.0, + "step": 24432 + }, + { + "epoch": 2.683175927959587, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.590998888015747, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7024786472320557, + "num_tokens": 632094069.0, + "step": 24433 + }, + { + "epoch": 2.683285745662201, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.422894239425659, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7050070762634277, + "num_tokens": 632123887.0, + "step": 24434 + }, + { + "epoch": 2.6833955633648143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.240046262741089, + "learning_rate": 1e-06, + "loss": 1.1583, + "mean_token_accuracy": 0.6722766757011414, + "num_tokens": 632151927.0, + "step": 24435 + }, + { + "epoch": 2.683505381067428, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9937868118286133, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7170809507369995, + "num_tokens": 632172273.0, + "step": 24436 + }, + { + "epoch": 2.683615198770042, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5314700603485107, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6998157501220703, + "num_tokens": 632200268.0, + "step": 24437 + }, + { + "epoch": 2.683725016472655, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.470017194747925, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6906604766845703, + "num_tokens": 632232815.0, + "step": 24438 + }, + { + "epoch": 2.683834834175269, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.78082013130188, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7068320512771606, + "num_tokens": 632256901.0, + "step": 24439 + }, + { + "epoch": 2.6839446518778827, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6102066040039062, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7089601159095764, + "num_tokens": 632282897.0, + "step": 24440 + }, + { + "epoch": 2.6840544695804964, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6709232330322266, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7109960317611694, + "num_tokens": 632307063.0, + "step": 24441 + }, + { + "epoch": 2.68416428728311, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.475059986114502, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7096751928329468, + "num_tokens": 632336231.0, + "step": 24442 + }, + { + "epoch": 2.6842741049857235, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4356024265289307, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7153341174125671, + "num_tokens": 632365393.0, + "step": 24443 + }, + { + "epoch": 2.6843839226883373, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.564749240875244, + "learning_rate": 1e-06, + "loss": 1.0997, + "mean_token_accuracy": 0.6795107126235962, + "num_tokens": 632392173.0, + "step": 24444 + }, + { + "epoch": 2.684493740390951, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5389158725738525, + "learning_rate": 1e-06, + "loss": 1.0794, + "mean_token_accuracy": 0.691033661365509, + "num_tokens": 632419194.0, + "step": 24445 + }, + { + "epoch": 2.6846035580935648, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.358855962753296, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7069394588470459, + "num_tokens": 632452581.0, + "step": 24446 + }, + { + "epoch": 2.6847133757961785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.255021333694458, + "learning_rate": 1e-06, + "loss": 1.0918, + "mean_token_accuracy": 0.6840667128562927, + "num_tokens": 632485790.0, + "step": 24447 + }, + { + "epoch": 2.684823193498792, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8571391105651855, + "learning_rate": 1e-06, + "loss": 0.926, + "mean_token_accuracy": 0.7264539003372192, + "num_tokens": 632504942.0, + "step": 24448 + }, + { + "epoch": 2.6849330112014056, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6615540981292725, + "learning_rate": 1e-06, + "loss": 0.9342, + "mean_token_accuracy": 0.7263088822364807, + "num_tokens": 632528163.0, + "step": 24449 + }, + { + "epoch": 2.6850428289040194, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5479822158813477, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.6967409253120422, + "num_tokens": 632556386.0, + "step": 24450 + }, + { + "epoch": 2.685152646606633, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9559218883514404, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7139434218406677, + "num_tokens": 632578355.0, + "step": 24451 + }, + { + "epoch": 2.685262464309247, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.433331251144409, + "learning_rate": 1e-06, + "loss": 0.9011, + "mean_token_accuracy": 0.733532190322876, + "num_tokens": 632603086.0, + "step": 24452 + }, + { + "epoch": 2.68537228201186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5336720943450928, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7109256982803345, + "num_tokens": 632631197.0, + "step": 24453 + }, + { + "epoch": 2.685482099714474, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5416369438171387, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6988694071769714, + "num_tokens": 632659157.0, + "step": 24454 + }, + { + "epoch": 2.6855919174170877, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6092376708984375, + "learning_rate": 1e-06, + "loss": 1.0711, + "mean_token_accuracy": 0.6886979341506958, + "num_tokens": 632684876.0, + "step": 24455 + }, + { + "epoch": 2.6857017351197014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.882465362548828, + "learning_rate": 1e-06, + "loss": 0.9366, + "mean_token_accuracy": 0.7233604192733765, + "num_tokens": 632704324.0, + "step": 24456 + }, + { + "epoch": 2.685811552822315, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4127278327941895, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.7132741808891296, + "num_tokens": 632732603.0, + "step": 24457 + }, + { + "epoch": 2.6859213705249285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.210019826889038, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7023270130157471, + "num_tokens": 632766272.0, + "step": 24458 + }, + { + "epoch": 2.6860311882275423, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4834675788879395, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.7063443660736084, + "num_tokens": 632793284.0, + "step": 24459 + }, + { + "epoch": 2.686141005930156, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4296746253967285, + "learning_rate": 1e-06, + "loss": 1.1433, + "mean_token_accuracy": 0.6762547492980957, + "num_tokens": 632822643.0, + "step": 24460 + }, + { + "epoch": 2.6862508236327693, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5656545162200928, + "learning_rate": 1e-06, + "loss": 0.9967, + "mean_token_accuracy": 0.6971883177757263, + "num_tokens": 632848903.0, + "step": 24461 + }, + { + "epoch": 2.686360641335383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.752333879470825, + "learning_rate": 1e-06, + "loss": 0.9354, + "mean_token_accuracy": 0.7304109334945679, + "num_tokens": 632871012.0, + "step": 24462 + }, + { + "epoch": 2.686470459037997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.615419864654541, + "learning_rate": 1e-06, + "loss": 0.9597, + "mean_token_accuracy": 0.7219228744506836, + "num_tokens": 632893130.0, + "step": 24463 + }, + { + "epoch": 2.6865802767406106, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.425372362136841, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6907711029052734, + "num_tokens": 632921317.0, + "step": 24464 + }, + { + "epoch": 2.6866900944432244, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4927470684051514, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7131458520889282, + "num_tokens": 632949586.0, + "step": 24465 + }, + { + "epoch": 2.6867999121458377, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7584853172302246, + "learning_rate": 1e-06, + "loss": 1.0782, + "mean_token_accuracy": 0.6850812435150146, + "num_tokens": 632975352.0, + "step": 24466 + }, + { + "epoch": 2.6869097298484514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4694347381591797, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7230274677276611, + "num_tokens": 633004096.0, + "step": 24467 + }, + { + "epoch": 2.687019547551065, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.342592477798462, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6868764758110046, + "num_tokens": 633035934.0, + "step": 24468 + }, + { + "epoch": 2.687129365253679, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.321944236755371, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7170600891113281, + "num_tokens": 633066967.0, + "step": 24469 + }, + { + "epoch": 2.6872391829562927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.764660358428955, + "learning_rate": 1e-06, + "loss": 1.027, + "mean_token_accuracy": 0.7026567459106445, + "num_tokens": 633091102.0, + "step": 24470 + }, + { + "epoch": 2.687349000658906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5007872581481934, + "learning_rate": 1e-06, + "loss": 0.9163, + "mean_token_accuracy": 0.7285235524177551, + "num_tokens": 633115293.0, + "step": 24471 + }, + { + "epoch": 2.68745881836152, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.780592918395996, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.713426411151886, + "num_tokens": 633138373.0, + "step": 24472 + }, + { + "epoch": 2.6875686360641335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.68711256980896, + "learning_rate": 1e-06, + "loss": 0.9929, + "mean_token_accuracy": 0.7147555351257324, + "num_tokens": 633162004.0, + "step": 24473 + }, + { + "epoch": 2.6876784537667473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6395208835601807, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.716551661491394, + "num_tokens": 633185919.0, + "step": 24474 + }, + { + "epoch": 2.687788271469361, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4953525066375732, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7034963965415955, + "num_tokens": 633214269.0, + "step": 24475 + }, + { + "epoch": 2.6878980891719744, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4017624855041504, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.7009381055831909, + "num_tokens": 633242678.0, + "step": 24476 + }, + { + "epoch": 2.688007906874588, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7495882511138916, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6938097476959229, + "num_tokens": 633264820.0, + "step": 24477 + }, + { + "epoch": 2.688117724577202, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3201496601104736, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6935151815414429, + "num_tokens": 633298596.0, + "step": 24478 + }, + { + "epoch": 2.6882275422798156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.460949182510376, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7199139595031738, + "num_tokens": 633325792.0, + "step": 24479 + }, + { + "epoch": 2.6883373599824294, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7718143463134766, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7164250612258911, + "num_tokens": 633348353.0, + "step": 24480 + }, + { + "epoch": 2.6884471776850427, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8097527027130127, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7190755605697632, + "num_tokens": 633371318.0, + "step": 24481 + }, + { + "epoch": 2.6885569953876565, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3801560401916504, + "learning_rate": 1e-06, + "loss": 0.9943, + "mean_token_accuracy": 0.707797646522522, + "num_tokens": 633400630.0, + "step": 24482 + }, + { + "epoch": 2.6886668130902702, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5765902996063232, + "learning_rate": 1e-06, + "loss": 0.9003, + "mean_token_accuracy": 0.7341176271438599, + "num_tokens": 633425392.0, + "step": 24483 + }, + { + "epoch": 2.6887766307928835, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7351953983306885, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7076947689056396, + "num_tokens": 633450601.0, + "step": 24484 + }, + { + "epoch": 2.6888864484954977, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.605954647064209, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7130401730537415, + "num_tokens": 633475397.0, + "step": 24485 + }, + { + "epoch": 2.688996266198111, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5521702766418457, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7065753936767578, + "num_tokens": 633500874.0, + "step": 24486 + }, + { + "epoch": 2.689106083900725, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.514998435974121, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6886193752288818, + "num_tokens": 633530432.0, + "step": 24487 + }, + { + "epoch": 2.6892159016033386, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5351357460021973, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7129873633384705, + "num_tokens": 633557693.0, + "step": 24488 + }, + { + "epoch": 2.689325719305952, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4250621795654297, + "learning_rate": 1e-06, + "loss": 0.9474, + "mean_token_accuracy": 0.7209276556968689, + "num_tokens": 633584872.0, + "step": 24489 + }, + { + "epoch": 2.6894355370085656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.916455030441284, + "learning_rate": 1e-06, + "loss": 0.9253, + "mean_token_accuracy": 0.7295025587081909, + "num_tokens": 633605096.0, + "step": 24490 + }, + { + "epoch": 2.6895453547111794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7762300968170166, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7159711718559265, + "num_tokens": 633628223.0, + "step": 24491 + }, + { + "epoch": 2.689655172413793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7640397548675537, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.7118393778800964, + "num_tokens": 633654837.0, + "step": 24492 + }, + { + "epoch": 2.689764990116407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.762016773223877, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.7029116153717041, + "num_tokens": 633678396.0, + "step": 24493 + }, + { + "epoch": 2.68987480781902, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.412883996963501, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6980481147766113, + "num_tokens": 633709747.0, + "step": 24494 + }, + { + "epoch": 2.689984625521634, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.579869270324707, + "learning_rate": 1e-06, + "loss": 1.1049, + "mean_token_accuracy": 0.6789761185646057, + "num_tokens": 633738224.0, + "step": 24495 + }, + { + "epoch": 2.6900944432242477, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.459477663040161, + "learning_rate": 1e-06, + "loss": 1.0963, + "mean_token_accuracy": 0.6812252998352051, + "num_tokens": 633765948.0, + "step": 24496 + }, + { + "epoch": 2.6902042609268615, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6884942054748535, + "learning_rate": 1e-06, + "loss": 0.9735, + "mean_token_accuracy": 0.7172428369522095, + "num_tokens": 633789191.0, + "step": 24497 + }, + { + "epoch": 2.6903140786294752, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0314910411834717, + "learning_rate": 1e-06, + "loss": 0.869, + "mean_token_accuracy": 0.7395660281181335, + "num_tokens": 633806814.0, + "step": 24498 + }, + { + "epoch": 2.6904238963320886, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5068671703338623, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7121333479881287, + "num_tokens": 633836908.0, + "step": 24499 + }, + { + "epoch": 2.6905337140347023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8991003036499023, + "learning_rate": 1e-06, + "loss": 0.9175, + "mean_token_accuracy": 0.7338355779647827, + "num_tokens": 633859780.0, + "step": 24500 + }, + { + "epoch": 2.690643531737316, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.356379508972168, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.6829313635826111, + "num_tokens": 633885859.0, + "step": 24501 + }, + { + "epoch": 2.69075334943993, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8039629459381104, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7064500451087952, + "num_tokens": 633909121.0, + "step": 24502 + }, + { + "epoch": 2.6908631671425436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.438328742980957, + "learning_rate": 1e-06, + "loss": 1.1096, + "mean_token_accuracy": 0.6782512664794922, + "num_tokens": 633938167.0, + "step": 24503 + }, + { + "epoch": 2.690972984845157, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4818146228790283, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7228796482086182, + "num_tokens": 633965109.0, + "step": 24504 + }, + { + "epoch": 2.6910828025477707, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6935081481933594, + "learning_rate": 1e-06, + "loss": 1.0116, + "mean_token_accuracy": 0.7061843276023865, + "num_tokens": 633988185.0, + "step": 24505 + }, + { + "epoch": 2.6911926202503844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.782276153564453, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.6921166181564331, + "num_tokens": 634012442.0, + "step": 24506 + }, + { + "epoch": 2.691302437952998, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7448344230651855, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7188681364059448, + "num_tokens": 634036849.0, + "step": 24507 + }, + { + "epoch": 2.691412255655612, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6636502742767334, + "learning_rate": 1e-06, + "loss": 1.0803, + "mean_token_accuracy": 0.6933348774909973, + "num_tokens": 634061913.0, + "step": 24508 + }, + { + "epoch": 2.6915220733582252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.858304977416992, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.6813665628433228, + "num_tokens": 634088505.0, + "step": 24509 + }, + { + "epoch": 2.691631891060839, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7088687419891357, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7033888101577759, + "num_tokens": 634119812.0, + "step": 24510 + }, + { + "epoch": 2.6917417087634528, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6358072757720947, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.7136977910995483, + "num_tokens": 634144470.0, + "step": 24511 + }, + { + "epoch": 2.691851526466066, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.303676128387451, + "learning_rate": 1e-06, + "loss": 1.0884, + "mean_token_accuracy": 0.6971130967140198, + "num_tokens": 634178419.0, + "step": 24512 + }, + { + "epoch": 2.69196134416868, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.603484630584717, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.6991510987281799, + "num_tokens": 634204048.0, + "step": 24513 + }, + { + "epoch": 2.6920711618712936, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8915226459503174, + "learning_rate": 1e-06, + "loss": 0.9439, + "mean_token_accuracy": 0.7324462532997131, + "num_tokens": 634224837.0, + "step": 24514 + }, + { + "epoch": 2.6921809795739073, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.732824325561523, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.6978524327278137, + "num_tokens": 634251640.0, + "step": 24515 + }, + { + "epoch": 2.692290797276521, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7394697666168213, + "learning_rate": 1e-06, + "loss": 1.1763, + "mean_token_accuracy": 0.6659101843833923, + "num_tokens": 634285192.0, + "step": 24516 + }, + { + "epoch": 2.6924006149791344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5238378047943115, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.7069556713104248, + "num_tokens": 634313160.0, + "step": 24517 + }, + { + "epoch": 2.692510432681748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7449097633361816, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7338082790374756, + "num_tokens": 634334500.0, + "step": 24518 + }, + { + "epoch": 2.692620250384362, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3887252807617188, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7050015926361084, + "num_tokens": 634361753.0, + "step": 24519 + }, + { + "epoch": 2.6927300680869757, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.549428939819336, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7210001945495605, + "num_tokens": 634387044.0, + "step": 24520 + }, + { + "epoch": 2.6928398857895894, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.441866397857666, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6915459632873535, + "num_tokens": 634416653.0, + "step": 24521 + }, + { + "epoch": 2.6929497034922028, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.690089702606201, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7190361022949219, + "num_tokens": 634440434.0, + "step": 24522 + }, + { + "epoch": 2.6930595211948165, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6417737007141113, + "learning_rate": 1e-06, + "loss": 1.0356, + "mean_token_accuracy": 0.7011715769767761, + "num_tokens": 634466287.0, + "step": 24523 + }, + { + "epoch": 2.6931693388974303, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.520822763442993, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.7195385098457336, + "num_tokens": 634491980.0, + "step": 24524 + }, + { + "epoch": 2.693279156600044, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5935752391815186, + "learning_rate": 1e-06, + "loss": 0.9472, + "mean_token_accuracy": 0.718387246131897, + "num_tokens": 634515161.0, + "step": 24525 + }, + { + "epoch": 2.693388974302658, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8184258937835693, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.7049810290336609, + "num_tokens": 634536852.0, + "step": 24526 + }, + { + "epoch": 2.693498792005271, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7534024715423584, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7161129713058472, + "num_tokens": 634561649.0, + "step": 24527 + }, + { + "epoch": 2.693608609707885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6123199462890625, + "learning_rate": 1e-06, + "loss": 0.9422, + "mean_token_accuracy": 0.7264487147331238, + "num_tokens": 634586384.0, + "step": 24528 + }, + { + "epoch": 2.6937184274104986, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.405834197998047, + "learning_rate": 1e-06, + "loss": 1.0096, + "mean_token_accuracy": 0.7077968120574951, + "num_tokens": 634617927.0, + "step": 24529 + }, + { + "epoch": 2.6938282451131124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4245240688323975, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7215784192085266, + "num_tokens": 634644921.0, + "step": 24530 + }, + { + "epoch": 2.693938062815726, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.919710397720337, + "learning_rate": 1e-06, + "loss": 0.9424, + "mean_token_accuracy": 0.7300810813903809, + "num_tokens": 634665892.0, + "step": 24531 + }, + { + "epoch": 2.6940478805183394, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1894922256469727, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7149728536605835, + "num_tokens": 634696236.0, + "step": 24532 + }, + { + "epoch": 2.694157698220953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.556699752807617, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.716812789440155, + "num_tokens": 634725664.0, + "step": 24533 + }, + { + "epoch": 2.694267515923567, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6174445152282715, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7115116119384766, + "num_tokens": 634749180.0, + "step": 24534 + }, + { + "epoch": 2.6943773336261807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.791067123413086, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7157340049743652, + "num_tokens": 634769158.0, + "step": 24535 + }, + { + "epoch": 2.6944871513287945, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7308263778686523, + "learning_rate": 1e-06, + "loss": 1.0541, + "mean_token_accuracy": 0.706015944480896, + "num_tokens": 634791663.0, + "step": 24536 + }, + { + "epoch": 2.6945969690314078, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7647852897644043, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7092243432998657, + "num_tokens": 634817045.0, + "step": 24537 + }, + { + "epoch": 2.6947067867340215, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.899038791656494, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7095484137535095, + "num_tokens": 634838364.0, + "step": 24538 + }, + { + "epoch": 2.6948166044366353, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.440288782119751, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6882007718086243, + "num_tokens": 634867696.0, + "step": 24539 + }, + { + "epoch": 2.6949264221392486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.05147385597229, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.7051827311515808, + "num_tokens": 634887761.0, + "step": 24540 + }, + { + "epoch": 2.6950362398418624, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6484591960906982, + "learning_rate": 1e-06, + "loss": 0.9997, + "mean_token_accuracy": 0.714352011680603, + "num_tokens": 634910623.0, + "step": 24541 + }, + { + "epoch": 2.695146057544476, + "ewc_loss": 1.0788440704345703e-05, + "grad_norm": 3.5432989597320557, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.707727313041687, + "num_tokens": 634933794.0, + "step": 24542 + }, + { + "epoch": 2.69525587524709, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.560967206954956, + "learning_rate": 1e-06, + "loss": 1.1194, + "mean_token_accuracy": 0.6798030734062195, + "num_tokens": 634963445.0, + "step": 24543 + }, + { + "epoch": 2.6953656929497036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7519731521606445, + "learning_rate": 1e-06, + "loss": 0.9605, + "mean_token_accuracy": 0.7209526300430298, + "num_tokens": 634984866.0, + "step": 24544 + }, + { + "epoch": 2.695475510652317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.576617479324341, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.715552031993866, + "num_tokens": 635012109.0, + "step": 24545 + }, + { + "epoch": 2.6955853283549307, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5445821285247803, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.6989452242851257, + "num_tokens": 635040466.0, + "step": 24546 + }, + { + "epoch": 2.6956951460575445, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6667630672454834, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7135207653045654, + "num_tokens": 635064335.0, + "step": 24547 + }, + { + "epoch": 2.695804963760158, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.515273094177246, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6998206377029419, + "num_tokens": 635091295.0, + "step": 24548 + }, + { + "epoch": 2.695914781462772, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.384166955947876, + "learning_rate": 1e-06, + "loss": 1.0543, + "mean_token_accuracy": 0.6921274065971375, + "num_tokens": 635122405.0, + "step": 24549 + }, + { + "epoch": 2.6960245991653853, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.84097957611084, + "learning_rate": 1e-06, + "loss": 0.8935, + "mean_token_accuracy": 0.7340365648269653, + "num_tokens": 635143309.0, + "step": 24550 + }, + { + "epoch": 2.696134416867999, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.445377826690674, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7132806777954102, + "num_tokens": 635172218.0, + "step": 24551 + }, + { + "epoch": 2.696244234570613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2517998218536377, + "learning_rate": 1e-06, + "loss": 1.0847, + "mean_token_accuracy": 0.6839176416397095, + "num_tokens": 635206805.0, + "step": 24552 + }, + { + "epoch": 2.6963540522732266, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6429004669189453, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.707211971282959, + "num_tokens": 635233987.0, + "step": 24553 + }, + { + "epoch": 2.6964638699758403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.837825298309326, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.687565803527832, + "num_tokens": 635257790.0, + "step": 24554 + }, + { + "epoch": 2.6965736876784536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.836090326309204, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7115492224693298, + "num_tokens": 635280021.0, + "step": 24555 + }, + { + "epoch": 2.6966835053810674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4362077713012695, + "learning_rate": 1e-06, + "loss": 1.0631, + "mean_token_accuracy": 0.691717267036438, + "num_tokens": 635309456.0, + "step": 24556 + }, + { + "epoch": 2.696793323083681, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7900960445404053, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7123104333877563, + "num_tokens": 635331597.0, + "step": 24557 + }, + { + "epoch": 2.696903140786295, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.894603729248047, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7096740007400513, + "num_tokens": 635352324.0, + "step": 24558 + }, + { + "epoch": 2.6970129584889087, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.7516391277313232, + "learning_rate": 1e-06, + "loss": 0.9379, + "mean_token_accuracy": 0.7206497192382812, + "num_tokens": 635373940.0, + "step": 24559 + }, + { + "epoch": 2.697122776191522, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.336925983428955, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.6968311071395874, + "num_tokens": 635407931.0, + "step": 24560 + }, + { + "epoch": 2.6972325938941357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4727587699890137, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7045489549636841, + "num_tokens": 635437344.0, + "step": 24561 + }, + { + "epoch": 2.6973424115967495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6901371479034424, + "learning_rate": 1e-06, + "loss": 1.049, + "mean_token_accuracy": 0.6949172019958496, + "num_tokens": 635462421.0, + "step": 24562 + }, + { + "epoch": 2.697452229299363, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.909433603286743, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.722545325756073, + "num_tokens": 635481273.0, + "step": 24563 + }, + { + "epoch": 2.697562047001977, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5968172550201416, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7029949426651001, + "num_tokens": 635505387.0, + "step": 24564 + }, + { + "epoch": 2.6976718647045903, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5410587787628174, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.6894932985305786, + "num_tokens": 635533811.0, + "step": 24565 + }, + { + "epoch": 2.697781682407204, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.558694362640381, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7079643607139587, + "num_tokens": 635561271.0, + "step": 24566 + }, + { + "epoch": 2.697891500109818, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.640105962753296, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.6998568773269653, + "num_tokens": 635584208.0, + "step": 24567 + }, + { + "epoch": 2.698001317812431, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6315786838531494, + "learning_rate": 1e-06, + "loss": 0.9226, + "mean_token_accuracy": 0.7314469218254089, + "num_tokens": 635611111.0, + "step": 24568 + }, + { + "epoch": 2.698111135515045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5394692420959473, + "learning_rate": 1e-06, + "loss": 0.9841, + "mean_token_accuracy": 0.7168274521827698, + "num_tokens": 635638897.0, + "step": 24569 + }, + { + "epoch": 2.6982209532176586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.35795259475708, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7117972373962402, + "num_tokens": 635670072.0, + "step": 24570 + }, + { + "epoch": 2.6983307709202724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9389896392822266, + "learning_rate": 1e-06, + "loss": 0.9218, + "mean_token_accuracy": 0.7251662015914917, + "num_tokens": 635688613.0, + "step": 24571 + }, + { + "epoch": 2.698440588622886, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7125468254089355, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7155725359916687, + "num_tokens": 635711296.0, + "step": 24572 + }, + { + "epoch": 2.6985504063254995, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5414397716522217, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7100240588188171, + "num_tokens": 635741297.0, + "step": 24573 + }, + { + "epoch": 2.6986602240281132, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.359809160232544, + "learning_rate": 1e-06, + "loss": 1.1036, + "mean_token_accuracy": 0.6977609395980835, + "num_tokens": 635776310.0, + "step": 24574 + }, + { + "epoch": 2.698770041730727, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.879565477371216, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.7029946446418762, + "num_tokens": 635799526.0, + "step": 24575 + }, + { + "epoch": 2.6988798594333407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.493424892425537, + "learning_rate": 1e-06, + "loss": 1.0615, + "mean_token_accuracy": 0.6947714686393738, + "num_tokens": 635827342.0, + "step": 24576 + }, + { + "epoch": 2.6989896771359545, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.588304042816162, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7090677618980408, + "num_tokens": 635853904.0, + "step": 24577 + }, + { + "epoch": 2.699099494838568, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.372598886489868, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7078484892845154, + "num_tokens": 635882300.0, + "step": 24578 + }, + { + "epoch": 2.6992093125411816, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4228014945983887, + "learning_rate": 1e-06, + "loss": 0.958, + "mean_token_accuracy": 0.7213301658630371, + "num_tokens": 635911264.0, + "step": 24579 + }, + { + "epoch": 2.6993191302437953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5358450412750244, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7113682627677917, + "num_tokens": 635937409.0, + "step": 24580 + }, + { + "epoch": 2.699428947946409, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5631825923919678, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.7016851305961609, + "num_tokens": 635962176.0, + "step": 24581 + }, + { + "epoch": 2.699538765649023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.92505145072937, + "learning_rate": 1e-06, + "loss": 0.9636, + "mean_token_accuracy": 0.7184423208236694, + "num_tokens": 635985408.0, + "step": 24582 + }, + { + "epoch": 2.699648583351636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7804088592529297, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7220081090927124, + "num_tokens": 636008460.0, + "step": 24583 + }, + { + "epoch": 2.69975840105425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.747291326522827, + "learning_rate": 1e-06, + "loss": 1.0552, + "mean_token_accuracy": 0.6938042044639587, + "num_tokens": 636033600.0, + "step": 24584 + }, + { + "epoch": 2.6998682187568637, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.43630051612854, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6807352304458618, + "num_tokens": 636064531.0, + "step": 24585 + }, + { + "epoch": 2.6999780364594774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6286239624023438, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7113890647888184, + "num_tokens": 636088199.0, + "step": 24586 + }, + { + "epoch": 2.700087854162091, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.56339955329895, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.712646484375, + "num_tokens": 636114960.0, + "step": 24587 + }, + { + "epoch": 2.7001976718647045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4692304134368896, + "learning_rate": 1e-06, + "loss": 1.0967, + "mean_token_accuracy": 0.6853215098381042, + "num_tokens": 636142518.0, + "step": 24588 + }, + { + "epoch": 2.7003074895673183, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4771358966827393, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7103224396705627, + "num_tokens": 636170558.0, + "step": 24589 + }, + { + "epoch": 2.700417307269932, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.759631872177124, + "learning_rate": 1e-06, + "loss": 0.96, + "mean_token_accuracy": 0.7118451595306396, + "num_tokens": 636195178.0, + "step": 24590 + }, + { + "epoch": 2.7005271249725453, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.823307991027832, + "learning_rate": 1e-06, + "loss": 1.1135, + "mean_token_accuracy": 0.6788216829299927, + "num_tokens": 636219507.0, + "step": 24591 + }, + { + "epoch": 2.700636942675159, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.439972162246704, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.6960239410400391, + "num_tokens": 636249262.0, + "step": 24592 + }, + { + "epoch": 2.700746760377773, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4316294193267822, + "learning_rate": 1e-06, + "loss": 1.1674, + "mean_token_accuracy": 0.6731691956520081, + "num_tokens": 636282270.0, + "step": 24593 + }, + { + "epoch": 2.7008565780803866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.577881336212158, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6852807998657227, + "num_tokens": 636309437.0, + "step": 24594 + }, + { + "epoch": 2.7009663957830004, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8330912590026855, + "learning_rate": 1e-06, + "loss": 0.9535, + "mean_token_accuracy": 0.7243294715881348, + "num_tokens": 636334045.0, + "step": 24595 + }, + { + "epoch": 2.7010762134856137, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7634754180908203, + "learning_rate": 1e-06, + "loss": 1.0289, + "mean_token_accuracy": 0.7053090333938599, + "num_tokens": 636356655.0, + "step": 24596 + }, + { + "epoch": 2.7011860311882274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6030123233795166, + "learning_rate": 1e-06, + "loss": 0.9246, + "mean_token_accuracy": 0.7303369045257568, + "num_tokens": 636380594.0, + "step": 24597 + }, + { + "epoch": 2.701295848890841, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4878642559051514, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.7014297246932983, + "num_tokens": 636409274.0, + "step": 24598 + }, + { + "epoch": 2.701405666593455, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7533953189849854, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7193238735198975, + "num_tokens": 636431881.0, + "step": 24599 + }, + { + "epoch": 2.7015154842960687, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6493725776672363, + "learning_rate": 1e-06, + "loss": 1.0435, + "mean_token_accuracy": 0.6948254108428955, + "num_tokens": 636456702.0, + "step": 24600 + }, + { + "epoch": 2.701625301998682, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7509610652923584, + "learning_rate": 1e-06, + "loss": 0.8877, + "mean_token_accuracy": 0.7408197522163391, + "num_tokens": 636477902.0, + "step": 24601 + }, + { + "epoch": 2.7017351197012958, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4858901500701904, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7059364318847656, + "num_tokens": 636506103.0, + "step": 24602 + }, + { + "epoch": 2.7018449374039095, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.485278367996216, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7091206312179565, + "num_tokens": 636534483.0, + "step": 24603 + }, + { + "epoch": 2.7019547551065233, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.594038486480713, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7143548727035522, + "num_tokens": 636560499.0, + "step": 24604 + }, + { + "epoch": 2.702064572809137, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7092292308807373, + "learning_rate": 1e-06, + "loss": 1.0515, + "mean_token_accuracy": 0.7037583589553833, + "num_tokens": 636587021.0, + "step": 24605 + }, + { + "epoch": 2.7021743905117503, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5712645053863525, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7102967500686646, + "num_tokens": 636611808.0, + "step": 24606 + }, + { + "epoch": 2.702284208214364, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.349907875061035, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.7057663202285767, + "num_tokens": 636642235.0, + "step": 24607 + }, + { + "epoch": 2.702394025916978, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.772313117980957, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.7000983953475952, + "num_tokens": 636668339.0, + "step": 24608 + }, + { + "epoch": 2.7025038436195916, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.60154390335083, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.6978253126144409, + "num_tokens": 636694031.0, + "step": 24609 + }, + { + "epoch": 2.7026136613222054, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.434466600418091, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.714409351348877, + "num_tokens": 636721196.0, + "step": 24610 + }, + { + "epoch": 2.7027234790248187, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.545379877090454, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7005173563957214, + "num_tokens": 636744077.0, + "step": 24611 + }, + { + "epoch": 2.7028332967274324, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6258764266967773, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.701861560344696, + "num_tokens": 636771047.0, + "step": 24612 + }, + { + "epoch": 2.702943114430046, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4336037635803223, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7215949296951294, + "num_tokens": 636797984.0, + "step": 24613 + }, + { + "epoch": 2.7030529321326595, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7364394664764404, + "learning_rate": 1e-06, + "loss": 0.9704, + "mean_token_accuracy": 0.7151626348495483, + "num_tokens": 636820040.0, + "step": 24614 + }, + { + "epoch": 2.7031627498352737, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5018246173858643, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7259774208068848, + "num_tokens": 636847213.0, + "step": 24615 + }, + { + "epoch": 2.703272567537887, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.574082612991333, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7094234824180603, + "num_tokens": 636873425.0, + "step": 24616 + }, + { + "epoch": 2.703382385240501, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6957690715789795, + "learning_rate": 1e-06, + "loss": 0.9849, + "mean_token_accuracy": 0.7094617486000061, + "num_tokens": 636895224.0, + "step": 24617 + }, + { + "epoch": 2.7034922029431145, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.468412160873413, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6864158511161804, + "num_tokens": 636925505.0, + "step": 24618 + }, + { + "epoch": 2.703602020645728, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4598801136016846, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7059999704360962, + "num_tokens": 636950985.0, + "step": 24619 + }, + { + "epoch": 2.7037118383483416, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.886202335357666, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7132536172866821, + "num_tokens": 636972902.0, + "step": 24620 + }, + { + "epoch": 2.7038216560509554, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.648836135864258, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7038394808769226, + "num_tokens": 636995027.0, + "step": 24621 + }, + { + "epoch": 2.703931473753569, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4027767181396484, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6902275085449219, + "num_tokens": 637025443.0, + "step": 24622 + }, + { + "epoch": 2.704041291456183, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.519782781600952, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7117436528205872, + "num_tokens": 637053295.0, + "step": 24623 + }, + { + "epoch": 2.704151109158796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.764273166656494, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7106842994689941, + "num_tokens": 637078599.0, + "step": 24624 + }, + { + "epoch": 2.70426092686141, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.344125747680664, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7000745534896851, + "num_tokens": 637109623.0, + "step": 24625 + }, + { + "epoch": 2.7043707445640237, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6460044384002686, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7041680216789246, + "num_tokens": 637132960.0, + "step": 24626 + }, + { + "epoch": 2.7044805622666375, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.693681478500366, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7193242907524109, + "num_tokens": 637159563.0, + "step": 24627 + }, + { + "epoch": 2.7045903799692512, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 4.558267593383789, + "learning_rate": 1e-06, + "loss": 0.9473, + "mean_token_accuracy": 0.721792459487915, + "num_tokens": 637189795.0, + "step": 24628 + }, + { + "epoch": 2.7047001976718645, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.481523036956787, + "learning_rate": 1e-06, + "loss": 1.0562, + "mean_token_accuracy": 0.6953638792037964, + "num_tokens": 637217854.0, + "step": 24629 + }, + { + "epoch": 2.7048100153744783, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3708655834198, + "learning_rate": 1e-06, + "loss": 0.8626, + "mean_token_accuracy": 0.7391735315322876, + "num_tokens": 637246068.0, + "step": 24630 + }, + { + "epoch": 2.704919833077092, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3596184253692627, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6927773952484131, + "num_tokens": 637275907.0, + "step": 24631 + }, + { + "epoch": 2.705029650779706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6927223205566406, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.704433798789978, + "num_tokens": 637298231.0, + "step": 24632 + }, + { + "epoch": 2.7051394684823196, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.452202558517456, + "learning_rate": 1e-06, + "loss": 0.9258, + "mean_token_accuracy": 0.719549298286438, + "num_tokens": 637324818.0, + "step": 24633 + }, + { + "epoch": 2.705249286184933, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3943586349487305, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7210161685943604, + "num_tokens": 637351693.0, + "step": 24634 + }, + { + "epoch": 2.7053591038875466, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6912992000579834, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.725874662399292, + "num_tokens": 637376108.0, + "step": 24635 + }, + { + "epoch": 2.7054689215901604, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.821547746658325, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.7036978602409363, + "num_tokens": 637402645.0, + "step": 24636 + }, + { + "epoch": 2.705578739292774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6587958335876465, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6964622735977173, + "num_tokens": 637427056.0, + "step": 24637 + }, + { + "epoch": 2.705688556995388, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4669907093048096, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6995006799697876, + "num_tokens": 637457793.0, + "step": 24638 + }, + { + "epoch": 2.705798374698001, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.563110828399658, + "learning_rate": 1e-06, + "loss": 0.9577, + "mean_token_accuracy": 0.7249168157577515, + "num_tokens": 637481962.0, + "step": 24639 + }, + { + "epoch": 2.705908192400615, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8205151557922363, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7226202487945557, + "num_tokens": 637502153.0, + "step": 24640 + }, + { + "epoch": 2.7060180101032287, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4755141735076904, + "learning_rate": 1e-06, + "loss": 1.0816, + "mean_token_accuracy": 0.6918985843658447, + "num_tokens": 637529072.0, + "step": 24641 + }, + { + "epoch": 2.706127827805842, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.904499053955078, + "learning_rate": 1e-06, + "loss": 0.9376, + "mean_token_accuracy": 0.726786196231842, + "num_tokens": 637549332.0, + "step": 24642 + }, + { + "epoch": 2.706237645508456, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8611721992492676, + "learning_rate": 1e-06, + "loss": 0.9134, + "mean_token_accuracy": 0.7364287972450256, + "num_tokens": 637571205.0, + "step": 24643 + }, + { + "epoch": 2.7063474632110696, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.56512188911438, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6974745988845825, + "num_tokens": 637599018.0, + "step": 24644 + }, + { + "epoch": 2.7064572809136833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.681058406829834, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6938391923904419, + "num_tokens": 637626618.0, + "step": 24645 + }, + { + "epoch": 2.706567098616297, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3527729511260986, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7033287286758423, + "num_tokens": 637655494.0, + "step": 24646 + }, + { + "epoch": 2.7066769163189104, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4948878288269043, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.7070109248161316, + "num_tokens": 637680457.0, + "step": 24647 + }, + { + "epoch": 2.706786734021524, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.492396593093872, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7099704146385193, + "num_tokens": 637708205.0, + "step": 24648 + }, + { + "epoch": 2.706896551724138, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.4999310970306396, + "learning_rate": 1e-06, + "loss": 0.9312, + "mean_token_accuracy": 0.7252828478813171, + "num_tokens": 637725835.0, + "step": 24649 + }, + { + "epoch": 2.7070063694267517, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.299574613571167, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7078030109405518, + "num_tokens": 637757318.0, + "step": 24650 + }, + { + "epoch": 2.7071161871293654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.338064193725586, + "learning_rate": 1e-06, + "loss": 1.0321, + "mean_token_accuracy": 0.7022203803062439, + "num_tokens": 637786045.0, + "step": 24651 + }, + { + "epoch": 2.7072260048319787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.498809814453125, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7109456062316895, + "num_tokens": 637814344.0, + "step": 24652 + }, + { + "epoch": 2.7073358225345925, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4226388931274414, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.697088360786438, + "num_tokens": 637851542.0, + "step": 24653 + }, + { + "epoch": 2.7074456402372062, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.481041193008423, + "learning_rate": 1e-06, + "loss": 1.1266, + "mean_token_accuracy": 0.675704836845398, + "num_tokens": 637880353.0, + "step": 24654 + }, + { + "epoch": 2.70755545793982, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5365946292877197, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7023961544036865, + "num_tokens": 637905596.0, + "step": 24655 + }, + { + "epoch": 2.7076652756424338, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5702664852142334, + "learning_rate": 1e-06, + "loss": 1.0036, + "mean_token_accuracy": 0.7033362984657288, + "num_tokens": 637930958.0, + "step": 24656 + }, + { + "epoch": 2.707775093345047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8214540481567383, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7180944681167603, + "num_tokens": 637952486.0, + "step": 24657 + }, + { + "epoch": 2.707884911047661, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.688311815261841, + "learning_rate": 1e-06, + "loss": 1.1257, + "mean_token_accuracy": 0.6848121881484985, + "num_tokens": 637980785.0, + "step": 24658 + }, + { + "epoch": 2.7079947287502746, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.329737663269043, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6964613199234009, + "num_tokens": 638012640.0, + "step": 24659 + }, + { + "epoch": 2.7081045464528883, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.624025344848633, + "learning_rate": 1e-06, + "loss": 0.9392, + "mean_token_accuracy": 0.726564884185791, + "num_tokens": 638035306.0, + "step": 24660 + }, + { + "epoch": 2.708214364155502, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.448657274246216, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7073836326599121, + "num_tokens": 638064193.0, + "step": 24661 + }, + { + "epoch": 2.7083241818581154, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6037395000457764, + "learning_rate": 1e-06, + "loss": 0.9355, + "mean_token_accuracy": 0.7234814167022705, + "num_tokens": 638088972.0, + "step": 24662 + }, + { + "epoch": 2.708433999560729, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.806960105895996, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7147073149681091, + "num_tokens": 638110260.0, + "step": 24663 + }, + { + "epoch": 2.708543817263343, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.766206741333008, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.7109826803207397, + "num_tokens": 638132932.0, + "step": 24664 + }, + { + "epoch": 2.7086536349659562, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.641308069229126, + "learning_rate": 1e-06, + "loss": 1.1405, + "mean_token_accuracy": 0.6804038882255554, + "num_tokens": 638159454.0, + "step": 24665 + }, + { + "epoch": 2.7087634526685704, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5965487957000732, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.714887797832489, + "num_tokens": 638183545.0, + "step": 24666 + }, + { + "epoch": 2.7088732703711838, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3549680709838867, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.7065896987915039, + "num_tokens": 638215187.0, + "step": 24667 + }, + { + "epoch": 2.7089830880737975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.701465606689453, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7025817632675171, + "num_tokens": 638240218.0, + "step": 24668 + }, + { + "epoch": 2.7090929057764113, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505563259124756, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7092812061309814, + "num_tokens": 638271344.0, + "step": 24669 + }, + { + "epoch": 2.7092027234790246, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3792335987091064, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.704588770866394, + "num_tokens": 638303577.0, + "step": 24670 + }, + { + "epoch": 2.7093125411816383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4905951023101807, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6922544240951538, + "num_tokens": 638330973.0, + "step": 24671 + }, + { + "epoch": 2.709422358884252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4298136234283447, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7090394496917725, + "num_tokens": 638357139.0, + "step": 24672 + }, + { + "epoch": 2.709532176586866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.3468422889709473, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.711942195892334, + "num_tokens": 638373815.0, + "step": 24673 + }, + { + "epoch": 2.7096419942894796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6169347763061523, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6823830604553223, + "num_tokens": 638399938.0, + "step": 24674 + }, + { + "epoch": 2.709751811992093, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5747413635253906, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.7172684073448181, + "num_tokens": 638422808.0, + "step": 24675 + }, + { + "epoch": 2.7098616296947067, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3664324283599854, + "learning_rate": 1e-06, + "loss": 0.9082, + "mean_token_accuracy": 0.736146867275238, + "num_tokens": 638450788.0, + "step": 24676 + }, + { + "epoch": 2.7099714473973204, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7069013118743896, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7065796256065369, + "num_tokens": 638478890.0, + "step": 24677 + }, + { + "epoch": 2.710081265099934, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3720314502716064, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7141724228858948, + "num_tokens": 638510335.0, + "step": 24678 + }, + { + "epoch": 2.710191082802548, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8394477367401123, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7080140113830566, + "num_tokens": 638533854.0, + "step": 24679 + }, + { + "epoch": 2.7103009005051613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.522308826446533, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6853660345077515, + "num_tokens": 638561507.0, + "step": 24680 + }, + { + "epoch": 2.710410718207775, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6745927333831787, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7106142044067383, + "num_tokens": 638586013.0, + "step": 24681 + }, + { + "epoch": 2.7105205359103888, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.608710765838623, + "learning_rate": 1e-06, + "loss": 0.916, + "mean_token_accuracy": 0.7314375638961792, + "num_tokens": 638609740.0, + "step": 24682 + }, + { + "epoch": 2.7106303536130025, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.568343162536621, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6998187303543091, + "num_tokens": 638636714.0, + "step": 24683 + }, + { + "epoch": 2.7107401713156163, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.579368829727173, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7208904027938843, + "num_tokens": 638661166.0, + "step": 24684 + }, + { + "epoch": 2.7108499890182296, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.43580961227417, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6971665620803833, + "num_tokens": 638689317.0, + "step": 24685 + }, + { + "epoch": 2.7109598067208434, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7089805603027344, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6827802062034607, + "num_tokens": 638714501.0, + "step": 24686 + }, + { + "epoch": 2.711069624423457, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6381499767303467, + "learning_rate": 1e-06, + "loss": 1.0604, + "mean_token_accuracy": 0.6924070119857788, + "num_tokens": 638739101.0, + "step": 24687 + }, + { + "epoch": 2.711179442126071, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.514744997024536, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.693150520324707, + "num_tokens": 638766889.0, + "step": 24688 + }, + { + "epoch": 2.7112892598286846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.713776111602783, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7089719176292419, + "num_tokens": 638789471.0, + "step": 24689 + }, + { + "epoch": 2.711399077531298, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.740598678588867, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7120843529701233, + "num_tokens": 638810942.0, + "step": 24690 + }, + { + "epoch": 2.7115088952339117, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.413522481918335, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7088948488235474, + "num_tokens": 638843824.0, + "step": 24691 + }, + { + "epoch": 2.7116187129365255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4133968353271484, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7068541049957275, + "num_tokens": 638874574.0, + "step": 24692 + }, + { + "epoch": 2.7117285306391388, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4042856693267822, + "learning_rate": 1e-06, + "loss": 0.991, + "mean_token_accuracy": 0.7048898935317993, + "num_tokens": 638903284.0, + "step": 24693 + }, + { + "epoch": 2.7118383483417525, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6363308429718018, + "learning_rate": 1e-06, + "loss": 1.0603, + "mean_token_accuracy": 0.6938211917877197, + "num_tokens": 638928789.0, + "step": 24694 + }, + { + "epoch": 2.7119481660443663, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4159200191497803, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6880528330802917, + "num_tokens": 638960128.0, + "step": 24695 + }, + { + "epoch": 2.71205798374698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4734108448028564, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7110004425048828, + "num_tokens": 638986557.0, + "step": 24696 + }, + { + "epoch": 2.712167801449594, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3923349380493164, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6918637156486511, + "num_tokens": 639019260.0, + "step": 24697 + }, + { + "epoch": 2.712277619152207, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.661251068115234, + "learning_rate": 1e-06, + "loss": 1.0359, + "mean_token_accuracy": 0.7036385536193848, + "num_tokens": 639051229.0, + "step": 24698 + }, + { + "epoch": 2.712387436854821, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.553575038909912, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7048767805099487, + "num_tokens": 639080680.0, + "step": 24699 + }, + { + "epoch": 2.7124972545574346, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.291077136993408, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.717911422252655, + "num_tokens": 639112770.0, + "step": 24700 + }, + { + "epoch": 2.7126070722600484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2838406562805176, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7086963653564453, + "num_tokens": 639132757.0, + "step": 24701 + }, + { + "epoch": 2.712716889962662, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.771446704864502, + "learning_rate": 1e-06, + "loss": 0.9721, + "mean_token_accuracy": 0.7217282652854919, + "num_tokens": 639155224.0, + "step": 24702 + }, + { + "epoch": 2.7128267076652754, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5676660537719727, + "learning_rate": 1e-06, + "loss": 0.987, + "mean_token_accuracy": 0.7036091089248657, + "num_tokens": 639180807.0, + "step": 24703 + }, + { + "epoch": 2.712936525367889, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.893899917602539, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.7006585597991943, + "num_tokens": 639200014.0, + "step": 24704 + }, + { + "epoch": 2.713046343070503, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.884624719619751, + "learning_rate": 1e-06, + "loss": 0.8956, + "mean_token_accuracy": 0.7309304475784302, + "num_tokens": 639219260.0, + "step": 24705 + }, + { + "epoch": 2.7131561607731167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4676170349121094, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.697190523147583, + "num_tokens": 639250650.0, + "step": 24706 + }, + { + "epoch": 2.7132659784757305, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.472360134124756, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7127425670623779, + "num_tokens": 639275641.0, + "step": 24707 + }, + { + "epoch": 2.713375796178344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6770451068878174, + "learning_rate": 1e-06, + "loss": 0.9843, + "mean_token_accuracy": 0.7097421884536743, + "num_tokens": 639297757.0, + "step": 24708 + }, + { + "epoch": 2.7134856138809575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4187240600585938, + "learning_rate": 1e-06, + "loss": 0.9924, + "mean_token_accuracy": 0.7077856063842773, + "num_tokens": 639325714.0, + "step": 24709 + }, + { + "epoch": 2.7135954315835713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8394248485565186, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7201855182647705, + "num_tokens": 639345003.0, + "step": 24710 + }, + { + "epoch": 2.713705249286185, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5128014087677, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7083292603492737, + "num_tokens": 639369979.0, + "step": 24711 + }, + { + "epoch": 2.713815066988799, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4697959423065186, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.7069675326347351, + "num_tokens": 639398053.0, + "step": 24712 + }, + { + "epoch": 2.713924884691412, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.36940860748291, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.6896141767501831, + "num_tokens": 639428209.0, + "step": 24713 + }, + { + "epoch": 2.714034702394026, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4686594009399414, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7179464101791382, + "num_tokens": 639454434.0, + "step": 24714 + }, + { + "epoch": 2.7141445200966396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.663264036178589, + "learning_rate": 1e-06, + "loss": 0.871, + "mean_token_accuracy": 0.7380264401435852, + "num_tokens": 639476779.0, + "step": 24715 + }, + { + "epoch": 2.7142543377992534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5157599449157715, + "learning_rate": 1e-06, + "loss": 1.0083, + "mean_token_accuracy": 0.7147886753082275, + "num_tokens": 639503392.0, + "step": 24716 + }, + { + "epoch": 2.714364155501867, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.562863826751709, + "learning_rate": 1e-06, + "loss": 0.9075, + "mean_token_accuracy": 0.7297091484069824, + "num_tokens": 639527342.0, + "step": 24717 + }, + { + "epoch": 2.7144739732044805, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7490947246551514, + "learning_rate": 1e-06, + "loss": 0.9901, + "mean_token_accuracy": 0.7208907604217529, + "num_tokens": 639550694.0, + "step": 24718 + }, + { + "epoch": 2.7145837909070942, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.303276300430298, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.711638331413269, + "num_tokens": 639585039.0, + "step": 24719 + }, + { + "epoch": 2.714693608609708, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.779754877090454, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.7197696566581726, + "num_tokens": 639607137.0, + "step": 24720 + }, + { + "epoch": 2.7148034263123213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.446857213973999, + "learning_rate": 1e-06, + "loss": 1.0011, + "mean_token_accuracy": 0.7088237404823303, + "num_tokens": 639635501.0, + "step": 24721 + }, + { + "epoch": 2.714913244014935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4421944618225098, + "learning_rate": 1e-06, + "loss": 1.0454, + "mean_token_accuracy": 0.6964441537857056, + "num_tokens": 639665461.0, + "step": 24722 + }, + { + "epoch": 2.715023061717549, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5170178413391113, + "learning_rate": 1e-06, + "loss": 1.0503, + "mean_token_accuracy": 0.6921331882476807, + "num_tokens": 639693491.0, + "step": 24723 + }, + { + "epoch": 2.7151328794201626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7932095527648926, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7101514935493469, + "num_tokens": 639715777.0, + "step": 24724 + }, + { + "epoch": 2.7152426971227763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617098331451416, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.7069747447967529, + "num_tokens": 639741348.0, + "step": 24725 + }, + { + "epoch": 2.7153525148253896, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.662749767303467, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.7117290496826172, + "num_tokens": 639764709.0, + "step": 24726 + }, + { + "epoch": 2.7154623325280034, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3219730854034424, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7216717004776001, + "num_tokens": 639793037.0, + "step": 24727 + }, + { + "epoch": 2.715572150230617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6178746223449707, + "learning_rate": 1e-06, + "loss": 1.0768, + "mean_token_accuracy": 0.6918237209320068, + "num_tokens": 639820215.0, + "step": 24728 + }, + { + "epoch": 2.715681967933231, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8137447834014893, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.7040080428123474, + "num_tokens": 639842429.0, + "step": 24729 + }, + { + "epoch": 2.7157917856358447, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.412224769592285, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7054476737976074, + "num_tokens": 639873513.0, + "step": 24730 + }, + { + "epoch": 2.715901603338458, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5409610271453857, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6841908097267151, + "num_tokens": 639899481.0, + "step": 24731 + }, + { + "epoch": 2.7160114210410717, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.553157091140747, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7057837843894958, + "num_tokens": 639924799.0, + "step": 24732 + }, + { + "epoch": 2.7161212387436855, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5728509426116943, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7007539868354797, + "num_tokens": 639951117.0, + "step": 24733 + }, + { + "epoch": 2.7162310564462993, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4746792316436768, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7134379148483276, + "num_tokens": 639974759.0, + "step": 24734 + }, + { + "epoch": 2.716340874148913, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.82726788520813, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.7015937566757202, + "num_tokens": 639997730.0, + "step": 24735 + }, + { + "epoch": 2.7164506918515263, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8459038734436035, + "learning_rate": 1e-06, + "loss": 0.9883, + "mean_token_accuracy": 0.7098333835601807, + "num_tokens": 640020029.0, + "step": 24736 + }, + { + "epoch": 2.71656050955414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7332470417022705, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7092363834381104, + "num_tokens": 640048122.0, + "step": 24737 + }, + { + "epoch": 2.716670327256754, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4631645679473877, + "learning_rate": 1e-06, + "loss": 1.0658, + "mean_token_accuracy": 0.6906740665435791, + "num_tokens": 640077578.0, + "step": 24738 + }, + { + "epoch": 2.7167801449593676, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.543893814086914, + "learning_rate": 1e-06, + "loss": 0.8786, + "mean_token_accuracy": 0.7378630042076111, + "num_tokens": 640102825.0, + "step": 24739 + }, + { + "epoch": 2.7168899626619814, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.557018995285034, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7182020545005798, + "num_tokens": 640132048.0, + "step": 24740 + }, + { + "epoch": 2.7169997803645947, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7492992877960205, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7085133790969849, + "num_tokens": 640156823.0, + "step": 24741 + }, + { + "epoch": 2.7171095980672084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.671783447265625, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6973750591278076, + "num_tokens": 640180505.0, + "step": 24742 + }, + { + "epoch": 2.717219415769822, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.923969030380249, + "learning_rate": 1e-06, + "loss": 0.893, + "mean_token_accuracy": 0.7359938621520996, + "num_tokens": 640199882.0, + "step": 24743 + }, + { + "epoch": 2.7173292334724355, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5690298080444336, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7176846265792847, + "num_tokens": 640228809.0, + "step": 24744 + }, + { + "epoch": 2.7174390511750497, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6462597846984863, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6919001936912537, + "num_tokens": 640256072.0, + "step": 24745 + }, + { + "epoch": 2.717548868877663, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.556384325027466, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7284420728683472, + "num_tokens": 640282519.0, + "step": 24746 + }, + { + "epoch": 2.7176586865802768, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.709132671356201, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.6975601315498352, + "num_tokens": 640307450.0, + "step": 24747 + }, + { + "epoch": 2.7177685042828905, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8279776573181152, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7079023718833923, + "num_tokens": 640329891.0, + "step": 24748 + }, + { + "epoch": 2.717878321985504, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.209198236465454, + "learning_rate": 1e-06, + "loss": 1.1335, + "mean_token_accuracy": 0.6792882680892944, + "num_tokens": 640364993.0, + "step": 24749 + }, + { + "epoch": 2.7179881396881176, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6451539993286133, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7089064121246338, + "num_tokens": 640392942.0, + "step": 24750 + }, + { + "epoch": 2.7180979573907313, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5471482276916504, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.6988055109977722, + "num_tokens": 640419676.0, + "step": 24751 + }, + { + "epoch": 2.718207775093345, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5210726261138916, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7124423980712891, + "num_tokens": 640448091.0, + "step": 24752 + }, + { + "epoch": 2.718317592795959, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.543653964996338, + "learning_rate": 1e-06, + "loss": 1.0713, + "mean_token_accuracy": 0.6947698593139648, + "num_tokens": 640477778.0, + "step": 24753 + }, + { + "epoch": 2.718427410498572, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9984819889068604, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7071641087532043, + "num_tokens": 640500398.0, + "step": 24754 + }, + { + "epoch": 2.718537228201186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.454962968826294, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.718920886516571, + "num_tokens": 640528757.0, + "step": 24755 + }, + { + "epoch": 2.7186470459037997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9342033863067627, + "learning_rate": 1e-06, + "loss": 1.0952, + "mean_token_accuracy": 0.6857393980026245, + "num_tokens": 640554774.0, + "step": 24756 + }, + { + "epoch": 2.7187568636064134, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5073330402374268, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.6979079246520996, + "num_tokens": 640582467.0, + "step": 24757 + }, + { + "epoch": 2.718866681309027, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3421058654785156, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.6999664306640625, + "num_tokens": 640614279.0, + "step": 24758 + }, + { + "epoch": 2.7189764990116405, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5839169025421143, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7016689777374268, + "num_tokens": 640642335.0, + "step": 24759 + }, + { + "epoch": 2.7190863167142543, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.580963611602783, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6887578368186951, + "num_tokens": 640669322.0, + "step": 24760 + }, + { + "epoch": 2.719196134416868, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0099973678588867, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7074095010757446, + "num_tokens": 640687288.0, + "step": 24761 + }, + { + "epoch": 2.719305952119482, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7882156372070312, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.705439031124115, + "num_tokens": 640710975.0, + "step": 24762 + }, + { + "epoch": 2.7194157698220955, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.221315622329712, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7051559686660767, + "num_tokens": 640747982.0, + "step": 24763 + }, + { + "epoch": 2.719525587524709, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.626622200012207, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7321497201919556, + "num_tokens": 640775056.0, + "step": 24764 + }, + { + "epoch": 2.7196354052273226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4641828536987305, + "learning_rate": 1e-06, + "loss": 1.0468, + "mean_token_accuracy": 0.6975953578948975, + "num_tokens": 640802794.0, + "step": 24765 + }, + { + "epoch": 2.7197452229299364, + "ewc_loss": 1.1026859283447266e-05, + "grad_norm": 80.5413589477539, + "learning_rate": 1e-06, + "loss": 1.0669, + "mean_token_accuracy": 0.6957840919494629, + "num_tokens": 640828970.0, + "step": 24766 + }, + { + "epoch": 2.71985504063255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7421517372131348, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.713537335395813, + "num_tokens": 640856830.0, + "step": 24767 + }, + { + "epoch": 2.719964858335164, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5152008533477783, + "learning_rate": 1e-06, + "loss": 1.0985, + "mean_token_accuracy": 0.6863486766815186, + "num_tokens": 640888052.0, + "step": 24768 + }, + { + "epoch": 2.720074676037777, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7161550521850586, + "learning_rate": 1e-06, + "loss": 0.9586, + "mean_token_accuracy": 0.721595048904419, + "num_tokens": 640912422.0, + "step": 24769 + }, + { + "epoch": 2.720184493740391, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8567159175872803, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7095458507537842, + "num_tokens": 640933340.0, + "step": 24770 + }, + { + "epoch": 2.7202943114430047, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7800426483154297, + "learning_rate": 1e-06, + "loss": 0.9304, + "mean_token_accuracy": 0.7281675934791565, + "num_tokens": 640953858.0, + "step": 24771 + }, + { + "epoch": 2.720404129145618, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.58353853225708, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7103224992752075, + "num_tokens": 640979591.0, + "step": 24772 + }, + { + "epoch": 2.720513946848232, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.531177282333374, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7089670896530151, + "num_tokens": 641005981.0, + "step": 24773 + }, + { + "epoch": 2.7206237645508455, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.569685459136963, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7223485112190247, + "num_tokens": 641030596.0, + "step": 24774 + }, + { + "epoch": 2.7207335822534593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7559051513671875, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6822587251663208, + "num_tokens": 641057299.0, + "step": 24775 + }, + { + "epoch": 2.720843399956073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6272857189178467, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.702674925327301, + "num_tokens": 641083248.0, + "step": 24776 + }, + { + "epoch": 2.7209532176586864, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6974682807922363, + "learning_rate": 1e-06, + "loss": 1.0988, + "mean_token_accuracy": 0.6824168562889099, + "num_tokens": 641110826.0, + "step": 24777 + }, + { + "epoch": 2.7210630353613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6527650356292725, + "learning_rate": 1e-06, + "loss": 0.9852, + "mean_token_accuracy": 0.7253684997558594, + "num_tokens": 641138748.0, + "step": 24778 + }, + { + "epoch": 2.721172853063914, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5849621295928955, + "learning_rate": 1e-06, + "loss": 1.0464, + "mean_token_accuracy": 0.6961884498596191, + "num_tokens": 641167288.0, + "step": 24779 + }, + { + "epoch": 2.7212826707665276, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4558210372924805, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6987894773483276, + "num_tokens": 641196385.0, + "step": 24780 + }, + { + "epoch": 2.7213924884691414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5524988174438477, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.7026077508926392, + "num_tokens": 641222366.0, + "step": 24781 + }, + { + "epoch": 2.7215023061717547, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3151650428771973, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7020966410636902, + "num_tokens": 641255955.0, + "step": 24782 + }, + { + "epoch": 2.7216121238743685, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617912530899048, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6926514506340027, + "num_tokens": 641280874.0, + "step": 24783 + }, + { + "epoch": 2.721721941576982, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9008290767669678, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.715308666229248, + "num_tokens": 641303592.0, + "step": 24784 + }, + { + "epoch": 2.721831759279596, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.41971755027771, + "learning_rate": 1e-06, + "loss": 1.1179, + "mean_token_accuracy": 0.6753545999526978, + "num_tokens": 641336893.0, + "step": 24785 + }, + { + "epoch": 2.7219415769822097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7944226264953613, + "learning_rate": 1e-06, + "loss": 0.9715, + "mean_token_accuracy": 0.7174336314201355, + "num_tokens": 641359312.0, + "step": 24786 + }, + { + "epoch": 2.722051394684823, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6854193210601807, + "learning_rate": 1e-06, + "loss": 0.9312, + "mean_token_accuracy": 0.7290964722633362, + "num_tokens": 641383693.0, + "step": 24787 + }, + { + "epoch": 2.722161212387437, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.68042254447937, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7129045724868774, + "num_tokens": 641407736.0, + "step": 24788 + }, + { + "epoch": 2.7222710300900506, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8729496002197266, + "learning_rate": 1e-06, + "loss": 1.0895, + "mean_token_accuracy": 0.677411675453186, + "num_tokens": 641433221.0, + "step": 24789 + }, + { + "epoch": 2.7223808477926643, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8083033561706543, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.703432559967041, + "num_tokens": 641456588.0, + "step": 24790 + }, + { + "epoch": 2.722490665495278, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5575082302093506, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6875700354576111, + "num_tokens": 641491270.0, + "step": 24791 + }, + { + "epoch": 2.7226004831978914, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5671424865722656, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6982696056365967, + "num_tokens": 641524657.0, + "step": 24792 + }, + { + "epoch": 2.722710300900505, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5969364643096924, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.686318039894104, + "num_tokens": 641551827.0, + "step": 24793 + }, + { + "epoch": 2.722820118603119, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.929898977279663, + "learning_rate": 1e-06, + "loss": 0.9227, + "mean_token_accuracy": 0.7332848906517029, + "num_tokens": 641570975.0, + "step": 24794 + }, + { + "epoch": 2.722929936305732, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.93766188621521, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7119789123535156, + "num_tokens": 641591883.0, + "step": 24795 + }, + { + "epoch": 2.7230397540083464, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7466256618499756, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7070341110229492, + "num_tokens": 641615483.0, + "step": 24796 + }, + { + "epoch": 2.7231495717109597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6081349849700928, + "learning_rate": 1e-06, + "loss": 0.9503, + "mean_token_accuracy": 0.7233903408050537, + "num_tokens": 641640463.0, + "step": 24797 + }, + { + "epoch": 2.7232593894135735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.528481960296631, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.6963376402854919, + "num_tokens": 641669341.0, + "step": 24798 + }, + { + "epoch": 2.7233692071161872, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5856668949127197, + "learning_rate": 1e-06, + "loss": 1.0842, + "mean_token_accuracy": 0.6914317011833191, + "num_tokens": 641695953.0, + "step": 24799 + }, + { + "epoch": 2.7234790248188006, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6764605045318604, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7024492025375366, + "num_tokens": 641720964.0, + "step": 24800 + }, + { + "epoch": 2.7235888425214143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2611544132232666, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6961263418197632, + "num_tokens": 641752724.0, + "step": 24801 + }, + { + "epoch": 2.723698660224028, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1029820442199707, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7237171530723572, + "num_tokens": 641774184.0, + "step": 24802 + }, + { + "epoch": 2.723808477926642, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4178507328033447, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6948983669281006, + "num_tokens": 641802550.0, + "step": 24803 + }, + { + "epoch": 2.7239182956292556, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.889965057373047, + "learning_rate": 1e-06, + "loss": 0.8931, + "mean_token_accuracy": 0.7299226522445679, + "num_tokens": 641822945.0, + "step": 24804 + }, + { + "epoch": 2.724028113331869, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6364595890045166, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7217605113983154, + "num_tokens": 641848120.0, + "step": 24805 + }, + { + "epoch": 2.7241379310344827, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4989023208618164, + "learning_rate": 1e-06, + "loss": 0.9454, + "mean_token_accuracy": 0.7265970706939697, + "num_tokens": 641874104.0, + "step": 24806 + }, + { + "epoch": 2.7242477487370964, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.621591329574585, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7163546681404114, + "num_tokens": 641898951.0, + "step": 24807 + }, + { + "epoch": 2.72435756643971, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.708570957183838, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7263176441192627, + "num_tokens": 641921362.0, + "step": 24808 + }, + { + "epoch": 2.724467384142324, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1900627613067627, + "learning_rate": 1e-06, + "loss": 0.952, + "mean_token_accuracy": 0.71761155128479, + "num_tokens": 641938341.0, + "step": 24809 + }, + { + "epoch": 2.7245772018449372, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.392817974090576, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6848629117012024, + "num_tokens": 641970748.0, + "step": 24810 + }, + { + "epoch": 2.724687019547551, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5432252883911133, + "learning_rate": 1e-06, + "loss": 1.063, + "mean_token_accuracy": 0.6886427402496338, + "num_tokens": 642001075.0, + "step": 24811 + }, + { + "epoch": 2.7247968372501647, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.529432773590088, + "learning_rate": 1e-06, + "loss": 0.9347, + "mean_token_accuracy": 0.7276386022567749, + "num_tokens": 642026261.0, + "step": 24812 + }, + { + "epoch": 2.7249066549527785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.512493371963501, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6968988180160522, + "num_tokens": 642055981.0, + "step": 24813 + }, + { + "epoch": 2.7250164726553923, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.506588935852051, + "learning_rate": 1e-06, + "loss": 1.0875, + "mean_token_accuracy": 0.6868183016777039, + "num_tokens": 642082195.0, + "step": 24814 + }, + { + "epoch": 2.7251262903580056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.367175817489624, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7066949009895325, + "num_tokens": 642114405.0, + "step": 24815 + }, + { + "epoch": 2.7252361080606193, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.621060848236084, + "learning_rate": 1e-06, + "loss": 1.0496, + "mean_token_accuracy": 0.6930108666419983, + "num_tokens": 642139251.0, + "step": 24816 + }, + { + "epoch": 2.725345925763233, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8587305545806885, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7176998853683472, + "num_tokens": 642163813.0, + "step": 24817 + }, + { + "epoch": 2.725455743465847, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.50480318069458, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7148972749710083, + "num_tokens": 642189645.0, + "step": 24818 + }, + { + "epoch": 2.7255655611684606, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.406891107559204, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7095547914505005, + "num_tokens": 642219610.0, + "step": 24819 + }, + { + "epoch": 2.725675378871074, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.980851173400879, + "learning_rate": 1e-06, + "loss": 0.9699, + "mean_token_accuracy": 0.7121893763542175, + "num_tokens": 642240897.0, + "step": 24820 + }, + { + "epoch": 2.7257851965736877, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5897490978240967, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.7023805975914001, + "num_tokens": 642270152.0, + "step": 24821 + }, + { + "epoch": 2.7258950142763014, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.333965301513672, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7096332311630249, + "num_tokens": 642303524.0, + "step": 24822 + }, + { + "epoch": 2.7260048319789147, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.793217182159424, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7049916982650757, + "num_tokens": 642326904.0, + "step": 24823 + }, + { + "epoch": 2.7261146496815285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5620946884155273, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7076340913772583, + "num_tokens": 642352381.0, + "step": 24824 + }, + { + "epoch": 2.7262244673841423, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505703926086426, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7063337564468384, + "num_tokens": 642378034.0, + "step": 24825 + }, + { + "epoch": 2.726334285086756, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5670113563537598, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.716132402420044, + "num_tokens": 642403044.0, + "step": 24826 + }, + { + "epoch": 2.7264441027893698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.635371685028076, + "learning_rate": 1e-06, + "loss": 1.0177, + "mean_token_accuracy": 0.7036552429199219, + "num_tokens": 642426545.0, + "step": 24827 + }, + { + "epoch": 2.726553920491983, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.623366355895996, + "learning_rate": 1e-06, + "loss": 0.9181, + "mean_token_accuracy": 0.7325237393379211, + "num_tokens": 642452117.0, + "step": 24828 + }, + { + "epoch": 2.726663738194597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6458888053894043, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.6897021532058716, + "num_tokens": 642478164.0, + "step": 24829 + }, + { + "epoch": 2.7267735558972106, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5025088787078857, + "learning_rate": 1e-06, + "loss": 1.1084, + "mean_token_accuracy": 0.6845278143882751, + "num_tokens": 642506938.0, + "step": 24830 + }, + { + "epoch": 2.7268833735998244, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.478959560394287, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6926605701446533, + "num_tokens": 642542506.0, + "step": 24831 + }, + { + "epoch": 2.726993191302438, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7539761066436768, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7139145135879517, + "num_tokens": 642563542.0, + "step": 24832 + }, + { + "epoch": 2.7271030090050514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4200615882873535, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7053999900817871, + "num_tokens": 642595227.0, + "step": 24833 + }, + { + "epoch": 2.727212826707665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0367391109466553, + "learning_rate": 1e-06, + "loss": 0.918, + "mean_token_accuracy": 0.7233630418777466, + "num_tokens": 642613641.0, + "step": 24834 + }, + { + "epoch": 2.727322644410279, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.525134801864624, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7039607763290405, + "num_tokens": 642640041.0, + "step": 24835 + }, + { + "epoch": 2.7274324621128927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8366329669952393, + "learning_rate": 1e-06, + "loss": 0.8912, + "mean_token_accuracy": 0.7395431399345398, + "num_tokens": 642658862.0, + "step": 24836 + }, + { + "epoch": 2.7275422798155065, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.602041006088257, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.7105955481529236, + "num_tokens": 642683284.0, + "step": 24837 + }, + { + "epoch": 2.7276520975181198, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4789297580718994, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7222521305084229, + "num_tokens": 642711859.0, + "step": 24838 + }, + { + "epoch": 2.7277619152207335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.237316370010376, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.6957372426986694, + "num_tokens": 642749230.0, + "step": 24839 + }, + { + "epoch": 2.7278717329233473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.59541392326355, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6930166482925415, + "num_tokens": 642775552.0, + "step": 24840 + }, + { + "epoch": 2.727981550625961, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.767265796661377, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.730099081993103, + "num_tokens": 642797175.0, + "step": 24841 + }, + { + "epoch": 2.728091368328575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6332428455352783, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7045522928237915, + "num_tokens": 642823312.0, + "step": 24842 + }, + { + "epoch": 2.728201186031188, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.962212085723877, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.702288031578064, + "num_tokens": 642851144.0, + "step": 24843 + }, + { + "epoch": 2.728311003733802, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6093688011169434, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7162735462188721, + "num_tokens": 642878449.0, + "step": 24844 + }, + { + "epoch": 2.7284208214364156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.753819704055786, + "learning_rate": 1e-06, + "loss": 0.911, + "mean_token_accuracy": 0.7326651215553284, + "num_tokens": 642900043.0, + "step": 24845 + }, + { + "epoch": 2.728530639139029, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.386915445327759, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.6838241815567017, + "num_tokens": 642933975.0, + "step": 24846 + }, + { + "epoch": 2.728640456841643, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9061081409454346, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7128930687904358, + "num_tokens": 642957037.0, + "step": 24847 + }, + { + "epoch": 2.7287502745442564, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.484489679336548, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7152577042579651, + "num_tokens": 642983748.0, + "step": 24848 + }, + { + "epoch": 2.72886009224687, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2884409427642822, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7081796526908875, + "num_tokens": 643015345.0, + "step": 24849 + }, + { + "epoch": 2.728969909949484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7281625270843506, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7043594121932983, + "num_tokens": 643037919.0, + "step": 24850 + }, + { + "epoch": 2.7290797276520973, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4453046321868896, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7218056321144104, + "num_tokens": 643066006.0, + "step": 24851 + }, + { + "epoch": 2.729189545354711, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.429877996444702, + "learning_rate": 1e-06, + "loss": 0.9432, + "mean_token_accuracy": 0.728880763053894, + "num_tokens": 643092663.0, + "step": 24852 + }, + { + "epoch": 2.729299363057325, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6532723903656006, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7067277431488037, + "num_tokens": 643122114.0, + "step": 24853 + }, + { + "epoch": 2.7294091807599385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.5052621364593506, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7168598771095276, + "num_tokens": 643143373.0, + "step": 24854 + }, + { + "epoch": 2.7295189984625523, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.152575969696045, + "learning_rate": 1e-06, + "loss": 0.9427, + "mean_token_accuracy": 0.7241411209106445, + "num_tokens": 643166868.0, + "step": 24855 + }, + { + "epoch": 2.7296288161651656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.565743923187256, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.6987212896347046, + "num_tokens": 643195898.0, + "step": 24856 + }, + { + "epoch": 2.7297386338677794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3485116958618164, + "learning_rate": 1e-06, + "loss": 1.0738, + "mean_token_accuracy": 0.6848390102386475, + "num_tokens": 643228559.0, + "step": 24857 + }, + { + "epoch": 2.729848451570393, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5884876251220703, + "learning_rate": 1e-06, + "loss": 0.9755, + "mean_token_accuracy": 0.7268160581588745, + "num_tokens": 643254623.0, + "step": 24858 + }, + { + "epoch": 2.729958269273007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6508636474609375, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7242641448974609, + "num_tokens": 643280531.0, + "step": 24859 + }, + { + "epoch": 2.7300680869756206, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.832437753677368, + "learning_rate": 1e-06, + "loss": 0.9602, + "mean_token_accuracy": 0.717235267162323, + "num_tokens": 643300152.0, + "step": 24860 + }, + { + "epoch": 2.730177904678234, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4702463150024414, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.710334062576294, + "num_tokens": 643327754.0, + "step": 24861 + }, + { + "epoch": 2.7302877223808477, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6924328804016113, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7151950597763062, + "num_tokens": 643352108.0, + "step": 24862 + }, + { + "epoch": 2.7303975400834615, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.429370641708374, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7095000743865967, + "num_tokens": 643380193.0, + "step": 24863 + }, + { + "epoch": 2.7305073577860752, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4861860275268555, + "learning_rate": 1e-06, + "loss": 1.0852, + "mean_token_accuracy": 0.6899616718292236, + "num_tokens": 643409522.0, + "step": 24864 + }, + { + "epoch": 2.730617175488689, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5214245319366455, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7119408249855042, + "num_tokens": 643437342.0, + "step": 24865 + }, + { + "epoch": 2.7307269931913023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.646923065185547, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7062854766845703, + "num_tokens": 643463951.0, + "step": 24866 + }, + { + "epoch": 2.730836810893916, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6987552642822266, + "learning_rate": 1e-06, + "loss": 0.9245, + "mean_token_accuracy": 0.7278194427490234, + "num_tokens": 643487187.0, + "step": 24867 + }, + { + "epoch": 2.73094662859653, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.973771333694458, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.6993573307991028, + "num_tokens": 643508957.0, + "step": 24868 + }, + { + "epoch": 2.7310564462991436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.721473455429077, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7072055339813232, + "num_tokens": 643534542.0, + "step": 24869 + }, + { + "epoch": 2.7311662640017573, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.58713960647583, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.710749089717865, + "num_tokens": 643560555.0, + "step": 24870 + }, + { + "epoch": 2.7312760817043706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.469280481338501, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7032705545425415, + "num_tokens": 643594435.0, + "step": 24871 + }, + { + "epoch": 2.7313858994069844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.387040853500366, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6895701885223389, + "num_tokens": 643623464.0, + "step": 24872 + }, + { + "epoch": 2.731495717109598, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.634880542755127, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7030172348022461, + "num_tokens": 643651590.0, + "step": 24873 + }, + { + "epoch": 2.7316055348122115, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 4.334460258483887, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.708245038986206, + "num_tokens": 643679139.0, + "step": 24874 + }, + { + "epoch": 2.7317153525148252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7766270637512207, + "learning_rate": 1e-06, + "loss": 1.0526, + "mean_token_accuracy": 0.6991209387779236, + "num_tokens": 643704449.0, + "step": 24875 + }, + { + "epoch": 2.731825170217439, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9074623584747314, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7144753336906433, + "num_tokens": 643727376.0, + "step": 24876 + }, + { + "epoch": 2.7319349879200527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.727720022201538, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7016439437866211, + "num_tokens": 643751934.0, + "step": 24877 + }, + { + "epoch": 2.7320448056226665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5276384353637695, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.70749431848526, + "num_tokens": 643778941.0, + "step": 24878 + }, + { + "epoch": 2.73215462332528, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.624392509460449, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.710058331489563, + "num_tokens": 643805336.0, + "step": 24879 + }, + { + "epoch": 2.7322644410278936, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.515695810317993, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.682873010635376, + "num_tokens": 643834946.0, + "step": 24880 + }, + { + "epoch": 2.7323742587305073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4919025897979736, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7006993889808655, + "num_tokens": 643864791.0, + "step": 24881 + }, + { + "epoch": 2.732484076433121, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5095856189727783, + "learning_rate": 1e-06, + "loss": 1.0667, + "mean_token_accuracy": 0.6958803534507751, + "num_tokens": 643896644.0, + "step": 24882 + }, + { + "epoch": 2.732593894135735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.517666816711426, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7181545495986938, + "num_tokens": 643924896.0, + "step": 24883 + }, + { + "epoch": 2.732703711838348, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4083783626556396, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7160338163375854, + "num_tokens": 643955376.0, + "step": 24884 + }, + { + "epoch": 2.732813529540962, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.450319290161133, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7242145538330078, + "num_tokens": 643984504.0, + "step": 24885 + }, + { + "epoch": 2.7329233472435757, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9082086086273193, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7043469548225403, + "num_tokens": 644010345.0, + "step": 24886 + }, + { + "epoch": 2.7330331649461894, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7581441402435303, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7110739946365356, + "num_tokens": 644034153.0, + "step": 24887 + }, + { + "epoch": 2.733142982648803, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5576698780059814, + "learning_rate": 1e-06, + "loss": 0.9614, + "mean_token_accuracy": 0.712704062461853, + "num_tokens": 644059787.0, + "step": 24888 + }, + { + "epoch": 2.7332528003514165, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.719054937362671, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.7226276993751526, + "num_tokens": 644082650.0, + "step": 24889 + }, + { + "epoch": 2.7333626180540302, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.672116756439209, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7263457179069519, + "num_tokens": 644106614.0, + "step": 24890 + }, + { + "epoch": 2.733472435756644, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7238264083862305, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6967933177947998, + "num_tokens": 644130761.0, + "step": 24891 + }, + { + "epoch": 2.7335822534592578, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8759100437164307, + "learning_rate": 1e-06, + "loss": 0.9382, + "mean_token_accuracy": 0.7226901650428772, + "num_tokens": 644154838.0, + "step": 24892 + }, + { + "epoch": 2.7336920711618715, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8285655975341797, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7186270356178284, + "num_tokens": 644179150.0, + "step": 24893 + }, + { + "epoch": 2.733801888864485, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.49794864654541, + "learning_rate": 1e-06, + "loss": 0.9539, + "mean_token_accuracy": 0.718372642993927, + "num_tokens": 644206990.0, + "step": 24894 + }, + { + "epoch": 2.7339117065670986, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6683051586151123, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6896603107452393, + "num_tokens": 644233162.0, + "step": 24895 + }, + { + "epoch": 2.7340215242697123, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.674074172973633, + "learning_rate": 1e-06, + "loss": 1.0717, + "mean_token_accuracy": 0.6873813271522522, + "num_tokens": 644256968.0, + "step": 24896 + }, + { + "epoch": 2.734131341972326, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4372432231903076, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.7045739889144897, + "num_tokens": 644284581.0, + "step": 24897 + }, + { + "epoch": 2.73424115967494, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.639309883117676, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.711207389831543, + "num_tokens": 644308226.0, + "step": 24898 + }, + { + "epoch": 2.734350977377553, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3780782222747803, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.6921079158782959, + "num_tokens": 644340434.0, + "step": 24899 + }, + { + "epoch": 2.734460795080167, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7859256267547607, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7111313343048096, + "num_tokens": 644371815.0, + "step": 24900 + }, + { + "epoch": 2.7345706127827807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5507895946502686, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7038906812667847, + "num_tokens": 644396875.0, + "step": 24901 + }, + { + "epoch": 2.734680430485394, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4921371936798096, + "learning_rate": 1e-06, + "loss": 1.0139, + "mean_token_accuracy": 0.7063413858413696, + "num_tokens": 644423665.0, + "step": 24902 + }, + { + "epoch": 2.7347902481880078, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3872084617614746, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6949989795684814, + "num_tokens": 644455080.0, + "step": 24903 + }, + { + "epoch": 2.7349000658906215, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5738861560821533, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.7229909896850586, + "num_tokens": 644478301.0, + "step": 24904 + }, + { + "epoch": 2.7350098835932353, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6685774326324463, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7148794531822205, + "num_tokens": 644500203.0, + "step": 24905 + }, + { + "epoch": 2.735119701295849, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.474665880203247, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.6980509161949158, + "num_tokens": 644528204.0, + "step": 24906 + }, + { + "epoch": 2.7352295189984623, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.836303234100342, + "learning_rate": 1e-06, + "loss": 0.9311, + "mean_token_accuracy": 0.7274969816207886, + "num_tokens": 644547240.0, + "step": 24907 + }, + { + "epoch": 2.735339336701076, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.62402606010437, + "learning_rate": 1e-06, + "loss": 0.9462, + "mean_token_accuracy": 0.7182701826095581, + "num_tokens": 644570855.0, + "step": 24908 + }, + { + "epoch": 2.73544915440369, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6165339946746826, + "learning_rate": 1e-06, + "loss": 1.0804, + "mean_token_accuracy": 0.6924795508384705, + "num_tokens": 644595127.0, + "step": 24909 + }, + { + "epoch": 2.7355589721063036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.422192335128784, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7116671204566956, + "num_tokens": 644625702.0, + "step": 24910 + }, + { + "epoch": 2.7356687898089174, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.812011957168579, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6920441389083862, + "num_tokens": 644649169.0, + "step": 24911 + }, + { + "epoch": 2.7357786075115307, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9749040603637695, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7238679528236389, + "num_tokens": 644670190.0, + "step": 24912 + }, + { + "epoch": 2.7358884252141444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6847565174102783, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.707419216632843, + "num_tokens": 644694420.0, + "step": 24913 + }, + { + "epoch": 2.735998242916758, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.821711778640747, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7015532851219177, + "num_tokens": 644717165.0, + "step": 24914 + }, + { + "epoch": 2.736108060619372, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5940752029418945, + "learning_rate": 1e-06, + "loss": 0.9746, + "mean_token_accuracy": 0.7167582511901855, + "num_tokens": 644745186.0, + "step": 24915 + }, + { + "epoch": 2.7362178783219857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.556098461151123, + "learning_rate": 1e-06, + "loss": 1.0555, + "mean_token_accuracy": 0.6960203647613525, + "num_tokens": 644773772.0, + "step": 24916 + }, + { + "epoch": 2.736327696024599, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6387948989868164, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6941452026367188, + "num_tokens": 644800001.0, + "step": 24917 + }, + { + "epoch": 2.736437513727213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0813543796539307, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7003689408302307, + "num_tokens": 644825060.0, + "step": 24918 + }, + { + "epoch": 2.7365473314298265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718470335006714, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7107372879981995, + "num_tokens": 644848268.0, + "step": 24919 + }, + { + "epoch": 2.7366571491324403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8304195404052734, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.715786874294281, + "num_tokens": 644868479.0, + "step": 24920 + }, + { + "epoch": 2.736766966835054, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2333731651306152, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.706771731376648, + "num_tokens": 644904886.0, + "step": 24921 + }, + { + "epoch": 2.7368767845376674, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2855072021484375, + "learning_rate": 1e-06, + "loss": 1.0719, + "mean_token_accuracy": 0.6907051205635071, + "num_tokens": 644942507.0, + "step": 24922 + }, + { + "epoch": 2.736986602240281, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0356955528259277, + "learning_rate": 1e-06, + "loss": 0.8968, + "mean_token_accuracy": 0.7357732057571411, + "num_tokens": 644960996.0, + "step": 24923 + }, + { + "epoch": 2.737096419942895, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.591907262802124, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7115558385848999, + "num_tokens": 644991209.0, + "step": 24924 + }, + { + "epoch": 2.737206237645508, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4453237056732178, + "learning_rate": 1e-06, + "loss": 1.1045, + "mean_token_accuracy": 0.691116988658905, + "num_tokens": 645021253.0, + "step": 24925 + }, + { + "epoch": 2.7373160553481224, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6380951404571533, + "learning_rate": 1e-06, + "loss": 0.9378, + "mean_token_accuracy": 0.7274125218391418, + "num_tokens": 645045743.0, + "step": 24926 + }, + { + "epoch": 2.7374258730507357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.563072443008423, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.707183837890625, + "num_tokens": 645070358.0, + "step": 24927 + }, + { + "epoch": 2.7375356907533495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.60109543800354, + "learning_rate": 1e-06, + "loss": 0.9412, + "mean_token_accuracy": 0.7268272638320923, + "num_tokens": 645093569.0, + "step": 24928 + }, + { + "epoch": 2.737645508455963, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.47879695892334, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.6974407434463501, + "num_tokens": 645123189.0, + "step": 24929 + }, + { + "epoch": 2.7377553261585765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5248148441314697, + "learning_rate": 1e-06, + "loss": 0.9742, + "mean_token_accuracy": 0.716639518737793, + "num_tokens": 645148340.0, + "step": 24930 + }, + { + "epoch": 2.7378651438611903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7893993854522705, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7291272282600403, + "num_tokens": 645170493.0, + "step": 24931 + }, + { + "epoch": 2.737974961563804, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0867598056793213, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6995646953582764, + "num_tokens": 645196556.0, + "step": 24932 + }, + { + "epoch": 2.738084779266418, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.507734775543213, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7058640122413635, + "num_tokens": 645222041.0, + "step": 24933 + }, + { + "epoch": 2.7381945969690316, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 2.830780506134033, + "learning_rate": 1e-06, + "loss": 0.9079, + "mean_token_accuracy": 0.7251729965209961, + "num_tokens": 645243118.0, + "step": 24934 + }, + { + "epoch": 2.738304414671645, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0636138916015625, + "learning_rate": 1e-06, + "loss": 0.9787, + "mean_token_accuracy": 0.7080973982810974, + "num_tokens": 645261831.0, + "step": 24935 + }, + { + "epoch": 2.7384142323742586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5564589500427246, + "learning_rate": 1e-06, + "loss": 1.0657, + "mean_token_accuracy": 0.6920861005783081, + "num_tokens": 645289555.0, + "step": 24936 + }, + { + "epoch": 2.7385240500768724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.897993564605713, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7155191898345947, + "num_tokens": 645311314.0, + "step": 24937 + }, + { + "epoch": 2.738633867779486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4269392490386963, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.6912806034088135, + "num_tokens": 645341909.0, + "step": 24938 + }, + { + "epoch": 2.7387436854821, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.850465774536133, + "learning_rate": 1e-06, + "loss": 1.0816, + "mean_token_accuracy": 0.6919999122619629, + "num_tokens": 645366844.0, + "step": 24939 + }, + { + "epoch": 2.738853503184713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.793797254562378, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.700669527053833, + "num_tokens": 645389199.0, + "step": 24940 + }, + { + "epoch": 2.738963320887327, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.394052505493164, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7061259746551514, + "num_tokens": 645423038.0, + "step": 24941 + }, + { + "epoch": 2.7390731385899407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4340717792510986, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7217225432395935, + "num_tokens": 645451731.0, + "step": 24942 + }, + { + "epoch": 2.7391829562925545, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8227903842926025, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7072840929031372, + "num_tokens": 645478334.0, + "step": 24943 + }, + { + "epoch": 2.7392927739951682, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8004963397979736, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.70513916015625, + "num_tokens": 645502011.0, + "step": 24944 + }, + { + "epoch": 2.7394025916977816, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.713071346282959, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.708619236946106, + "num_tokens": 645526537.0, + "step": 24945 + }, + { + "epoch": 2.7395124094003953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.775113582611084, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.7395420074462891, + "num_tokens": 645555576.0, + "step": 24946 + }, + { + "epoch": 2.739622227103009, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.884723424911499, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.713576078414917, + "num_tokens": 645579414.0, + "step": 24947 + }, + { + "epoch": 2.739732044805623, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.852897882461548, + "learning_rate": 1e-06, + "loss": 0.9313, + "mean_token_accuracy": 0.7192487120628357, + "num_tokens": 645600487.0, + "step": 24948 + }, + { + "epoch": 2.7398418625082366, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5659806728363037, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7028618454933167, + "num_tokens": 645627786.0, + "step": 24949 + }, + { + "epoch": 2.73995168021085, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.866753339767456, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7199653387069702, + "num_tokens": 645650574.0, + "step": 24950 + }, + { + "epoch": 2.7400614979134637, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.416623592376709, + "learning_rate": 1e-06, + "loss": 1.0876, + "mean_token_accuracy": 0.6870085597038269, + "num_tokens": 645687087.0, + "step": 24951 + }, + { + "epoch": 2.7401713156160774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5683765411376953, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7129147052764893, + "num_tokens": 645710254.0, + "step": 24952 + }, + { + "epoch": 2.7402811333186907, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.436821699142456, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.6986855864524841, + "num_tokens": 645740072.0, + "step": 24953 + }, + { + "epoch": 2.7403909510213045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0174336433410645, + "learning_rate": 1e-06, + "loss": 0.9716, + "mean_token_accuracy": 0.7177561521530151, + "num_tokens": 645758717.0, + "step": 24954 + }, + { + "epoch": 2.7405007687239182, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6163430213928223, + "learning_rate": 1e-06, + "loss": 1.0477, + "mean_token_accuracy": 0.6920600533485413, + "num_tokens": 645783376.0, + "step": 24955 + }, + { + "epoch": 2.740610586426532, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5118141174316406, + "learning_rate": 1e-06, + "loss": 1.0621, + "mean_token_accuracy": 0.6940160989761353, + "num_tokens": 645810540.0, + "step": 24956 + }, + { + "epoch": 2.7407204041291457, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.699554443359375, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.7056944370269775, + "num_tokens": 645836400.0, + "step": 24957 + }, + { + "epoch": 2.740830221831759, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3836145401000977, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6923938989639282, + "num_tokens": 645865217.0, + "step": 24958 + }, + { + "epoch": 2.740940039534373, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3824622631073, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6965879201889038, + "num_tokens": 645897124.0, + "step": 24959 + }, + { + "epoch": 2.7410498572369866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7252206802368164, + "learning_rate": 1e-06, + "loss": 0.9005, + "mean_token_accuracy": 0.7346581220626831, + "num_tokens": 645919008.0, + "step": 24960 + }, + { + "epoch": 2.7411596749396003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8003461360931396, + "learning_rate": 1e-06, + "loss": 1.0384, + "mean_token_accuracy": 0.7030096650123596, + "num_tokens": 645944119.0, + "step": 24961 + }, + { + "epoch": 2.741269492642214, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3911867141723633, + "learning_rate": 1e-06, + "loss": 1.0727, + "mean_token_accuracy": 0.6886705160140991, + "num_tokens": 645976846.0, + "step": 24962 + }, + { + "epoch": 2.7413793103448274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.650634288787842, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7128112316131592, + "num_tokens": 646000486.0, + "step": 24963 + }, + { + "epoch": 2.741489128047441, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6170103549957275, + "learning_rate": 1e-06, + "loss": 1.0763, + "mean_token_accuracy": 0.6878223419189453, + "num_tokens": 646028394.0, + "step": 24964 + }, + { + "epoch": 2.741598945750055, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5962014198303223, + "learning_rate": 1e-06, + "loss": 1.0756, + "mean_token_accuracy": 0.6927846670150757, + "num_tokens": 646058019.0, + "step": 24965 + }, + { + "epoch": 2.7417087634526687, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5609278678894043, + "learning_rate": 1e-06, + "loss": 1.0684, + "mean_token_accuracy": 0.6980985403060913, + "num_tokens": 646082838.0, + "step": 24966 + }, + { + "epoch": 2.7418185811552824, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.788806915283203, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7153345346450806, + "num_tokens": 646112665.0, + "step": 24967 + }, + { + "epoch": 2.7419283988578957, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3153176307678223, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6962319612503052, + "num_tokens": 646143820.0, + "step": 24968 + }, + { + "epoch": 2.7420382165605095, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6372900009155273, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6955173015594482, + "num_tokens": 646170636.0, + "step": 24969 + }, + { + "epoch": 2.7421480342631233, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9320247173309326, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6982094049453735, + "num_tokens": 646192243.0, + "step": 24970 + }, + { + "epoch": 2.742257851965737, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.315211057662964, + "learning_rate": 1e-06, + "loss": 1.0959, + "mean_token_accuracy": 0.6870648860931396, + "num_tokens": 646224106.0, + "step": 24971 + }, + { + "epoch": 2.7423676696683508, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.417375326156616, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.6970070600509644, + "num_tokens": 646251003.0, + "step": 24972 + }, + { + "epoch": 2.742477487370964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5103743076324463, + "learning_rate": 1e-06, + "loss": 1.0724, + "mean_token_accuracy": 0.689564049243927, + "num_tokens": 646278161.0, + "step": 24973 + }, + { + "epoch": 2.742587305073578, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.735316514968872, + "learning_rate": 1e-06, + "loss": 0.8791, + "mean_token_accuracy": 0.7388389110565186, + "num_tokens": 646300023.0, + "step": 24974 + }, + { + "epoch": 2.7426971227761916, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.524811029434204, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.6961686611175537, + "num_tokens": 646330269.0, + "step": 24975 + }, + { + "epoch": 2.742806940478805, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.45975923538208, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6963300108909607, + "num_tokens": 646359177.0, + "step": 24976 + }, + { + "epoch": 2.742916758181419, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.086688756942749, + "learning_rate": 1e-06, + "loss": 0.989, + "mean_token_accuracy": 0.7091054916381836, + "num_tokens": 646378780.0, + "step": 24977 + }, + { + "epoch": 2.7430265758840324, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.634176731109619, + "learning_rate": 1e-06, + "loss": 0.9994, + "mean_token_accuracy": 0.7153843641281128, + "num_tokens": 646408380.0, + "step": 24978 + }, + { + "epoch": 2.743136393586646, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.498242139816284, + "learning_rate": 1e-06, + "loss": 1.0549, + "mean_token_accuracy": 0.7078375816345215, + "num_tokens": 646436526.0, + "step": 24979 + }, + { + "epoch": 2.74324621128926, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.824728012084961, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.7253782749176025, + "num_tokens": 646457486.0, + "step": 24980 + }, + { + "epoch": 2.7433560289918733, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9091010093688965, + "learning_rate": 1e-06, + "loss": 0.992, + "mean_token_accuracy": 0.7120192646980286, + "num_tokens": 646478215.0, + "step": 24981 + }, + { + "epoch": 2.743465846694487, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6933810710906982, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.7065200805664062, + "num_tokens": 646502529.0, + "step": 24982 + }, + { + "epoch": 2.7435756643971008, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4625003337860107, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.7006809115409851, + "num_tokens": 646532780.0, + "step": 24983 + }, + { + "epoch": 2.7436854820997145, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6423425674438477, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7217873334884644, + "num_tokens": 646555900.0, + "step": 24984 + }, + { + "epoch": 2.7437952998023283, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2229628562927246, + "learning_rate": 1e-06, + "loss": 1.1054, + "mean_token_accuracy": 0.6822019815444946, + "num_tokens": 646590148.0, + "step": 24985 + }, + { + "epoch": 2.7439051175049416, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4964053630828857, + "learning_rate": 1e-06, + "loss": 0.9352, + "mean_token_accuracy": 0.7247315049171448, + "num_tokens": 646615531.0, + "step": 24986 + }, + { + "epoch": 2.7440149352075554, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.777576446533203, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7204495668411255, + "num_tokens": 646640800.0, + "step": 24987 + }, + { + "epoch": 2.744124752910169, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8988091945648193, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7091175317764282, + "num_tokens": 646664533.0, + "step": 24988 + }, + { + "epoch": 2.744234570612783, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.627654552459717, + "learning_rate": 1e-06, + "loss": 0.9221, + "mean_token_accuracy": 0.7280027866363525, + "num_tokens": 646687570.0, + "step": 24989 + }, + { + "epoch": 2.7443443883153966, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6086292266845703, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6918874979019165, + "num_tokens": 646713050.0, + "step": 24990 + }, + { + "epoch": 2.74445420601801, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.850121021270752, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7226783037185669, + "num_tokens": 646736710.0, + "step": 24991 + }, + { + "epoch": 2.7445640237206237, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.790839195251465, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7127211093902588, + "num_tokens": 646760872.0, + "step": 24992 + }, + { + "epoch": 2.7446738414232374, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.733685255050659, + "learning_rate": 1e-06, + "loss": 0.9629, + "mean_token_accuracy": 0.7279938459396362, + "num_tokens": 646783717.0, + "step": 24993 + }, + { + "epoch": 2.744783659125851, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4853076934814453, + "learning_rate": 1e-06, + "loss": 1.0352, + "mean_token_accuracy": 0.6986125707626343, + "num_tokens": 646810412.0, + "step": 24994 + }, + { + "epoch": 2.744893476828465, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7274670600891113, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7008500099182129, + "num_tokens": 646833771.0, + "step": 24995 + }, + { + "epoch": 2.7450032945310783, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.684286594390869, + "learning_rate": 1e-06, + "loss": 1.0005, + "mean_token_accuracy": 0.7116097807884216, + "num_tokens": 646858231.0, + "step": 24996 + }, + { + "epoch": 2.745113112233692, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9575185775756836, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7056892514228821, + "num_tokens": 646879355.0, + "step": 24997 + }, + { + "epoch": 2.745222929936306, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5040745735168457, + "learning_rate": 1e-06, + "loss": 1.1187, + "mean_token_accuracy": 0.6780335903167725, + "num_tokens": 646910891.0, + "step": 24998 + }, + { + "epoch": 2.7453327476389195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.575892686843872, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7203288078308105, + "num_tokens": 646938572.0, + "step": 24999 + }, + { + "epoch": 2.7454425653415333, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.544536590576172, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7183760404586792, + "num_tokens": 646964803.0, + "step": 25000 + }, + { + "epoch": 2.7455523830441466, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3263967037200928, + "learning_rate": 1e-06, + "loss": 1.0461, + "mean_token_accuracy": 0.702431321144104, + "num_tokens": 646997019.0, + "step": 25001 + }, + { + "epoch": 2.7456622007467604, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8447153568267822, + "learning_rate": 1e-06, + "loss": 0.9572, + "mean_token_accuracy": 0.7182263135910034, + "num_tokens": 647020212.0, + "step": 25002 + }, + { + "epoch": 2.745772018449374, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7424542903900146, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6978768706321716, + "num_tokens": 647049845.0, + "step": 25003 + }, + { + "epoch": 2.7458818361519874, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6942532062530518, + "learning_rate": 1e-06, + "loss": 1.1199, + "mean_token_accuracy": 0.6899091005325317, + "num_tokens": 647078205.0, + "step": 25004 + }, + { + "epoch": 2.745991653854601, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3355531692504883, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.6940008401870728, + "num_tokens": 647113634.0, + "step": 25005 + }, + { + "epoch": 2.746101471557215, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7224209308624268, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7050018310546875, + "num_tokens": 647136875.0, + "step": 25006 + }, + { + "epoch": 2.7462112892598287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.537783622741699, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7122015953063965, + "num_tokens": 647162932.0, + "step": 25007 + }, + { + "epoch": 2.7463211069624425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5432825088500977, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6924485564231873, + "num_tokens": 647190846.0, + "step": 25008 + }, + { + "epoch": 2.746430924665056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4647605419158936, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7048797607421875, + "num_tokens": 647219098.0, + "step": 25009 + }, + { + "epoch": 2.7465407423676695, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4746174812316895, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7049096822738647, + "num_tokens": 647243920.0, + "step": 25010 + }, + { + "epoch": 2.7466505600702833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6298983097076416, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7108044028282166, + "num_tokens": 647268324.0, + "step": 25011 + }, + { + "epoch": 2.746760377772897, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.833219289779663, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7132774591445923, + "num_tokens": 647291575.0, + "step": 25012 + }, + { + "epoch": 2.746870195475511, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.660012722015381, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.7320395708084106, + "num_tokens": 647314862.0, + "step": 25013 + }, + { + "epoch": 2.746980013178124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5947341918945312, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.701502799987793, + "num_tokens": 647338860.0, + "step": 25014 + }, + { + "epoch": 2.747089830880738, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9534454345703125, + "learning_rate": 1e-06, + "loss": 1.0335, + "mean_token_accuracy": 0.7040719985961914, + "num_tokens": 647367882.0, + "step": 25015 + }, + { + "epoch": 2.7471996485833516, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.566800832748413, + "learning_rate": 1e-06, + "loss": 1.0676, + "mean_token_accuracy": 0.6870315074920654, + "num_tokens": 647394841.0, + "step": 25016 + }, + { + "epoch": 2.7473094662859654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.595963478088379, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7111930847167969, + "num_tokens": 647419836.0, + "step": 25017 + }, + { + "epoch": 2.747419283988579, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.622922420501709, + "learning_rate": 1e-06, + "loss": 0.9823, + "mean_token_accuracy": 0.7124453783035278, + "num_tokens": 647445445.0, + "step": 25018 + }, + { + "epoch": 2.7475291016911925, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.242567539215088, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.716871976852417, + "num_tokens": 647476707.0, + "step": 25019 + }, + { + "epoch": 2.7476389193938062, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.9443607330322266, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.720904529094696, + "num_tokens": 647493810.0, + "step": 25020 + }, + { + "epoch": 2.74774873709642, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5968947410583496, + "learning_rate": 1e-06, + "loss": 0.9541, + "mean_token_accuracy": 0.718259334564209, + "num_tokens": 647521427.0, + "step": 25021 + }, + { + "epoch": 2.7478585547990337, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.738939046859741, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.7036044597625732, + "num_tokens": 647549213.0, + "step": 25022 + }, + { + "epoch": 2.7479683725016475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3327476978302, + "learning_rate": 1e-06, + "loss": 1.1469, + "mean_token_accuracy": 0.6696415543556213, + "num_tokens": 647581008.0, + "step": 25023 + }, + { + "epoch": 2.748078190204261, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6513752937316895, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7180942296981812, + "num_tokens": 647605790.0, + "step": 25024 + }, + { + "epoch": 2.7481880079068746, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.482069492340088, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7097384929656982, + "num_tokens": 647631816.0, + "step": 25025 + }, + { + "epoch": 2.7482978256094883, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5815279483795166, + "learning_rate": 1e-06, + "loss": 0.9504, + "mean_token_accuracy": 0.721850574016571, + "num_tokens": 647656111.0, + "step": 25026 + }, + { + "epoch": 2.7484076433121016, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8225772380828857, + "learning_rate": 1e-06, + "loss": 1.0189, + "mean_token_accuracy": 0.7100429534912109, + "num_tokens": 647677258.0, + "step": 25027 + }, + { + "epoch": 2.748517461014716, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0703282356262207, + "learning_rate": 1e-06, + "loss": 1.0105, + "mean_token_accuracy": 0.7079223990440369, + "num_tokens": 647697016.0, + "step": 25028 + }, + { + "epoch": 2.748627278717329, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.662822723388672, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7032793760299683, + "num_tokens": 647723095.0, + "step": 25029 + }, + { + "epoch": 2.748737096419943, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.813565731048584, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.6944329738616943, + "num_tokens": 647747672.0, + "step": 25030 + }, + { + "epoch": 2.7488469141225567, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6211867332458496, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.704920768737793, + "num_tokens": 647774007.0, + "step": 25031 + }, + { + "epoch": 2.74895673182517, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3955042362213135, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6982589364051819, + "num_tokens": 647804272.0, + "step": 25032 + }, + { + "epoch": 2.7490665495277837, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.774181365966797, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.7179082632064819, + "num_tokens": 647827428.0, + "step": 25033 + }, + { + "epoch": 2.7491763672303975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7354257106781006, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7069540023803711, + "num_tokens": 647849150.0, + "step": 25034 + }, + { + "epoch": 2.7492861849330112, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.538011312484741, + "learning_rate": 1e-06, + "loss": 0.9545, + "mean_token_accuracy": 0.7254502773284912, + "num_tokens": 647876113.0, + "step": 25035 + }, + { + "epoch": 2.749396002635625, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.504336357116699, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6994832754135132, + "num_tokens": 647903846.0, + "step": 25036 + }, + { + "epoch": 2.7495058203382383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2913174629211426, + "learning_rate": 1e-06, + "loss": 1.1027, + "mean_token_accuracy": 0.6850553154945374, + "num_tokens": 647938368.0, + "step": 25037 + }, + { + "epoch": 2.749615638040852, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.367525100708008, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7009099125862122, + "num_tokens": 647970228.0, + "step": 25038 + }, + { + "epoch": 2.749725455743466, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6597607135772705, + "learning_rate": 1e-06, + "loss": 1.0185, + "mean_token_accuracy": 0.6969586610794067, + "num_tokens": 647999975.0, + "step": 25039 + }, + { + "epoch": 2.7498352734460796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8176870346069336, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7079582214355469, + "num_tokens": 648025289.0, + "step": 25040 + }, + { + "epoch": 2.7499450911486933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.465599775314331, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.712487518787384, + "num_tokens": 648056141.0, + "step": 25041 + }, + { + "epoch": 2.7500549088513067, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.413938283920288, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7113035917282104, + "num_tokens": 648087984.0, + "step": 25042 + }, + { + "epoch": 2.7501647265539204, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3686609268188477, + "learning_rate": 1e-06, + "loss": 0.9421, + "mean_token_accuracy": 0.7219980955123901, + "num_tokens": 648118362.0, + "step": 25043 + }, + { + "epoch": 2.750274544256534, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.477485179901123, + "learning_rate": 1e-06, + "loss": 0.9023, + "mean_token_accuracy": 0.7342177629470825, + "num_tokens": 648142816.0, + "step": 25044 + }, + { + "epoch": 2.750384361959148, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4130868911743164, + "learning_rate": 1e-06, + "loss": 0.9101, + "mean_token_accuracy": 0.7306714653968811, + "num_tokens": 648170513.0, + "step": 25045 + }, + { + "epoch": 2.7504941796617617, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5079452991485596, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.715853750705719, + "num_tokens": 648195166.0, + "step": 25046 + }, + { + "epoch": 2.750603997364375, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.431232452392578, + "learning_rate": 1e-06, + "loss": 1.0091, + "mean_token_accuracy": 0.7047194838523865, + "num_tokens": 648222767.0, + "step": 25047 + }, + { + "epoch": 2.7507138150669888, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2801270484924316, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7077252864837646, + "num_tokens": 648255722.0, + "step": 25048 + }, + { + "epoch": 2.7508236327696025, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.555466413497925, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7267084121704102, + "num_tokens": 648280043.0, + "step": 25049 + }, + { + "epoch": 2.7509334504722163, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6187386512756348, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.684948742389679, + "num_tokens": 648307814.0, + "step": 25050 + }, + { + "epoch": 2.75104326817483, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.654933214187622, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6944023966789246, + "num_tokens": 648336217.0, + "step": 25051 + }, + { + "epoch": 2.7511530858774433, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.435821056365967, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7257927060127258, + "num_tokens": 648362643.0, + "step": 25052 + }, + { + "epoch": 2.751262903580057, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4872806072235107, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6946984529495239, + "num_tokens": 648389974.0, + "step": 25053 + }, + { + "epoch": 2.751372721282671, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4590117931365967, + "learning_rate": 1e-06, + "loss": 1.0825, + "mean_token_accuracy": 0.6889733076095581, + "num_tokens": 648418638.0, + "step": 25054 + }, + { + "epoch": 2.751482538985284, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.75850248336792, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7130294442176819, + "num_tokens": 648440188.0, + "step": 25055 + }, + { + "epoch": 2.7515923566878984, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6685070991516113, + "learning_rate": 1e-06, + "loss": 0.9127, + "mean_token_accuracy": 0.7306607365608215, + "num_tokens": 648463041.0, + "step": 25056 + }, + { + "epoch": 2.7517021743905117, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8272364139556885, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7170182466506958, + "num_tokens": 648489706.0, + "step": 25057 + }, + { + "epoch": 2.7518119920931254, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.659106731414795, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7016990184783936, + "num_tokens": 648513222.0, + "step": 25058 + }, + { + "epoch": 2.751921809795739, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6134767532348633, + "learning_rate": 1e-06, + "loss": 0.9701, + "mean_token_accuracy": 0.7199063301086426, + "num_tokens": 648538625.0, + "step": 25059 + }, + { + "epoch": 2.7520316274983525, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8268868923187256, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7111858129501343, + "num_tokens": 648562289.0, + "step": 25060 + }, + { + "epoch": 2.7521414452009663, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.457857847213745, + "learning_rate": 1e-06, + "loss": 1.0771, + "mean_token_accuracy": 0.6922141313552856, + "num_tokens": 648595217.0, + "step": 25061 + }, + { + "epoch": 2.75225126290358, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7164251804351807, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7079917192459106, + "num_tokens": 648620006.0, + "step": 25062 + }, + { + "epoch": 2.7523610806061938, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5581817626953125, + "learning_rate": 1e-06, + "loss": 1.1267, + "mean_token_accuracy": 0.6681994795799255, + "num_tokens": 648648475.0, + "step": 25063 + }, + { + "epoch": 2.7524708983088075, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.842078685760498, + "learning_rate": 1e-06, + "loss": 1.0038, + "mean_token_accuracy": 0.7101237773895264, + "num_tokens": 648672852.0, + "step": 25064 + }, + { + "epoch": 2.752580716011421, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5222795009613037, + "learning_rate": 1e-06, + "loss": 0.9273, + "mean_token_accuracy": 0.7343146204948425, + "num_tokens": 648698109.0, + "step": 25065 + }, + { + "epoch": 2.7526905337140346, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5115365982055664, + "learning_rate": 1e-06, + "loss": 1.0212, + "mean_token_accuracy": 0.7024797201156616, + "num_tokens": 648725701.0, + "step": 25066 + }, + { + "epoch": 2.7528003514166484, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.806105852127075, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.7347348928451538, + "num_tokens": 648749283.0, + "step": 25067 + }, + { + "epoch": 2.752910169119262, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.919919967651367, + "learning_rate": 1e-06, + "loss": 0.9153, + "mean_token_accuracy": 0.7342907190322876, + "num_tokens": 648771991.0, + "step": 25068 + }, + { + "epoch": 2.753019986821876, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6431996822357178, + "learning_rate": 1e-06, + "loss": 0.9436, + "mean_token_accuracy": 0.7226779460906982, + "num_tokens": 648795677.0, + "step": 25069 + }, + { + "epoch": 2.753129804524489, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6899731159210205, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7118810415267944, + "num_tokens": 648818931.0, + "step": 25070 + }, + { + "epoch": 2.753239622227103, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5971696376800537, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6917798519134521, + "num_tokens": 648847331.0, + "step": 25071 + }, + { + "epoch": 2.7533494399297167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.817392587661743, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6987614631652832, + "num_tokens": 648868817.0, + "step": 25072 + }, + { + "epoch": 2.7534592576323305, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 17.06941032409668, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.6986070871353149, + "num_tokens": 648895614.0, + "step": 25073 + }, + { + "epoch": 2.753569075334944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9124419689178467, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7147471904754639, + "num_tokens": 648921781.0, + "step": 25074 + }, + { + "epoch": 2.7536788930375575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4025018215179443, + "learning_rate": 1e-06, + "loss": 1.0683, + "mean_token_accuracy": 0.6894440054893494, + "num_tokens": 648952615.0, + "step": 25075 + }, + { + "epoch": 2.7537887107401713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9391849040985107, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7176870107650757, + "num_tokens": 648976681.0, + "step": 25076 + }, + { + "epoch": 2.753898528442785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2122433185577393, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7166674137115479, + "num_tokens": 649001451.0, + "step": 25077 + }, + { + "epoch": 2.754008346145399, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5093698501586914, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.685782790184021, + "num_tokens": 649030563.0, + "step": 25078 + }, + { + "epoch": 2.7541181638480126, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.809800148010254, + "learning_rate": 1e-06, + "loss": 1.0193, + "mean_token_accuracy": 0.7002972960472107, + "num_tokens": 649052981.0, + "step": 25079 + }, + { + "epoch": 2.754227981550626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4325203895568848, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.7100592851638794, + "num_tokens": 649081091.0, + "step": 25080 + }, + { + "epoch": 2.7543377992532396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.608360528945923, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7092170715332031, + "num_tokens": 649105045.0, + "step": 25081 + }, + { + "epoch": 2.7544476169558534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.516563653945923, + "learning_rate": 1e-06, + "loss": 0.9594, + "mean_token_accuracy": 0.7191413640975952, + "num_tokens": 649133027.0, + "step": 25082 + }, + { + "epoch": 2.7545574346584667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.856130599975586, + "learning_rate": 1e-06, + "loss": 0.937, + "mean_token_accuracy": 0.7228973507881165, + "num_tokens": 649151761.0, + "step": 25083 + }, + { + "epoch": 2.7546672523610805, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.657936096191406, + "learning_rate": 1e-06, + "loss": 1.0098, + "mean_token_accuracy": 0.7046746611595154, + "num_tokens": 649183703.0, + "step": 25084 + }, + { + "epoch": 2.754777070063694, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718000650405884, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.6991704702377319, + "num_tokens": 649209334.0, + "step": 25085 + }, + { + "epoch": 2.754886887766308, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8069159984588623, + "learning_rate": 1e-06, + "loss": 0.946, + "mean_token_accuracy": 0.7267407178878784, + "num_tokens": 649235028.0, + "step": 25086 + }, + { + "epoch": 2.7549967054689217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.954371690750122, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7083964347839355, + "num_tokens": 649258176.0, + "step": 25087 + }, + { + "epoch": 2.755106523171535, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1628851890563965, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7159811854362488, + "num_tokens": 649281539.0, + "step": 25088 + }, + { + "epoch": 2.755216340874149, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3921406269073486, + "learning_rate": 1e-06, + "loss": 1.1192, + "mean_token_accuracy": 0.6758006811141968, + "num_tokens": 649312403.0, + "step": 25089 + }, + { + "epoch": 2.7553261585767626, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.597099781036377, + "learning_rate": 1e-06, + "loss": 1.0859, + "mean_token_accuracy": 0.6884214878082275, + "num_tokens": 649341352.0, + "step": 25090 + }, + { + "epoch": 2.7554359762793763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0747358798980713, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7126787900924683, + "num_tokens": 649363178.0, + "step": 25091 + }, + { + "epoch": 2.75554579398199, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.003159523010254, + "learning_rate": 1e-06, + "loss": 0.923, + "mean_token_accuracy": 0.7254555821418762, + "num_tokens": 649381050.0, + "step": 25092 + }, + { + "epoch": 2.7556556116846034, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7478606700897217, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7153018712997437, + "num_tokens": 649405317.0, + "step": 25093 + }, + { + "epoch": 2.755765429387217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7402753829956055, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7162231802940369, + "num_tokens": 649430816.0, + "step": 25094 + }, + { + "epoch": 2.755875247089831, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.549387216567993, + "learning_rate": 1e-06, + "loss": 0.9167, + "mean_token_accuracy": 0.7311901450157166, + "num_tokens": 649456318.0, + "step": 25095 + }, + { + "epoch": 2.7559850647924446, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.698035955429077, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7092931866645813, + "num_tokens": 649480182.0, + "step": 25096 + }, + { + "epoch": 2.7560948824950584, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6995468139648438, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7128796577453613, + "num_tokens": 649504133.0, + "step": 25097 + }, + { + "epoch": 2.7562047001976717, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6150002479553223, + "learning_rate": 1e-06, + "loss": 1.0598, + "mean_token_accuracy": 0.7019001841545105, + "num_tokens": 649533342.0, + "step": 25098 + }, + { + "epoch": 2.7563145179002855, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3543894290924072, + "learning_rate": 1e-06, + "loss": 1.0685, + "mean_token_accuracy": 0.6900050044059753, + "num_tokens": 649567140.0, + "step": 25099 + }, + { + "epoch": 2.7564243356028992, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5842478275299072, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7112659215927124, + "num_tokens": 649591643.0, + "step": 25100 + }, + { + "epoch": 2.756534153305513, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.2412211894989014, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.689091682434082, + "num_tokens": 649624483.0, + "step": 25101 + }, + { + "epoch": 2.7566439710081267, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5355539321899414, + "learning_rate": 1e-06, + "loss": 0.8527, + "mean_token_accuracy": 0.7482115626335144, + "num_tokens": 649650361.0, + "step": 25102 + }, + { + "epoch": 2.75675378871074, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9100871086120605, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.6949750781059265, + "num_tokens": 649670929.0, + "step": 25103 + }, + { + "epoch": 2.756863606413354, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.243795156478882, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7093931436538696, + "num_tokens": 649704762.0, + "step": 25104 + }, + { + "epoch": 2.7569734241159676, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.753283739089966, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.6856681108474731, + "num_tokens": 649728888.0, + "step": 25105 + }, + { + "epoch": 2.757083241818581, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5683112144470215, + "learning_rate": 1e-06, + "loss": 1.121, + "mean_token_accuracy": 0.6780074834823608, + "num_tokens": 649756537.0, + "step": 25106 + }, + { + "epoch": 2.757193059521195, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8371081352233887, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7069652080535889, + "num_tokens": 649784537.0, + "step": 25107 + }, + { + "epoch": 2.7573028772238084, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6220877170562744, + "learning_rate": 1e-06, + "loss": 1.094, + "mean_token_accuracy": 0.6831737160682678, + "num_tokens": 649812101.0, + "step": 25108 + }, + { + "epoch": 2.757412694926422, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.493079900741577, + "learning_rate": 1e-06, + "loss": 0.9513, + "mean_token_accuracy": 0.7185577154159546, + "num_tokens": 649840318.0, + "step": 25109 + }, + { + "epoch": 2.757522512629036, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.617828607559204, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7210714817047119, + "num_tokens": 649865001.0, + "step": 25110 + }, + { + "epoch": 2.7576323303316492, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6173853874206543, + "learning_rate": 1e-06, + "loss": 0.9262, + "mean_token_accuracy": 0.7252652645111084, + "num_tokens": 649889271.0, + "step": 25111 + }, + { + "epoch": 2.757742148034263, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.510019063949585, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6982113122940063, + "num_tokens": 649916816.0, + "step": 25112 + }, + { + "epoch": 2.7578519657368767, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8537564277648926, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.6900969743728638, + "num_tokens": 649941801.0, + "step": 25113 + }, + { + "epoch": 2.7579617834394905, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7845678329467773, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7081387042999268, + "num_tokens": 649964154.0, + "step": 25114 + }, + { + "epoch": 2.7580716011421043, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.400108575820923, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.698296070098877, + "num_tokens": 649994431.0, + "step": 25115 + }, + { + "epoch": 2.7581814188447176, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.493257522583008, + "learning_rate": 1e-06, + "loss": 0.9937, + "mean_token_accuracy": 0.712215781211853, + "num_tokens": 650023704.0, + "step": 25116 + }, + { + "epoch": 2.7582912365473313, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7457120418548584, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7033960819244385, + "num_tokens": 650046884.0, + "step": 25117 + }, + { + "epoch": 2.758401054249945, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6075196266174316, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7266978025436401, + "num_tokens": 650071453.0, + "step": 25118 + }, + { + "epoch": 2.758510871952559, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4878087043762207, + "learning_rate": 1e-06, + "loss": 0.9439, + "mean_token_accuracy": 0.7233147621154785, + "num_tokens": 650096453.0, + "step": 25119 + }, + { + "epoch": 2.7586206896551726, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6545467376708984, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.7061270475387573, + "num_tokens": 650120391.0, + "step": 25120 + }, + { + "epoch": 2.758730507357786, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.696186065673828, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.6993013620376587, + "num_tokens": 650144385.0, + "step": 25121 + }, + { + "epoch": 2.7588403250603997, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.3348751068115234, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6937246918678284, + "num_tokens": 650169422.0, + "step": 25122 + }, + { + "epoch": 2.7589501427630134, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.593661069869995, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7052006721496582, + "num_tokens": 650194483.0, + "step": 25123 + }, + { + "epoch": 2.759059960465627, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6016716957092285, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.6994940042495728, + "num_tokens": 650220021.0, + "step": 25124 + }, + { + "epoch": 2.759169778168241, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6346747875213623, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7051740288734436, + "num_tokens": 650246245.0, + "step": 25125 + }, + { + "epoch": 2.7592795958708543, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.588834524154663, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6950573921203613, + "num_tokens": 650272698.0, + "step": 25126 + }, + { + "epoch": 2.759389413573468, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7280967235565186, + "learning_rate": 1e-06, + "loss": 0.9887, + "mean_token_accuracy": 0.7120985984802246, + "num_tokens": 650297928.0, + "step": 25127 + }, + { + "epoch": 2.7594992312760818, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.682077646255493, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.6968758702278137, + "num_tokens": 650323221.0, + "step": 25128 + }, + { + "epoch": 2.7596090489786955, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3330039978027344, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7172318696975708, + "num_tokens": 650354418.0, + "step": 25129 + }, + { + "epoch": 2.7597188666813093, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6425812244415283, + "learning_rate": 1e-06, + "loss": 0.8823, + "mean_token_accuracy": 0.7346118688583374, + "num_tokens": 650376365.0, + "step": 25130 + }, + { + "epoch": 2.7598286843839226, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5715954303741455, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6940508484840393, + "num_tokens": 650400045.0, + "step": 25131 + }, + { + "epoch": 2.7599385020865363, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9661388397216797, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7124311327934265, + "num_tokens": 650418615.0, + "step": 25132 + }, + { + "epoch": 2.76004831978915, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.513988733291626, + "learning_rate": 1e-06, + "loss": 0.9244, + "mean_token_accuracy": 0.7298973798751831, + "num_tokens": 650443488.0, + "step": 25133 + }, + { + "epoch": 2.7601581374917634, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4221460819244385, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.700444221496582, + "num_tokens": 650473019.0, + "step": 25134 + }, + { + "epoch": 2.760267955194377, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5982208251953125, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.7156636714935303, + "num_tokens": 650499245.0, + "step": 25135 + }, + { + "epoch": 2.760377772896991, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7166285514831543, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7082499861717224, + "num_tokens": 650522784.0, + "step": 25136 + }, + { + "epoch": 2.7604875905996047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.766937017440796, + "learning_rate": 1e-06, + "loss": 0.9676, + "mean_token_accuracy": 0.7119370698928833, + "num_tokens": 650543810.0, + "step": 25137 + }, + { + "epoch": 2.7605974083022184, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.727346420288086, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7195094227790833, + "num_tokens": 650565417.0, + "step": 25138 + }, + { + "epoch": 2.7607072260048318, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.447458505630493, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.712066650390625, + "num_tokens": 650596977.0, + "step": 25139 + }, + { + "epoch": 2.7608170437074455, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6256392002105713, + "learning_rate": 1e-06, + "loss": 1.0258, + "mean_token_accuracy": 0.6979715824127197, + "num_tokens": 650621860.0, + "step": 25140 + }, + { + "epoch": 2.7609268614100593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6570725440979004, + "learning_rate": 1e-06, + "loss": 1.0206, + "mean_token_accuracy": 0.7068402767181396, + "num_tokens": 650646498.0, + "step": 25141 + }, + { + "epoch": 2.761036679112673, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0552899837493896, + "learning_rate": 1e-06, + "loss": 0.9204, + "mean_token_accuracy": 0.728705883026123, + "num_tokens": 650664077.0, + "step": 25142 + }, + { + "epoch": 2.761146496815287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.784846067428589, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.7019045948982239, + "num_tokens": 650686151.0, + "step": 25143 + }, + { + "epoch": 2.7612563145179, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.672041416168213, + "learning_rate": 1e-06, + "loss": 1.1128, + "mean_token_accuracy": 0.6784051656723022, + "num_tokens": 650714203.0, + "step": 25144 + }, + { + "epoch": 2.761366132220514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.91039776802063, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7307063937187195, + "num_tokens": 650734054.0, + "step": 25145 + }, + { + "epoch": 2.7614759499231276, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.234166145324707, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.687292754650116, + "num_tokens": 650761726.0, + "step": 25146 + }, + { + "epoch": 2.7615857676257414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8197107315063477, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7107795476913452, + "num_tokens": 650784206.0, + "step": 25147 + }, + { + "epoch": 2.761695585328355, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.798999786376953, + "learning_rate": 1e-06, + "loss": 1.0065, + "mean_token_accuracy": 0.6990728378295898, + "num_tokens": 650807736.0, + "step": 25148 + }, + { + "epoch": 2.7618054030309684, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7537484169006348, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7183825969696045, + "num_tokens": 650830542.0, + "step": 25149 + }, + { + "epoch": 2.761915220733582, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.561598539352417, + "learning_rate": 1e-06, + "loss": 0.9622, + "mean_token_accuracy": 0.7158535718917847, + "num_tokens": 650859543.0, + "step": 25150 + }, + { + "epoch": 2.762025038436196, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.730318546295166, + "learning_rate": 1e-06, + "loss": 0.9475, + "mean_token_accuracy": 0.720065712928772, + "num_tokens": 650881841.0, + "step": 25151 + }, + { + "epoch": 2.7621348561388097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.68520188331604, + "learning_rate": 1e-06, + "loss": 0.9259, + "mean_token_accuracy": 0.7234829068183899, + "num_tokens": 650906005.0, + "step": 25152 + }, + { + "epoch": 2.7622446738414235, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3922927379608154, + "learning_rate": 1e-06, + "loss": 0.9789, + "mean_token_accuracy": 0.7147879600524902, + "num_tokens": 650933378.0, + "step": 25153 + }, + { + "epoch": 2.762354491544037, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.556708812713623, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.6915737390518188, + "num_tokens": 650960849.0, + "step": 25154 + }, + { + "epoch": 2.7624643092466505, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3502449989318848, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7058308720588684, + "num_tokens": 650990245.0, + "step": 25155 + }, + { + "epoch": 2.7625741269492643, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9738686084747314, + "learning_rate": 1e-06, + "loss": 0.947, + "mean_token_accuracy": 0.7184909582138062, + "num_tokens": 651007513.0, + "step": 25156 + }, + { + "epoch": 2.7626839446518776, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7341158390045166, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.7095234394073486, + "num_tokens": 651033184.0, + "step": 25157 + }, + { + "epoch": 2.762793762354492, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5571885108947754, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7070982456207275, + "num_tokens": 651061072.0, + "step": 25158 + }, + { + "epoch": 2.762903580057105, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6859452724456787, + "learning_rate": 1e-06, + "loss": 1.122, + "mean_token_accuracy": 0.679573655128479, + "num_tokens": 651086402.0, + "step": 25159 + }, + { + "epoch": 2.763013397759719, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.601276397705078, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.7049462199211121, + "num_tokens": 651110355.0, + "step": 25160 + }, + { + "epoch": 2.7631232154623326, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.59716534614563, + "learning_rate": 1e-06, + "loss": 0.8199, + "mean_token_accuracy": 0.7618987560272217, + "num_tokens": 651132937.0, + "step": 25161 + }, + { + "epoch": 2.763233033164946, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.720215320587158, + "learning_rate": 1e-06, + "loss": 1.0518, + "mean_token_accuracy": 0.7065999507904053, + "num_tokens": 651159254.0, + "step": 25162 + }, + { + "epoch": 2.7633428508675597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6499712467193604, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.695257306098938, + "num_tokens": 651183771.0, + "step": 25163 + }, + { + "epoch": 2.7634526685701735, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.541923761367798, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.7101922631263733, + "num_tokens": 651210647.0, + "step": 25164 + }, + { + "epoch": 2.763562486272787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7384300231933594, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7055286169052124, + "num_tokens": 651232450.0, + "step": 25165 + }, + { + "epoch": 2.763672303975401, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.808244705200195, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.6879478693008423, + "num_tokens": 651262363.0, + "step": 25166 + }, + { + "epoch": 2.7637821216780143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.546229600906372, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6996785402297974, + "num_tokens": 651294423.0, + "step": 25167 + }, + { + "epoch": 2.763891939380628, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6252381801605225, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.6983630657196045, + "num_tokens": 651320244.0, + "step": 25168 + }, + { + "epoch": 2.764001757083242, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5411393642425537, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7074693441390991, + "num_tokens": 651345228.0, + "step": 25169 + }, + { + "epoch": 2.7641115747858556, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5069587230682373, + "learning_rate": 1e-06, + "loss": 0.9557, + "mean_token_accuracy": 0.7158468961715698, + "num_tokens": 651370252.0, + "step": 25170 + }, + { + "epoch": 2.7642213924884693, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4148013591766357, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7189528942108154, + "num_tokens": 651399536.0, + "step": 25171 + }, + { + "epoch": 2.7643312101910826, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4504587650299072, + "learning_rate": 1e-06, + "loss": 1.0906, + "mean_token_accuracy": 0.6865293979644775, + "num_tokens": 651427551.0, + "step": 25172 + }, + { + "epoch": 2.7644410278936964, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.500685930252075, + "learning_rate": 1e-06, + "loss": 1.0656, + "mean_token_accuracy": 0.6980780363082886, + "num_tokens": 651456709.0, + "step": 25173 + }, + { + "epoch": 2.76455084559631, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.918442726135254, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7035707831382751, + "num_tokens": 651483211.0, + "step": 25174 + }, + { + "epoch": 2.764660663298924, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9008049964904785, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7216180562973022, + "num_tokens": 651503020.0, + "step": 25175 + }, + { + "epoch": 2.7647704810015377, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6411304473876953, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7048075199127197, + "num_tokens": 651527231.0, + "step": 25176 + }, + { + "epoch": 2.764880298704151, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4923675060272217, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7201203107833862, + "num_tokens": 651554015.0, + "step": 25177 + }, + { + "epoch": 2.7649901164067647, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5940165519714355, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7095679044723511, + "num_tokens": 651580250.0, + "step": 25178 + }, + { + "epoch": 2.7650999341093785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.652358055114746, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.713305652141571, + "num_tokens": 651606831.0, + "step": 25179 + }, + { + "epoch": 2.7652097518119922, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6557679176330566, + "learning_rate": 1e-06, + "loss": 1.0453, + "mean_token_accuracy": 0.7091314196586609, + "num_tokens": 651629974.0, + "step": 25180 + }, + { + "epoch": 2.765319569514606, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4706621170043945, + "learning_rate": 1e-06, + "loss": 1.0032, + "mean_token_accuracy": 0.7104427814483643, + "num_tokens": 651657057.0, + "step": 25181 + }, + { + "epoch": 2.7654293872172193, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.517893075942993, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7222145795822144, + "num_tokens": 651682385.0, + "step": 25182 + }, + { + "epoch": 2.765539204919833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6228840351104736, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.701735258102417, + "num_tokens": 651706546.0, + "step": 25183 + }, + { + "epoch": 2.765649022622447, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7306487560272217, + "learning_rate": 1e-06, + "loss": 0.9466, + "mean_token_accuracy": 0.7280791401863098, + "num_tokens": 651734169.0, + "step": 25184 + }, + { + "epoch": 2.76575884032506, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5618650913238525, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.713349461555481, + "num_tokens": 651760196.0, + "step": 25185 + }, + { + "epoch": 2.765868658027674, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3951220512390137, + "learning_rate": 1e-06, + "loss": 1.1075, + "mean_token_accuracy": 0.6953117251396179, + "num_tokens": 651791156.0, + "step": 25186 + }, + { + "epoch": 2.7659784757302877, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.015028715133667, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7293044924736023, + "num_tokens": 651809139.0, + "step": 25187 + }, + { + "epoch": 2.7660882934329014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7965457439422607, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7214375734329224, + "num_tokens": 651829919.0, + "step": 25188 + }, + { + "epoch": 2.766198111135515, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.435326099395752, + "learning_rate": 1e-06, + "loss": 1.0801, + "mean_token_accuracy": 0.691394567489624, + "num_tokens": 651859312.0, + "step": 25189 + }, + { + "epoch": 2.7663079288381285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7918524742126465, + "learning_rate": 1e-06, + "loss": 0.9848, + "mean_token_accuracy": 0.7072048187255859, + "num_tokens": 651889055.0, + "step": 25190 + }, + { + "epoch": 2.7664177465407422, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.912753105163574, + "learning_rate": 1e-06, + "loss": 0.8379, + "mean_token_accuracy": 0.7464572191238403, + "num_tokens": 651906880.0, + "step": 25191 + }, + { + "epoch": 2.766527564243356, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8426945209503174, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.7182668447494507, + "num_tokens": 651929096.0, + "step": 25192 + }, + { + "epoch": 2.7666373819459698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0873684883117676, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.6929757595062256, + "num_tokens": 651951758.0, + "step": 25193 + }, + { + "epoch": 2.7667471996485835, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.622633218765259, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.704465925693512, + "num_tokens": 651976696.0, + "step": 25194 + }, + { + "epoch": 2.766857017351197, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5106871128082275, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7203381061553955, + "num_tokens": 652004851.0, + "step": 25195 + }, + { + "epoch": 2.7669668350538106, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.736482620239258, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.7065092325210571, + "num_tokens": 652027235.0, + "step": 25196 + }, + { + "epoch": 2.7670766527564243, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5742058753967285, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.7038989067077637, + "num_tokens": 652053645.0, + "step": 25197 + }, + { + "epoch": 2.767186470459038, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7049691677093506, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6956213116645813, + "num_tokens": 652080875.0, + "step": 25198 + }, + { + "epoch": 2.767296288161652, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.804014205932617, + "learning_rate": 1e-06, + "loss": 1.0507, + "mean_token_accuracy": 0.6961139440536499, + "num_tokens": 652105946.0, + "step": 25199 + }, + { + "epoch": 2.767406105864265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.702120304107666, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7088255882263184, + "num_tokens": 652132432.0, + "step": 25200 + }, + { + "epoch": 2.767515923566879, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5937607288360596, + "learning_rate": 1e-06, + "loss": 1.1057, + "mean_token_accuracy": 0.6792899370193481, + "num_tokens": 652168571.0, + "step": 25201 + }, + { + "epoch": 2.7676257412694927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.698636293411255, + "learning_rate": 1e-06, + "loss": 0.9931, + "mean_token_accuracy": 0.7144179344177246, + "num_tokens": 652192561.0, + "step": 25202 + }, + { + "epoch": 2.7677355589721064, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.726494789123535, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.706591010093689, + "num_tokens": 652216199.0, + "step": 25203 + }, + { + "epoch": 2.76784537667472, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7056775093078613, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6920703053474426, + "num_tokens": 652240440.0, + "step": 25204 + }, + { + "epoch": 2.7679551943773335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6692473888397217, + "learning_rate": 1e-06, + "loss": 1.0587, + "mean_token_accuracy": 0.7025705575942993, + "num_tokens": 652264582.0, + "step": 25205 + }, + { + "epoch": 2.7680650120799473, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.475950002670288, + "learning_rate": 1e-06, + "loss": 0.9384, + "mean_token_accuracy": 0.7264400720596313, + "num_tokens": 652291585.0, + "step": 25206 + }, + { + "epoch": 2.768174829782561, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.465426445007324, + "learning_rate": 1e-06, + "loss": 1.0994, + "mean_token_accuracy": 0.6889489889144897, + "num_tokens": 652323196.0, + "step": 25207 + }, + { + "epoch": 2.7682846474851748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4094855785369873, + "learning_rate": 1e-06, + "loss": 1.0981, + "mean_token_accuracy": 0.6810996532440186, + "num_tokens": 652355899.0, + "step": 25208 + }, + { + "epoch": 2.7683944651877885, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.606384038925171, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.6822172999382019, + "num_tokens": 652381810.0, + "step": 25209 + }, + { + "epoch": 2.768504282890402, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8085505962371826, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7256208658218384, + "num_tokens": 652403246.0, + "step": 25210 + }, + { + "epoch": 2.7686141005930156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.871462345123291, + "learning_rate": 1e-06, + "loss": 1.045, + "mean_token_accuracy": 0.6913636922836304, + "num_tokens": 652427886.0, + "step": 25211 + }, + { + "epoch": 2.7687239182956294, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4676167964935303, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7127453088760376, + "num_tokens": 652454926.0, + "step": 25212 + }, + { + "epoch": 2.7688337359982427, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.512110471725464, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7240115404129028, + "num_tokens": 652481219.0, + "step": 25213 + }, + { + "epoch": 2.7689435537008564, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.9299545288085938, + "learning_rate": 1e-06, + "loss": 1.0414, + "mean_token_accuracy": 0.6969757080078125, + "num_tokens": 652507642.0, + "step": 25214 + }, + { + "epoch": 2.76905337140347, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5913305282592773, + "learning_rate": 1e-06, + "loss": 0.9146, + "mean_token_accuracy": 0.7367466688156128, + "num_tokens": 652530668.0, + "step": 25215 + }, + { + "epoch": 2.769163189106084, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6722805500030518, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7129205465316772, + "num_tokens": 652552515.0, + "step": 25216 + }, + { + "epoch": 2.7692730068086977, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7624311447143555, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7261549830436707, + "num_tokens": 652572944.0, + "step": 25217 + }, + { + "epoch": 2.769382824511311, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.419344663619995, + "learning_rate": 1e-06, + "loss": 1.0079, + "mean_token_accuracy": 0.7014051675796509, + "num_tokens": 652601639.0, + "step": 25218 + }, + { + "epoch": 2.7694926422139248, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6087186336517334, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6886337399482727, + "num_tokens": 652626592.0, + "step": 25219 + }, + { + "epoch": 2.7696024599165385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.419482707977295, + "learning_rate": 1e-06, + "loss": 1.102, + "mean_token_accuracy": 0.6795880794525146, + "num_tokens": 652655322.0, + "step": 25220 + }, + { + "epoch": 2.7697122776191523, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6317179203033447, + "learning_rate": 1e-06, + "loss": 0.9438, + "mean_token_accuracy": 0.725935161113739, + "num_tokens": 652679234.0, + "step": 25221 + }, + { + "epoch": 2.769822095321766, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9243416786193848, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7087839841842651, + "num_tokens": 652699212.0, + "step": 25222 + }, + { + "epoch": 2.7699319130243794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5517258644104004, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7150627374649048, + "num_tokens": 652726972.0, + "step": 25223 + }, + { + "epoch": 2.770041730726993, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7365517616271973, + "learning_rate": 1e-06, + "loss": 0.9576, + "mean_token_accuracy": 0.7273170948028564, + "num_tokens": 652748924.0, + "step": 25224 + }, + { + "epoch": 2.770151548429607, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6352391242980957, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.701022744178772, + "num_tokens": 652772025.0, + "step": 25225 + }, + { + "epoch": 2.7702613661322206, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.881514310836792, + "learning_rate": 1e-06, + "loss": 0.9398, + "mean_token_accuracy": 0.7256169319152832, + "num_tokens": 652798692.0, + "step": 25226 + }, + { + "epoch": 2.7703711838348344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.599694013595581, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.703255295753479, + "num_tokens": 652830187.0, + "step": 25227 + }, + { + "epoch": 2.7704810015374477, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.720102310180664, + "learning_rate": 1e-06, + "loss": 0.9553, + "mean_token_accuracy": 0.7172249555587769, + "num_tokens": 652853772.0, + "step": 25228 + }, + { + "epoch": 2.7705908192400615, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.788409471511841, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7073935866355896, + "num_tokens": 652879406.0, + "step": 25229 + }, + { + "epoch": 2.770700636942675, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6142921447753906, + "learning_rate": 1e-06, + "loss": 0.8962, + "mean_token_accuracy": 0.7273266911506653, + "num_tokens": 652904322.0, + "step": 25230 + }, + { + "epoch": 2.770810454645289, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.47279953956604, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7133809328079224, + "num_tokens": 652932569.0, + "step": 25231 + }, + { + "epoch": 2.7709202723479027, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.589390277862549, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.7066145539283752, + "num_tokens": 652957935.0, + "step": 25232 + }, + { + "epoch": 2.771030090050516, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4811818599700928, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.701881468296051, + "num_tokens": 652983410.0, + "step": 25233 + }, + { + "epoch": 2.77113990775313, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6671745777130127, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.6972600221633911, + "num_tokens": 653009975.0, + "step": 25234 + }, + { + "epoch": 2.7712497254557436, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.436192750930786, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7228344082832336, + "num_tokens": 653036431.0, + "step": 25235 + }, + { + "epoch": 2.771359543158357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7014317512512207, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7232098579406738, + "num_tokens": 653058546.0, + "step": 25236 + }, + { + "epoch": 2.771469360860971, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.736161708831787, + "learning_rate": 1e-06, + "loss": 0.8892, + "mean_token_accuracy": 0.738073468208313, + "num_tokens": 653083335.0, + "step": 25237 + }, + { + "epoch": 2.7715791785635844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5287327766418457, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.722139835357666, + "num_tokens": 653110027.0, + "step": 25238 + }, + { + "epoch": 2.771688996266198, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4950921535491943, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7148869037628174, + "num_tokens": 653136265.0, + "step": 25239 + }, + { + "epoch": 2.771798813968812, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7261834144592285, + "learning_rate": 1e-06, + "loss": 1.0142, + "mean_token_accuracy": 0.7034886479377747, + "num_tokens": 653159842.0, + "step": 25240 + }, + { + "epoch": 2.771908631671425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.514342784881592, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7064517140388489, + "num_tokens": 653186815.0, + "step": 25241 + }, + { + "epoch": 2.772018449374039, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.654514789581299, + "learning_rate": 1e-06, + "loss": 1.0987, + "mean_token_accuracy": 0.6962733268737793, + "num_tokens": 653211266.0, + "step": 25242 + }, + { + "epoch": 2.7721282670766527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.643968343734741, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.6973623037338257, + "num_tokens": 653240510.0, + "step": 25243 + }, + { + "epoch": 2.7722380847792665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0915277004241943, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6935351490974426, + "num_tokens": 653261176.0, + "step": 25244 + }, + { + "epoch": 2.7723479024818802, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5880489349365234, + "learning_rate": 1e-06, + "loss": 1.0843, + "mean_token_accuracy": 0.6890237927436829, + "num_tokens": 653290730.0, + "step": 25245 + }, + { + "epoch": 2.7724577201844935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5491347312927246, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.697548508644104, + "num_tokens": 653316082.0, + "step": 25246 + }, + { + "epoch": 2.7725675378871073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5379207134246826, + "learning_rate": 1e-06, + "loss": 0.9542, + "mean_token_accuracy": 0.7206398844718933, + "num_tokens": 653341875.0, + "step": 25247 + }, + { + "epoch": 2.772677355589721, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9800662994384766, + "learning_rate": 1e-06, + "loss": 0.922, + "mean_token_accuracy": 0.7266163229942322, + "num_tokens": 653366545.0, + "step": 25248 + }, + { + "epoch": 2.772787173292335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.742541790008545, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.7013031840324402, + "num_tokens": 653391297.0, + "step": 25249 + }, + { + "epoch": 2.7728969909949486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1927595138549805, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.7103868126869202, + "num_tokens": 653414622.0, + "step": 25250 + }, + { + "epoch": 2.773006808697562, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6416983604431152, + "learning_rate": 1e-06, + "loss": 0.9251, + "mean_token_accuracy": 0.7280457615852356, + "num_tokens": 653438389.0, + "step": 25251 + }, + { + "epoch": 2.7731166264001756, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.606565475463867, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.7030104398727417, + "num_tokens": 653462007.0, + "step": 25252 + }, + { + "epoch": 2.7732264441027894, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.425337791442871, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6979787349700928, + "num_tokens": 653491402.0, + "step": 25253 + }, + { + "epoch": 2.773336261805403, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4264745712280273, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6884572505950928, + "num_tokens": 653521194.0, + "step": 25254 + }, + { + "epoch": 2.773446079508017, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.772773027420044, + "learning_rate": 1e-06, + "loss": 1.0494, + "mean_token_accuracy": 0.7002319097518921, + "num_tokens": 653546587.0, + "step": 25255 + }, + { + "epoch": 2.7735558972106302, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5407097339630127, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.7027312517166138, + "num_tokens": 653572844.0, + "step": 25256 + }, + { + "epoch": 2.773665714913244, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3453190326690674, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.69612717628479, + "num_tokens": 653604279.0, + "step": 25257 + }, + { + "epoch": 2.7737755326158577, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3988800048828125, + "learning_rate": 1e-06, + "loss": 1.0317, + "mean_token_accuracy": 0.6990172266960144, + "num_tokens": 653630499.0, + "step": 25258 + }, + { + "epoch": 2.7738853503184715, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6284070014953613, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7067962288856506, + "num_tokens": 653652860.0, + "step": 25259 + }, + { + "epoch": 2.7739951680210853, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7406418323516846, + "learning_rate": 1e-06, + "loss": 1.0253, + "mean_token_accuracy": 0.698147177696228, + "num_tokens": 653678067.0, + "step": 25260 + }, + { + "epoch": 2.7741049857236986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7276902198791504, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7036343812942505, + "num_tokens": 653705602.0, + "step": 25261 + }, + { + "epoch": 2.7742148034263123, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.751812219619751, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.733208179473877, + "num_tokens": 653728381.0, + "step": 25262 + }, + { + "epoch": 2.774324621128926, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7038657665252686, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7034540176391602, + "num_tokens": 653752289.0, + "step": 25263 + }, + { + "epoch": 2.7744344388315394, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9183552265167236, + "learning_rate": 1e-06, + "loss": 0.9417, + "mean_token_accuracy": 0.7210487723350525, + "num_tokens": 653776572.0, + "step": 25264 + }, + { + "epoch": 2.774544256534153, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.701258897781372, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7145166397094727, + "num_tokens": 653801743.0, + "step": 25265 + }, + { + "epoch": 2.774654074236767, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5962576866149902, + "learning_rate": 1e-06, + "loss": 1.0661, + "mean_token_accuracy": 0.6900120377540588, + "num_tokens": 653828938.0, + "step": 25266 + }, + { + "epoch": 2.7747638919393807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5624241828918457, + "learning_rate": 1e-06, + "loss": 1.0233, + "mean_token_accuracy": 0.7056819796562195, + "num_tokens": 653855618.0, + "step": 25267 + }, + { + "epoch": 2.7748737096419944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6987719535827637, + "learning_rate": 1e-06, + "loss": 0.9558, + "mean_token_accuracy": 0.7211729288101196, + "num_tokens": 653878047.0, + "step": 25268 + }, + { + "epoch": 2.7749835273446077, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.343775749206543, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6832695603370667, + "num_tokens": 653910661.0, + "step": 25269 + }, + { + "epoch": 2.7750933450472215, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8193039894104004, + "learning_rate": 1e-06, + "loss": 0.9483, + "mean_token_accuracy": 0.7222452163696289, + "num_tokens": 653931833.0, + "step": 25270 + }, + { + "epoch": 2.7752031627498353, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6521010398864746, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.700459897518158, + "num_tokens": 653956526.0, + "step": 25271 + }, + { + "epoch": 2.775312980452449, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.620161771774292, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.701454222202301, + "num_tokens": 653981615.0, + "step": 25272 + }, + { + "epoch": 2.7754227981550628, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.595099925994873, + "learning_rate": 1e-06, + "loss": 0.9853, + "mean_token_accuracy": 0.7148860096931458, + "num_tokens": 654009960.0, + "step": 25273 + }, + { + "epoch": 2.775532615857676, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7112843990325928, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.706081748008728, + "num_tokens": 654032444.0, + "step": 25274 + }, + { + "epoch": 2.77564243356029, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.092866897583008, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7063769102096558, + "num_tokens": 654051166.0, + "step": 25275 + }, + { + "epoch": 2.7757522512629036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.736457347869873, + "learning_rate": 1e-06, + "loss": 1.0423, + "mean_token_accuracy": 0.6920729279518127, + "num_tokens": 654076902.0, + "step": 25276 + }, + { + "epoch": 2.7758620689655173, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.909123420715332, + "learning_rate": 1e-06, + "loss": 0.9203, + "mean_token_accuracy": 0.7354937791824341, + "num_tokens": 654096938.0, + "step": 25277 + }, + { + "epoch": 2.775971886668131, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6156399250030518, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7164438962936401, + "num_tokens": 654122767.0, + "step": 25278 + }, + { + "epoch": 2.7760817043707444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7253081798553467, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7088744640350342, + "num_tokens": 654145677.0, + "step": 25279 + }, + { + "epoch": 2.776191522073358, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.720580816268921, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7123334407806396, + "num_tokens": 654169535.0, + "step": 25280 + }, + { + "epoch": 2.776301339775972, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.790590763092041, + "learning_rate": 1e-06, + "loss": 0.9006, + "mean_token_accuracy": 0.7406666874885559, + "num_tokens": 654192822.0, + "step": 25281 + }, + { + "epoch": 2.7764111574785857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.368157386779785, + "learning_rate": 1e-06, + "loss": 1.0312, + "mean_token_accuracy": 0.700971245765686, + "num_tokens": 654225419.0, + "step": 25282 + }, + { + "epoch": 2.7765209751811994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5197627544403076, + "learning_rate": 1e-06, + "loss": 0.9732, + "mean_token_accuracy": 0.7117583751678467, + "num_tokens": 654249585.0, + "step": 25283 + }, + { + "epoch": 2.7766307928838128, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.632596492767334, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7071386575698853, + "num_tokens": 654273537.0, + "step": 25284 + }, + { + "epoch": 2.7767406105864265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.470186233520508, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.6970573663711548, + "num_tokens": 654303510.0, + "step": 25285 + }, + { + "epoch": 2.7768504282890403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3699803352355957, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6935648918151855, + "num_tokens": 654333327.0, + "step": 25286 + }, + { + "epoch": 2.7769602459916536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.991828680038452, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7088272571563721, + "num_tokens": 654353996.0, + "step": 25287 + }, + { + "epoch": 2.777070063694268, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.976006031036377, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7226357460021973, + "num_tokens": 654374873.0, + "step": 25288 + }, + { + "epoch": 2.777179881396881, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8425378799438477, + "learning_rate": 1e-06, + "loss": 0.948, + "mean_token_accuracy": 0.723125696182251, + "num_tokens": 654397303.0, + "step": 25289 + }, + { + "epoch": 2.777289699099495, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.760572910308838, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.6949939727783203, + "num_tokens": 654421715.0, + "step": 25290 + }, + { + "epoch": 2.7773995168021086, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5914034843444824, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7003545761108398, + "num_tokens": 654446136.0, + "step": 25291 + }, + { + "epoch": 2.777509334504722, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6127490997314453, + "learning_rate": 1e-06, + "loss": 0.8866, + "mean_token_accuracy": 0.7351269721984863, + "num_tokens": 654469152.0, + "step": 25292 + }, + { + "epoch": 2.7776191522073357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.604275703430176, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7462071180343628, + "num_tokens": 654495377.0, + "step": 25293 + }, + { + "epoch": 2.7777289699099494, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8744616508483887, + "learning_rate": 1e-06, + "loss": 1.018, + "mean_token_accuracy": 0.7056422829627991, + "num_tokens": 654516281.0, + "step": 25294 + }, + { + "epoch": 2.777838787612563, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.705467462539673, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.714237630367279, + "num_tokens": 654540311.0, + "step": 25295 + }, + { + "epoch": 2.777948605315177, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.845954656600952, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7121159434318542, + "num_tokens": 654563102.0, + "step": 25296 + }, + { + "epoch": 2.7780584230177903, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3725337982177734, + "learning_rate": 1e-06, + "loss": 1.0773, + "mean_token_accuracy": 0.6943215727806091, + "num_tokens": 654592510.0, + "step": 25297 + }, + { + "epoch": 2.778168240720404, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5997180938720703, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7156223058700562, + "num_tokens": 654615826.0, + "step": 25298 + }, + { + "epoch": 2.778278058423018, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718503952026367, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6978906393051147, + "num_tokens": 654641014.0, + "step": 25299 + }, + { + "epoch": 2.7783878761256315, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5367965698242188, + "learning_rate": 1e-06, + "loss": 0.9684, + "mean_token_accuracy": 0.7186356782913208, + "num_tokens": 654669484.0, + "step": 25300 + }, + { + "epoch": 2.7784976938282453, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7519009113311768, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7018704414367676, + "num_tokens": 654691943.0, + "step": 25301 + }, + { + "epoch": 2.7786075115308586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4106709957122803, + "learning_rate": 1e-06, + "loss": 0.9891, + "mean_token_accuracy": 0.7177609801292419, + "num_tokens": 654724098.0, + "step": 25302 + }, + { + "epoch": 2.7787173292334724, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.54019832611084, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.7016463279724121, + "num_tokens": 654751649.0, + "step": 25303 + }, + { + "epoch": 2.778827146936086, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2490150928497314, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7188612818717957, + "num_tokens": 654774086.0, + "step": 25304 + }, + { + "epoch": 2.7789369646387, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5558297634124756, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.690036952495575, + "num_tokens": 654800126.0, + "step": 25305 + }, + { + "epoch": 2.7790467823413136, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5232393741607666, + "learning_rate": 1e-06, + "loss": 1.0736, + "mean_token_accuracy": 0.6971802711486816, + "num_tokens": 654828206.0, + "step": 25306 + }, + { + "epoch": 2.779156600043927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0301673412323, + "learning_rate": 1e-06, + "loss": 0.8561, + "mean_token_accuracy": 0.7485281229019165, + "num_tokens": 654847164.0, + "step": 25307 + }, + { + "epoch": 2.7792664177465407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8161919116973877, + "learning_rate": 1e-06, + "loss": 0.965, + "mean_token_accuracy": 0.7159997224807739, + "num_tokens": 654869336.0, + "step": 25308 + }, + { + "epoch": 2.7793762354491545, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.546238899230957, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7175061106681824, + "num_tokens": 654897779.0, + "step": 25309 + }, + { + "epoch": 2.779486053151768, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9608912467956543, + "learning_rate": 1e-06, + "loss": 0.9806, + "mean_token_accuracy": 0.7084559202194214, + "num_tokens": 654917220.0, + "step": 25310 + }, + { + "epoch": 2.779595870854382, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7940359115600586, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7028459906578064, + "num_tokens": 654939692.0, + "step": 25311 + }, + { + "epoch": 2.7797056885569953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6555051803588867, + "learning_rate": 1e-06, + "loss": 1.0993, + "mean_token_accuracy": 0.6787185668945312, + "num_tokens": 654965842.0, + "step": 25312 + }, + { + "epoch": 2.779815506259609, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4889602661132812, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.6942154169082642, + "num_tokens": 654993830.0, + "step": 25313 + }, + { + "epoch": 2.779925323962223, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.863438844680786, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7120271921157837, + "num_tokens": 655015958.0, + "step": 25314 + }, + { + "epoch": 2.780035141664836, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4693737030029297, + "learning_rate": 1e-06, + "loss": 0.9306, + "mean_token_accuracy": 0.7296594977378845, + "num_tokens": 655044444.0, + "step": 25315 + }, + { + "epoch": 2.78014495936745, + "ewc_loss": 1.1026859283447266e-05, + "grad_norm": 80.53854370117188, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7173383235931396, + "num_tokens": 655071755.0, + "step": 25316 + }, + { + "epoch": 2.7802547770700636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7313897609710693, + "learning_rate": 1e-06, + "loss": 1.1212, + "mean_token_accuracy": 0.6788663864135742, + "num_tokens": 655099814.0, + "step": 25317 + }, + { + "epoch": 2.7803645947726774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.516188144683838, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7232099771499634, + "num_tokens": 655129052.0, + "step": 25318 + }, + { + "epoch": 2.780474412475291, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.683964729309082, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7246373295783997, + "num_tokens": 655153983.0, + "step": 25319 + }, + { + "epoch": 2.7805842301779045, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5623579025268555, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.696820855140686, + "num_tokens": 655180614.0, + "step": 25320 + }, + { + "epoch": 2.780694047880518, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4255993366241455, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6937665939331055, + "num_tokens": 655209629.0, + "step": 25321 + }, + { + "epoch": 2.780803865583132, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5879173278808594, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7163234353065491, + "num_tokens": 655234356.0, + "step": 25322 + }, + { + "epoch": 2.7809136832857457, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.437089681625366, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7060134410858154, + "num_tokens": 655263774.0, + "step": 25323 + }, + { + "epoch": 2.7810235009883595, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.561598062515259, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7140446901321411, + "num_tokens": 655290867.0, + "step": 25324 + }, + { + "epoch": 2.781133318690973, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4347476959228516, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6955006122589111, + "num_tokens": 655319273.0, + "step": 25325 + }, + { + "epoch": 2.7812431363935866, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.566831588745117, + "learning_rate": 1e-06, + "loss": 0.9423, + "mean_token_accuracy": 0.7272188663482666, + "num_tokens": 655344099.0, + "step": 25326 + }, + { + "epoch": 2.7813529540962003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3722636699676514, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.68416428565979, + "num_tokens": 655376303.0, + "step": 25327 + }, + { + "epoch": 2.781462771798814, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4446353912353516, + "learning_rate": 1e-06, + "loss": 1.017, + "mean_token_accuracy": 0.7141239047050476, + "num_tokens": 655404148.0, + "step": 25328 + }, + { + "epoch": 2.781572589501428, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.784712553024292, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.6983453631401062, + "num_tokens": 655425556.0, + "step": 25329 + }, + { + "epoch": 2.781682407204041, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.646010637283325, + "learning_rate": 1e-06, + "loss": 0.9318, + "mean_token_accuracy": 0.7278422713279724, + "num_tokens": 655448683.0, + "step": 25330 + }, + { + "epoch": 2.781792224906655, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5029923915863037, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7148224711418152, + "num_tokens": 655475223.0, + "step": 25331 + }, + { + "epoch": 2.7819020426092687, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7214739322662354, + "learning_rate": 1e-06, + "loss": 0.9563, + "mean_token_accuracy": 0.7187546491622925, + "num_tokens": 655503829.0, + "step": 25332 + }, + { + "epoch": 2.7820118603118824, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.504103422164917, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.7063115835189819, + "num_tokens": 655530470.0, + "step": 25333 + }, + { + "epoch": 2.782121678014496, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8674635887145996, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.7080353498458862, + "num_tokens": 655553107.0, + "step": 25334 + }, + { + "epoch": 2.7822314957171095, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.597625970840454, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.710228443145752, + "num_tokens": 655579577.0, + "step": 25335 + }, + { + "epoch": 2.7823413134197232, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7045860290527344, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.720668613910675, + "num_tokens": 655604063.0, + "step": 25336 + }, + { + "epoch": 2.782451131122337, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6200485229492188, + "learning_rate": 1e-06, + "loss": 0.9562, + "mean_token_accuracy": 0.7238210439682007, + "num_tokens": 655629811.0, + "step": 25337 + }, + { + "epoch": 2.7825609488249503, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.566527843475342, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.692511796951294, + "num_tokens": 655658096.0, + "step": 25338 + }, + { + "epoch": 2.7826707665275645, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.911329984664917, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6971386671066284, + "num_tokens": 655686339.0, + "step": 25339 + }, + { + "epoch": 2.782780584230178, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7694756984710693, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6912829875946045, + "num_tokens": 655710433.0, + "step": 25340 + }, + { + "epoch": 2.7828904019327916, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5371198654174805, + "learning_rate": 1e-06, + "loss": 1.0303, + "mean_token_accuracy": 0.7043001651763916, + "num_tokens": 655738374.0, + "step": 25341 + }, + { + "epoch": 2.7830002196354053, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6324756145477295, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7090905904769897, + "num_tokens": 655763612.0, + "step": 25342 + }, + { + "epoch": 2.7831100373380186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.410114049911499, + "learning_rate": 1e-06, + "loss": 1.0334, + "mean_token_accuracy": 0.6956870555877686, + "num_tokens": 655794128.0, + "step": 25343 + }, + { + "epoch": 2.7832198550406324, + "ewc_loss": 1.0788440704345703e-05, + "grad_norm": 3.7333688735961914, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7172146439552307, + "num_tokens": 655817587.0, + "step": 25344 + }, + { + "epoch": 2.783329672743246, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.270051956176758, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7114152312278748, + "num_tokens": 655840349.0, + "step": 25345 + }, + { + "epoch": 2.78343949044586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6954729557037354, + "learning_rate": 1e-06, + "loss": 0.9674, + "mean_token_accuracy": 0.7173880338668823, + "num_tokens": 655863924.0, + "step": 25346 + }, + { + "epoch": 2.7835493081484737, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.322017192840576, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7111412286758423, + "num_tokens": 655886611.0, + "step": 25347 + }, + { + "epoch": 2.783659125851087, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.574695110321045, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.709468424320221, + "num_tokens": 655914869.0, + "step": 25348 + }, + { + "epoch": 2.7837689435537007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5660924911499023, + "learning_rate": 1e-06, + "loss": 0.9601, + "mean_token_accuracy": 0.7158561944961548, + "num_tokens": 655941675.0, + "step": 25349 + }, + { + "epoch": 2.7838787612563145, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.40669846534729, + "learning_rate": 1e-06, + "loss": 1.1134, + "mean_token_accuracy": 0.6867598295211792, + "num_tokens": 655971798.0, + "step": 25350 + }, + { + "epoch": 2.7839885789589283, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4961416721343994, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7292174100875854, + "num_tokens": 655995942.0, + "step": 25351 + }, + { + "epoch": 2.784098396661542, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2327330112457275, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6911259889602661, + "num_tokens": 656030686.0, + "step": 25352 + }, + { + "epoch": 2.7842082143641553, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.481126308441162, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7066543102264404, + "num_tokens": 656060557.0, + "step": 25353 + }, + { + "epoch": 2.784318032066769, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.543006420135498, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7282536625862122, + "num_tokens": 656086004.0, + "step": 25354 + }, + { + "epoch": 2.784427849769383, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.580766201019287, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7008081674575806, + "num_tokens": 656111463.0, + "step": 25355 + }, + { + "epoch": 2.7845376674719966, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6389009952545166, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7085516452789307, + "num_tokens": 656136462.0, + "step": 25356 + }, + { + "epoch": 2.7846474851746104, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7841055393218994, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7078357338905334, + "num_tokens": 656160109.0, + "step": 25357 + }, + { + "epoch": 2.7847573028772237, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.610212564468384, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7000797986984253, + "num_tokens": 656184727.0, + "step": 25358 + }, + { + "epoch": 2.7848671205798374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8344502449035645, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7129850387573242, + "num_tokens": 656204424.0, + "step": 25359 + }, + { + "epoch": 2.784976938282451, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5658273696899414, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6982716917991638, + "num_tokens": 656231785.0, + "step": 25360 + }, + { + "epoch": 2.785086755985065, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.776102304458618, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.708391010761261, + "num_tokens": 656255960.0, + "step": 25361 + }, + { + "epoch": 2.7851965736876787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.307468891143799, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6833934783935547, + "num_tokens": 656289877.0, + "step": 25362 + }, + { + "epoch": 2.785306391390292, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.628514528274536, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6931197047233582, + "num_tokens": 656313410.0, + "step": 25363 + }, + { + "epoch": 2.7854162090929058, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.760544776916504, + "learning_rate": 1e-06, + "loss": 1.076, + "mean_token_accuracy": 0.6894617080688477, + "num_tokens": 656338139.0, + "step": 25364 + }, + { + "epoch": 2.7855260267955195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.573694944381714, + "learning_rate": 1e-06, + "loss": 1.0426, + "mean_token_accuracy": 0.6951481103897095, + "num_tokens": 656365724.0, + "step": 25365 + }, + { + "epoch": 2.785635844498133, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1393990516662598, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7114952802658081, + "num_tokens": 656397010.0, + "step": 25366 + }, + { + "epoch": 2.7857456622007466, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5404257774353027, + "learning_rate": 1e-06, + "loss": 1.1242, + "mean_token_accuracy": 0.6783034801483154, + "num_tokens": 656426045.0, + "step": 25367 + }, + { + "epoch": 2.7858554799033604, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.953420639038086, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7254776954650879, + "num_tokens": 656445907.0, + "step": 25368 + }, + { + "epoch": 2.785965297605974, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.661412477493286, + "learning_rate": 1e-06, + "loss": 1.0316, + "mean_token_accuracy": 0.6985648274421692, + "num_tokens": 656471183.0, + "step": 25369 + }, + { + "epoch": 2.786075115308588, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5035362243652344, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7036814093589783, + "num_tokens": 656499016.0, + "step": 25370 + }, + { + "epoch": 2.786184933011201, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6538703441619873, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.7029228210449219, + "num_tokens": 656523225.0, + "step": 25371 + }, + { + "epoch": 2.786294750713815, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.643768310546875, + "learning_rate": 1e-06, + "loss": 1.1416, + "mean_token_accuracy": 0.6870101690292358, + "num_tokens": 656548239.0, + "step": 25372 + }, + { + "epoch": 2.7864045684164287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.550063133239746, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7142201662063599, + "num_tokens": 656574403.0, + "step": 25373 + }, + { + "epoch": 2.7865143861190425, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.664062023162842, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7178941369056702, + "num_tokens": 656597325.0, + "step": 25374 + }, + { + "epoch": 2.786624203821656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.607174873352051, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7034375071525574, + "num_tokens": 656624767.0, + "step": 25375 + }, + { + "epoch": 2.7867340215242695, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6174306869506836, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7095954418182373, + "num_tokens": 656650865.0, + "step": 25376 + }, + { + "epoch": 2.7868438392268833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.471790313720703, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7035622000694275, + "num_tokens": 656679434.0, + "step": 25377 + }, + { + "epoch": 2.786953656929497, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 2.359884023666382, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7146303057670593, + "num_tokens": 656707461.0, + "step": 25378 + }, + { + "epoch": 2.787063474632111, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6908040046691895, + "learning_rate": 1e-06, + "loss": 1.0166, + "mean_token_accuracy": 0.713646411895752, + "num_tokens": 656731067.0, + "step": 25379 + }, + { + "epoch": 2.7871732923347246, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7374119758605957, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.6859002113342285, + "num_tokens": 656756016.0, + "step": 25380 + }, + { + "epoch": 2.787283110037338, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.745532989501953, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7231181859970093, + "num_tokens": 656776780.0, + "step": 25381 + }, + { + "epoch": 2.7873929277399516, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5479767322540283, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7024532556533813, + "num_tokens": 656802819.0, + "step": 25382 + }, + { + "epoch": 2.7875027454425654, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.77777361869812, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7054400444030762, + "num_tokens": 656825854.0, + "step": 25383 + }, + { + "epoch": 2.787612563145179, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8422560691833496, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7202318906784058, + "num_tokens": 656846330.0, + "step": 25384 + }, + { + "epoch": 2.787722380847793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7796590328216553, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7216008901596069, + "num_tokens": 656870508.0, + "step": 25385 + }, + { + "epoch": 2.787832198550406, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7204113006591797, + "learning_rate": 1e-06, + "loss": 0.9283, + "mean_token_accuracy": 0.7299859523773193, + "num_tokens": 656894746.0, + "step": 25386 + }, + { + "epoch": 2.78794201625302, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.494652271270752, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7176305055618286, + "num_tokens": 656924634.0, + "step": 25387 + }, + { + "epoch": 2.7880518339556337, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3222813606262207, + "learning_rate": 1e-06, + "loss": 0.9906, + "mean_token_accuracy": 0.7188920378684998, + "num_tokens": 656954425.0, + "step": 25388 + }, + { + "epoch": 2.7881616516582475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.544822931289673, + "learning_rate": 1e-06, + "loss": 1.0237, + "mean_token_accuracy": 0.7005835175514221, + "num_tokens": 656981932.0, + "step": 25389 + }, + { + "epoch": 2.7882714693608612, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.454862594604492, + "learning_rate": 1e-06, + "loss": 1.1461, + "mean_token_accuracy": 0.6734911203384399, + "num_tokens": 657014208.0, + "step": 25390 + }, + { + "epoch": 2.7883812870634745, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.313908100128174, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.704719603061676, + "num_tokens": 657043290.0, + "step": 25391 + }, + { + "epoch": 2.7884911047660883, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.143113374710083, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7174220085144043, + "num_tokens": 657064427.0, + "step": 25392 + }, + { + "epoch": 2.788600922468702, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5585899353027344, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7088665962219238, + "num_tokens": 657089835.0, + "step": 25393 + }, + { + "epoch": 2.7887107401713154, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.431403160095215, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.7061365246772766, + "num_tokens": 657117918.0, + "step": 25394 + }, + { + "epoch": 2.788820557873929, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.418991804122925, + "learning_rate": 1e-06, + "loss": 1.1164, + "mean_token_accuracy": 0.6786162257194519, + "num_tokens": 657150714.0, + "step": 25395 + }, + { + "epoch": 2.788930375576543, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.628063678741455, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7069516777992249, + "num_tokens": 657175174.0, + "step": 25396 + }, + { + "epoch": 2.7890401932791566, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.636366844177246, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6981561183929443, + "num_tokens": 657202885.0, + "step": 25397 + }, + { + "epoch": 2.7891500109817704, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.526488780975342, + "learning_rate": 1e-06, + "loss": 1.0311, + "mean_token_accuracy": 0.705329179763794, + "num_tokens": 657231845.0, + "step": 25398 + }, + { + "epoch": 2.7892598286843837, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4296047687530518, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7083368301391602, + "num_tokens": 657261026.0, + "step": 25399 + }, + { + "epoch": 2.7893696463869975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5116264820098877, + "learning_rate": 1e-06, + "loss": 1.0758, + "mean_token_accuracy": 0.6969581246376038, + "num_tokens": 657291041.0, + "step": 25400 + }, + { + "epoch": 2.7894794640896112, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4078054428100586, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7104783654212952, + "num_tokens": 657319078.0, + "step": 25401 + }, + { + "epoch": 2.789589281792225, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.876190423965454, + "learning_rate": 1e-06, + "loss": 0.8827, + "mean_token_accuracy": 0.7379778623580933, + "num_tokens": 657338584.0, + "step": 25402 + }, + { + "epoch": 2.7896990994948387, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6508166790008545, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6996403932571411, + "num_tokens": 657361728.0, + "step": 25403 + }, + { + "epoch": 2.789808917197452, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6619653701782227, + "learning_rate": 1e-06, + "loss": 0.9461, + "mean_token_accuracy": 0.717155933380127, + "num_tokens": 657384977.0, + "step": 25404 + }, + { + "epoch": 2.789918734900066, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.407356023788452, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6916305422782898, + "num_tokens": 657415883.0, + "step": 25405 + }, + { + "epoch": 2.7900285526026796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.767991065979004, + "learning_rate": 1e-06, + "loss": 0.8748, + "mean_token_accuracy": 0.7456405162811279, + "num_tokens": 657438089.0, + "step": 25406 + }, + { + "epoch": 2.7901383703052933, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5365092754364014, + "learning_rate": 1e-06, + "loss": 1.0547, + "mean_token_accuracy": 0.6967148780822754, + "num_tokens": 657467505.0, + "step": 25407 + }, + { + "epoch": 2.790248188007907, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6854135990142822, + "learning_rate": 1e-06, + "loss": 1.0932, + "mean_token_accuracy": 0.6895333528518677, + "num_tokens": 657496368.0, + "step": 25408 + }, + { + "epoch": 2.7903580057105204, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3449060916900635, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7070847749710083, + "num_tokens": 657529212.0, + "step": 25409 + }, + { + "epoch": 2.790467823413134, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.579331636428833, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7143214344978333, + "num_tokens": 657556847.0, + "step": 25410 + }, + { + "epoch": 2.790577641115748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4734110832214355, + "learning_rate": 1e-06, + "loss": 1.0841, + "mean_token_accuracy": 0.6890235543251038, + "num_tokens": 657587255.0, + "step": 25411 + }, + { + "epoch": 2.7906874588183617, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8041484355926514, + "learning_rate": 1e-06, + "loss": 0.9088, + "mean_token_accuracy": 0.7316510677337646, + "num_tokens": 657607807.0, + "step": 25412 + }, + { + "epoch": 2.7907972765209754, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.74546217918396, + "learning_rate": 1e-06, + "loss": 0.9255, + "mean_token_accuracy": 0.7366757392883301, + "num_tokens": 657632519.0, + "step": 25413 + }, + { + "epoch": 2.7909070942235887, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5897574424743652, + "learning_rate": 1e-06, + "loss": 1.1089, + "mean_token_accuracy": 0.6798962950706482, + "num_tokens": 657660615.0, + "step": 25414 + }, + { + "epoch": 2.7910169119262025, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.507549285888672, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6886417269706726, + "num_tokens": 657688699.0, + "step": 25415 + }, + { + "epoch": 2.7911267296288162, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.4503872394561768, + "learning_rate": 1e-06, + "loss": 0.9911, + "mean_token_accuracy": 0.7119187116622925, + "num_tokens": 657717203.0, + "step": 25416 + }, + { + "epoch": 2.7912365473314296, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.638354539871216, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7040542364120483, + "num_tokens": 657743949.0, + "step": 25417 + }, + { + "epoch": 2.7913463650340438, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.31937837600708, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.7013546228408813, + "num_tokens": 657774713.0, + "step": 25418 + }, + { + "epoch": 2.791456182736657, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5896453857421875, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7352070212364197, + "num_tokens": 657799743.0, + "step": 25419 + }, + { + "epoch": 2.791566000439271, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.441119909286499, + "learning_rate": 1e-06, + "loss": 0.9187, + "mean_token_accuracy": 0.7237260937690735, + "num_tokens": 657829342.0, + "step": 25420 + }, + { + "epoch": 2.7916758181418846, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.542447805404663, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.7319416999816895, + "num_tokens": 657854566.0, + "step": 25421 + }, + { + "epoch": 2.791785635844498, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.7487432956695557, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7200421690940857, + "num_tokens": 657879366.0, + "step": 25422 + }, + { + "epoch": 2.7918954535471117, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.597425699234009, + "learning_rate": 1e-06, + "loss": 0.9952, + "mean_token_accuracy": 0.7092518210411072, + "num_tokens": 657905414.0, + "step": 25423 + }, + { + "epoch": 2.7920052712497254, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5676474571228027, + "learning_rate": 1e-06, + "loss": 1.0405, + "mean_token_accuracy": 0.7115103006362915, + "num_tokens": 657931013.0, + "step": 25424 + }, + { + "epoch": 2.792115088952339, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.482262134552002, + "learning_rate": 1e-06, + "loss": 1.0936, + "mean_token_accuracy": 0.6850061416625977, + "num_tokens": 657959293.0, + "step": 25425 + }, + { + "epoch": 2.792224906654953, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4881083965301514, + "learning_rate": 1e-06, + "loss": 1.0623, + "mean_token_accuracy": 0.686913251876831, + "num_tokens": 657985708.0, + "step": 25426 + }, + { + "epoch": 2.7923347243575662, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6273186206817627, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7075976729393005, + "num_tokens": 658013079.0, + "step": 25427 + }, + { + "epoch": 2.79244454206018, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.578472852706909, + "learning_rate": 1e-06, + "loss": 1.0596, + "mean_token_accuracy": 0.696349024772644, + "num_tokens": 658042567.0, + "step": 25428 + }, + { + "epoch": 2.7925543597627938, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4644546508789062, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7112504243850708, + "num_tokens": 658072626.0, + "step": 25429 + }, + { + "epoch": 2.7926641774654075, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.4180524349212646, + "learning_rate": 1e-06, + "loss": 1.0561, + "mean_token_accuracy": 0.6891480684280396, + "num_tokens": 658102681.0, + "step": 25430 + }, + { + "epoch": 2.7927739951680213, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6555869579315186, + "learning_rate": 1e-06, + "loss": 0.9114, + "mean_token_accuracy": 0.7300463914871216, + "num_tokens": 658125645.0, + "step": 25431 + }, + { + "epoch": 2.7928838128706346, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6442644596099854, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.7159937620162964, + "num_tokens": 658148908.0, + "step": 25432 + }, + { + "epoch": 2.7929936305732483, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.650326728820801, + "learning_rate": 1e-06, + "loss": 1.0286, + "mean_token_accuracy": 0.6977106332778931, + "num_tokens": 658171824.0, + "step": 25433 + }, + { + "epoch": 2.793103448275862, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.5457687377929688, + "learning_rate": 1e-06, + "loss": 0.9599, + "mean_token_accuracy": 0.717700719833374, + "num_tokens": 658196709.0, + "step": 25434 + }, + { + "epoch": 2.793213265978476, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.9846644401550293, + "learning_rate": 1e-06, + "loss": 0.9367, + "mean_token_accuracy": 0.7274607419967651, + "num_tokens": 658216649.0, + "step": 25435 + }, + { + "epoch": 2.7933230836810896, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.581890106201172, + "learning_rate": 1e-06, + "loss": 0.9421, + "mean_token_accuracy": 0.7264997363090515, + "num_tokens": 658241009.0, + "step": 25436 + }, + { + "epoch": 2.793432901383703, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7005066871643066, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.6992205381393433, + "num_tokens": 658267432.0, + "step": 25437 + }, + { + "epoch": 2.7935427190863167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5684711933135986, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.707154393196106, + "num_tokens": 658294948.0, + "step": 25438 + }, + { + "epoch": 2.7936525367889304, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6414830684661865, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7232197523117065, + "num_tokens": 658319781.0, + "step": 25439 + }, + { + "epoch": 2.793762354491544, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8748981952667236, + "learning_rate": 1e-06, + "loss": 1.0399, + "mean_token_accuracy": 0.7013993859291077, + "num_tokens": 658347955.0, + "step": 25440 + }, + { + "epoch": 2.793872172194158, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8595821857452393, + "learning_rate": 1e-06, + "loss": 0.9044, + "mean_token_accuracy": 0.7320740818977356, + "num_tokens": 658370227.0, + "step": 25441 + }, + { + "epoch": 2.7939819898967713, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5828895568847656, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7111480832099915, + "num_tokens": 658399478.0, + "step": 25442 + }, + { + "epoch": 2.794091807599385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9624059200286865, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7120543718338013, + "num_tokens": 658419334.0, + "step": 25443 + }, + { + "epoch": 2.794201625301999, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4542696475982666, + "learning_rate": 1e-06, + "loss": 0.9752, + "mean_token_accuracy": 0.7248016595840454, + "num_tokens": 658446821.0, + "step": 25444 + }, + { + "epoch": 2.794311443004612, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.461406946182251, + "learning_rate": 1e-06, + "loss": 0.9014, + "mean_token_accuracy": 0.7358428239822388, + "num_tokens": 658472890.0, + "step": 25445 + }, + { + "epoch": 2.794421260707226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7138421535491943, + "learning_rate": 1e-06, + "loss": 1.0248, + "mean_token_accuracy": 0.7054005861282349, + "num_tokens": 658501453.0, + "step": 25446 + }, + { + "epoch": 2.7945310784098396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.566070795059204, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7265310287475586, + "num_tokens": 658525789.0, + "step": 25447 + }, + { + "epoch": 2.7946408961124534, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7739675045013428, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7094314694404602, + "num_tokens": 658547999.0, + "step": 25448 + }, + { + "epoch": 2.794750713815067, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.903822898864746, + "learning_rate": 1e-06, + "loss": 0.8969, + "mean_token_accuracy": 0.7317246794700623, + "num_tokens": 658567804.0, + "step": 25449 + }, + { + "epoch": 2.7948605315176804, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6168062686920166, + "learning_rate": 1e-06, + "loss": 1.0002, + "mean_token_accuracy": 0.7011993527412415, + "num_tokens": 658595068.0, + "step": 25450 + }, + { + "epoch": 2.794970349220294, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.727642774581909, + "learning_rate": 1e-06, + "loss": 1.0308, + "mean_token_accuracy": 0.703994631767273, + "num_tokens": 658621032.0, + "step": 25451 + }, + { + "epoch": 2.795080166922908, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7379441261291504, + "learning_rate": 1e-06, + "loss": 0.9667, + "mean_token_accuracy": 0.7140692472457886, + "num_tokens": 658641357.0, + "step": 25452 + }, + { + "epoch": 2.7951899846255217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.606358766555786, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6859307885169983, + "num_tokens": 658671439.0, + "step": 25453 + }, + { + "epoch": 2.7952998023281355, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6796436309814453, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7062485218048096, + "num_tokens": 658696053.0, + "step": 25454 + }, + { + "epoch": 2.7954096200307488, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.768192768096924, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7202625274658203, + "num_tokens": 658721063.0, + "step": 25455 + }, + { + "epoch": 2.7955194377333625, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.679063320159912, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7267030477523804, + "num_tokens": 658746630.0, + "step": 25456 + }, + { + "epoch": 2.7956292554359763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4546420574188232, + "learning_rate": 1e-06, + "loss": 0.9729, + "mean_token_accuracy": 0.7183547616004944, + "num_tokens": 658771703.0, + "step": 25457 + }, + { + "epoch": 2.79573907313859, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4601922035217285, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7148165702819824, + "num_tokens": 658800801.0, + "step": 25458 + }, + { + "epoch": 2.795848890841204, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.358912229537964, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.70087069272995, + "num_tokens": 658832820.0, + "step": 25459 + }, + { + "epoch": 2.795958708543817, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.326493501663208, + "learning_rate": 1e-06, + "loss": 1.1086, + "mean_token_accuracy": 0.6786356568336487, + "num_tokens": 658868258.0, + "step": 25460 + }, + { + "epoch": 2.796068526246431, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7264761924743652, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7079139947891235, + "num_tokens": 658891424.0, + "step": 25461 + }, + { + "epoch": 2.7961783439490446, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.569920539855957, + "learning_rate": 1e-06, + "loss": 1.0887, + "mean_token_accuracy": 0.6813951134681702, + "num_tokens": 658919905.0, + "step": 25462 + }, + { + "epoch": 2.7962881616516584, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8382060527801514, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7081928253173828, + "num_tokens": 658945164.0, + "step": 25463 + }, + { + "epoch": 2.796397979354272, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8175902366638184, + "learning_rate": 1e-06, + "loss": 0.8994, + "mean_token_accuracy": 0.7308365106582642, + "num_tokens": 658967595.0, + "step": 25464 + }, + { + "epoch": 2.7965077970568855, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.40527606010437, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7084237337112427, + "num_tokens": 658999509.0, + "step": 25465 + }, + { + "epoch": 2.796617614759499, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.456385850906372, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.7035852074623108, + "num_tokens": 659028209.0, + "step": 25466 + }, + { + "epoch": 2.796727432462113, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.622959613800049, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7104895114898682, + "num_tokens": 659054733.0, + "step": 25467 + }, + { + "epoch": 2.7968372501647263, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2083072662353516, + "learning_rate": 1e-06, + "loss": 1.1187, + "mean_token_accuracy": 0.6731311082839966, + "num_tokens": 659086566.0, + "step": 25468 + }, + { + "epoch": 2.7969470678673405, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.308562994003296, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.6922663450241089, + "num_tokens": 659113076.0, + "step": 25469 + }, + { + "epoch": 2.797056885569954, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4760048389434814, + "learning_rate": 1e-06, + "loss": 0.9314, + "mean_token_accuracy": 0.7224390506744385, + "num_tokens": 659140510.0, + "step": 25470 + }, + { + "epoch": 2.7971667032725676, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.523491621017456, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7057805061340332, + "num_tokens": 659171065.0, + "step": 25471 + }, + { + "epoch": 2.7972765209751813, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4551525115966797, + "learning_rate": 1e-06, + "loss": 1.0505, + "mean_token_accuracy": 0.6917873620986938, + "num_tokens": 659198950.0, + "step": 25472 + }, + { + "epoch": 2.7973863386777946, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.3194169998168945, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.6904832720756531, + "num_tokens": 659228794.0, + "step": 25473 + }, + { + "epoch": 2.7974961563804084, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.707535743713379, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7017389535903931, + "num_tokens": 659250785.0, + "step": 25474 + }, + { + "epoch": 2.797605974083022, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6059441566467285, + "learning_rate": 1e-06, + "loss": 0.8525, + "mean_token_accuracy": 0.7429753541946411, + "num_tokens": 659273650.0, + "step": 25475 + }, + { + "epoch": 2.797715791785636, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.833899736404419, + "learning_rate": 1e-06, + "loss": 0.902, + "mean_token_accuracy": 0.7302583456039429, + "num_tokens": 659297636.0, + "step": 25476 + }, + { + "epoch": 2.7978256094882497, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6151552200317383, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7082458138465881, + "num_tokens": 659325961.0, + "step": 25477 + }, + { + "epoch": 2.797935427190863, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.385714054107666, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6901631355285645, + "num_tokens": 659355236.0, + "step": 25478 + }, + { + "epoch": 2.7980452448934767, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6902058124542236, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7226449847221375, + "num_tokens": 659380549.0, + "step": 25479 + }, + { + "epoch": 2.7981550625960905, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.2636218070983887, + "learning_rate": 1e-06, + "loss": 0.9182, + "mean_token_accuracy": 0.7274976968765259, + "num_tokens": 659404170.0, + "step": 25480 + }, + { + "epoch": 2.7982648802987042, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.472597599029541, + "learning_rate": 1e-06, + "loss": 0.9263, + "mean_token_accuracy": 0.7344211339950562, + "num_tokens": 659429358.0, + "step": 25481 + }, + { + "epoch": 2.798374698001318, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.760535717010498, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7143770456314087, + "num_tokens": 659452045.0, + "step": 25482 + }, + { + "epoch": 2.7984845157039313, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.548919677734375, + "learning_rate": 1e-06, + "loss": 1.0342, + "mean_token_accuracy": 0.7036023139953613, + "num_tokens": 659482183.0, + "step": 25483 + }, + { + "epoch": 2.798594333406545, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 3.418222665786743, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7004165053367615, + "num_tokens": 659506878.0, + "step": 25484 + }, + { + "epoch": 2.798704151109159, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5492656230926514, + "learning_rate": 1e-06, + "loss": 1.0313, + "mean_token_accuracy": 0.7073832154273987, + "num_tokens": 659534891.0, + "step": 25485 + }, + { + "epoch": 2.7988139688117726, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.446542263031006, + "learning_rate": 1e-06, + "loss": 0.971, + "mean_token_accuracy": 0.7148915529251099, + "num_tokens": 659563039.0, + "step": 25486 + }, + { + "epoch": 2.7989237865143863, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.669642448425293, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6977345943450928, + "num_tokens": 659587458.0, + "step": 25487 + }, + { + "epoch": 2.7990336042169996, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6858959197998047, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7125104665756226, + "num_tokens": 659610684.0, + "step": 25488 + }, + { + "epoch": 2.7991434219196134, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7829244136810303, + "learning_rate": 1e-06, + "loss": 0.9196, + "mean_token_accuracy": 0.7303091287612915, + "num_tokens": 659631616.0, + "step": 25489 + }, + { + "epoch": 2.799253239622227, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.788630962371826, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.698712170124054, + "num_tokens": 659659811.0, + "step": 25490 + }, + { + "epoch": 2.799363057324841, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8303773403167725, + "learning_rate": 1e-06, + "loss": 0.9277, + "mean_token_accuracy": 0.7279758453369141, + "num_tokens": 659681866.0, + "step": 25491 + }, + { + "epoch": 2.7994728750274547, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.563535451889038, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7058898210525513, + "num_tokens": 659708529.0, + "step": 25492 + }, + { + "epoch": 2.799582692730068, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5227296352386475, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7126920223236084, + "num_tokens": 659737239.0, + "step": 25493 + }, + { + "epoch": 2.7996925104326817, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.864445447921753, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.717254638671875, + "num_tokens": 659758319.0, + "step": 25494 + }, + { + "epoch": 2.7998023281352955, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7017266750335693, + "learning_rate": 1e-06, + "loss": 1.0282, + "mean_token_accuracy": 0.7041823863983154, + "num_tokens": 659781992.0, + "step": 25495 + }, + { + "epoch": 2.799912145837909, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8063106536865234, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7070152759552002, + "num_tokens": 659803204.0, + "step": 25496 + }, + { + "epoch": 2.8000219635405226, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.666532278060913, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.7052284479141235, + "num_tokens": 659827306.0, + "step": 25497 + }, + { + "epoch": 2.8001317812431363, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9312386512756348, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7095638513565063, + "num_tokens": 659847819.0, + "step": 25498 + }, + { + "epoch": 2.80024159894575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 6.698991775512695, + "learning_rate": 1e-06, + "loss": 1.0582, + "mean_token_accuracy": 0.7010862827301025, + "num_tokens": 659871600.0, + "step": 25499 + }, + { + "epoch": 2.800351416648364, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.484706163406372, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.7159804105758667, + "num_tokens": 659900805.0, + "step": 25500 + }, + { + "epoch": 2.800461234350977, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4663403034210205, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7035982608795166, + "num_tokens": 659929084.0, + "step": 25501 + }, + { + "epoch": 2.800571052053591, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2824440002441406, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6971202492713928, + "num_tokens": 659961327.0, + "step": 25502 + }, + { + "epoch": 2.8006808697562047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7112796306610107, + "learning_rate": 1e-06, + "loss": 1.0728, + "mean_token_accuracy": 0.6872564554214478, + "num_tokens": 659986817.0, + "step": 25503 + }, + { + "epoch": 2.8007906874588184, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.926048517227173, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.6972783803939819, + "num_tokens": 660008085.0, + "step": 25504 + }, + { + "epoch": 2.800900505161432, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4425246715545654, + "learning_rate": 1e-06, + "loss": 0.9712, + "mean_token_accuracy": 0.724526047706604, + "num_tokens": 660034970.0, + "step": 25505 + }, + { + "epoch": 2.8010103228640455, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8682405948638916, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7176339626312256, + "num_tokens": 660056318.0, + "step": 25506 + }, + { + "epoch": 2.8011201405666593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6832473278045654, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7269185781478882, + "num_tokens": 660079209.0, + "step": 25507 + }, + { + "epoch": 2.801229958269273, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4892444610595703, + "learning_rate": 1e-06, + "loss": 1.0945, + "mean_token_accuracy": 0.6840551495552063, + "num_tokens": 660107530.0, + "step": 25508 + }, + { + "epoch": 2.8013397759718868, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.742114305496216, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.7161783576011658, + "num_tokens": 660131087.0, + "step": 25509 + }, + { + "epoch": 2.8014495936745005, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4751429557800293, + "learning_rate": 1e-06, + "loss": 1.1003, + "mean_token_accuracy": 0.6952629685401917, + "num_tokens": 660162306.0, + "step": 25510 + }, + { + "epoch": 2.801559411377114, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.558964490890503, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7169627547264099, + "num_tokens": 660186485.0, + "step": 25511 + }, + { + "epoch": 2.8016692290797276, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.674016237258911, + "learning_rate": 1e-06, + "loss": 1.0267, + "mean_token_accuracy": 0.7058591842651367, + "num_tokens": 660210493.0, + "step": 25512 + }, + { + "epoch": 2.8017790467823414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.658834457397461, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6921408176422119, + "num_tokens": 660233809.0, + "step": 25513 + }, + { + "epoch": 2.801888864484955, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.412238836288452, + "learning_rate": 1e-06, + "loss": 0.9983, + "mean_token_accuracy": 0.7094357013702393, + "num_tokens": 660262261.0, + "step": 25514 + }, + { + "epoch": 2.801998682187569, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.85465145111084, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.7211632132530212, + "num_tokens": 660285282.0, + "step": 25515 + }, + { + "epoch": 2.802108499890182, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.594616174697876, + "learning_rate": 1e-06, + "loss": 0.9803, + "mean_token_accuracy": 0.714506208896637, + "num_tokens": 660312026.0, + "step": 25516 + }, + { + "epoch": 2.802218317592796, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.632200241088867, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7017714977264404, + "num_tokens": 660336797.0, + "step": 25517 + }, + { + "epoch": 2.8023281352954097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4180469512939453, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7098301649093628, + "num_tokens": 660363648.0, + "step": 25518 + }, + { + "epoch": 2.802437952998023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6859776973724365, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7047510147094727, + "num_tokens": 660388352.0, + "step": 25519 + }, + { + "epoch": 2.802547770700637, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.652707576751709, + "learning_rate": 1e-06, + "loss": 1.0357, + "mean_token_accuracy": 0.7057087421417236, + "num_tokens": 660412677.0, + "step": 25520 + }, + { + "epoch": 2.8026575884032505, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.446959972381592, + "learning_rate": 1e-06, + "loss": 1.1153, + "mean_token_accuracy": 0.6819308996200562, + "num_tokens": 660442384.0, + "step": 25521 + }, + { + "epoch": 2.8027674061058643, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4990005493164062, + "learning_rate": 1e-06, + "loss": 0.9431, + "mean_token_accuracy": 0.7221859693527222, + "num_tokens": 660471277.0, + "step": 25522 + }, + { + "epoch": 2.802877223808478, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.475931167602539, + "learning_rate": 1e-06, + "loss": 1.0392, + "mean_token_accuracy": 0.7087636590003967, + "num_tokens": 660499776.0, + "step": 25523 + }, + { + "epoch": 2.8029870415110913, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.394502639770508, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.6967003345489502, + "num_tokens": 660529542.0, + "step": 25524 + }, + { + "epoch": 2.803096859213705, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.451547622680664, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.703844428062439, + "num_tokens": 660557255.0, + "step": 25525 + }, + { + "epoch": 2.803206676916319, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7684779167175293, + "learning_rate": 1e-06, + "loss": 0.9643, + "mean_token_accuracy": 0.7163400650024414, + "num_tokens": 660577448.0, + "step": 25526 + }, + { + "epoch": 2.8033164946189326, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.461230754852295, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.7001940011978149, + "num_tokens": 660611802.0, + "step": 25527 + }, + { + "epoch": 2.8034263123215464, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9703521728515625, + "learning_rate": 1e-06, + "loss": 0.883, + "mean_token_accuracy": 0.7382485866546631, + "num_tokens": 660632030.0, + "step": 25528 + }, + { + "epoch": 2.8035361300241597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.915330171585083, + "learning_rate": 1e-06, + "loss": 1.0625, + "mean_token_accuracy": 0.6941787600517273, + "num_tokens": 660656917.0, + "step": 25529 + }, + { + "epoch": 2.8036459477267734, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0721449851989746, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7282481789588928, + "num_tokens": 660677591.0, + "step": 25530 + }, + { + "epoch": 2.803755765429387, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5717360973358154, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7258371114730835, + "num_tokens": 660704066.0, + "step": 25531 + }, + { + "epoch": 2.803865583132001, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.421539545059204, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6977199912071228, + "num_tokens": 660736158.0, + "step": 25532 + }, + { + "epoch": 2.8039754008346147, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5531160831451416, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.717340886592865, + "num_tokens": 660764931.0, + "step": 25533 + }, + { + "epoch": 2.804085218537228, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.706212043762207, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.6958572864532471, + "num_tokens": 660791878.0, + "step": 25534 + }, + { + "epoch": 2.804195036239842, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.610306739807129, + "learning_rate": 1e-06, + "loss": 0.9651, + "mean_token_accuracy": 0.720188558101654, + "num_tokens": 660818303.0, + "step": 25535 + }, + { + "epoch": 2.8043048539424555, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5199356079101562, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.7065131664276123, + "num_tokens": 660847686.0, + "step": 25536 + }, + { + "epoch": 2.8044146716450693, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5614800453186035, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.7068222761154175, + "num_tokens": 660874618.0, + "step": 25537 + }, + { + "epoch": 2.804524489347683, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.780843734741211, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7291766405105591, + "num_tokens": 660895920.0, + "step": 25538 + }, + { + "epoch": 2.8046343070502964, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.889524459838867, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7148176431655884, + "num_tokens": 660918930.0, + "step": 25539 + }, + { + "epoch": 2.80474412475291, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505568742752075, + "learning_rate": 1e-06, + "loss": 0.9321, + "mean_token_accuracy": 0.7279075384140015, + "num_tokens": 660946257.0, + "step": 25540 + }, + { + "epoch": 2.804853942455524, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6887760162353516, + "learning_rate": 1e-06, + "loss": 1.0599, + "mean_token_accuracy": 0.6971209049224854, + "num_tokens": 660970851.0, + "step": 25541 + }, + { + "epoch": 2.8049637601581376, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4725983142852783, + "learning_rate": 1e-06, + "loss": 1.1304, + "mean_token_accuracy": 0.6798572540283203, + "num_tokens": 661001842.0, + "step": 25542 + }, + { + "epoch": 2.8050735778607514, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4453983306884766, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.7036167979240417, + "num_tokens": 661033093.0, + "step": 25543 + }, + { + "epoch": 2.8051833955633647, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.757997989654541, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7178632616996765, + "num_tokens": 661055272.0, + "step": 25544 + }, + { + "epoch": 2.8052932132659785, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5344386100769043, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.711565375328064, + "num_tokens": 661080432.0, + "step": 25545 + }, + { + "epoch": 2.8054030309685922, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.718000888824463, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.719806969165802, + "num_tokens": 661105508.0, + "step": 25546 + }, + { + "epoch": 2.8055128486712055, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5256311893463135, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7042322158813477, + "num_tokens": 661135946.0, + "step": 25547 + }, + { + "epoch": 2.8056226663738193, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7704508304595947, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7030371427536011, + "num_tokens": 661159454.0, + "step": 25548 + }, + { + "epoch": 2.805732484076433, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.601655960083008, + "learning_rate": 1e-06, + "loss": 0.9027, + "mean_token_accuracy": 0.7405350208282471, + "num_tokens": 661182410.0, + "step": 25549 + }, + { + "epoch": 2.805842301779047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9709179401397705, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7217914462089539, + "num_tokens": 661201229.0, + "step": 25550 + }, + { + "epoch": 2.8059521194816606, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.498520612716675, + "learning_rate": 1e-06, + "loss": 1.092, + "mean_token_accuracy": 0.691889762878418, + "num_tokens": 661229079.0, + "step": 25551 + }, + { + "epoch": 2.806061937184274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6984682083129883, + "learning_rate": 1e-06, + "loss": 0.9946, + "mean_token_accuracy": 0.7073038816452026, + "num_tokens": 661252597.0, + "step": 25552 + }, + { + "epoch": 2.8061717548868876, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5329976081848145, + "learning_rate": 1e-06, + "loss": 1.0743, + "mean_token_accuracy": 0.6841060519218445, + "num_tokens": 661280899.0, + "step": 25553 + }, + { + "epoch": 2.8062815725895014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5485377311706543, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.709416925907135, + "num_tokens": 661309258.0, + "step": 25554 + }, + { + "epoch": 2.806391390292115, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6828222274780273, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6890422105789185, + "num_tokens": 661334630.0, + "step": 25555 + }, + { + "epoch": 2.806501207994729, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4663138389587402, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7120669484138489, + "num_tokens": 661361697.0, + "step": 25556 + }, + { + "epoch": 2.806611025697342, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.652482032775879, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7111383676528931, + "num_tokens": 661387248.0, + "step": 25557 + }, + { + "epoch": 2.806720843399956, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.547656774520874, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7064837217330933, + "num_tokens": 661412647.0, + "step": 25558 + }, + { + "epoch": 2.8068306611025697, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5714778900146484, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7065346240997314, + "num_tokens": 661438448.0, + "step": 25559 + }, + { + "epoch": 2.8069404788051835, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.722254514694214, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7107786536216736, + "num_tokens": 661463863.0, + "step": 25560 + }, + { + "epoch": 2.8070502965077972, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5666873455047607, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.7016345262527466, + "num_tokens": 661492452.0, + "step": 25561 + }, + { + "epoch": 2.8071601142104106, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.683837890625, + "learning_rate": 1e-06, + "loss": 0.9794, + "mean_token_accuracy": 0.7184871435165405, + "num_tokens": 661514392.0, + "step": 25562 + }, + { + "epoch": 2.8072699319130243, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.569068193435669, + "learning_rate": 1e-06, + "loss": 0.9357, + "mean_token_accuracy": 0.720970630645752, + "num_tokens": 661540750.0, + "step": 25563 + }, + { + "epoch": 2.807379749615638, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5352423191070557, + "learning_rate": 1e-06, + "loss": 1.0787, + "mean_token_accuracy": 0.6910036206245422, + "num_tokens": 661568038.0, + "step": 25564 + }, + { + "epoch": 2.807489567318252, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.62166690826416, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7147165536880493, + "num_tokens": 661594930.0, + "step": 25565 + }, + { + "epoch": 2.8075993850208656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3298757076263428, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7024825811386108, + "num_tokens": 661623558.0, + "step": 25566 + }, + { + "epoch": 2.807709202723479, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.476228713989258, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7130420207977295, + "num_tokens": 661652206.0, + "step": 25567 + }, + { + "epoch": 2.8078190204260927, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3842968940734863, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.6972070932388306, + "num_tokens": 661683089.0, + "step": 25568 + }, + { + "epoch": 2.8079288381287064, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5771195888519287, + "learning_rate": 1e-06, + "loss": 0.9194, + "mean_token_accuracy": 0.7272419333457947, + "num_tokens": 661707272.0, + "step": 25569 + }, + { + "epoch": 2.80803865583132, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.709315061569214, + "learning_rate": 1e-06, + "loss": 0.9488, + "mean_token_accuracy": 0.7184474468231201, + "num_tokens": 661730598.0, + "step": 25570 + }, + { + "epoch": 2.808148473533934, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.045144557952881, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7238346338272095, + "num_tokens": 661753754.0, + "step": 25571 + }, + { + "epoch": 2.8082582912365472, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5569891929626465, + "learning_rate": 1e-06, + "loss": 0.9854, + "mean_token_accuracy": 0.7138671875, + "num_tokens": 661779277.0, + "step": 25572 + }, + { + "epoch": 2.808368108939161, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7301242351531982, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7158364057540894, + "num_tokens": 661804152.0, + "step": 25573 + }, + { + "epoch": 2.8084779266417748, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7081384658813477, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6911009550094604, + "num_tokens": 661826690.0, + "step": 25574 + }, + { + "epoch": 2.808587744344388, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.329946517944336, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6979177594184875, + "num_tokens": 661858275.0, + "step": 25575 + }, + { + "epoch": 2.808697562047002, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.64028263092041, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.6945039629936218, + "num_tokens": 661885693.0, + "step": 25576 + }, + { + "epoch": 2.8088073797496156, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.543309450149536, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7044084668159485, + "num_tokens": 661910365.0, + "step": 25577 + }, + { + "epoch": 2.8089171974522293, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8489410877227783, + "learning_rate": 1e-06, + "loss": 0.9358, + "mean_token_accuracy": 0.7234253883361816, + "num_tokens": 661930340.0, + "step": 25578 + }, + { + "epoch": 2.809027015154843, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.836501121520996, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7086660861968994, + "num_tokens": 661950143.0, + "step": 25579 + }, + { + "epoch": 2.8091368328574564, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1760799884796143, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7237679362297058, + "num_tokens": 661972186.0, + "step": 25580 + }, + { + "epoch": 2.80924665056007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.448547601699829, + "learning_rate": 1e-06, + "loss": 0.9874, + "mean_token_accuracy": 0.715822160243988, + "num_tokens": 661998612.0, + "step": 25581 + }, + { + "epoch": 2.809356468262684, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.596670627593994, + "learning_rate": 1e-06, + "loss": 0.9974, + "mean_token_accuracy": 0.7132648229598999, + "num_tokens": 662027212.0, + "step": 25582 + }, + { + "epoch": 2.8094662859652977, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5359678268432617, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7178834676742554, + "num_tokens": 662052504.0, + "step": 25583 + }, + { + "epoch": 2.8095761036679114, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.740586996078491, + "learning_rate": 1e-06, + "loss": 0.923, + "mean_token_accuracy": 0.7300857901573181, + "num_tokens": 662075237.0, + "step": 25584 + }, + { + "epoch": 2.8096859213705248, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.567504644393921, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7028964161872864, + "num_tokens": 662100237.0, + "step": 25585 + }, + { + "epoch": 2.8097957390731385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.593059539794922, + "learning_rate": 1e-06, + "loss": 0.9168, + "mean_token_accuracy": 0.7358624935150146, + "num_tokens": 662125680.0, + "step": 25586 + }, + { + "epoch": 2.8099055567757523, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.924309253692627, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7016347646713257, + "num_tokens": 662146244.0, + "step": 25587 + }, + { + "epoch": 2.810015374478366, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6630043983459473, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7153352499008179, + "num_tokens": 662177330.0, + "step": 25588 + }, + { + "epoch": 2.81012519218098, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.799124240875244, + "learning_rate": 1e-06, + "loss": 0.9334, + "mean_token_accuracy": 0.726514458656311, + "num_tokens": 662197208.0, + "step": 25589 + }, + { + "epoch": 2.810235009883593, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9152026176452637, + "learning_rate": 1e-06, + "loss": 0.8742, + "mean_token_accuracy": 0.7409647703170776, + "num_tokens": 662216227.0, + "step": 25590 + }, + { + "epoch": 2.810344827586207, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.596587657928467, + "learning_rate": 1e-06, + "loss": 1.0484, + "mean_token_accuracy": 0.698415994644165, + "num_tokens": 662241684.0, + "step": 25591 + }, + { + "epoch": 2.8104546452888206, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.844130516052246, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7144461870193481, + "num_tokens": 662262461.0, + "step": 25592 + }, + { + "epoch": 2.8105644629914344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6534268856048584, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7151815295219421, + "num_tokens": 662287571.0, + "step": 25593 + }, + { + "epoch": 2.810674280694048, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5126569271087646, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7048611640930176, + "num_tokens": 662317521.0, + "step": 25594 + }, + { + "epoch": 2.8107840983966614, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.365532398223877, + "learning_rate": 1e-06, + "loss": 1.1074, + "mean_token_accuracy": 0.687497615814209, + "num_tokens": 662346423.0, + "step": 25595 + }, + { + "epoch": 2.810893916099275, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9084982872009277, + "learning_rate": 1e-06, + "loss": 1.0049, + "mean_token_accuracy": 0.7024375200271606, + "num_tokens": 662367305.0, + "step": 25596 + }, + { + "epoch": 2.811003733801889, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.585738182067871, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7077807784080505, + "num_tokens": 662392026.0, + "step": 25597 + }, + { + "epoch": 2.8111135515045023, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6516964435577393, + "learning_rate": 1e-06, + "loss": 0.9298, + "mean_token_accuracy": 0.7261887788772583, + "num_tokens": 662417973.0, + "step": 25598 + }, + { + "epoch": 2.8112233692071165, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5686142444610596, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7122089266777039, + "num_tokens": 662449961.0, + "step": 25599 + }, + { + "epoch": 2.8113331869097298, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.689598321914673, + "learning_rate": 1e-06, + "loss": 1.0439, + "mean_token_accuracy": 0.7027484178543091, + "num_tokens": 662474543.0, + "step": 25600 + }, + { + "epoch": 2.8114430046123435, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5090739727020264, + "learning_rate": 1e-06, + "loss": 1.081, + "mean_token_accuracy": 0.6914771795272827, + "num_tokens": 662503803.0, + "step": 25601 + }, + { + "epoch": 2.8115528223149573, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.34561824798584, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7106708288192749, + "num_tokens": 662538210.0, + "step": 25602 + }, + { + "epoch": 2.8116626400175706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6903016567230225, + "learning_rate": 1e-06, + "loss": 0.8989, + "mean_token_accuracy": 0.7325291037559509, + "num_tokens": 662559687.0, + "step": 25603 + }, + { + "epoch": 2.8117724577201844, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.660874605178833, + "learning_rate": 1e-06, + "loss": 0.9272, + "mean_token_accuracy": 0.7285867929458618, + "num_tokens": 662583448.0, + "step": 25604 + }, + { + "epoch": 2.811882275422798, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.567620277404785, + "learning_rate": 1e-06, + "loss": 1.0261, + "mean_token_accuracy": 0.7050381302833557, + "num_tokens": 662608705.0, + "step": 25605 + }, + { + "epoch": 2.811992093125412, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7976865768432617, + "learning_rate": 1e-06, + "loss": 1.0629, + "mean_token_accuracy": 0.685791015625, + "num_tokens": 662630715.0, + "step": 25606 + }, + { + "epoch": 2.8121019108280256, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5421946048736572, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6969228386878967, + "num_tokens": 662657962.0, + "step": 25607 + }, + { + "epoch": 2.812211728530639, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6758460998535156, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.7101324200630188, + "num_tokens": 662684548.0, + "step": 25608 + }, + { + "epoch": 2.8123215462332527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.539311170578003, + "learning_rate": 1e-06, + "loss": 0.9501, + "mean_token_accuracy": 0.7222375869750977, + "num_tokens": 662710245.0, + "step": 25609 + }, + { + "epoch": 2.8124313639358665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.633789300918579, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7175912857055664, + "num_tokens": 662735937.0, + "step": 25610 + }, + { + "epoch": 2.81254118163848, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9198620319366455, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7109345197677612, + "num_tokens": 662755845.0, + "step": 25611 + }, + { + "epoch": 2.812650999341094, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.667576313018799, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7018182277679443, + "num_tokens": 662781038.0, + "step": 25612 + }, + { + "epoch": 2.8127608170437073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.733783483505249, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.7055359482765198, + "num_tokens": 662805811.0, + "step": 25613 + }, + { + "epoch": 2.812870634746321, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.737725019454956, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7226999998092651, + "num_tokens": 662826943.0, + "step": 25614 + }, + { + "epoch": 2.812980452448935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4377663135528564, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7027950882911682, + "num_tokens": 662854053.0, + "step": 25615 + }, + { + "epoch": 2.8130902701515486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.46431827545166, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6977381706237793, + "num_tokens": 662882986.0, + "step": 25616 + }, + { + "epoch": 2.8132000878541623, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7286393642425537, + "learning_rate": 1e-06, + "loss": 1.0128, + "mean_token_accuracy": 0.7112372517585754, + "num_tokens": 662913476.0, + "step": 25617 + }, + { + "epoch": 2.8133099055567756, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6188251972198486, + "learning_rate": 1e-06, + "loss": 0.9569, + "mean_token_accuracy": 0.7212168574333191, + "num_tokens": 662940969.0, + "step": 25618 + }, + { + "epoch": 2.8134197232593894, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5311965942382812, + "learning_rate": 1e-06, + "loss": 0.9115, + "mean_token_accuracy": 0.7256460189819336, + "num_tokens": 662969481.0, + "step": 25619 + }, + { + "epoch": 2.813529540962003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0313916206359863, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7003542184829712, + "num_tokens": 662990900.0, + "step": 25620 + }, + { + "epoch": 2.813639358664617, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.485633373260498, + "learning_rate": 1e-06, + "loss": 1.1214, + "mean_token_accuracy": 0.6799525618553162, + "num_tokens": 663025348.0, + "step": 25621 + }, + { + "epoch": 2.8137491763672307, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4320545196533203, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.7077065706253052, + "num_tokens": 663055529.0, + "step": 25622 + }, + { + "epoch": 2.813858994069844, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5203635692596436, + "learning_rate": 1e-06, + "loss": 0.9504, + "mean_token_accuracy": 0.7181764245033264, + "num_tokens": 663082171.0, + "step": 25623 + }, + { + "epoch": 2.8139688117724577, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.903262138366699, + "learning_rate": 1e-06, + "loss": 0.9337, + "mean_token_accuracy": 0.7167159914970398, + "num_tokens": 663102817.0, + "step": 25624 + }, + { + "epoch": 2.8140786294750715, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6051273345947266, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6994526386260986, + "num_tokens": 663126762.0, + "step": 25625 + }, + { + "epoch": 2.814188447177685, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.843801736831665, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7099618911743164, + "num_tokens": 663149265.0, + "step": 25626 + }, + { + "epoch": 2.8142982648802986, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.757944345474243, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7255690097808838, + "num_tokens": 663171640.0, + "step": 25627 + }, + { + "epoch": 2.8144080825829123, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2863705158233643, + "learning_rate": 1e-06, + "loss": 0.9531, + "mean_token_accuracy": 0.7194809317588806, + "num_tokens": 663205859.0, + "step": 25628 + }, + { + "epoch": 2.814517900285526, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5483078956604004, + "learning_rate": 1e-06, + "loss": 0.944, + "mean_token_accuracy": 0.7327595353126526, + "num_tokens": 663229362.0, + "step": 25629 + }, + { + "epoch": 2.81462771798814, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9447178840637207, + "learning_rate": 1e-06, + "loss": 0.9453, + "mean_token_accuracy": 0.7254205346107483, + "num_tokens": 663251463.0, + "step": 25630 + }, + { + "epoch": 2.814737535690753, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6924123764038086, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.7020226716995239, + "num_tokens": 663275502.0, + "step": 25631 + }, + { + "epoch": 2.814847353393367, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.456350326538086, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7045910358428955, + "num_tokens": 663305986.0, + "step": 25632 + }, + { + "epoch": 2.8149571710959806, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6466848850250244, + "learning_rate": 1e-06, + "loss": 0.9178, + "mean_token_accuracy": 0.7337713241577148, + "num_tokens": 663330428.0, + "step": 25633 + }, + { + "epoch": 2.8150669887985944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.600691556930542, + "learning_rate": 1e-06, + "loss": 1.0137, + "mean_token_accuracy": 0.7082222104072571, + "num_tokens": 663356398.0, + "step": 25634 + }, + { + "epoch": 2.815176806501208, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.890197277069092, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7068704962730408, + "num_tokens": 663376647.0, + "step": 25635 + }, + { + "epoch": 2.8152866242038215, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7925384044647217, + "learning_rate": 1e-06, + "loss": 0.9784, + "mean_token_accuracy": 0.7163275480270386, + "num_tokens": 663399079.0, + "step": 25636 + }, + { + "epoch": 2.8153964419064352, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4881792068481445, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7177746295928955, + "num_tokens": 663424783.0, + "step": 25637 + }, + { + "epoch": 2.815506259609049, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.378530740737915, + "learning_rate": 1e-06, + "loss": 1.0247, + "mean_token_accuracy": 0.699872612953186, + "num_tokens": 663456579.0, + "step": 25638 + }, + { + "epoch": 2.8156160773116627, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6321988105773926, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.6986815929412842, + "num_tokens": 663481581.0, + "step": 25639 + }, + { + "epoch": 2.8157258950142765, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7519311904907227, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7124239206314087, + "num_tokens": 663504593.0, + "step": 25640 + }, + { + "epoch": 2.81583571271689, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.57478666305542, + "learning_rate": 1e-06, + "loss": 0.9913, + "mean_token_accuracy": 0.7166440486907959, + "num_tokens": 663529251.0, + "step": 25641 + }, + { + "epoch": 2.8159455304195036, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.472121238708496, + "learning_rate": 1e-06, + "loss": 0.9286, + "mean_token_accuracy": 0.7310703992843628, + "num_tokens": 663553956.0, + "step": 25642 + }, + { + "epoch": 2.8160553481221173, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7192957401275635, + "learning_rate": 1e-06, + "loss": 0.9452, + "mean_token_accuracy": 0.7195116877555847, + "num_tokens": 663575866.0, + "step": 25643 + }, + { + "epoch": 2.816165165824731, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3608882427215576, + "learning_rate": 1e-06, + "loss": 1.0187, + "mean_token_accuracy": 0.7027122378349304, + "num_tokens": 663609903.0, + "step": 25644 + }, + { + "epoch": 2.816274983527345, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.550062894821167, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.706020712852478, + "num_tokens": 663637331.0, + "step": 25645 + }, + { + "epoch": 2.816384801229958, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4280261993408203, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7165040969848633, + "num_tokens": 663667723.0, + "step": 25646 + }, + { + "epoch": 2.816494618932572, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.660053253173828, + "learning_rate": 1e-06, + "loss": 1.0422, + "mean_token_accuracy": 0.6982450485229492, + "num_tokens": 663691961.0, + "step": 25647 + }, + { + "epoch": 2.8166044366351857, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1243112087249756, + "learning_rate": 1e-06, + "loss": 0.9174, + "mean_token_accuracy": 0.7347854971885681, + "num_tokens": 663712363.0, + "step": 25648 + }, + { + "epoch": 2.816714254337799, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5642073154449463, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6964574456214905, + "num_tokens": 663744232.0, + "step": 25649 + }, + { + "epoch": 2.816824072040413, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.460577964782715, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.7038429379463196, + "num_tokens": 663771402.0, + "step": 25650 + }, + { + "epoch": 2.8169338897430265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5750460624694824, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.708289623260498, + "num_tokens": 663798841.0, + "step": 25651 + }, + { + "epoch": 2.8170437074456403, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.040148973464966, + "learning_rate": 1e-06, + "loss": 0.9995, + "mean_token_accuracy": 0.7036510705947876, + "num_tokens": 663825270.0, + "step": 25652 + }, + { + "epoch": 2.817153525148254, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2153215408325195, + "learning_rate": 1e-06, + "loss": 1.078, + "mean_token_accuracy": 0.6936937570571899, + "num_tokens": 663856581.0, + "step": 25653 + }, + { + "epoch": 2.8172633428508673, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4356465339660645, + "learning_rate": 1e-06, + "loss": 0.9197, + "mean_token_accuracy": 0.7348482012748718, + "num_tokens": 663882578.0, + "step": 25654 + }, + { + "epoch": 2.817373160553481, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.502270221710205, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7030264139175415, + "num_tokens": 663908507.0, + "step": 25655 + }, + { + "epoch": 2.817482978256095, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5458364486694336, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7169477939605713, + "num_tokens": 663934012.0, + "step": 25656 + }, + { + "epoch": 2.8175927959587086, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.254661798477173, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6985859870910645, + "num_tokens": 663971194.0, + "step": 25657 + }, + { + "epoch": 2.8177026136613224, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.564882278442383, + "learning_rate": 1e-06, + "loss": 0.9409, + "mean_token_accuracy": 0.7221022248268127, + "num_tokens": 663996352.0, + "step": 25658 + }, + { + "epoch": 2.8178124313639357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8090476989746094, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6971908807754517, + "num_tokens": 664016776.0, + "step": 25659 + }, + { + "epoch": 2.8179222490665494, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4927635192871094, + "learning_rate": 1e-06, + "loss": 0.9166, + "mean_token_accuracy": 0.7341421246528625, + "num_tokens": 664042536.0, + "step": 25660 + }, + { + "epoch": 2.818032066769163, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4497416019439697, + "learning_rate": 1e-06, + "loss": 0.9163, + "mean_token_accuracy": 0.7373255491256714, + "num_tokens": 664070228.0, + "step": 25661 + }, + { + "epoch": 2.818141884471777, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.809014081954956, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7187232971191406, + "num_tokens": 664093475.0, + "step": 25662 + }, + { + "epoch": 2.8182517021743907, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.855247735977173, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7097920775413513, + "num_tokens": 664114673.0, + "step": 25663 + }, + { + "epoch": 2.818361519877004, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.568234920501709, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.709107518196106, + "num_tokens": 664139000.0, + "step": 25664 + }, + { + "epoch": 2.8184713375796178, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5185842514038086, + "learning_rate": 1e-06, + "loss": 1.0567, + "mean_token_accuracy": 0.6989679336547852, + "num_tokens": 664169452.0, + "step": 25665 + }, + { + "epoch": 2.8185811552822315, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.035313129425049, + "learning_rate": 1e-06, + "loss": 0.9754, + "mean_token_accuracy": 0.7162885665893555, + "num_tokens": 664190664.0, + "step": 25666 + }, + { + "epoch": 2.8186909729848453, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6874868869781494, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.707821786403656, + "num_tokens": 664214414.0, + "step": 25667 + }, + { + "epoch": 2.818800790687459, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.376570224761963, + "learning_rate": 1e-06, + "loss": 1.0028, + "mean_token_accuracy": 0.7080641984939575, + "num_tokens": 664243671.0, + "step": 25668 + }, + { + "epoch": 2.8189106083900723, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.376187562942505, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.6976920366287231, + "num_tokens": 664272682.0, + "step": 25669 + }, + { + "epoch": 2.819020426092686, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.762216806411743, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7073280811309814, + "num_tokens": 664297167.0, + "step": 25670 + }, + { + "epoch": 2.8191302437953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4646804332733154, + "learning_rate": 1e-06, + "loss": 1.0662, + "mean_token_accuracy": 0.6957137584686279, + "num_tokens": 664327436.0, + "step": 25671 + }, + { + "epoch": 2.8192400614979136, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.53798508644104, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7077295780181885, + "num_tokens": 664355794.0, + "step": 25672 + }, + { + "epoch": 2.8193498792005274, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4182381629943848, + "learning_rate": 1e-06, + "loss": 1.0842, + "mean_token_accuracy": 0.6864618062973022, + "num_tokens": 664385123.0, + "step": 25673 + }, + { + "epoch": 2.8194596969031407, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.4958086013793945, + "learning_rate": 1e-06, + "loss": 0.8683, + "mean_token_accuracy": 0.7445856928825378, + "num_tokens": 664399166.0, + "step": 25674 + }, + { + "epoch": 2.8195695146057544, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5367233753204346, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6963160037994385, + "num_tokens": 664426734.0, + "step": 25675 + }, + { + "epoch": 2.819679332308368, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.809154510498047, + "learning_rate": 1e-06, + "loss": 0.9288, + "mean_token_accuracy": 0.722240149974823, + "num_tokens": 664449383.0, + "step": 25676 + }, + { + "epoch": 2.8197891500109815, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.405165910720825, + "learning_rate": 1e-06, + "loss": 0.9842, + "mean_token_accuracy": 0.7181432247161865, + "num_tokens": 664478943.0, + "step": 25677 + }, + { + "epoch": 2.8198989677135953, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5380728244781494, + "learning_rate": 1e-06, + "loss": 1.0409, + "mean_token_accuracy": 0.703944206237793, + "num_tokens": 664504458.0, + "step": 25678 + }, + { + "epoch": 2.820008785416209, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3327713012695312, + "learning_rate": 1e-06, + "loss": 1.0613, + "mean_token_accuracy": 0.6958141326904297, + "num_tokens": 664535890.0, + "step": 25679 + }, + { + "epoch": 2.820118603118823, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6264548301696777, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.717684268951416, + "num_tokens": 664558345.0, + "step": 25680 + }, + { + "epoch": 2.8202284208214365, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.607980251312256, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7131980657577515, + "num_tokens": 664580299.0, + "step": 25681 + }, + { + "epoch": 2.82033823852405, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.033701181411743, + "learning_rate": 1e-06, + "loss": 0.9986, + "mean_token_accuracy": 0.711228609085083, + "num_tokens": 664599077.0, + "step": 25682 + }, + { + "epoch": 2.8204480562266636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8041110038757324, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7109489440917969, + "num_tokens": 664624121.0, + "step": 25683 + }, + { + "epoch": 2.8205578739292774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2297489643096924, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.7118187546730042, + "num_tokens": 664660951.0, + "step": 25684 + }, + { + "epoch": 2.820667691631891, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9830267429351807, + "learning_rate": 1e-06, + "loss": 0.9244, + "mean_token_accuracy": 0.7349708080291748, + "num_tokens": 664684410.0, + "step": 25685 + }, + { + "epoch": 2.820777509334505, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7579362392425537, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7080879807472229, + "num_tokens": 664706851.0, + "step": 25686 + }, + { + "epoch": 2.820887327037118, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6565613746643066, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7040655016899109, + "num_tokens": 664731863.0, + "step": 25687 + }, + { + "epoch": 2.820997144739732, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8692779541015625, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7182148098945618, + "num_tokens": 664752228.0, + "step": 25688 + }, + { + "epoch": 2.8211069624423457, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.659952163696289, + "learning_rate": 1e-06, + "loss": 0.8558, + "mean_token_accuracy": 0.7399086356163025, + "num_tokens": 664773933.0, + "step": 25689 + }, + { + "epoch": 2.8212167801449595, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.42997145652771, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7123820781707764, + "num_tokens": 664802609.0, + "step": 25690 + }, + { + "epoch": 2.8213265978475732, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.005326986312866, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.699268102645874, + "num_tokens": 664821200.0, + "step": 25691 + }, + { + "epoch": 2.8214364155501865, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6348876953125, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6956007480621338, + "num_tokens": 664848426.0, + "step": 25692 + }, + { + "epoch": 2.8215462332528003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.418501138687134, + "learning_rate": 1e-06, + "loss": 1.0033, + "mean_token_accuracy": 0.7124802470207214, + "num_tokens": 664874057.0, + "step": 25693 + }, + { + "epoch": 2.821656050955414, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7652413845062256, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7105507850646973, + "num_tokens": 664897031.0, + "step": 25694 + }, + { + "epoch": 2.821765868658028, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5910634994506836, + "learning_rate": 1e-06, + "loss": 1.105, + "mean_token_accuracy": 0.680260419845581, + "num_tokens": 664922382.0, + "step": 25695 + }, + { + "epoch": 2.8218756863606416, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7716307640075684, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7094993591308594, + "num_tokens": 664946332.0, + "step": 25696 + }, + { + "epoch": 2.821985504063255, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5945987701416016, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.7099893093109131, + "num_tokens": 664970614.0, + "step": 25697 + }, + { + "epoch": 2.8220953217658686, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6089656352996826, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.704259991645813, + "num_tokens": 664997642.0, + "step": 25698 + }, + { + "epoch": 2.8222051394684824, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.387298345565796, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.7206760048866272, + "num_tokens": 665025423.0, + "step": 25699 + }, + { + "epoch": 2.8223149571710957, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4268698692321777, + "learning_rate": 1e-06, + "loss": 1.0328, + "mean_token_accuracy": 0.7042258977890015, + "num_tokens": 665053357.0, + "step": 25700 + }, + { + "epoch": 2.82242477487371, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6061151027679443, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7158844470977783, + "num_tokens": 665078191.0, + "step": 25701 + }, + { + "epoch": 2.822534592576323, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.73279571533203, + "learning_rate": 1e-06, + "loss": 0.9074, + "mean_token_accuracy": 0.731360673904419, + "num_tokens": 665100123.0, + "step": 25702 + }, + { + "epoch": 2.822644410278937, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 5.375677585601807, + "learning_rate": 1e-06, + "loss": 1.1369, + "mean_token_accuracy": 0.6767115592956543, + "num_tokens": 665130091.0, + "step": 25703 + }, + { + "epoch": 2.8227542279815507, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.1615452766418457, + "learning_rate": 1e-06, + "loss": 0.9922, + "mean_token_accuracy": 0.7105871438980103, + "num_tokens": 665148247.0, + "step": 25704 + }, + { + "epoch": 2.822864045684164, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5437703132629395, + "learning_rate": 1e-06, + "loss": 1.0953, + "mean_token_accuracy": 0.6888200044631958, + "num_tokens": 665178671.0, + "step": 25705 + }, + { + "epoch": 2.822973863386778, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.415742874145508, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7094908952713013, + "num_tokens": 665206942.0, + "step": 25706 + }, + { + "epoch": 2.8230836810893916, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8020262718200684, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7045645117759705, + "num_tokens": 665228016.0, + "step": 25707 + }, + { + "epoch": 2.8231934987920053, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6694791316986084, + "learning_rate": 1e-06, + "loss": 1.0432, + "mean_token_accuracy": 0.7049598693847656, + "num_tokens": 665252564.0, + "step": 25708 + }, + { + "epoch": 2.823303316494619, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.528867483139038, + "learning_rate": 1e-06, + "loss": 1.0816, + "mean_token_accuracy": 0.6910779476165771, + "num_tokens": 665279318.0, + "step": 25709 + }, + { + "epoch": 2.8234131341972324, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.517866373062134, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7200018167495728, + "num_tokens": 665305386.0, + "step": 25710 + }, + { + "epoch": 2.823522951899846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.701871633529663, + "learning_rate": 1e-06, + "loss": 0.8478, + "mean_token_accuracy": 0.7506734132766724, + "num_tokens": 665327383.0, + "step": 25711 + }, + { + "epoch": 2.82363276960246, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4670658111572266, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6933103203773499, + "num_tokens": 665356505.0, + "step": 25712 + }, + { + "epoch": 2.8237425873050737, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6361706256866455, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6916105151176453, + "num_tokens": 665382325.0, + "step": 25713 + }, + { + "epoch": 2.8238524050076874, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.499929428100586, + "learning_rate": 1e-06, + "loss": 0.9686, + "mean_token_accuracy": 0.7206968665122986, + "num_tokens": 665408629.0, + "step": 25714 + }, + { + "epoch": 2.8239622227103007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.472912549972534, + "learning_rate": 1e-06, + "loss": 1.0336, + "mean_token_accuracy": 0.7029283046722412, + "num_tokens": 665436323.0, + "step": 25715 + }, + { + "epoch": 2.8240720404129145, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6085267066955566, + "learning_rate": 1e-06, + "loss": 0.8937, + "mean_token_accuracy": 0.7472953796386719, + "num_tokens": 665463777.0, + "step": 25716 + }, + { + "epoch": 2.8241818581155282, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8285646438598633, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7100176811218262, + "num_tokens": 665485836.0, + "step": 25717 + }, + { + "epoch": 2.824291675818142, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9657697677612305, + "learning_rate": 1e-06, + "loss": 1.0528, + "mean_token_accuracy": 0.7019460201263428, + "num_tokens": 665510241.0, + "step": 25718 + }, + { + "epoch": 2.8244014935207558, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5051565170288086, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6912119388580322, + "num_tokens": 665539111.0, + "step": 25719 + }, + { + "epoch": 2.824511311223369, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.407155752182007, + "learning_rate": 1e-06, + "loss": 1.1197, + "mean_token_accuracy": 0.6881234049797058, + "num_tokens": 665572159.0, + "step": 25720 + }, + { + "epoch": 2.824621128925983, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.55600643157959, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6960914134979248, + "num_tokens": 665600417.0, + "step": 25721 + }, + { + "epoch": 2.8247309466285966, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7058510780334473, + "learning_rate": 1e-06, + "loss": 1.0623, + "mean_token_accuracy": 0.6984611749649048, + "num_tokens": 665625111.0, + "step": 25722 + }, + { + "epoch": 2.8248407643312103, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.975860834121704, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7148042917251587, + "num_tokens": 665645082.0, + "step": 25723 + }, + { + "epoch": 2.824950582033824, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.639765977859497, + "learning_rate": 1e-06, + "loss": 1.0569, + "mean_token_accuracy": 0.6948224306106567, + "num_tokens": 665675922.0, + "step": 25724 + }, + { + "epoch": 2.8250603997364374, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.647932529449463, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.722689688205719, + "num_tokens": 665701672.0, + "step": 25725 + }, + { + "epoch": 2.825170217439051, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.674150228500366, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.7202236652374268, + "num_tokens": 665726620.0, + "step": 25726 + }, + { + "epoch": 2.825280035141665, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.498141050338745, + "learning_rate": 1e-06, + "loss": 1.0624, + "mean_token_accuracy": 0.6912434101104736, + "num_tokens": 665752886.0, + "step": 25727 + }, + { + "epoch": 2.8253898528442782, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 4.4313764572143555, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.704913854598999, + "num_tokens": 665776049.0, + "step": 25728 + }, + { + "epoch": 2.825499670546892, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5944018363952637, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7114619016647339, + "num_tokens": 665801166.0, + "step": 25729 + }, + { + "epoch": 2.8256094882495058, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505478858947754, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6864340901374817, + "num_tokens": 665829909.0, + "step": 25730 + }, + { + "epoch": 2.8257193059521195, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8498690128326416, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7115633487701416, + "num_tokens": 665851708.0, + "step": 25731 + }, + { + "epoch": 2.8258291236547333, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6473591327667236, + "learning_rate": 1e-06, + "loss": 1.0184, + "mean_token_accuracy": 0.7036672234535217, + "num_tokens": 665875189.0, + "step": 25732 + }, + { + "epoch": 2.8259389413573466, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6197197437286377, + "learning_rate": 1e-06, + "loss": 1.0153, + "mean_token_accuracy": 0.7016795873641968, + "num_tokens": 665897904.0, + "step": 25733 + }, + { + "epoch": 2.8260487590599603, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6573939323425293, + "learning_rate": 1e-06, + "loss": 0.9048, + "mean_token_accuracy": 0.7300460338592529, + "num_tokens": 665921091.0, + "step": 25734 + }, + { + "epoch": 2.826158576762574, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3593709468841553, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7246805429458618, + "num_tokens": 665948538.0, + "step": 25735 + }, + { + "epoch": 2.826268394465188, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7305099964141846, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.6983758211135864, + "num_tokens": 665974966.0, + "step": 25736 + }, + { + "epoch": 2.8263782121678016, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.879417896270752, + "learning_rate": 1e-06, + "loss": 0.9845, + "mean_token_accuracy": 0.7147121429443359, + "num_tokens": 665994760.0, + "step": 25737 + }, + { + "epoch": 2.826488029870415, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7428505420684814, + "learning_rate": 1e-06, + "loss": 1.0278, + "mean_token_accuracy": 0.7062881588935852, + "num_tokens": 666021963.0, + "step": 25738 + }, + { + "epoch": 2.8265978475730287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.464078187942505, + "learning_rate": 1e-06, + "loss": 0.9675, + "mean_token_accuracy": 0.7126717567443848, + "num_tokens": 666050709.0, + "step": 25739 + }, + { + "epoch": 2.8267076652756424, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4745547771453857, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6953048706054688, + "num_tokens": 666080470.0, + "step": 25740 + }, + { + "epoch": 2.826817482978256, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5848464965820312, + "learning_rate": 1e-06, + "loss": 1.0192, + "mean_token_accuracy": 0.7007172703742981, + "num_tokens": 666105655.0, + "step": 25741 + }, + { + "epoch": 2.82692730068087, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.790142774581909, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7092273235321045, + "num_tokens": 666129224.0, + "step": 25742 + }, + { + "epoch": 2.8270371183834833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.595094919204712, + "learning_rate": 1e-06, + "loss": 0.9747, + "mean_token_accuracy": 0.7091094255447388, + "num_tokens": 666153247.0, + "step": 25743 + }, + { + "epoch": 2.827146936086097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8319873809814453, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7065648436546326, + "num_tokens": 666175056.0, + "step": 25744 + }, + { + "epoch": 2.8272567537887108, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.432483434677124, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6993966102600098, + "num_tokens": 666207477.0, + "step": 25745 + }, + { + "epoch": 2.8273665714913245, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.588230848312378, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.6994720697402954, + "num_tokens": 666233368.0, + "step": 25746 + }, + { + "epoch": 2.8274763891939383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5766475200653076, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7052531242370605, + "num_tokens": 666258159.0, + "step": 25747 + }, + { + "epoch": 2.8275862068965516, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.537867546081543, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7068161964416504, + "num_tokens": 666284219.0, + "step": 25748 + }, + { + "epoch": 2.8276960245991654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.803262948989868, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7129451632499695, + "num_tokens": 666305589.0, + "step": 25749 + }, + { + "epoch": 2.827805842301779, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.664832353591919, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.694217324256897, + "num_tokens": 666330272.0, + "step": 25750 + }, + { + "epoch": 2.827915660004393, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5847692489624023, + "learning_rate": 1e-06, + "loss": 1.0099, + "mean_token_accuracy": 0.7119398713111877, + "num_tokens": 666356032.0, + "step": 25751 + }, + { + "epoch": 2.8280254777070066, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.292530059814453, + "learning_rate": 1e-06, + "loss": 0.9156, + "mean_token_accuracy": 0.7268512845039368, + "num_tokens": 666372562.0, + "step": 25752 + }, + { + "epoch": 2.82813529540962, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8646976947784424, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7125941514968872, + "num_tokens": 666395419.0, + "step": 25753 + }, + { + "epoch": 2.8282451131122337, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536757469177246, + "learning_rate": 1e-06, + "loss": 0.9305, + "mean_token_accuracy": 0.7378917932510376, + "num_tokens": 666420506.0, + "step": 25754 + }, + { + "epoch": 2.8283549308148475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5326309204101562, + "learning_rate": 1e-06, + "loss": 1.0819, + "mean_token_accuracy": 0.6896483302116394, + "num_tokens": 666450363.0, + "step": 25755 + }, + { + "epoch": 2.8284647485174608, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6671783924102783, + "learning_rate": 1e-06, + "loss": 0.9575, + "mean_token_accuracy": 0.719597339630127, + "num_tokens": 666472731.0, + "step": 25756 + }, + { + "epoch": 2.8285745662200745, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.616533041000366, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.6995410323143005, + "num_tokens": 666498792.0, + "step": 25757 + }, + { + "epoch": 2.8286843839226883, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.462453842163086, + "learning_rate": 1e-06, + "loss": 1.0557, + "mean_token_accuracy": 0.6966254115104675, + "num_tokens": 666526752.0, + "step": 25758 + }, + { + "epoch": 2.828794201625302, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7488489151000977, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7363325357437134, + "num_tokens": 666548945.0, + "step": 25759 + }, + { + "epoch": 2.828904019327916, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.958763837814331, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6961930990219116, + "num_tokens": 666569835.0, + "step": 25760 + }, + { + "epoch": 2.829013837030529, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6432933807373047, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7098420262336731, + "num_tokens": 666593958.0, + "step": 25761 + }, + { + "epoch": 2.829123654733143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6545755863189697, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7015010118484497, + "num_tokens": 666618422.0, + "step": 25762 + }, + { + "epoch": 2.8292334724357566, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.86686110496521, + "learning_rate": 1e-06, + "loss": 0.9161, + "mean_token_accuracy": 0.7327753305435181, + "num_tokens": 666639019.0, + "step": 25763 + }, + { + "epoch": 2.8293432901383704, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.746601104736328, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7070420980453491, + "num_tokens": 666662132.0, + "step": 25764 + }, + { + "epoch": 2.829453107840984, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 3.2568631172180176, + "learning_rate": 1e-06, + "loss": 0.9867, + "mean_token_accuracy": 0.7197601199150085, + "num_tokens": 666680745.0, + "step": 25765 + }, + { + "epoch": 2.8295629255435975, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5056262016296387, + "learning_rate": 1e-06, + "loss": 0.9423, + "mean_token_accuracy": 0.7253434062004089, + "num_tokens": 666708013.0, + "step": 25766 + }, + { + "epoch": 2.829672743246211, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.620603084564209, + "learning_rate": 1e-06, + "loss": 1.1414, + "mean_token_accuracy": 0.6722615957260132, + "num_tokens": 666738869.0, + "step": 25767 + }, + { + "epoch": 2.829782560948825, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.54801082611084, + "learning_rate": 1e-06, + "loss": 1.0182, + "mean_token_accuracy": 0.7130910754203796, + "num_tokens": 666764646.0, + "step": 25768 + }, + { + "epoch": 2.8298923786514387, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.660158157348633, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6920185089111328, + "num_tokens": 666792890.0, + "step": 25769 + }, + { + "epoch": 2.8300021963540525, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.798802375793457, + "learning_rate": 1e-06, + "loss": 0.9613, + "mean_token_accuracy": 0.7205463647842407, + "num_tokens": 666817196.0, + "step": 25770 + }, + { + "epoch": 2.830112014056666, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.759152412414551, + "learning_rate": 1e-06, + "loss": 0.8604, + "mean_token_accuracy": 0.7448438405990601, + "num_tokens": 666837965.0, + "step": 25771 + }, + { + "epoch": 2.8302218317592795, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7461109161376953, + "learning_rate": 1e-06, + "loss": 1.0894, + "mean_token_accuracy": 0.7027384042739868, + "num_tokens": 666861412.0, + "step": 25772 + }, + { + "epoch": 2.8303316494618933, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8599987030029297, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.702847957611084, + "num_tokens": 666883947.0, + "step": 25773 + }, + { + "epoch": 2.830441467164507, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.002580404281616, + "learning_rate": 1e-06, + "loss": 0.953, + "mean_token_accuracy": 0.7290796041488647, + "num_tokens": 666904341.0, + "step": 25774 + }, + { + "epoch": 2.830551284867121, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.798945665359497, + "learning_rate": 1e-06, + "loss": 0.9807, + "mean_token_accuracy": 0.7148867845535278, + "num_tokens": 666926188.0, + "step": 25775 + }, + { + "epoch": 2.830661102569734, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.983442544937134, + "learning_rate": 1e-06, + "loss": 1.002, + "mean_token_accuracy": 0.7125094532966614, + "num_tokens": 666950234.0, + "step": 25776 + }, + { + "epoch": 2.830770920272348, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.547745704650879, + "learning_rate": 1e-06, + "loss": 1.0223, + "mean_token_accuracy": 0.7025704979896545, + "num_tokens": 666976412.0, + "step": 25777 + }, + { + "epoch": 2.8308807379749616, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5714681148529053, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6822754144668579, + "num_tokens": 667005457.0, + "step": 25778 + }, + { + "epoch": 2.830990555677575, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7667293548583984, + "learning_rate": 1e-06, + "loss": 0.9727, + "mean_token_accuracy": 0.7155424356460571, + "num_tokens": 667028380.0, + "step": 25779 + }, + { + "epoch": 2.831100373380189, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.408958911895752, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.6911128163337708, + "num_tokens": 667059785.0, + "step": 25780 + }, + { + "epoch": 2.8312101910828025, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.772367000579834, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7159427404403687, + "num_tokens": 667083512.0, + "step": 25781 + }, + { + "epoch": 2.8313200087854162, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8713276386260986, + "learning_rate": 1e-06, + "loss": 0.8602, + "mean_token_accuracy": 0.737881064414978, + "num_tokens": 667104526.0, + "step": 25782 + }, + { + "epoch": 2.83142982648803, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8057498931884766, + "learning_rate": 1e-06, + "loss": 1.061, + "mean_token_accuracy": 0.6932995915412903, + "num_tokens": 667125859.0, + "step": 25783 + }, + { + "epoch": 2.8315396441906433, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.794996976852417, + "learning_rate": 1e-06, + "loss": 0.9859, + "mean_token_accuracy": 0.7131296396255493, + "num_tokens": 667151027.0, + "step": 25784 + }, + { + "epoch": 2.831649461893257, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3329055309295654, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.724325954914093, + "num_tokens": 667186120.0, + "step": 25785 + }, + { + "epoch": 2.831759279595871, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7835845947265625, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.7007697820663452, + "num_tokens": 667209178.0, + "step": 25786 + }, + { + "epoch": 2.8318690972984846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5663814544677734, + "learning_rate": 1e-06, + "loss": 0.9482, + "mean_token_accuracy": 0.7182454466819763, + "num_tokens": 667233824.0, + "step": 25787 + }, + { + "epoch": 2.8319789150010983, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8888628482818604, + "learning_rate": 1e-06, + "loss": 0.9653, + "mean_token_accuracy": 0.7210418581962585, + "num_tokens": 667255284.0, + "step": 25788 + }, + { + "epoch": 2.8320887327037116, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.400294065475464, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6918954253196716, + "num_tokens": 667284434.0, + "step": 25789 + }, + { + "epoch": 2.8321985504063254, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.607797145843506, + "learning_rate": 1e-06, + "loss": 1.0365, + "mean_token_accuracy": 0.7049602270126343, + "num_tokens": 667307963.0, + "step": 25790 + }, + { + "epoch": 2.832308368108939, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.78725266456604, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7125468254089355, + "num_tokens": 667330400.0, + "step": 25791 + }, + { + "epoch": 2.832418185811553, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7103588581085205, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.6973329782485962, + "num_tokens": 667353480.0, + "step": 25792 + }, + { + "epoch": 2.8325280035141667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6432714462280273, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7201427817344666, + "num_tokens": 667378113.0, + "step": 25793 + }, + { + "epoch": 2.83263782121678, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4689061641693115, + "learning_rate": 1e-06, + "loss": 0.9109, + "mean_token_accuracy": 0.7301678657531738, + "num_tokens": 667405448.0, + "step": 25794 + }, + { + "epoch": 2.8327476389193937, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4430625438690186, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7276099324226379, + "num_tokens": 667431979.0, + "step": 25795 + }, + { + "epoch": 2.8328574566220075, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.82731294631958, + "learning_rate": 1e-06, + "loss": 0.9396, + "mean_token_accuracy": 0.7216723561286926, + "num_tokens": 667454702.0, + "step": 25796 + }, + { + "epoch": 2.8329672743246213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.884197235107422, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.7008152008056641, + "num_tokens": 667476501.0, + "step": 25797 + }, + { + "epoch": 2.833077092027235, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6499125957489014, + "learning_rate": 1e-06, + "loss": 1.0101, + "mean_token_accuracy": 0.7070584297180176, + "num_tokens": 667502412.0, + "step": 25798 + }, + { + "epoch": 2.8331869097298483, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.109576463699341, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.7326910495758057, + "num_tokens": 667521563.0, + "step": 25799 + }, + { + "epoch": 2.833296727432462, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.559812307357788, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7066440582275391, + "num_tokens": 667548691.0, + "step": 25800 + }, + { + "epoch": 2.833406545135076, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7054390907287598, + "learning_rate": 1e-06, + "loss": 1.0645, + "mean_token_accuracy": 0.693006157875061, + "num_tokens": 667575700.0, + "step": 25801 + }, + { + "epoch": 2.8335163628376896, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6891369819641113, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7094847559928894, + "num_tokens": 667599917.0, + "step": 25802 + }, + { + "epoch": 2.8336261805403034, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4618237018585205, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.6943278312683105, + "num_tokens": 667630484.0, + "step": 25803 + }, + { + "epoch": 2.8337359982429167, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.301536798477173, + "learning_rate": 1e-06, + "loss": 1.099, + "mean_token_accuracy": 0.6839454174041748, + "num_tokens": 667662177.0, + "step": 25804 + }, + { + "epoch": 2.8338458159455304, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5452802181243896, + "learning_rate": 1e-06, + "loss": 1.0794, + "mean_token_accuracy": 0.6919983625411987, + "num_tokens": 667687948.0, + "step": 25805 + }, + { + "epoch": 2.833955633648144, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7962822914123535, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7046817541122437, + "num_tokens": 667711620.0, + "step": 25806 + }, + { + "epoch": 2.8340654513507575, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7793197631835938, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7320060729980469, + "num_tokens": 667733656.0, + "step": 25807 + }, + { + "epoch": 2.8341752690533712, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2226974964141846, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.7008910775184631, + "num_tokens": 667766584.0, + "step": 25808 + }, + { + "epoch": 2.834285086755985, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4906179904937744, + "learning_rate": 1e-06, + "loss": 1.0527, + "mean_token_accuracy": 0.6942685842514038, + "num_tokens": 667795665.0, + "step": 25809 + }, + { + "epoch": 2.8343949044585988, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.62406849861145, + "learning_rate": 1e-06, + "loss": 0.9889, + "mean_token_accuracy": 0.7103822827339172, + "num_tokens": 667820599.0, + "step": 25810 + }, + { + "epoch": 2.8345047221612125, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5474298000335693, + "learning_rate": 1e-06, + "loss": 0.9788, + "mean_token_accuracy": 0.7170678973197937, + "num_tokens": 667849042.0, + "step": 25811 + }, + { + "epoch": 2.834614539863826, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.635451078414917, + "learning_rate": 1e-06, + "loss": 0.9245, + "mean_token_accuracy": 0.7291488647460938, + "num_tokens": 667877786.0, + "step": 25812 + }, + { + "epoch": 2.8347243575664396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8577868938446045, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7240815758705139, + "num_tokens": 667900474.0, + "step": 25813 + }, + { + "epoch": 2.8348341752690533, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.452044725418091, + "learning_rate": 1e-06, + "loss": 0.9778, + "mean_token_accuracy": 0.7159984111785889, + "num_tokens": 667930778.0, + "step": 25814 + }, + { + "epoch": 2.834943992971667, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6107120513916016, + "learning_rate": 1e-06, + "loss": 0.9447, + "mean_token_accuracy": 0.7290112972259521, + "num_tokens": 667956985.0, + "step": 25815 + }, + { + "epoch": 2.835053810674281, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7070255279541016, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.736772894859314, + "num_tokens": 667976498.0, + "step": 25816 + }, + { + "epoch": 2.835163628376894, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.836885690689087, + "learning_rate": 1e-06, + "loss": 1.1171, + "mean_token_accuracy": 0.6806514859199524, + "num_tokens": 667999562.0, + "step": 25817 + }, + { + "epoch": 2.835273446079508, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.658740758895874, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.705903947353363, + "num_tokens": 668024673.0, + "step": 25818 + }, + { + "epoch": 2.8353832637821217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9434425830841064, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7226234674453735, + "num_tokens": 668044984.0, + "step": 25819 + }, + { + "epoch": 2.8354930814847354, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.099400758743286, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.72495436668396, + "num_tokens": 668062676.0, + "step": 25820 + }, + { + "epoch": 2.835602899187349, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5009851455688477, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6907130479812622, + "num_tokens": 668095964.0, + "step": 25821 + }, + { + "epoch": 2.8357127168899625, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5994651317596436, + "learning_rate": 1e-06, + "loss": 1.023, + "mean_token_accuracy": 0.7038850784301758, + "num_tokens": 668120610.0, + "step": 25822 + }, + { + "epoch": 2.8358225345925763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3402953147888184, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6857304573059082, + "num_tokens": 668154666.0, + "step": 25823 + }, + { + "epoch": 2.83593235229519, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6215569972991943, + "learning_rate": 1e-06, + "loss": 1.0285, + "mean_token_accuracy": 0.692385196685791, + "num_tokens": 668179455.0, + "step": 25824 + }, + { + "epoch": 2.836042169997804, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4845941066741943, + "learning_rate": 1e-06, + "loss": 0.9371, + "mean_token_accuracy": 0.7240356206893921, + "num_tokens": 668204743.0, + "step": 25825 + }, + { + "epoch": 2.8361519877004175, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6190226078033447, + "learning_rate": 1e-06, + "loss": 0.9908, + "mean_token_accuracy": 0.7100169658660889, + "num_tokens": 668229429.0, + "step": 25826 + }, + { + "epoch": 2.836261805403031, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7740039825439453, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.723739504814148, + "num_tokens": 668252642.0, + "step": 25827 + }, + { + "epoch": 2.8363716231056446, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.744137763977051, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.7113402485847473, + "num_tokens": 668274927.0, + "step": 25828 + }, + { + "epoch": 2.8364814408082584, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7122201919555664, + "learning_rate": 1e-06, + "loss": 1.0074, + "mean_token_accuracy": 0.7055046558380127, + "num_tokens": 668300529.0, + "step": 25829 + }, + { + "epoch": 2.8365912585108717, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.668610095977783, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.7205365300178528, + "num_tokens": 668323105.0, + "step": 25830 + }, + { + "epoch": 2.836701076213486, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.821948528289795, + "learning_rate": 1e-06, + "loss": 0.9709, + "mean_token_accuracy": 0.7084476947784424, + "num_tokens": 668343240.0, + "step": 25831 + }, + { + "epoch": 2.836810893916099, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.806128740310669, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.7002626657485962, + "num_tokens": 668366235.0, + "step": 25832 + }, + { + "epoch": 2.836920711618713, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6651909351348877, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6863179802894592, + "num_tokens": 668390790.0, + "step": 25833 + }, + { + "epoch": 2.8370305293213267, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5867979526519775, + "learning_rate": 1e-06, + "loss": 1.0167, + "mean_token_accuracy": 0.7038354277610779, + "num_tokens": 668417367.0, + "step": 25834 + }, + { + "epoch": 2.83714034702394, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9795420169830322, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.715114951133728, + "num_tokens": 668436434.0, + "step": 25835 + }, + { + "epoch": 2.837250164726554, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.661823034286499, + "learning_rate": 1e-06, + "loss": 0.928, + "mean_token_accuracy": 0.725716233253479, + "num_tokens": 668462158.0, + "step": 25836 + }, + { + "epoch": 2.8373599824291675, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8714375495910645, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7032127976417542, + "num_tokens": 668484016.0, + "step": 25837 + }, + { + "epoch": 2.8374698001317813, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6171176433563232, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7063207626342773, + "num_tokens": 668509697.0, + "step": 25838 + }, + { + "epoch": 2.837579617834395, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7342450618743896, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.7098785638809204, + "num_tokens": 668534839.0, + "step": 25839 + }, + { + "epoch": 2.8376894355370084, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.803156614303589, + "learning_rate": 1e-06, + "loss": 1.0135, + "mean_token_accuracy": 0.7030923366546631, + "num_tokens": 668558902.0, + "step": 25840 + }, + { + "epoch": 2.837799253239622, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5118424892425537, + "learning_rate": 1e-06, + "loss": 1.0725, + "mean_token_accuracy": 0.696525514125824, + "num_tokens": 668589644.0, + "step": 25841 + }, + { + "epoch": 2.837909070942236, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3437907695770264, + "learning_rate": 1e-06, + "loss": 1.0307, + "mean_token_accuracy": 0.6973673105239868, + "num_tokens": 668623296.0, + "step": 25842 + }, + { + "epoch": 2.8380188886448496, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4467968940734863, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6917827129364014, + "num_tokens": 668652015.0, + "step": 25843 + }, + { + "epoch": 2.8381287063474634, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9121947288513184, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7015455365180969, + "num_tokens": 668681762.0, + "step": 25844 + }, + { + "epoch": 2.8382385240500767, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.577791690826416, + "learning_rate": 1e-06, + "loss": 0.8977, + "mean_token_accuracy": 0.7340739369392395, + "num_tokens": 668706177.0, + "step": 25845 + }, + { + "epoch": 2.8383483417526905, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.031362533569336, + "learning_rate": 1e-06, + "loss": 0.8832, + "mean_token_accuracy": 0.7393824458122253, + "num_tokens": 668724696.0, + "step": 25846 + }, + { + "epoch": 2.838458159455304, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6157896518707275, + "learning_rate": 1e-06, + "loss": 0.9556, + "mean_token_accuracy": 0.720502495765686, + "num_tokens": 668748726.0, + "step": 25847 + }, + { + "epoch": 2.838567977157918, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7959177494049072, + "learning_rate": 1e-06, + "loss": 0.9502, + "mean_token_accuracy": 0.7232422828674316, + "num_tokens": 668771087.0, + "step": 25848 + }, + { + "epoch": 2.8386777948605317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617166519165039, + "learning_rate": 1e-06, + "loss": 1.0837, + "mean_token_accuracy": 0.700406551361084, + "num_tokens": 668796119.0, + "step": 25849 + }, + { + "epoch": 2.838787612563145, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.694190740585327, + "learning_rate": 1e-06, + "loss": 0.9972, + "mean_token_accuracy": 0.7078831195831299, + "num_tokens": 668821536.0, + "step": 25850 + }, + { + "epoch": 2.838897430265759, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8966197967529297, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7085081338882446, + "num_tokens": 668849118.0, + "step": 25851 + }, + { + "epoch": 2.8390072479683726, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5899617671966553, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6897762417793274, + "num_tokens": 668878656.0, + "step": 25852 + }, + { + "epoch": 2.8391170656709863, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.976837396621704, + "learning_rate": 1e-06, + "loss": 0.8728, + "mean_token_accuracy": 0.7419557571411133, + "num_tokens": 668897345.0, + "step": 25853 + }, + { + "epoch": 2.8392268833736, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.708310604095459, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7017462253570557, + "num_tokens": 668921684.0, + "step": 25854 + }, + { + "epoch": 2.8393367010762134, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4719533920288086, + "learning_rate": 1e-06, + "loss": 1.0004, + "mean_token_accuracy": 0.7047923803329468, + "num_tokens": 668948081.0, + "step": 25855 + }, + { + "epoch": 2.839446518778827, + "ewc_loss": 1.1026859283447266e-05, + "grad_norm": 80.53672790527344, + "learning_rate": 1e-06, + "loss": 1.131, + "mean_token_accuracy": 0.6778309345245361, + "num_tokens": 668978585.0, + "step": 25856 + }, + { + "epoch": 2.839556336481441, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4030730724334717, + "learning_rate": 1e-06, + "loss": 1.0418, + "mean_token_accuracy": 0.6916571855545044, + "num_tokens": 669009970.0, + "step": 25857 + }, + { + "epoch": 2.839666154184054, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.3151299953460693, + "learning_rate": 1e-06, + "loss": 1.02, + "mean_token_accuracy": 0.7021033763885498, + "num_tokens": 669036112.0, + "step": 25858 + }, + { + "epoch": 2.839775971886668, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6707253456115723, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7132289409637451, + "num_tokens": 669062067.0, + "step": 25859 + }, + { + "epoch": 2.8398857895892817, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5346338748931885, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7215293645858765, + "num_tokens": 669089372.0, + "step": 25860 + }, + { + "epoch": 2.8399956072918955, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.479356288909912, + "learning_rate": 1e-06, + "loss": 1.008, + "mean_token_accuracy": 0.7112302780151367, + "num_tokens": 669115387.0, + "step": 25861 + }, + { + "epoch": 2.8401054249945092, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.663600206375122, + "learning_rate": 1e-06, + "loss": 1.0309, + "mean_token_accuracy": 0.6995917558670044, + "num_tokens": 669139887.0, + "step": 25862 + }, + { + "epoch": 2.8402152426971226, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.488530158996582, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6857722401618958, + "num_tokens": 669169332.0, + "step": 25863 + }, + { + "epoch": 2.8403250603997363, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.002791404724121, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.7056734561920166, + "num_tokens": 669193611.0, + "step": 25864 + }, + { + "epoch": 2.84043487810235, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6143386363983154, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7275469303131104, + "num_tokens": 669217802.0, + "step": 25865 + }, + { + "epoch": 2.840544695804964, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6195321083068848, + "learning_rate": 1e-06, + "loss": 0.937, + "mean_token_accuracy": 0.7219095230102539, + "num_tokens": 669241755.0, + "step": 25866 + }, + { + "epoch": 2.8406545135075776, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8062498569488525, + "learning_rate": 1e-06, + "loss": 1.1048, + "mean_token_accuracy": 0.6839410662651062, + "num_tokens": 669263832.0, + "step": 25867 + }, + { + "epoch": 2.840764331210191, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7633912563323975, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7199506163597107, + "num_tokens": 669286704.0, + "step": 25868 + }, + { + "epoch": 2.8408741489128047, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5534017086029053, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7090606689453125, + "num_tokens": 669313161.0, + "step": 25869 + }, + { + "epoch": 2.8409839666154184, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0421910285949707, + "learning_rate": 1e-06, + "loss": 1.0618, + "mean_token_accuracy": 0.6945362091064453, + "num_tokens": 669334300.0, + "step": 25870 + }, + { + "epoch": 2.841093784318032, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5573487281799316, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7070901393890381, + "num_tokens": 669362793.0, + "step": 25871 + }, + { + "epoch": 2.841203602020646, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.445803165435791, + "learning_rate": 1e-06, + "loss": 1.0764, + "mean_token_accuracy": 0.7077729105949402, + "num_tokens": 669391541.0, + "step": 25872 + }, + { + "epoch": 2.8413134197232592, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4619266986846924, + "learning_rate": 1e-06, + "loss": 0.9694, + "mean_token_accuracy": 0.7180957198143005, + "num_tokens": 669418882.0, + "step": 25873 + }, + { + "epoch": 2.841423237425873, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.698460340499878, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.6972484588623047, + "num_tokens": 669441562.0, + "step": 25874 + }, + { + "epoch": 2.8415330551284868, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.571014404296875, + "learning_rate": 1e-06, + "loss": 1.1139, + "mean_token_accuracy": 0.6819307208061218, + "num_tokens": 669468471.0, + "step": 25875 + }, + { + "epoch": 2.8416428728311005, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6933817863464355, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6965123414993286, + "num_tokens": 669493183.0, + "step": 25876 + }, + { + "epoch": 2.8417526905337143, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.762763500213623, + "learning_rate": 1e-06, + "loss": 1.0362, + "mean_token_accuracy": 0.7009739279747009, + "num_tokens": 669516414.0, + "step": 25877 + }, + { + "epoch": 2.8418625082363276, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.539036750793457, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.690824031829834, + "num_tokens": 669542733.0, + "step": 25878 + }, + { + "epoch": 2.8419723259389413, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.675968647003174, + "learning_rate": 1e-06, + "loss": 0.9828, + "mean_token_accuracy": 0.71977299451828, + "num_tokens": 669566008.0, + "step": 25879 + }, + { + "epoch": 2.842082143641555, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.547865390777588, + "learning_rate": 1e-06, + "loss": 0.9899, + "mean_token_accuracy": 0.7113392949104309, + "num_tokens": 669593634.0, + "step": 25880 + }, + { + "epoch": 2.8421919613441684, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4395811557769775, + "learning_rate": 1e-06, + "loss": 0.9711, + "mean_token_accuracy": 0.7157027721405029, + "num_tokens": 669621603.0, + "step": 25881 + }, + { + "epoch": 2.8423017790467826, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7567126750946045, + "learning_rate": 1e-06, + "loss": 0.9959, + "mean_token_accuracy": 0.7195572853088379, + "num_tokens": 669644770.0, + "step": 25882 + }, + { + "epoch": 2.842411596749396, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.60327410697937, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.710405707359314, + "num_tokens": 669669975.0, + "step": 25883 + }, + { + "epoch": 2.8425214144520097, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.663810968399048, + "learning_rate": 1e-06, + "loss": 1.0872, + "mean_token_accuracy": 0.6822121143341064, + "num_tokens": 669693620.0, + "step": 25884 + }, + { + "epoch": 2.8426312321546234, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5001001358032227, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7040748000144958, + "num_tokens": 669720800.0, + "step": 25885 + }, + { + "epoch": 2.8427410498572367, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.71858549118042, + "learning_rate": 1e-06, + "loss": 0.939, + "mean_token_accuracy": 0.7289127111434937, + "num_tokens": 669743147.0, + "step": 25886 + }, + { + "epoch": 2.8428508675598505, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6510329246520996, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6990470290184021, + "num_tokens": 669773579.0, + "step": 25887 + }, + { + "epoch": 2.8429606852624643, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5852904319763184, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.7015210390090942, + "num_tokens": 669801002.0, + "step": 25888 + }, + { + "epoch": 2.843070502965078, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.643996238708496, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.727278470993042, + "num_tokens": 669823367.0, + "step": 25889 + }, + { + "epoch": 2.8431803206676918, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.526606559753418, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7102248072624207, + "num_tokens": 669853913.0, + "step": 25890 + }, + { + "epoch": 2.843290138370305, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.365149974822998, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.694543719291687, + "num_tokens": 669885033.0, + "step": 25891 + }, + { + "epoch": 2.843399956072919, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4369046688079834, + "learning_rate": 1e-06, + "loss": 1.052, + "mean_token_accuracy": 0.7039506435394287, + "num_tokens": 669915585.0, + "step": 25892 + }, + { + "epoch": 2.8435097737755326, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.701481342315674, + "learning_rate": 1e-06, + "loss": 1.0298, + "mean_token_accuracy": 0.7012158632278442, + "num_tokens": 669940756.0, + "step": 25893 + }, + { + "epoch": 2.8436195914781464, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.745605707168579, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7179482579231262, + "num_tokens": 669961158.0, + "step": 25894 + }, + { + "epoch": 2.84372940918076, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.795769214630127, + "learning_rate": 1e-06, + "loss": 0.9915, + "mean_token_accuracy": 0.7107542157173157, + "num_tokens": 669984561.0, + "step": 25895 + }, + { + "epoch": 2.8438392268833734, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5560145378112793, + "learning_rate": 1e-06, + "loss": 1.0115, + "mean_token_accuracy": 0.7044437527656555, + "num_tokens": 670008493.0, + "step": 25896 + }, + { + "epoch": 2.843949044585987, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5051937103271484, + "learning_rate": 1e-06, + "loss": 1.0588, + "mean_token_accuracy": 0.6951718330383301, + "num_tokens": 670033800.0, + "step": 25897 + }, + { + "epoch": 2.844058862288601, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.541154623031616, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.7061099410057068, + "num_tokens": 670060141.0, + "step": 25898 + }, + { + "epoch": 2.8441686799912147, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5848329067230225, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.7168340682983398, + "num_tokens": 670087456.0, + "step": 25899 + }, + { + "epoch": 2.8442784976938285, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3890814781188965, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7074731588363647, + "num_tokens": 670118073.0, + "step": 25900 + }, + { + "epoch": 2.8443883153964418, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.54429292678833, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7120436429977417, + "num_tokens": 670142476.0, + "step": 25901 + }, + { + "epoch": 2.8444981330990555, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5442404747009277, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7069709300994873, + "num_tokens": 670168641.0, + "step": 25902 + }, + { + "epoch": 2.8446079508016693, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.378274917602539, + "learning_rate": 1e-06, + "loss": 1.0697, + "mean_token_accuracy": 0.7013311386108398, + "num_tokens": 670201852.0, + "step": 25903 + }, + { + "epoch": 2.844717768504283, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7276346683502197, + "learning_rate": 1e-06, + "loss": 0.9382, + "mean_token_accuracy": 0.7220016121864319, + "num_tokens": 670224372.0, + "step": 25904 + }, + { + "epoch": 2.844827586206897, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.539440155029297, + "learning_rate": 1e-06, + "loss": 1.0462, + "mean_token_accuracy": 0.7068358659744263, + "num_tokens": 670252213.0, + "step": 25905 + }, + { + "epoch": 2.84493740390951, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.576246738433838, + "learning_rate": 1e-06, + "loss": 1.0664, + "mean_token_accuracy": 0.6937448978424072, + "num_tokens": 670280173.0, + "step": 25906 + }, + { + "epoch": 2.845047221612124, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4901974201202393, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7266435027122498, + "num_tokens": 670310822.0, + "step": 25907 + }, + { + "epoch": 2.8451570393147376, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.321626663208008, + "learning_rate": 1e-06, + "loss": 0.9935, + "mean_token_accuracy": 0.7166093587875366, + "num_tokens": 670338498.0, + "step": 25908 + }, + { + "epoch": 2.845266857017351, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8567001819610596, + "learning_rate": 1e-06, + "loss": 0.9067, + "mean_token_accuracy": 0.7382228374481201, + "num_tokens": 670360003.0, + "step": 25909 + }, + { + "epoch": 2.8453766747199647, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3642358779907227, + "learning_rate": 1e-06, + "loss": 1.0333, + "mean_token_accuracy": 0.698512852191925, + "num_tokens": 670392819.0, + "step": 25910 + }, + { + "epoch": 2.8454864924225785, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.511892557144165, + "learning_rate": 1e-06, + "loss": 1.0354, + "mean_token_accuracy": 0.699028730392456, + "num_tokens": 670420464.0, + "step": 25911 + }, + { + "epoch": 2.845596310125192, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.366323947906494, + "learning_rate": 1e-06, + "loss": 1.0951, + "mean_token_accuracy": 0.6851195096969604, + "num_tokens": 670451696.0, + "step": 25912 + }, + { + "epoch": 2.845706127827806, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.508298635482788, + "learning_rate": 1e-06, + "loss": 1.0326, + "mean_token_accuracy": 0.7040501236915588, + "num_tokens": 670480806.0, + "step": 25913 + }, + { + "epoch": 2.8458159455304193, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0141470432281494, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7180517911911011, + "num_tokens": 670502025.0, + "step": 25914 + }, + { + "epoch": 2.845925763233033, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4220969676971436, + "learning_rate": 1e-06, + "loss": 1.0441, + "mean_token_accuracy": 0.705168604850769, + "num_tokens": 670533258.0, + "step": 25915 + }, + { + "epoch": 2.846035580935647, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.516954183578491, + "learning_rate": 1e-06, + "loss": 1.013, + "mean_token_accuracy": 0.7049879431724548, + "num_tokens": 670560018.0, + "step": 25916 + }, + { + "epoch": 2.8461453986382605, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5353500843048096, + "learning_rate": 1e-06, + "loss": 1.1266, + "mean_token_accuracy": 0.6786479949951172, + "num_tokens": 670590028.0, + "step": 25917 + }, + { + "epoch": 2.8462552163408743, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.679476022720337, + "learning_rate": 1e-06, + "loss": 0.9386, + "mean_token_accuracy": 0.7232319116592407, + "num_tokens": 670615101.0, + "step": 25918 + }, + { + "epoch": 2.8463650340434876, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.2756001949310303, + "learning_rate": 1e-06, + "loss": 1.0061, + "mean_token_accuracy": 0.7245634198188782, + "num_tokens": 670646292.0, + "step": 25919 + }, + { + "epoch": 2.8464748517461014, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.710555076599121, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7110523581504822, + "num_tokens": 670667824.0, + "step": 25920 + }, + { + "epoch": 2.846584669448715, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.827044725418091, + "learning_rate": 1e-06, + "loss": 1.028, + "mean_token_accuracy": 0.7077646255493164, + "num_tokens": 670691045.0, + "step": 25921 + }, + { + "epoch": 2.846694487151329, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3909246921539307, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.7046552300453186, + "num_tokens": 670722992.0, + "step": 25922 + }, + { + "epoch": 2.8468043048539426, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4200456142425537, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7035893201828003, + "num_tokens": 670751400.0, + "step": 25923 + }, + { + "epoch": 2.846914122556556, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.400902271270752, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6863478422164917, + "num_tokens": 670780341.0, + "step": 25924 + }, + { + "epoch": 2.8470239402591697, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.639580249786377, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.7024644017219543, + "num_tokens": 670809210.0, + "step": 25925 + }, + { + "epoch": 2.8471337579617835, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4203734397888184, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6998611092567444, + "num_tokens": 670837346.0, + "step": 25926 + }, + { + "epoch": 2.8472435756643972, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.366542339324951, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.724523663520813, + "num_tokens": 670866039.0, + "step": 25927 + }, + { + "epoch": 2.847353393367011, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7009551525115967, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7063577771186829, + "num_tokens": 670889048.0, + "step": 25928 + }, + { + "epoch": 2.8474632110696243, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3781607151031494, + "learning_rate": 1e-06, + "loss": 1.0877, + "mean_token_accuracy": 0.6791515946388245, + "num_tokens": 670918102.0, + "step": 25929 + }, + { + "epoch": 2.847573028772238, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.610196352005005, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.7103744745254517, + "num_tokens": 670944479.0, + "step": 25930 + }, + { + "epoch": 2.847682846474852, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.528661012649536, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.7148134708404541, + "num_tokens": 670972970.0, + "step": 25931 + }, + { + "epoch": 2.8477926641774656, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7893662452697754, + "learning_rate": 1e-06, + "loss": 1.0218, + "mean_token_accuracy": 0.699350118637085, + "num_tokens": 670997133.0, + "step": 25932 + }, + { + "epoch": 2.8479024818800793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7803237438201904, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7004216909408569, + "num_tokens": 671023043.0, + "step": 25933 + }, + { + "epoch": 2.8480122995826926, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8285157680511475, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6968035697937012, + "num_tokens": 671046145.0, + "step": 25934 + }, + { + "epoch": 2.8481221172853064, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.629178285598755, + "learning_rate": 1e-06, + "loss": 1.0689, + "mean_token_accuracy": 0.6848217248916626, + "num_tokens": 671073990.0, + "step": 25935 + }, + { + "epoch": 2.84823193498792, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.708624839782715, + "learning_rate": 1e-06, + "loss": 0.9289, + "mean_token_accuracy": 0.7239587306976318, + "num_tokens": 671096700.0, + "step": 25936 + }, + { + "epoch": 2.8483417526905335, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.639561414718628, + "learning_rate": 1e-06, + "loss": 1.0209, + "mean_token_accuracy": 0.6975242495536804, + "num_tokens": 671121925.0, + "step": 25937 + }, + { + "epoch": 2.8484515703931472, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4591801166534424, + "learning_rate": 1e-06, + "loss": 1.0014, + "mean_token_accuracy": 0.7072983980178833, + "num_tokens": 671149683.0, + "step": 25938 + }, + { + "epoch": 2.848561388095761, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6364874839782715, + "learning_rate": 1e-06, + "loss": 1.011, + "mean_token_accuracy": 0.7076224088668823, + "num_tokens": 671174455.0, + "step": 25939 + }, + { + "epoch": 2.8486712057983747, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6063382625579834, + "learning_rate": 1e-06, + "loss": 0.9037, + "mean_token_accuracy": 0.7355508208274841, + "num_tokens": 671197544.0, + "step": 25940 + }, + { + "epoch": 2.8487810235009885, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.529038190841675, + "learning_rate": 1e-06, + "loss": 1.041, + "mean_token_accuracy": 0.6917989253997803, + "num_tokens": 671227099.0, + "step": 25941 + }, + { + "epoch": 2.848890841203602, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.293881416320801, + "learning_rate": 1e-06, + "loss": 0.9237, + "mean_token_accuracy": 0.7301174402236938, + "num_tokens": 671247278.0, + "step": 25942 + }, + { + "epoch": 2.8490006589062156, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9812722206115723, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.6981725096702576, + "num_tokens": 671270021.0, + "step": 25943 + }, + { + "epoch": 2.8491104766088293, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5561232566833496, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7238601446151733, + "num_tokens": 671297070.0, + "step": 25944 + }, + { + "epoch": 2.849220294311443, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.9504292011260986, + "learning_rate": 1e-06, + "loss": 1.0364, + "mean_token_accuracy": 0.69500732421875, + "num_tokens": 671327388.0, + "step": 25945 + }, + { + "epoch": 2.849330112014057, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8012161254882812, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7042667865753174, + "num_tokens": 671357515.0, + "step": 25946 + }, + { + "epoch": 2.84943992971667, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.490062713623047, + "learning_rate": 1e-06, + "loss": 0.9413, + "mean_token_accuracy": 0.7200120687484741, + "num_tokens": 671383917.0, + "step": 25947 + }, + { + "epoch": 2.849549747419284, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.060270309448242, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7120016813278198, + "num_tokens": 671405249.0, + "step": 25948 + }, + { + "epoch": 2.8496595651218977, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.553206205368042, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6914482116699219, + "num_tokens": 671434360.0, + "step": 25949 + }, + { + "epoch": 2.8497693828245114, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.482903480529785, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7056897282600403, + "num_tokens": 671463945.0, + "step": 25950 + }, + { + "epoch": 2.849879200527125, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.663214921951294, + "learning_rate": 1e-06, + "loss": 0.9471, + "mean_token_accuracy": 0.7244241237640381, + "num_tokens": 671488614.0, + "step": 25951 + }, + { + "epoch": 2.8499890182297385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5647127628326416, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.704785943031311, + "num_tokens": 671516910.0, + "step": 25952 + }, + { + "epoch": 2.8500988359323522, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.663663148880005, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7138782739639282, + "num_tokens": 671542002.0, + "step": 25953 + }, + { + "epoch": 2.850208653634966, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8828368186950684, + "learning_rate": 1e-06, + "loss": 1.0532, + "mean_token_accuracy": 0.7042235136032104, + "num_tokens": 671564639.0, + "step": 25954 + }, + { + "epoch": 2.8503184713375798, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.669671058654785, + "learning_rate": 1e-06, + "loss": 1.0487, + "mean_token_accuracy": 0.6997530460357666, + "num_tokens": 671591875.0, + "step": 25955 + }, + { + "epoch": 2.8504282890401935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.041442632675171, + "learning_rate": 1e-06, + "loss": 1.0396, + "mean_token_accuracy": 0.6942800283432007, + "num_tokens": 671618230.0, + "step": 25956 + }, + { + "epoch": 2.850538106742807, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.319369316101074, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7090946435928345, + "num_tokens": 671650632.0, + "step": 25957 + }, + { + "epoch": 2.8506479244454206, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7959461212158203, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.6977190375328064, + "num_tokens": 671674132.0, + "step": 25958 + }, + { + "epoch": 2.8507577421480343, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6286494731903076, + "learning_rate": 1e-06, + "loss": 1.0636, + "mean_token_accuracy": 0.6955308318138123, + "num_tokens": 671701303.0, + "step": 25959 + }, + { + "epoch": 2.8508675598506477, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.613767147064209, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6942753195762634, + "num_tokens": 671731503.0, + "step": 25960 + }, + { + "epoch": 2.850977377553262, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5456128120422363, + "learning_rate": 1e-06, + "loss": 0.9196, + "mean_token_accuracy": 0.7313582301139832, + "num_tokens": 671758936.0, + "step": 25961 + }, + { + "epoch": 2.851087195255875, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.470283031463623, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.6978940963745117, + "num_tokens": 671789371.0, + "step": 25962 + }, + { + "epoch": 2.851197012958489, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.609262704849243, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7161707878112793, + "num_tokens": 671815168.0, + "step": 25963 + }, + { + "epoch": 2.8513068306611027, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8257317543029785, + "learning_rate": 1e-06, + "loss": 0.8896, + "mean_token_accuracy": 0.7369022369384766, + "num_tokens": 671835725.0, + "step": 25964 + }, + { + "epoch": 2.851416648363716, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4997386932373047, + "learning_rate": 1e-06, + "loss": 1.073, + "mean_token_accuracy": 0.6907103657722473, + "num_tokens": 671864361.0, + "step": 25965 + }, + { + "epoch": 2.8515264660663298, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.499176263809204, + "learning_rate": 1e-06, + "loss": 1.0865, + "mean_token_accuracy": 0.6953549385070801, + "num_tokens": 671893204.0, + "step": 25966 + }, + { + "epoch": 2.8516362837689435, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8231120109558105, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7095531225204468, + "num_tokens": 671917250.0, + "step": 25967 + }, + { + "epoch": 2.8517461014715573, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4860098361968994, + "learning_rate": 1e-06, + "loss": 1.0491, + "mean_token_accuracy": 0.6948172450065613, + "num_tokens": 671948757.0, + "step": 25968 + }, + { + "epoch": 2.851855919174171, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8659420013427734, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7056999206542969, + "num_tokens": 671971315.0, + "step": 25969 + }, + { + "epoch": 2.8519657368767843, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7757437229156494, + "learning_rate": 1e-06, + "loss": 1.1252, + "mean_token_accuracy": 0.6804901361465454, + "num_tokens": 672001667.0, + "step": 25970 + }, + { + "epoch": 2.852075554579398, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9060118198394775, + "learning_rate": 1e-06, + "loss": 0.9481, + "mean_token_accuracy": 0.7277852892875671, + "num_tokens": 672021785.0, + "step": 25971 + }, + { + "epoch": 2.852185372282012, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7714426517486572, + "learning_rate": 1e-06, + "loss": 0.9996, + "mean_token_accuracy": 0.7102811336517334, + "num_tokens": 672047617.0, + "step": 25972 + }, + { + "epoch": 2.8522951899846256, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8248729705810547, + "learning_rate": 1e-06, + "loss": 1.0244, + "mean_token_accuracy": 0.7034536004066467, + "num_tokens": 672068906.0, + "step": 25973 + }, + { + "epoch": 2.8524050076872394, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.481189012527466, + "learning_rate": 1e-06, + "loss": 1.0368, + "mean_token_accuracy": 0.6995131969451904, + "num_tokens": 672096120.0, + "step": 25974 + }, + { + "epoch": 2.8525148253898527, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5040123462677, + "learning_rate": 1e-06, + "loss": 1.0257, + "mean_token_accuracy": 0.718791127204895, + "num_tokens": 672123519.0, + "step": 25975 + }, + { + "epoch": 2.8526246430924664, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.842165231704712, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7050048112869263, + "num_tokens": 672148463.0, + "step": 25976 + }, + { + "epoch": 2.85273446079508, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.519196033477783, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.7435632944107056, + "num_tokens": 672175014.0, + "step": 25977 + }, + { + "epoch": 2.852844278497694, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8022093772888184, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7077789306640625, + "num_tokens": 672196646.0, + "step": 25978 + }, + { + "epoch": 2.8529540962003077, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0645244121551514, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7085732221603394, + "num_tokens": 672223961.0, + "step": 25979 + }, + { + "epoch": 2.853063913902921, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4710636138916016, + "learning_rate": 1e-06, + "loss": 0.9909, + "mean_token_accuracy": 0.7133077383041382, + "num_tokens": 672252712.0, + "step": 25980 + }, + { + "epoch": 2.853173731605535, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7016525268554688, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.708938479423523, + "num_tokens": 672276394.0, + "step": 25981 + }, + { + "epoch": 2.8532835493081485, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6052324771881104, + "learning_rate": 1e-06, + "loss": 0.9204, + "mean_token_accuracy": 0.7313562631607056, + "num_tokens": 672300503.0, + "step": 25982 + }, + { + "epoch": 2.8533933670107623, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9609639644622803, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.7151857614517212, + "num_tokens": 672322330.0, + "step": 25983 + }, + { + "epoch": 2.853503184713376, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.800508737564087, + "learning_rate": 1e-06, + "loss": 0.9971, + "mean_token_accuracy": 0.7098509073257446, + "num_tokens": 672345273.0, + "step": 25984 + }, + { + "epoch": 2.8536130024159894, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.72383713722229, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.7002589702606201, + "num_tokens": 672372616.0, + "step": 25985 + }, + { + "epoch": 2.853722820118603, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7496883869171143, + "learning_rate": 1e-06, + "loss": 0.8388, + "mean_token_accuracy": 0.751352071762085, + "num_tokens": 672394953.0, + "step": 25986 + }, + { + "epoch": 2.853832637821217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5417189598083496, + "learning_rate": 1e-06, + "loss": 1.062, + "mean_token_accuracy": 0.6994578838348389, + "num_tokens": 672423046.0, + "step": 25987 + }, + { + "epoch": 2.85394245552383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 8.095911026000977, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7170895338058472, + "num_tokens": 672443008.0, + "step": 25988 + }, + { + "epoch": 2.854052273226444, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.739389181137085, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7097815275192261, + "num_tokens": 672467994.0, + "step": 25989 + }, + { + "epoch": 2.8541620909290577, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4207420349121094, + "learning_rate": 1e-06, + "loss": 1.0874, + "mean_token_accuracy": 0.685606837272644, + "num_tokens": 672497637.0, + "step": 25990 + }, + { + "epoch": 2.8542719086316715, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5705528259277344, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.7004508972167969, + "num_tokens": 672522773.0, + "step": 25991 + }, + { + "epoch": 2.854381726334285, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5733115673065186, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7072702050209045, + "num_tokens": 672550357.0, + "step": 25992 + }, + { + "epoch": 2.8544915440368985, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5746707916259766, + "learning_rate": 1e-06, + "loss": 0.9061, + "mean_token_accuracy": 0.7360143065452576, + "num_tokens": 672572533.0, + "step": 25993 + }, + { + "epoch": 2.8546013617395123, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3776016235351562, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.710065484046936, + "num_tokens": 672599089.0, + "step": 25994 + }, + { + "epoch": 2.854711179442126, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4616894721984863, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.7194402813911438, + "num_tokens": 672626113.0, + "step": 25995 + }, + { + "epoch": 2.85482099714474, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5581154823303223, + "learning_rate": 1e-06, + "loss": 1.036, + "mean_token_accuracy": 0.7036807537078857, + "num_tokens": 672651555.0, + "step": 25996 + }, + { + "epoch": 2.8549308148473536, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.820028781890869, + "learning_rate": 1e-06, + "loss": 1.0402, + "mean_token_accuracy": 0.7002724409103394, + "num_tokens": 672674340.0, + "step": 25997 + }, + { + "epoch": 2.855040632549967, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2722396850585938, + "learning_rate": 1e-06, + "loss": 0.9248, + "mean_token_accuracy": 0.7297372817993164, + "num_tokens": 672704680.0, + "step": 25998 + }, + { + "epoch": 2.8551504502525806, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.521043539047241, + "learning_rate": 1e-06, + "loss": 1.0688, + "mean_token_accuracy": 0.6896106600761414, + "num_tokens": 672731569.0, + "step": 25999 + }, + { + "epoch": 2.8552602679551944, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 7.881146430969238, + "learning_rate": 1e-06, + "loss": 1.0947, + "mean_token_accuracy": 0.687356173992157, + "num_tokens": 672763196.0, + "step": 26000 + }, + { + "epoch": 2.855370085657808, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.071059226989746, + "learning_rate": 1e-06, + "loss": 0.8819, + "mean_token_accuracy": 0.7405069470405579, + "num_tokens": 672784965.0, + "step": 26001 + }, + { + "epoch": 2.855479903360422, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.480957508087158, + "learning_rate": 1e-06, + "loss": 1.0957, + "mean_token_accuracy": 0.6839119791984558, + "num_tokens": 672816672.0, + "step": 26002 + }, + { + "epoch": 2.855589721063035, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5976130962371826, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.7290138006210327, + "num_tokens": 672840746.0, + "step": 26003 + }, + { + "epoch": 2.855699538765649, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5836148262023926, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7223331928253174, + "num_tokens": 672866472.0, + "step": 26004 + }, + { + "epoch": 2.8558093564682627, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3712069988250732, + "learning_rate": 1e-06, + "loss": 1.0176, + "mean_token_accuracy": 0.7023303508758545, + "num_tokens": 672894541.0, + "step": 26005 + }, + { + "epoch": 2.8559191741708765, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6050524711608887, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.7137330770492554, + "num_tokens": 672919064.0, + "step": 26006 + }, + { + "epoch": 2.8560289918734902, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.505573034286499, + "learning_rate": 1e-06, + "loss": 1.0411, + "mean_token_accuracy": 0.7039060592651367, + "num_tokens": 672946396.0, + "step": 26007 + }, + { + "epoch": 2.8561388095761036, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.315091609954834, + "learning_rate": 1e-06, + "loss": 1.0144, + "mean_token_accuracy": 0.7028216123580933, + "num_tokens": 672975469.0, + "step": 26008 + }, + { + "epoch": 2.8562486272787173, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.320469617843628, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7315669059753418, + "num_tokens": 673005393.0, + "step": 26009 + }, + { + "epoch": 2.856358444981331, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9273743629455566, + "learning_rate": 1e-06, + "loss": 0.905, + "mean_token_accuracy": 0.7305420637130737, + "num_tokens": 673023441.0, + "step": 26010 + }, + { + "epoch": 2.8564682626839444, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.694993257522583, + "learning_rate": 1e-06, + "loss": 0.8924, + "mean_token_accuracy": 0.7345792055130005, + "num_tokens": 673049411.0, + "step": 26011 + }, + { + "epoch": 2.8565780803865586, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.918259382247925, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7116904854774475, + "num_tokens": 673071869.0, + "step": 26012 + }, + { + "epoch": 2.856687898089172, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6603147983551025, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6884164810180664, + "num_tokens": 673098439.0, + "step": 26013 + }, + { + "epoch": 2.8567977157917857, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.758460521697998, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.7097887992858887, + "num_tokens": 673120897.0, + "step": 26014 + }, + { + "epoch": 2.8569075334943994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.959144353866577, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7111524343490601, + "num_tokens": 673141537.0, + "step": 26015 + }, + { + "epoch": 2.8570173511970127, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4995346069335938, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.706221342086792, + "num_tokens": 673169449.0, + "step": 26016 + }, + { + "epoch": 2.8571271688996265, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4572150707244873, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7138292789459229, + "num_tokens": 673197705.0, + "step": 26017 + }, + { + "epoch": 2.8572369866022402, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.517573833465576, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6936035752296448, + "num_tokens": 673227243.0, + "step": 26018 + }, + { + "epoch": 2.857346804304854, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.649374485015869, + "learning_rate": 1e-06, + "loss": 1.0228, + "mean_token_accuracy": 0.6985350847244263, + "num_tokens": 673252030.0, + "step": 26019 + }, + { + "epoch": 2.8574566220074678, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.550475835800171, + "learning_rate": 1e-06, + "loss": 0.9719, + "mean_token_accuracy": 0.7162494659423828, + "num_tokens": 673280208.0, + "step": 26020 + }, + { + "epoch": 2.857566439710081, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.490756034851074, + "learning_rate": 1e-06, + "loss": 1.0419, + "mean_token_accuracy": 0.6964784860610962, + "num_tokens": 673311114.0, + "step": 26021 + }, + { + "epoch": 2.857676257412695, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.560589075088501, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7170611619949341, + "num_tokens": 673336523.0, + "step": 26022 + }, + { + "epoch": 2.8577860751153086, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6206753253936768, + "learning_rate": 1e-06, + "loss": 1.0677, + "mean_token_accuracy": 0.695442795753479, + "num_tokens": 673363109.0, + "step": 26023 + }, + { + "epoch": 2.8578958928179223, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.761624813079834, + "learning_rate": 1e-06, + "loss": 0.9636, + "mean_token_accuracy": 0.7277753353118896, + "num_tokens": 673388866.0, + "step": 26024 + }, + { + "epoch": 2.858005710520536, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6545398235321045, + "learning_rate": 1e-06, + "loss": 0.9973, + "mean_token_accuracy": 0.7064769268035889, + "num_tokens": 673413250.0, + "step": 26025 + }, + { + "epoch": 2.8581155282231494, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.826904773712158, + "learning_rate": 1e-06, + "loss": 0.9649, + "mean_token_accuracy": 0.7142277956008911, + "num_tokens": 673433658.0, + "step": 26026 + }, + { + "epoch": 2.858225345925763, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.596109390258789, + "learning_rate": 1e-06, + "loss": 0.9312, + "mean_token_accuracy": 0.7223158478736877, + "num_tokens": 673453861.0, + "step": 26027 + }, + { + "epoch": 2.858335163628377, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7135322093963623, + "learning_rate": 1e-06, + "loss": 1.1457, + "mean_token_accuracy": 0.6686117053031921, + "num_tokens": 673484844.0, + "step": 26028 + }, + { + "epoch": 2.8584449813309907, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.66900634765625, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7186751365661621, + "num_tokens": 673510058.0, + "step": 26029 + }, + { + "epoch": 2.8585547990336044, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5168540477752686, + "learning_rate": 1e-06, + "loss": 1.0084, + "mean_token_accuracy": 0.709558367729187, + "num_tokens": 673537680.0, + "step": 26030 + }, + { + "epoch": 2.8586646167362177, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5730836391448975, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.712956428527832, + "num_tokens": 673562118.0, + "step": 26031 + }, + { + "epoch": 2.8587744344388315, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.63107967376709, + "learning_rate": 1e-06, + "loss": 0.9186, + "mean_token_accuracy": 0.7269462943077087, + "num_tokens": 673587339.0, + "step": 26032 + }, + { + "epoch": 2.8588842521414453, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5626344680786133, + "learning_rate": 1e-06, + "loss": 0.9888, + "mean_token_accuracy": 0.710155725479126, + "num_tokens": 673612062.0, + "step": 26033 + }, + { + "epoch": 2.858994069844059, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.472874402999878, + "learning_rate": 1e-06, + "loss": 0.9702, + "mean_token_accuracy": 0.7154535055160522, + "num_tokens": 673639397.0, + "step": 26034 + }, + { + "epoch": 2.8591038875466728, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.1949591636657715, + "learning_rate": 1e-06, + "loss": 0.9642, + "mean_token_accuracy": 0.7187316417694092, + "num_tokens": 673672041.0, + "step": 26035 + }, + { + "epoch": 2.859213705249286, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8189640045166016, + "learning_rate": 1e-06, + "loss": 1.088, + "mean_token_accuracy": 0.7006044983863831, + "num_tokens": 673696084.0, + "step": 26036 + }, + { + "epoch": 2.8593235229519, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.488619804382324, + "learning_rate": 1e-06, + "loss": 1.0227, + "mean_token_accuracy": 0.7014150619506836, + "num_tokens": 673725454.0, + "step": 26037 + }, + { + "epoch": 2.8594333406545136, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8146634101867676, + "learning_rate": 1e-06, + "loss": 0.9759, + "mean_token_accuracy": 0.7128179669380188, + "num_tokens": 673746551.0, + "step": 26038 + }, + { + "epoch": 2.859543158357127, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7263338565826416, + "learning_rate": 1e-06, + "loss": 0.9723, + "mean_token_accuracy": 0.7191100120544434, + "num_tokens": 673772122.0, + "step": 26039 + }, + { + "epoch": 2.8596529760597407, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.520268440246582, + "learning_rate": 1e-06, + "loss": 1.0197, + "mean_token_accuracy": 0.708914577960968, + "num_tokens": 673797623.0, + "step": 26040 + }, + { + "epoch": 2.8597627937623544, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3558428287506104, + "learning_rate": 1e-06, + "loss": 1.0262, + "mean_token_accuracy": 0.7007328271865845, + "num_tokens": 673828902.0, + "step": 26041 + }, + { + "epoch": 2.859872611464968, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4162404537200928, + "learning_rate": 1e-06, + "loss": 1.015, + "mean_token_accuracy": 0.7061845064163208, + "num_tokens": 673859008.0, + "step": 26042 + }, + { + "epoch": 2.859982429167582, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7010886669158936, + "learning_rate": 1e-06, + "loss": 0.9346, + "mean_token_accuracy": 0.7216453552246094, + "num_tokens": 673881917.0, + "step": 26043 + }, + { + "epoch": 2.8600922468701953, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.417782783508301, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.713076651096344, + "num_tokens": 673912750.0, + "step": 26044 + }, + { + "epoch": 2.860202064572809, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7716684341430664, + "learning_rate": 1e-06, + "loss": 1.0254, + "mean_token_accuracy": 0.7084335088729858, + "num_tokens": 673934835.0, + "step": 26045 + }, + { + "epoch": 2.8603118822754228, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5809435844421387, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7054572701454163, + "num_tokens": 673963165.0, + "step": 26046 + }, + { + "epoch": 2.8604216999780365, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.811673879623413, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7193200588226318, + "num_tokens": 673984440.0, + "step": 26047 + }, + { + "epoch": 2.8605315176806503, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.482445240020752, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7214491367340088, + "num_tokens": 674009921.0, + "step": 26048 + }, + { + "epoch": 2.8606413353832636, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6812357902526855, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.6973322629928589, + "num_tokens": 674034168.0, + "step": 26049 + }, + { + "epoch": 2.8607511530858774, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.822035312652588, + "learning_rate": 1e-06, + "loss": 0.9356, + "mean_token_accuracy": 0.7253232002258301, + "num_tokens": 674054595.0, + "step": 26050 + }, + { + "epoch": 2.860860970788491, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.122131586074829, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7077903747558594, + "num_tokens": 674073090.0, + "step": 26051 + }, + { + "epoch": 2.860970788491105, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5118765830993652, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.7036271095275879, + "num_tokens": 674102260.0, + "step": 26052 + }, + { + "epoch": 2.8610806061937186, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.836332321166992, + "learning_rate": 1e-06, + "loss": 1.0283, + "mean_token_accuracy": 0.6985836625099182, + "num_tokens": 674123703.0, + "step": 26053 + }, + { + "epoch": 2.861190423896332, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5159175395965576, + "learning_rate": 1e-06, + "loss": 1.0696, + "mean_token_accuracy": 0.6933308839797974, + "num_tokens": 674151101.0, + "step": 26054 + }, + { + "epoch": 2.8613002415989457, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.691765785217285, + "learning_rate": 1e-06, + "loss": 1.0343, + "mean_token_accuracy": 0.6954768896102905, + "num_tokens": 674176981.0, + "step": 26055 + }, + { + "epoch": 2.8614100593015594, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6989927291870117, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6954089999198914, + "num_tokens": 674202164.0, + "step": 26056 + }, + { + "epoch": 2.861519877004173, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.756164312362671, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7187714576721191, + "num_tokens": 674227273.0, + "step": 26057 + }, + { + "epoch": 2.861629694706787, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6721038818359375, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7179683446884155, + "num_tokens": 674251889.0, + "step": 26058 + }, + { + "epoch": 2.8617395124094003, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505718946456909, + "learning_rate": 1e-06, + "loss": 1.0815, + "mean_token_accuracy": 0.6906534433364868, + "num_tokens": 674282868.0, + "step": 26059 + }, + { + "epoch": 2.861849330112014, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5383105278015137, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7039773464202881, + "num_tokens": 674310603.0, + "step": 26060 + }, + { + "epoch": 2.861959147814628, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.85939359664917, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7085457444190979, + "num_tokens": 674333678.0, + "step": 26061 + }, + { + "epoch": 2.862068965517241, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5447757244110107, + "learning_rate": 1e-06, + "loss": 0.9458, + "mean_token_accuracy": 0.7289880514144897, + "num_tokens": 674361458.0, + "step": 26062 + }, + { + "epoch": 2.8621787832198553, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7236733436584473, + "learning_rate": 1e-06, + "loss": 0.9515, + "mean_token_accuracy": 0.7274991273880005, + "num_tokens": 674384351.0, + "step": 26063 + }, + { + "epoch": 2.8622886009224686, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4542720317840576, + "learning_rate": 1e-06, + "loss": 1.0612, + "mean_token_accuracy": 0.6887230277061462, + "num_tokens": 674411895.0, + "step": 26064 + }, + { + "epoch": 2.8623984186250824, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.618884563446045, + "learning_rate": 1e-06, + "loss": 1.0919, + "mean_token_accuracy": 0.6805851459503174, + "num_tokens": 674441370.0, + "step": 26065 + }, + { + "epoch": 2.862508236327696, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.390648603439331, + "learning_rate": 1e-06, + "loss": 1.0646, + "mean_token_accuracy": 0.6881159543991089, + "num_tokens": 674473057.0, + "step": 26066 + }, + { + "epoch": 2.8626180540303094, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5129754543304443, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7152995467185974, + "num_tokens": 674500205.0, + "step": 26067 + }, + { + "epoch": 2.862727871732923, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.692821979522705, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7092283964157104, + "num_tokens": 674522500.0, + "step": 26068 + }, + { + "epoch": 2.862837689435537, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4589927196502686, + "learning_rate": 1e-06, + "loss": 1.048, + "mean_token_accuracy": 0.691896378993988, + "num_tokens": 674553971.0, + "step": 26069 + }, + { + "epoch": 2.8629475071381507, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.347003221511841, + "learning_rate": 1e-06, + "loss": 1.0514, + "mean_token_accuracy": 0.6917552947998047, + "num_tokens": 674583947.0, + "step": 26070 + }, + { + "epoch": 2.8630573248407645, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5849227905273438, + "learning_rate": 1e-06, + "loss": 1.0443, + "mean_token_accuracy": 0.7000510692596436, + "num_tokens": 674609286.0, + "step": 26071 + }, + { + "epoch": 2.863167142543378, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0653538703918457, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.711033284664154, + "num_tokens": 674627776.0, + "step": 26072 + }, + { + "epoch": 2.8632769602459915, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5255634784698486, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7138130068778992, + "num_tokens": 674652947.0, + "step": 26073 + }, + { + "epoch": 2.8633867779486053, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4539647102355957, + "learning_rate": 1e-06, + "loss": 1.051, + "mean_token_accuracy": 0.6953767538070679, + "num_tokens": 674680767.0, + "step": 26074 + }, + { + "epoch": 2.863496595651219, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5356554985046387, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.6960809230804443, + "num_tokens": 674706645.0, + "step": 26075 + }, + { + "epoch": 2.863606413353833, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4425406455993652, + "learning_rate": 1e-06, + "loss": 1.0273, + "mean_token_accuracy": 0.7054687142372131, + "num_tokens": 674735371.0, + "step": 26076 + }, + { + "epoch": 2.863716231056446, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5649309158325195, + "learning_rate": 1e-06, + "loss": 1.0692, + "mean_token_accuracy": 0.6956853866577148, + "num_tokens": 674761652.0, + "step": 26077 + }, + { + "epoch": 2.86382604875906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.405977964401245, + "learning_rate": 1e-06, + "loss": 0.9895, + "mean_token_accuracy": 0.7190202474594116, + "num_tokens": 674790968.0, + "step": 26078 + }, + { + "epoch": 2.8639358664616736, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.587738037109375, + "learning_rate": 1e-06, + "loss": 1.15, + "mean_token_accuracy": 0.6790460348129272, + "num_tokens": 674820329.0, + "step": 26079 + }, + { + "epoch": 2.8640456841642874, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5676116943359375, + "learning_rate": 1e-06, + "loss": 0.9863, + "mean_token_accuracy": 0.7191609144210815, + "num_tokens": 674843544.0, + "step": 26080 + }, + { + "epoch": 2.864155501866901, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7713160514831543, + "learning_rate": 1e-06, + "loss": 0.9266, + "mean_token_accuracy": 0.7322689294815063, + "num_tokens": 674866950.0, + "step": 26081 + }, + { + "epoch": 2.8642653195695145, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6755659580230713, + "learning_rate": 1e-06, + "loss": 1.1024, + "mean_token_accuracy": 0.6806780099868774, + "num_tokens": 674892458.0, + "step": 26082 + }, + { + "epoch": 2.8643751372721282, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 20.56781005859375, + "learning_rate": 1e-06, + "loss": 1.0154, + "mean_token_accuracy": 0.7094969749450684, + "num_tokens": 674912625.0, + "step": 26083 + }, + { + "epoch": 2.864484954974742, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6853067874908447, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7037535905838013, + "num_tokens": 674937917.0, + "step": 26084 + }, + { + "epoch": 2.8645947726773557, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.555379867553711, + "learning_rate": 1e-06, + "loss": 0.9284, + "mean_token_accuracy": 0.7261725664138794, + "num_tokens": 674963788.0, + "step": 26085 + }, + { + "epoch": 2.8647045903799695, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7704017162323, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.718919038772583, + "num_tokens": 674985862.0, + "step": 26086 + }, + { + "epoch": 2.864814408082583, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9346656799316406, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7075014114379883, + "num_tokens": 675006808.0, + "step": 26087 + }, + { + "epoch": 2.8649242257851966, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.532331943511963, + "learning_rate": 1e-06, + "loss": 0.9824, + "mean_token_accuracy": 0.7115505933761597, + "num_tokens": 675033623.0, + "step": 26088 + }, + { + "epoch": 2.8650340434878103, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7941834926605225, + "learning_rate": 1e-06, + "loss": 0.923, + "mean_token_accuracy": 0.7226711511611938, + "num_tokens": 675053083.0, + "step": 26089 + }, + { + "epoch": 2.8651438611904236, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.433683395385742, + "learning_rate": 1e-06, + "loss": 1.0374, + "mean_token_accuracy": 0.7015104293823242, + "num_tokens": 675082485.0, + "step": 26090 + }, + { + "epoch": 2.8652536788930374, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8071813583374023, + "learning_rate": 1e-06, + "loss": 1.0933, + "mean_token_accuracy": 0.6890416741371155, + "num_tokens": 675105349.0, + "step": 26091 + }, + { + "epoch": 2.865363496595651, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.920177698135376, + "learning_rate": 1e-06, + "loss": 1.0095, + "mean_token_accuracy": 0.7077733278274536, + "num_tokens": 675129468.0, + "step": 26092 + }, + { + "epoch": 2.865473314298265, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5320589542388916, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7077595591545105, + "num_tokens": 675158711.0, + "step": 26093 + }, + { + "epoch": 2.8655831320008787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4388668537139893, + "learning_rate": 1e-06, + "loss": 0.9175, + "mean_token_accuracy": 0.7353941798210144, + "num_tokens": 675186157.0, + "step": 26094 + }, + { + "epoch": 2.865692949703492, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.466644287109375, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7030248641967773, + "num_tokens": 675213161.0, + "step": 26095 + }, + { + "epoch": 2.8658027674061057, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.378190040588379, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.7021732330322266, + "num_tokens": 675243061.0, + "step": 26096 + }, + { + "epoch": 2.8659125851087195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6783087253570557, + "learning_rate": 1e-06, + "loss": 0.9799, + "mean_token_accuracy": 0.7221343517303467, + "num_tokens": 675267805.0, + "step": 26097 + }, + { + "epoch": 2.8660224028113332, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8469297885894775, + "learning_rate": 1e-06, + "loss": 0.9348, + "mean_token_accuracy": 0.7255914807319641, + "num_tokens": 675288863.0, + "step": 26098 + }, + { + "epoch": 2.866132220513947, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6854703426361084, + "learning_rate": 1e-06, + "loss": 0.8806, + "mean_token_accuracy": 0.7389373779296875, + "num_tokens": 675312823.0, + "step": 26099 + }, + { + "epoch": 2.8662420382165603, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3183605670928955, + "learning_rate": 1e-06, + "loss": 0.9992, + "mean_token_accuracy": 0.7063831090927124, + "num_tokens": 675343825.0, + "step": 26100 + }, + { + "epoch": 2.866351855919174, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.742302656173706, + "learning_rate": 1e-06, + "loss": 1.0202, + "mean_token_accuracy": 0.6995628476142883, + "num_tokens": 675364877.0, + "step": 26101 + }, + { + "epoch": 2.866461673621788, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5139288902282715, + "learning_rate": 1e-06, + "loss": 0.9634, + "mean_token_accuracy": 0.710085928440094, + "num_tokens": 675390248.0, + "step": 26102 + }, + { + "epoch": 2.8665714913244016, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.452907085418701, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7079422473907471, + "num_tokens": 675418776.0, + "step": 26103 + }, + { + "epoch": 2.8666813090270153, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2751991748809814, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.725613534450531, + "num_tokens": 675452449.0, + "step": 26104 + }, + { + "epoch": 2.8667911267296287, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.730045795440674, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7041386365890503, + "num_tokens": 675472964.0, + "step": 26105 + }, + { + "epoch": 2.8669009444322424, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5724592208862305, + "learning_rate": 1e-06, + "loss": 1.0136, + "mean_token_accuracy": 0.7063564658164978, + "num_tokens": 675499433.0, + "step": 26106 + }, + { + "epoch": 2.867010762134856, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3483285903930664, + "learning_rate": 1e-06, + "loss": 1.0931, + "mean_token_accuracy": 0.6891638040542603, + "num_tokens": 675531158.0, + "step": 26107 + }, + { + "epoch": 2.86712057983747, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7395410537719727, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7232604026794434, + "num_tokens": 675552905.0, + "step": 26108 + }, + { + "epoch": 2.8672303975400837, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.671912908554077, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7108182311058044, + "num_tokens": 675577409.0, + "step": 26109 + }, + { + "epoch": 2.867340215242697, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.611717462539673, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7242966890335083, + "num_tokens": 675600582.0, + "step": 26110 + }, + { + "epoch": 2.8674500329453108, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6300714015960693, + "learning_rate": 1e-06, + "loss": 0.9549, + "mean_token_accuracy": 0.7292259931564331, + "num_tokens": 675623921.0, + "step": 26111 + }, + { + "epoch": 2.8675598506479245, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9487273693084717, + "learning_rate": 1e-06, + "loss": 0.89, + "mean_token_accuracy": 0.7378448247909546, + "num_tokens": 675644555.0, + "step": 26112 + }, + { + "epoch": 2.8676696683505383, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8573052883148193, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7076478004455566, + "num_tokens": 675667029.0, + "step": 26113 + }, + { + "epoch": 2.867779486053152, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.475493907928467, + "learning_rate": 1e-06, + "loss": 1.108, + "mean_token_accuracy": 0.6882561445236206, + "num_tokens": 675702160.0, + "step": 26114 + }, + { + "epoch": 2.8678893037557653, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 7.984523296356201, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7267861366271973, + "num_tokens": 675723715.0, + "step": 26115 + }, + { + "epoch": 2.867999121458379, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.619417905807495, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7041946053504944, + "num_tokens": 675749148.0, + "step": 26116 + }, + { + "epoch": 2.868108939160993, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.556962013244629, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6974823474884033, + "num_tokens": 675776758.0, + "step": 26117 + }, + { + "epoch": 2.868218756863606, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7180111408233643, + "learning_rate": 1e-06, + "loss": 1.0222, + "mean_token_accuracy": 0.6999313831329346, + "num_tokens": 675800674.0, + "step": 26118 + }, + { + "epoch": 2.86832857456622, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6858110427856445, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6921623349189758, + "num_tokens": 675827345.0, + "step": 26119 + }, + { + "epoch": 2.8684383922688337, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.605260133743286, + "learning_rate": 1e-06, + "loss": 0.9993, + "mean_token_accuracy": 0.7085409760475159, + "num_tokens": 675861466.0, + "step": 26120 + }, + { + "epoch": 2.8685482099714474, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6847267150878906, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7158344984054565, + "num_tokens": 675884135.0, + "step": 26121 + }, + { + "epoch": 2.868658027674061, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6287059783935547, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7088642716407776, + "num_tokens": 675908730.0, + "step": 26122 + }, + { + "epoch": 2.8687678453766745, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.606595516204834, + "learning_rate": 1e-06, + "loss": 0.9113, + "mean_token_accuracy": 0.7297235727310181, + "num_tokens": 675932288.0, + "step": 26123 + }, + { + "epoch": 2.8688776630792883, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.797929048538208, + "learning_rate": 1e-06, + "loss": 0.9487, + "mean_token_accuracy": 0.7241211533546448, + "num_tokens": 675954517.0, + "step": 26124 + }, + { + "epoch": 2.868987480781902, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5018181800842285, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7135162949562073, + "num_tokens": 675980211.0, + "step": 26125 + }, + { + "epoch": 2.869097298484516, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5017478466033936, + "learning_rate": 1e-06, + "loss": 0.9361, + "mean_token_accuracy": 0.7282232642173767, + "num_tokens": 676005009.0, + "step": 26126 + }, + { + "epoch": 2.8692071161871295, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5808215141296387, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.7091109752655029, + "num_tokens": 676030630.0, + "step": 26127 + }, + { + "epoch": 2.869316933889743, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6078708171844482, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7213841676712036, + "num_tokens": 676055394.0, + "step": 26128 + }, + { + "epoch": 2.8694267515923566, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.628688097000122, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.719334602355957, + "num_tokens": 676081325.0, + "step": 26129 + }, + { + "epoch": 2.8695365692949704, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0440940856933594, + "learning_rate": 1e-06, + "loss": 1.0476, + "mean_token_accuracy": 0.6905003190040588, + "num_tokens": 676107605.0, + "step": 26130 + }, + { + "epoch": 2.869646386997584, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8131356239318848, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7100296020507812, + "num_tokens": 676130392.0, + "step": 26131 + }, + { + "epoch": 2.869756204700198, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.546314001083374, + "learning_rate": 1e-06, + "loss": 1.0104, + "mean_token_accuracy": 0.7032327055931091, + "num_tokens": 676155618.0, + "step": 26132 + }, + { + "epoch": 2.869866022402811, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5607433319091797, + "learning_rate": 1e-06, + "loss": 0.9809, + "mean_token_accuracy": 0.7118076086044312, + "num_tokens": 676180193.0, + "step": 26133 + }, + { + "epoch": 2.869975840105425, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.542682409286499, + "learning_rate": 1e-06, + "loss": 1.1088, + "mean_token_accuracy": 0.6785663366317749, + "num_tokens": 676207552.0, + "step": 26134 + }, + { + "epoch": 2.8700856578080387, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7611072063446045, + "learning_rate": 1e-06, + "loss": 0.9817, + "mean_token_accuracy": 0.7072421908378601, + "num_tokens": 676231269.0, + "step": 26135 + }, + { + "epoch": 2.8701954755106525, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8495872020721436, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7288041710853577, + "num_tokens": 676251030.0, + "step": 26136 + }, + { + "epoch": 2.870305293213266, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.576213836669922, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7264730334281921, + "num_tokens": 676277159.0, + "step": 26137 + }, + { + "epoch": 2.8704151109158795, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.618403911590576, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7104369401931763, + "num_tokens": 676302300.0, + "step": 26138 + }, + { + "epoch": 2.8705249286184933, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 8.045122146606445, + "learning_rate": 1e-06, + "loss": 0.9399, + "mean_token_accuracy": 0.7236638069152832, + "num_tokens": 676322622.0, + "step": 26139 + }, + { + "epoch": 2.870634746321107, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.75474214553833, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.6972384452819824, + "num_tokens": 676347935.0, + "step": 26140 + }, + { + "epoch": 2.8707445640237204, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5310404300689697, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6867588758468628, + "num_tokens": 676379047.0, + "step": 26141 + }, + { + "epoch": 2.8708543817263346, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5380032062530518, + "learning_rate": 1e-06, + "loss": 1.1428, + "mean_token_accuracy": 0.6693884134292603, + "num_tokens": 676408422.0, + "step": 26142 + }, + { + "epoch": 2.870964199428948, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.652543306350708, + "learning_rate": 1e-06, + "loss": 1.0196, + "mean_token_accuracy": 0.6994717121124268, + "num_tokens": 676435121.0, + "step": 26143 + }, + { + "epoch": 2.8710740171315616, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.595799207687378, + "learning_rate": 1e-06, + "loss": 1.0608, + "mean_token_accuracy": 0.6966245174407959, + "num_tokens": 676460750.0, + "step": 26144 + }, + { + "epoch": 2.8711838348341754, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7155065536499023, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7234532237052917, + "num_tokens": 676482900.0, + "step": 26145 + }, + { + "epoch": 2.8712936525367887, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4590303897857666, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7061643600463867, + "num_tokens": 676510402.0, + "step": 26146 + }, + { + "epoch": 2.8714034702394025, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3980093002319336, + "learning_rate": 1e-06, + "loss": 0.9795, + "mean_token_accuracy": 0.7170060873031616, + "num_tokens": 676541557.0, + "step": 26147 + }, + { + "epoch": 2.871513287942016, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5228638648986816, + "learning_rate": 1e-06, + "loss": 0.9199, + "mean_token_accuracy": 0.7274425625801086, + "num_tokens": 676567699.0, + "step": 26148 + }, + { + "epoch": 2.87162310564463, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6576197147369385, + "learning_rate": 1e-06, + "loss": 0.9639, + "mean_token_accuracy": 0.7130695581436157, + "num_tokens": 676590430.0, + "step": 26149 + }, + { + "epoch": 2.8717329233472437, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7306134700775146, + "learning_rate": 1e-06, + "loss": 1.0201, + "mean_token_accuracy": 0.7026109099388123, + "num_tokens": 676616324.0, + "step": 26150 + }, + { + "epoch": 2.871842741049857, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.369436740875244, + "learning_rate": 1e-06, + "loss": 0.9949, + "mean_token_accuracy": 0.7119792699813843, + "num_tokens": 676645348.0, + "step": 26151 + }, + { + "epoch": 2.871952558752471, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.052881956100464, + "learning_rate": 1e-06, + "loss": 0.9618, + "mean_token_accuracy": 0.7214671969413757, + "num_tokens": 676664263.0, + "step": 26152 + }, + { + "epoch": 2.8720623764550846, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.515331268310547, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.713408350944519, + "num_tokens": 676689852.0, + "step": 26153 + }, + { + "epoch": 2.8721721941576983, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6341562271118164, + "learning_rate": 1e-06, + "loss": 1.0833, + "mean_token_accuracy": 0.6882159113883972, + "num_tokens": 676716524.0, + "step": 26154 + }, + { + "epoch": 2.872282011860312, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5282764434814453, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7181422710418701, + "num_tokens": 676741984.0, + "step": 26155 + }, + { + "epoch": 2.8723918295629254, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.399937868118286, + "learning_rate": 1e-06, + "loss": 1.0755, + "mean_token_accuracy": 0.6882603168487549, + "num_tokens": 676774168.0, + "step": 26156 + }, + { + "epoch": 2.872501647265539, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4444804191589355, + "learning_rate": 1e-06, + "loss": 1.0805, + "mean_token_accuracy": 0.6959488391876221, + "num_tokens": 676806321.0, + "step": 26157 + }, + { + "epoch": 2.872611464968153, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5928258895874023, + "learning_rate": 1e-06, + "loss": 0.9552, + "mean_token_accuracy": 0.7210568189620972, + "num_tokens": 676830146.0, + "step": 26158 + }, + { + "epoch": 2.8727212826707667, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.43415904045105, + "learning_rate": 1e-06, + "loss": 0.943, + "mean_token_accuracy": 0.7238463163375854, + "num_tokens": 676856918.0, + "step": 26159 + }, + { + "epoch": 2.8728311003733804, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7323391437530518, + "learning_rate": 1e-06, + "loss": 0.955, + "mean_token_accuracy": 0.7172352075576782, + "num_tokens": 676877334.0, + "step": 26160 + }, + { + "epoch": 2.8729409180759937, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6300532817840576, + "learning_rate": 1e-06, + "loss": 0.9589, + "mean_token_accuracy": 0.7073565721511841, + "num_tokens": 676902466.0, + "step": 26161 + }, + { + "epoch": 2.8730507357786075, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8282413482666016, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7138974666595459, + "num_tokens": 676923813.0, + "step": 26162 + }, + { + "epoch": 2.8731605534812212, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7999629974365234, + "learning_rate": 1e-06, + "loss": 1.0042, + "mean_token_accuracy": 0.7123794555664062, + "num_tokens": 676946898.0, + "step": 26163 + }, + { + "epoch": 2.873270371183835, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.981069803237915, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7047762274742126, + "num_tokens": 676972284.0, + "step": 26164 + }, + { + "epoch": 2.8733801888864487, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6233441829681396, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.706775426864624, + "num_tokens": 676997213.0, + "step": 26165 + }, + { + "epoch": 2.873490006589062, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7313599586486816, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7019592523574829, + "num_tokens": 677020832.0, + "step": 26166 + }, + { + "epoch": 2.873599824291676, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.805612087249756, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6966869831085205, + "num_tokens": 677044474.0, + "step": 26167 + }, + { + "epoch": 2.8737096419942896, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5205299854278564, + "learning_rate": 1e-06, + "loss": 0.9462, + "mean_token_accuracy": 0.7179996967315674, + "num_tokens": 677071226.0, + "step": 26168 + }, + { + "epoch": 2.873819459696903, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7962214946746826, + "learning_rate": 1e-06, + "loss": 0.899, + "mean_token_accuracy": 0.7329698801040649, + "num_tokens": 677094413.0, + "step": 26169 + }, + { + "epoch": 2.8739292773995166, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.434624195098877, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6953390836715698, + "num_tokens": 677128378.0, + "step": 26170 + }, + { + "epoch": 2.8740390951021304, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7885994911193848, + "learning_rate": 1e-06, + "loss": 1.0329, + "mean_token_accuracy": 0.6994806528091431, + "num_tokens": 677151911.0, + "step": 26171 + }, + { + "epoch": 2.874148912804744, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4093973636627197, + "learning_rate": 1e-06, + "loss": 0.9876, + "mean_token_accuracy": 0.7198590040206909, + "num_tokens": 677180932.0, + "step": 26172 + }, + { + "epoch": 2.874258730507358, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6058802604675293, + "learning_rate": 1e-06, + "loss": 0.9113, + "mean_token_accuracy": 0.7276901602745056, + "num_tokens": 677204864.0, + "step": 26173 + }, + { + "epoch": 2.8743685482099712, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 2.5185329914093018, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7038223147392273, + "num_tokens": 677229322.0, + "step": 26174 + }, + { + "epoch": 2.874478365912585, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.944432020187378, + "learning_rate": 1e-06, + "loss": 1.0119, + "mean_token_accuracy": 0.7089414000511169, + "num_tokens": 677251165.0, + "step": 26175 + }, + { + "epoch": 2.8745881836151987, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7049083709716797, + "learning_rate": 1e-06, + "loss": 1.0347, + "mean_token_accuracy": 0.7078245878219604, + "num_tokens": 677276506.0, + "step": 26176 + }, + { + "epoch": 2.8746980013178125, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6455657482147217, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6960598230361938, + "num_tokens": 677303683.0, + "step": 26177 + }, + { + "epoch": 2.8748078190204263, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5069239139556885, + "learning_rate": 1e-06, + "loss": 0.9561, + "mean_token_accuracy": 0.7250733375549316, + "num_tokens": 677329909.0, + "step": 26178 + }, + { + "epoch": 2.8749176367230396, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.773200035095215, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7052347660064697, + "num_tokens": 677352625.0, + "step": 26179 + }, + { + "epoch": 2.8750274544256533, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5318241119384766, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7143481969833374, + "num_tokens": 677377575.0, + "step": 26180 + }, + { + "epoch": 2.875137272128267, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.2875759601593018, + "learning_rate": 1e-06, + "loss": 1.0243, + "mean_token_accuracy": 0.6977924704551697, + "num_tokens": 677413143.0, + "step": 26181 + }, + { + "epoch": 2.875247089830881, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6823909282684326, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7053179740905762, + "num_tokens": 677437420.0, + "step": 26182 + }, + { + "epoch": 2.8753569075334946, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7027251720428467, + "learning_rate": 1e-06, + "loss": 1.123, + "mean_token_accuracy": 0.6806782484054565, + "num_tokens": 677460817.0, + "step": 26183 + }, + { + "epoch": 2.875466725236108, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.211425542831421, + "learning_rate": 1e-06, + "loss": 1.136, + "mean_token_accuracy": 0.6757145524024963, + "num_tokens": 677490990.0, + "step": 26184 + }, + { + "epoch": 2.8755765429387217, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5318515300750732, + "learning_rate": 1e-06, + "loss": 1.1176, + "mean_token_accuracy": 0.6775949597358704, + "num_tokens": 677519083.0, + "step": 26185 + }, + { + "epoch": 2.8756863606413354, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.824199914932251, + "learning_rate": 1e-06, + "loss": 1.0663, + "mean_token_accuracy": 0.6985855102539062, + "num_tokens": 677541676.0, + "step": 26186 + }, + { + "epoch": 2.875796178343949, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.2863099575042725, + "learning_rate": 1e-06, + "loss": 1.0972, + "mean_token_accuracy": 0.6857648491859436, + "num_tokens": 677574898.0, + "step": 26187 + }, + { + "epoch": 2.875905996046563, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7597672939300537, + "learning_rate": 1e-06, + "loss": 1.0236, + "mean_token_accuracy": 0.7035884857177734, + "num_tokens": 677597511.0, + "step": 26188 + }, + { + "epoch": 2.8760158137491763, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6933817863464355, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7145248651504517, + "num_tokens": 677621779.0, + "step": 26189 + }, + { + "epoch": 2.87612563145179, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.831855058670044, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.6972131729125977, + "num_tokens": 677645561.0, + "step": 26190 + }, + { + "epoch": 2.8762354491544038, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6396424770355225, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7153448462486267, + "num_tokens": 677669700.0, + "step": 26191 + }, + { + "epoch": 2.876345266857017, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.541247844696045, + "learning_rate": 1e-06, + "loss": 1.0539, + "mean_token_accuracy": 0.6888704299926758, + "num_tokens": 677698827.0, + "step": 26192 + }, + { + "epoch": 2.8764550845596313, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4688782691955566, + "learning_rate": 1e-06, + "loss": 1.0067, + "mean_token_accuracy": 0.7047235369682312, + "num_tokens": 677729257.0, + "step": 26193 + }, + { + "epoch": 2.8765649022622446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5553297996520996, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6944464445114136, + "num_tokens": 677755695.0, + "step": 26194 + }, + { + "epoch": 2.8766747199648584, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5578627586364746, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.7016650438308716, + "num_tokens": 677783088.0, + "step": 26195 + }, + { + "epoch": 2.876784537667472, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3998937606811523, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7068381905555725, + "num_tokens": 677814594.0, + "step": 26196 + }, + { + "epoch": 2.8768943553700854, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7035093307495117, + "learning_rate": 1e-06, + "loss": 0.9214, + "mean_token_accuracy": 0.7286341190338135, + "num_tokens": 677836354.0, + "step": 26197 + }, + { + "epoch": 2.877004173072699, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5574615001678467, + "learning_rate": 1e-06, + "loss": 0.9492, + "mean_token_accuracy": 0.7200610637664795, + "num_tokens": 677865061.0, + "step": 26198 + }, + { + "epoch": 2.877113990775313, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6426684856414795, + "learning_rate": 1e-06, + "loss": 1.1228, + "mean_token_accuracy": 0.6817433834075928, + "num_tokens": 677893485.0, + "step": 26199 + }, + { + "epoch": 2.8772238084779267, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6272850036621094, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7041945457458496, + "num_tokens": 677917057.0, + "step": 26200 + }, + { + "epoch": 2.8773336261805404, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4265925884246826, + "learning_rate": 1e-06, + "loss": 0.9596, + "mean_token_accuracy": 0.7232726216316223, + "num_tokens": 677946716.0, + "step": 26201 + }, + { + "epoch": 2.8774434438831538, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5910065174102783, + "learning_rate": 1e-06, + "loss": 1.0798, + "mean_token_accuracy": 0.685688853263855, + "num_tokens": 677974654.0, + "step": 26202 + }, + { + "epoch": 2.8775532615857675, + "ewc_loss": 1.0728836059570312e-05, + "grad_norm": 5.07299280166626, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6982448101043701, + "num_tokens": 678000009.0, + "step": 26203 + }, + { + "epoch": 2.8776630792883813, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4969165325164795, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7118273973464966, + "num_tokens": 678027105.0, + "step": 26204 + }, + { + "epoch": 2.877772896990995, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3113009929656982, + "learning_rate": 1e-06, + "loss": 1.0022, + "mean_token_accuracy": 0.7086408734321594, + "num_tokens": 678057760.0, + "step": 26205 + }, + { + "epoch": 2.877882714693609, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5723989009857178, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7084225416183472, + "num_tokens": 678084386.0, + "step": 26206 + }, + { + "epoch": 2.877992532396222, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5387237071990967, + "learning_rate": 1e-06, + "loss": 1.0215, + "mean_token_accuracy": 0.7016944885253906, + "num_tokens": 678110250.0, + "step": 26207 + }, + { + "epoch": 2.878102350098836, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.518465518951416, + "learning_rate": 1e-06, + "loss": 1.0866, + "mean_token_accuracy": 0.6881092190742493, + "num_tokens": 678138519.0, + "step": 26208 + }, + { + "epoch": 2.8782121678014496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.742666006088257, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.70319664478302, + "num_tokens": 678160459.0, + "step": 26209 + }, + { + "epoch": 2.8783219855040634, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.655677556991577, + "learning_rate": 1e-06, + "loss": 0.9523, + "mean_token_accuracy": 0.7164022326469421, + "num_tokens": 678183192.0, + "step": 26210 + }, + { + "epoch": 2.878431803206677, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.125958204269409, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7116451263427734, + "num_tokens": 678212127.0, + "step": 26211 + }, + { + "epoch": 2.8785416209092904, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.45865535736084, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.714009702205658, + "num_tokens": 678240686.0, + "step": 26212 + }, + { + "epoch": 2.878651438611904, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5172269344329834, + "learning_rate": 1e-06, + "loss": 0.9196, + "mean_token_accuracy": 0.7369353771209717, + "num_tokens": 678270121.0, + "step": 26213 + }, + { + "epoch": 2.878761256314518, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5225329399108887, + "learning_rate": 1e-06, + "loss": 1.1363, + "mean_token_accuracy": 0.6787221431732178, + "num_tokens": 678300259.0, + "step": 26214 + }, + { + "epoch": 2.8788710740171317, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5654447078704834, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.6979498267173767, + "num_tokens": 678325595.0, + "step": 26215 + }, + { + "epoch": 2.8789808917197455, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4906203746795654, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7024239897727966, + "num_tokens": 678351866.0, + "step": 26216 + }, + { + "epoch": 2.879090709422359, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.549988031387329, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.6977459788322449, + "num_tokens": 678380016.0, + "step": 26217 + }, + { + "epoch": 2.8792005271249725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.584282875061035, + "learning_rate": 1e-06, + "loss": 0.9654, + "mean_token_accuracy": 0.7217919826507568, + "num_tokens": 678406271.0, + "step": 26218 + }, + { + "epoch": 2.8793103448275863, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.593207597732544, + "learning_rate": 1e-06, + "loss": 0.9805, + "mean_token_accuracy": 0.7180073261260986, + "num_tokens": 678433377.0, + "step": 26219 + }, + { + "epoch": 2.8794201625301996, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5840160846710205, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7141648530960083, + "num_tokens": 678457691.0, + "step": 26220 + }, + { + "epoch": 2.8795299802328134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.442566394805908, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6933261752128601, + "num_tokens": 678486304.0, + "step": 26221 + }, + { + "epoch": 2.879639797935427, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.645479440689087, + "learning_rate": 1e-06, + "loss": 1.0572, + "mean_token_accuracy": 0.6963245868682861, + "num_tokens": 678512215.0, + "step": 26222 + }, + { + "epoch": 2.879749615638041, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.431072235107422, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7104469537734985, + "num_tokens": 678537965.0, + "step": 26223 + }, + { + "epoch": 2.8798594333406546, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.410496711730957, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7159493565559387, + "num_tokens": 678567305.0, + "step": 26224 + }, + { + "epoch": 2.879969251043268, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7991034984588623, + "learning_rate": 1e-06, + "loss": 0.9793, + "mean_token_accuracy": 0.7129325866699219, + "num_tokens": 678588741.0, + "step": 26225 + }, + { + "epoch": 2.8800790687458817, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5963265895843506, + "learning_rate": 1e-06, + "loss": 1.046, + "mean_token_accuracy": 0.6994031667709351, + "num_tokens": 678614744.0, + "step": 26226 + }, + { + "epoch": 2.8801888864484955, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9805173873901367, + "learning_rate": 1e-06, + "loss": 0.8882, + "mean_token_accuracy": 0.7268844246864319, + "num_tokens": 678633366.0, + "step": 26227 + }, + { + "epoch": 2.8802987041511092, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8114540576934814, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7206258773803711, + "num_tokens": 678655254.0, + "step": 26228 + }, + { + "epoch": 2.880408521853723, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5965793132781982, + "learning_rate": 1e-06, + "loss": 0.9225, + "mean_token_accuracy": 0.7274394631385803, + "num_tokens": 678678121.0, + "step": 26229 + }, + { + "epoch": 2.8805183395563363, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3970894813537598, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7056692242622375, + "num_tokens": 678708706.0, + "step": 26230 + }, + { + "epoch": 2.88062815725895, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.509431838989258, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6951560974121094, + "num_tokens": 678736911.0, + "step": 26231 + }, + { + "epoch": 2.880737974961564, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6342740058898926, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7055401802062988, + "num_tokens": 678760187.0, + "step": 26232 + }, + { + "epoch": 2.8808477926641776, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7127397060394287, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.712902307510376, + "num_tokens": 678783427.0, + "step": 26233 + }, + { + "epoch": 2.8809576103667913, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.703024387359619, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.7123831510543823, + "num_tokens": 678806767.0, + "step": 26234 + }, + { + "epoch": 2.8810674280694046, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6810061931610107, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7033770084381104, + "num_tokens": 678831097.0, + "step": 26235 + }, + { + "epoch": 2.8811772457720184, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.448307991027832, + "learning_rate": 1e-06, + "loss": 1.0709, + "mean_token_accuracy": 0.6958308815956116, + "num_tokens": 678859565.0, + "step": 26236 + }, + { + "epoch": 2.881287063474632, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4740328788757324, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7106560468673706, + "num_tokens": 678887636.0, + "step": 26237 + }, + { + "epoch": 2.881396881177246, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5200343132019043, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7094149589538574, + "num_tokens": 678915873.0, + "step": 26238 + }, + { + "epoch": 2.8815066988798597, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.578042984008789, + "learning_rate": 1e-06, + "loss": 0.9926, + "mean_token_accuracy": 0.7072102427482605, + "num_tokens": 678941675.0, + "step": 26239 + }, + { + "epoch": 2.881616516582473, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7237327098846436, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7148294448852539, + "num_tokens": 678965197.0, + "step": 26240 + }, + { + "epoch": 2.8817263342850867, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7478277683258057, + "learning_rate": 1e-06, + "loss": 1.0783, + "mean_token_accuracy": 0.6855237483978271, + "num_tokens": 678990892.0, + "step": 26241 + }, + { + "epoch": 2.8818361519877005, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5961108207702637, + "learning_rate": 1e-06, + "loss": 0.9278, + "mean_token_accuracy": 0.7203280925750732, + "num_tokens": 679012837.0, + "step": 26242 + }, + { + "epoch": 2.881945969690314, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3612060546875, + "learning_rate": 1e-06, + "loss": 1.1169, + "mean_token_accuracy": 0.6861584186553955, + "num_tokens": 679045032.0, + "step": 26243 + }, + { + "epoch": 2.882055787392928, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.855473518371582, + "learning_rate": 1e-06, + "loss": 0.9663, + "mean_token_accuracy": 0.7129112482070923, + "num_tokens": 679064203.0, + "step": 26244 + }, + { + "epoch": 2.8821656050955413, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.359133005142212, + "learning_rate": 1e-06, + "loss": 1.0111, + "mean_token_accuracy": 0.704597532749176, + "num_tokens": 679091818.0, + "step": 26245 + }, + { + "epoch": 2.882275422798155, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5842669010162354, + "learning_rate": 1e-06, + "loss": 0.887, + "mean_token_accuracy": 0.7383267283439636, + "num_tokens": 679117480.0, + "step": 26246 + }, + { + "epoch": 2.882385240500769, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.39137864112854, + "learning_rate": 1e-06, + "loss": 1.0221, + "mean_token_accuracy": 0.7000758051872253, + "num_tokens": 679147118.0, + "step": 26247 + }, + { + "epoch": 2.882495058203382, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.693551540374756, + "learning_rate": 1e-06, + "loss": 1.054, + "mean_token_accuracy": 0.6958169937133789, + "num_tokens": 679171613.0, + "step": 26248 + }, + { + "epoch": 2.882604875905996, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.449300765991211, + "learning_rate": 1e-06, + "loss": 1.0378, + "mean_token_accuracy": 0.7079580426216125, + "num_tokens": 679200715.0, + "step": 26249 + }, + { + "epoch": 2.8827146936086097, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4605531692504883, + "learning_rate": 1e-06, + "loss": 1.019, + "mean_token_accuracy": 0.7168596386909485, + "num_tokens": 679229449.0, + "step": 26250 + }, + { + "epoch": 2.8828245113112234, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5172269344329834, + "learning_rate": 1e-06, + "loss": 1.1072, + "mean_token_accuracy": 0.6849128603935242, + "num_tokens": 679255059.0, + "step": 26251 + }, + { + "epoch": 2.882934329013837, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8689146041870117, + "learning_rate": 1e-06, + "loss": 0.9905, + "mean_token_accuracy": 0.7127945423126221, + "num_tokens": 679274261.0, + "step": 26252 + }, + { + "epoch": 2.8830441467164505, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3818931579589844, + "learning_rate": 1e-06, + "loss": 0.9714, + "mean_token_accuracy": 0.7336574196815491, + "num_tokens": 679302410.0, + "step": 26253 + }, + { + "epoch": 2.8831539644190642, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6865792274475098, + "learning_rate": 1e-06, + "loss": 0.9086, + "mean_token_accuracy": 0.7349820137023926, + "num_tokens": 679327577.0, + "step": 26254 + }, + { + "epoch": 2.883263782121678, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5393176078796387, + "learning_rate": 1e-06, + "loss": 0.9317, + "mean_token_accuracy": 0.7210342884063721, + "num_tokens": 679353635.0, + "step": 26255 + }, + { + "epoch": 2.8833735998242918, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6650586128234863, + "learning_rate": 1e-06, + "loss": 1.0812, + "mean_token_accuracy": 0.6981405019760132, + "num_tokens": 679382008.0, + "step": 26256 + }, + { + "epoch": 2.8834834175269055, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.424966335296631, + "learning_rate": 1e-06, + "loss": 1.0775, + "mean_token_accuracy": 0.6957824230194092, + "num_tokens": 679410828.0, + "step": 26257 + }, + { + "epoch": 2.883593235229519, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3783137798309326, + "learning_rate": 1e-06, + "loss": 1.0165, + "mean_token_accuracy": 0.7035305500030518, + "num_tokens": 679443533.0, + "step": 26258 + }, + { + "epoch": 2.8837030529321326, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8848531246185303, + "learning_rate": 1e-06, + "loss": 0.9451, + "mean_token_accuracy": 0.7178577184677124, + "num_tokens": 679463114.0, + "step": 26259 + }, + { + "epoch": 2.8838128706347463, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6055564880371094, + "learning_rate": 1e-06, + "loss": 0.8892, + "mean_token_accuracy": 0.7358834147453308, + "num_tokens": 679490246.0, + "step": 26260 + }, + { + "epoch": 2.88392268833736, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.471454381942749, + "learning_rate": 1e-06, + "loss": 0.9933, + "mean_token_accuracy": 0.7076976299285889, + "num_tokens": 679517195.0, + "step": 26261 + }, + { + "epoch": 2.884032506039974, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6911606788635254, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7192909717559814, + "num_tokens": 679540984.0, + "step": 26262 + }, + { + "epoch": 2.884142323742587, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.710345983505249, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.6989746689796448, + "num_tokens": 679564476.0, + "step": 26263 + }, + { + "epoch": 2.884252141445201, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5336616039276123, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7069947719573975, + "num_tokens": 679591224.0, + "step": 26264 + }, + { + "epoch": 2.8843619591478147, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.489285469055176, + "learning_rate": 1e-06, + "loss": 0.9094, + "mean_token_accuracy": 0.7301813960075378, + "num_tokens": 679618867.0, + "step": 26265 + }, + { + "epoch": 2.8844717768504284, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4124817848205566, + "learning_rate": 1e-06, + "loss": 1.0467, + "mean_token_accuracy": 0.6959846019744873, + "num_tokens": 679647701.0, + "step": 26266 + }, + { + "epoch": 2.884581594553042, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.888789176940918, + "learning_rate": 1e-06, + "loss": 0.9097, + "mean_token_accuracy": 0.7266459465026855, + "num_tokens": 679666694.0, + "step": 26267 + }, + { + "epoch": 2.8846914122556555, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.672152042388916, + "learning_rate": 1e-06, + "loss": 0.9689, + "mean_token_accuracy": 0.7238104939460754, + "num_tokens": 679690306.0, + "step": 26268 + }, + { + "epoch": 2.8848012299582693, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 16.755054473876953, + "learning_rate": 1e-06, + "loss": 1.0064, + "mean_token_accuracy": 0.7033064365386963, + "num_tokens": 679710856.0, + "step": 26269 + }, + { + "epoch": 2.884911047660883, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5488526821136475, + "learning_rate": 1e-06, + "loss": 1.0458, + "mean_token_accuracy": 0.692133903503418, + "num_tokens": 679740890.0, + "step": 26270 + }, + { + "epoch": 2.8850208653634963, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.495720863342285, + "learning_rate": 1e-06, + "loss": 1.0048, + "mean_token_accuracy": 0.7071592807769775, + "num_tokens": 679772541.0, + "step": 26271 + }, + { + "epoch": 2.88513068306611, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.089437484741211, + "learning_rate": 1e-06, + "loss": 0.9382, + "mean_token_accuracy": 0.7173327207565308, + "num_tokens": 679792942.0, + "step": 26272 + }, + { + "epoch": 2.885240500768724, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.451035976409912, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.6929123401641846, + "num_tokens": 679821926.0, + "step": 26273 + }, + { + "epoch": 2.8853503184713376, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5587706565856934, + "learning_rate": 1e-06, + "loss": 0.9544, + "mean_token_accuracy": 0.7196818590164185, + "num_tokens": 679846469.0, + "step": 26274 + }, + { + "epoch": 2.8854601361739514, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3417375087738037, + "learning_rate": 1e-06, + "loss": 1.0421, + "mean_token_accuracy": 0.6940499544143677, + "num_tokens": 679876950.0, + "step": 26275 + }, + { + "epoch": 2.8855699538765647, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3957455158233643, + "learning_rate": 1e-06, + "loss": 1.0481, + "mean_token_accuracy": 0.6890462636947632, + "num_tokens": 679907480.0, + "step": 26276 + }, + { + "epoch": 2.8856797715791784, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6997439861297607, + "learning_rate": 1e-06, + "loss": 0.9856, + "mean_token_accuracy": 0.7178494930267334, + "num_tokens": 679932416.0, + "step": 26277 + }, + { + "epoch": 2.885789589281792, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9379003047943115, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7069657444953918, + "num_tokens": 679956498.0, + "step": 26278 + }, + { + "epoch": 2.885899406984406, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5893096923828125, + "learning_rate": 1e-06, + "loss": 1.0251, + "mean_token_accuracy": 0.699786901473999, + "num_tokens": 679983183.0, + "step": 26279 + }, + { + "epoch": 2.8860092246870197, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3246994018554688, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7323100566864014, + "num_tokens": 680013070.0, + "step": 26280 + }, + { + "epoch": 2.886119042389633, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6805310249328613, + "learning_rate": 1e-06, + "loss": 1.1058, + "mean_token_accuracy": 0.6950415968894958, + "num_tokens": 680038849.0, + "step": 26281 + }, + { + "epoch": 2.8862288600922468, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.599008560180664, + "learning_rate": 1e-06, + "loss": 0.9928, + "mean_token_accuracy": 0.7058003544807434, + "num_tokens": 680064328.0, + "step": 26282 + }, + { + "epoch": 2.8863386777948605, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5399036407470703, + "learning_rate": 1e-06, + "loss": 1.0926, + "mean_token_accuracy": 0.6897439360618591, + "num_tokens": 680091408.0, + "step": 26283 + }, + { + "epoch": 2.8864484954974743, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.386880397796631, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.715988278388977, + "num_tokens": 680123767.0, + "step": 26284 + }, + { + "epoch": 2.886558313200088, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.975980758666992, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7008914947509766, + "num_tokens": 680143812.0, + "step": 26285 + }, + { + "epoch": 2.8866681309027014, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6018152236938477, + "learning_rate": 1e-06, + "loss": 0.9624, + "mean_token_accuracy": 0.7333109378814697, + "num_tokens": 680169260.0, + "step": 26286 + }, + { + "epoch": 2.886777948605315, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.546534776687622, + "learning_rate": 1e-06, + "loss": 0.9756, + "mean_token_accuracy": 0.7204917669296265, + "num_tokens": 680193558.0, + "step": 26287 + }, + { + "epoch": 2.886887766307929, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5873773097991943, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.6998076438903809, + "num_tokens": 680220597.0, + "step": 26288 + }, + { + "epoch": 2.8869975840105426, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.942612648010254, + "learning_rate": 1e-06, + "loss": 0.9608, + "mean_token_accuracy": 0.7248885035514832, + "num_tokens": 680240737.0, + "step": 26289 + }, + { + "epoch": 2.8871074017131564, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561795473098755, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7198672890663147, + "num_tokens": 680266014.0, + "step": 26290 + }, + { + "epoch": 2.8872172194157697, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.636406660079956, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.7119077444076538, + "num_tokens": 680290115.0, + "step": 26291 + }, + { + "epoch": 2.8873270371183835, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5504603385925293, + "learning_rate": 1e-06, + "loss": 1.0109, + "mean_token_accuracy": 0.7017645835876465, + "num_tokens": 680318250.0, + "step": 26292 + }, + { + "epoch": 2.887436854820997, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6626789569854736, + "learning_rate": 1e-06, + "loss": 1.016, + "mean_token_accuracy": 0.705958902835846, + "num_tokens": 680342706.0, + "step": 26293 + }, + { + "epoch": 2.887546672523611, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4889726638793945, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6934572458267212, + "num_tokens": 680373243.0, + "step": 26294 + }, + { + "epoch": 2.8876564902262247, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.2740767002105713, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.705782413482666, + "num_tokens": 680404986.0, + "step": 26295 + }, + { + "epoch": 2.887766307928838, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6034066677093506, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.7003260850906372, + "num_tokens": 680430597.0, + "step": 26296 + }, + { + "epoch": 2.887876125631452, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.451268434524536, + "learning_rate": 1e-06, + "loss": 0.9433, + "mean_token_accuracy": 0.727212131023407, + "num_tokens": 680458734.0, + "step": 26297 + }, + { + "epoch": 2.8879859433340656, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3086791038513184, + "learning_rate": 1e-06, + "loss": 1.1018, + "mean_token_accuracy": 0.6862562298774719, + "num_tokens": 680490034.0, + "step": 26298 + }, + { + "epoch": 2.888095761036679, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5271108150482178, + "learning_rate": 1e-06, + "loss": 0.9329, + "mean_token_accuracy": 0.7240293622016907, + "num_tokens": 680515547.0, + "step": 26299 + }, + { + "epoch": 2.8882055787392926, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8390114307403564, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7110006809234619, + "num_tokens": 680538587.0, + "step": 26300 + }, + { + "epoch": 2.8883153964419064, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5747008323669434, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.702038586139679, + "num_tokens": 680565636.0, + "step": 26301 + }, + { + "epoch": 2.88842521414452, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.561185836791992, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7170089483261108, + "num_tokens": 680589704.0, + "step": 26302 + }, + { + "epoch": 2.888535031847134, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8278744220733643, + "learning_rate": 1e-06, + "loss": 1.0214, + "mean_token_accuracy": 0.7006821632385254, + "num_tokens": 680611431.0, + "step": 26303 + }, + { + "epoch": 2.888644849549747, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.799744129180908, + "learning_rate": 1e-06, + "loss": 1.0383, + "mean_token_accuracy": 0.6976175308227539, + "num_tokens": 680634011.0, + "step": 26304 + }, + { + "epoch": 2.888754667252361, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.757498264312744, + "learning_rate": 1e-06, + "loss": 0.9362, + "mean_token_accuracy": 0.7281864881515503, + "num_tokens": 680657458.0, + "step": 26305 + }, + { + "epoch": 2.8888644849549747, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.57059645652771, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.7370632290840149, + "num_tokens": 680683301.0, + "step": 26306 + }, + { + "epoch": 2.8889743026575885, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.518090009689331, + "learning_rate": 1e-06, + "loss": 1.0741, + "mean_token_accuracy": 0.6898725032806396, + "num_tokens": 680709705.0, + "step": 26307 + }, + { + "epoch": 2.8890841203602022, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4681668281555176, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.699518084526062, + "num_tokens": 680736867.0, + "step": 26308 + }, + { + "epoch": 2.8891939380628155, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.596262216567993, + "learning_rate": 1e-06, + "loss": 1.0125, + "mean_token_accuracy": 0.7093909978866577, + "num_tokens": 680761365.0, + "step": 26309 + }, + { + "epoch": 2.8893037557654293, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.695441484451294, + "learning_rate": 1e-06, + "loss": 1.055, + "mean_token_accuracy": 0.6919959783554077, + "num_tokens": 680785840.0, + "step": 26310 + }, + { + "epoch": 2.889413573468043, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6410300731658936, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7099128365516663, + "num_tokens": 680811875.0, + "step": 26311 + }, + { + "epoch": 2.889523391170657, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7499582767486572, + "learning_rate": 1e-06, + "loss": 0.889, + "mean_token_accuracy": 0.7438482046127319, + "num_tokens": 680834686.0, + "step": 26312 + }, + { + "epoch": 2.8896332088732706, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4364311695098877, + "learning_rate": 1e-06, + "loss": 1.0501, + "mean_token_accuracy": 0.6956024765968323, + "num_tokens": 680863684.0, + "step": 26313 + }, + { + "epoch": 2.889743026575884, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.62565541267395, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7139343023300171, + "num_tokens": 680886106.0, + "step": 26314 + }, + { + "epoch": 2.8898528442784976, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.460427761077881, + "learning_rate": 1e-06, + "loss": 1.0162, + "mean_token_accuracy": 0.7020004391670227, + "num_tokens": 680914405.0, + "step": 26315 + }, + { + "epoch": 2.8899626619811114, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.390779972076416, + "learning_rate": 1e-06, + "loss": 1.0274, + "mean_token_accuracy": 0.7057667374610901, + "num_tokens": 680943760.0, + "step": 26316 + }, + { + "epoch": 2.890072479683725, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4890382289886475, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7202681303024292, + "num_tokens": 680971377.0, + "step": 26317 + }, + { + "epoch": 2.890182297386339, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4236316680908203, + "learning_rate": 1e-06, + "loss": 1.0188, + "mean_token_accuracy": 0.7010962963104248, + "num_tokens": 681002931.0, + "step": 26318 + }, + { + "epoch": 2.8902921150889522, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 2.7127389907836914, + "learning_rate": 1e-06, + "loss": 1.0355, + "mean_token_accuracy": 0.7059595584869385, + "num_tokens": 681030516.0, + "step": 26319 + }, + { + "epoch": 2.890401932791566, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4250118732452393, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6833556294441223, + "num_tokens": 681059702.0, + "step": 26320 + }, + { + "epoch": 2.8905117504941797, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4930050373077393, + "learning_rate": 1e-06, + "loss": 1.1597, + "mean_token_accuracy": 0.6658616065979004, + "num_tokens": 681092970.0, + "step": 26321 + }, + { + "epoch": 2.890621568196793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6643199920654297, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.725477933883667, + "num_tokens": 681116302.0, + "step": 26322 + }, + { + "epoch": 2.8907313858994073, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.607992649078369, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.7014204263687134, + "num_tokens": 681140960.0, + "step": 26323 + }, + { + "epoch": 2.8908412036020206, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6748106479644775, + "learning_rate": 1e-06, + "loss": 1.0047, + "mean_token_accuracy": 0.7038216590881348, + "num_tokens": 681166078.0, + "step": 26324 + }, + { + "epoch": 2.8909510213046343, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.777292013168335, + "learning_rate": 1e-06, + "loss": 1.1231, + "mean_token_accuracy": 0.6877814531326294, + "num_tokens": 681194378.0, + "step": 26325 + }, + { + "epoch": 2.891060839007248, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.598116159439087, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7036769986152649, + "num_tokens": 681220387.0, + "step": 26326 + }, + { + "epoch": 2.8911706567098614, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6175167560577393, + "learning_rate": 1e-06, + "loss": 1.0406, + "mean_token_accuracy": 0.699225664138794, + "num_tokens": 681246376.0, + "step": 26327 + }, + { + "epoch": 2.891280474412475, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.405365228652954, + "learning_rate": 1e-06, + "loss": 1.0806, + "mean_token_accuracy": 0.6915347576141357, + "num_tokens": 681277320.0, + "step": 26328 + }, + { + "epoch": 2.891390292115089, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5516974925994873, + "learning_rate": 1e-06, + "loss": 0.9521, + "mean_token_accuracy": 0.7243713736534119, + "num_tokens": 681301160.0, + "step": 26329 + }, + { + "epoch": 2.8915001098177027, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8276526927948, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7186342477798462, + "num_tokens": 681322226.0, + "step": 26330 + }, + { + "epoch": 2.8916099275203164, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.749688148498535, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7186442613601685, + "num_tokens": 681345895.0, + "step": 26331 + }, + { + "epoch": 2.8917197452229297, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7526912689208984, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7187422513961792, + "num_tokens": 681369058.0, + "step": 26332 + }, + { + "epoch": 2.8918295629255435, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8431832790374756, + "learning_rate": 1e-06, + "loss": 1.0732, + "mean_token_accuracy": 0.6999054551124573, + "num_tokens": 681393988.0, + "step": 26333 + }, + { + "epoch": 2.8919393806281573, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6706507205963135, + "learning_rate": 1e-06, + "loss": 0.9477, + "mean_token_accuracy": 0.7217923998832703, + "num_tokens": 681419271.0, + "step": 26334 + }, + { + "epoch": 2.892049198330771, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6441707611083984, + "learning_rate": 1e-06, + "loss": 1.0583, + "mean_token_accuracy": 0.6924057006835938, + "num_tokens": 681445181.0, + "step": 26335 + }, + { + "epoch": 2.8921590160333848, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.2774176597595215, + "learning_rate": 1e-06, + "loss": 1.067, + "mean_token_accuracy": 0.6938235759735107, + "num_tokens": 681478406.0, + "step": 26336 + }, + { + "epoch": 2.892268833735998, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7445764541625977, + "learning_rate": 1e-06, + "loss": 1.0371, + "mean_token_accuracy": 0.704161524772644, + "num_tokens": 681502338.0, + "step": 26337 + }, + { + "epoch": 2.892378651438612, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.552962303161621, + "learning_rate": 1e-06, + "loss": 1.089, + "mean_token_accuracy": 0.6855794787406921, + "num_tokens": 681531407.0, + "step": 26338 + }, + { + "epoch": 2.8924884691412256, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2420339584350586, + "learning_rate": 1e-06, + "loss": 1.0521, + "mean_token_accuracy": 0.6907366514205933, + "num_tokens": 681563462.0, + "step": 26339 + }, + { + "epoch": 2.8925982868438394, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7774362564086914, + "learning_rate": 1e-06, + "loss": 0.9449, + "mean_token_accuracy": 0.7268500328063965, + "num_tokens": 681584214.0, + "step": 26340 + }, + { + "epoch": 2.892708104546453, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.506147623062134, + "learning_rate": 1e-06, + "loss": 0.969, + "mean_token_accuracy": 0.7203527688980103, + "num_tokens": 681613791.0, + "step": 26341 + }, + { + "epoch": 2.8928179222490664, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.655764102935791, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.696216344833374, + "num_tokens": 681646841.0, + "step": 26342 + }, + { + "epoch": 2.89292773995168, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7317774295806885, + "learning_rate": 1e-06, + "loss": 1.0948, + "mean_token_accuracy": 0.6900326013565063, + "num_tokens": 681672239.0, + "step": 26343 + }, + { + "epoch": 2.893037557654294, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9124598503112793, + "learning_rate": 1e-06, + "loss": 1.0239, + "mean_token_accuracy": 0.7024473547935486, + "num_tokens": 681692934.0, + "step": 26344 + }, + { + "epoch": 2.8931473753569077, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.544675350189209, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.7278621792793274, + "num_tokens": 681716216.0, + "step": 26345 + }, + { + "epoch": 2.8932571930595214, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.526179552078247, + "learning_rate": 1e-06, + "loss": 1.0186, + "mean_token_accuracy": 0.7034387588500977, + "num_tokens": 681745947.0, + "step": 26346 + }, + { + "epoch": 2.8933670107621348, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.2387619018554688, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7196558713912964, + "num_tokens": 681764836.0, + "step": 26347 + }, + { + "epoch": 2.8934768284647485, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.711782932281494, + "learning_rate": 1e-06, + "loss": 1.0627, + "mean_token_accuracy": 0.6995314359664917, + "num_tokens": 681790826.0, + "step": 26348 + }, + { + "epoch": 2.8935866461673623, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.469374418258667, + "learning_rate": 1e-06, + "loss": 0.9835, + "mean_token_accuracy": 0.7159371376037598, + "num_tokens": 681818482.0, + "step": 26349 + }, + { + "epoch": 2.8936964638699756, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6428446769714355, + "learning_rate": 1e-06, + "loss": 1.0795, + "mean_token_accuracy": 0.6902985572814941, + "num_tokens": 681843663.0, + "step": 26350 + }, + { + "epoch": 2.8938062815725893, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5505261421203613, + "learning_rate": 1e-06, + "loss": 1.0152, + "mean_token_accuracy": 0.709392786026001, + "num_tokens": 681870011.0, + "step": 26351 + }, + { + "epoch": 2.893916099275203, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.558851480484009, + "learning_rate": 1e-06, + "loss": 0.9445, + "mean_token_accuracy": 0.7224729061126709, + "num_tokens": 681894347.0, + "step": 26352 + }, + { + "epoch": 2.894025916977817, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4541335105895996, + "learning_rate": 1e-06, + "loss": 0.9478, + "mean_token_accuracy": 0.7197361588478088, + "num_tokens": 681921147.0, + "step": 26353 + }, + { + "epoch": 2.8941357346804306, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.418944835662842, + "learning_rate": 1e-06, + "loss": 0.9248, + "mean_token_accuracy": 0.721278727054596, + "num_tokens": 681946221.0, + "step": 26354 + }, + { + "epoch": 2.894245552383044, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4255874156951904, + "learning_rate": 1e-06, + "loss": 1.0998, + "mean_token_accuracy": 0.6799219846725464, + "num_tokens": 681977249.0, + "step": 26355 + }, + { + "epoch": 2.8943553700856577, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.58823299407959, + "learning_rate": 1e-06, + "loss": 1.0302, + "mean_token_accuracy": 0.7054418325424194, + "num_tokens": 682004527.0, + "step": 26356 + }, + { + "epoch": 2.8944651877882714, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6730849742889404, + "learning_rate": 1e-06, + "loss": 0.9104, + "mean_token_accuracy": 0.7311022877693176, + "num_tokens": 682028877.0, + "step": 26357 + }, + { + "epoch": 2.894575005490885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.534992218017578, + "learning_rate": 1e-06, + "loss": 0.9525, + "mean_token_accuracy": 0.7184209823608398, + "num_tokens": 682055359.0, + "step": 26358 + }, + { + "epoch": 2.894684823193499, + "ewc_loss": 1.0728836059570312e-05, + "grad_norm": 4.396067142486572, + "learning_rate": 1e-06, + "loss": 0.9682, + "mean_token_accuracy": 0.7169999480247498, + "num_tokens": 682077733.0, + "step": 26359 + }, + { + "epoch": 2.8947946408961123, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.867770195007324, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.6999126076698303, + "num_tokens": 682099772.0, + "step": 26360 + }, + { + "epoch": 2.894904458598726, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.617985725402832, + "learning_rate": 1e-06, + "loss": 0.9827, + "mean_token_accuracy": 0.7189851999282837, + "num_tokens": 682127088.0, + "step": 26361 + }, + { + "epoch": 2.89501427630134, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7774412631988525, + "learning_rate": 1e-06, + "loss": 0.9955, + "mean_token_accuracy": 0.7131981253623962, + "num_tokens": 682149027.0, + "step": 26362 + }, + { + "epoch": 2.8951240940039535, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.661579132080078, + "learning_rate": 1e-06, + "loss": 1.0571, + "mean_token_accuracy": 0.6998952627182007, + "num_tokens": 682172011.0, + "step": 26363 + }, + { + "epoch": 2.8952339117065673, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5524909496307373, + "learning_rate": 1e-06, + "loss": 1.0746, + "mean_token_accuracy": 0.6925017833709717, + "num_tokens": 682199704.0, + "step": 26364 + }, + { + "epoch": 2.8953437294091806, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8635802268981934, + "learning_rate": 1e-06, + "loss": 0.8585, + "mean_token_accuracy": 0.7464759349822998, + "num_tokens": 682219918.0, + "step": 26365 + }, + { + "epoch": 2.8954535471117944, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8946285247802734, + "learning_rate": 1e-06, + "loss": 0.95, + "mean_token_accuracy": 0.7223720550537109, + "num_tokens": 682241287.0, + "step": 26366 + }, + { + "epoch": 2.895563364814408, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5503244400024414, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.702981173992157, + "num_tokens": 682266925.0, + "step": 26367 + }, + { + "epoch": 2.895673182517022, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.917402982711792, + "learning_rate": 1e-06, + "loss": 0.9728, + "mean_token_accuracy": 0.7234955430030823, + "num_tokens": 682287503.0, + "step": 26368 + }, + { + "epoch": 2.8957830002196356, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5420873165130615, + "learning_rate": 1e-06, + "loss": 0.9182, + "mean_token_accuracy": 0.7321656942367554, + "num_tokens": 682311142.0, + "step": 26369 + }, + { + "epoch": 2.895892817922249, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3553905487060547, + "learning_rate": 1e-06, + "loss": 1.0455, + "mean_token_accuracy": 0.6938139200210571, + "num_tokens": 682346342.0, + "step": 26370 + }, + { + "epoch": 2.8960026356248627, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6173932552337646, + "learning_rate": 1e-06, + "loss": 1.0044, + "mean_token_accuracy": 0.7044086456298828, + "num_tokens": 682370628.0, + "step": 26371 + }, + { + "epoch": 2.8961124533274765, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.831301689147949, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7095798850059509, + "num_tokens": 682393547.0, + "step": 26372 + }, + { + "epoch": 2.89622227103009, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.48097562789917, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7180934548377991, + "num_tokens": 682421741.0, + "step": 26373 + }, + { + "epoch": 2.896332088732704, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.588225841522217, + "learning_rate": 1e-06, + "loss": 1.0242, + "mean_token_accuracy": 0.7036919593811035, + "num_tokens": 682449528.0, + "step": 26374 + }, + { + "epoch": 2.8964419064353173, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6317269802093506, + "learning_rate": 1e-06, + "loss": 1.0553, + "mean_token_accuracy": 0.6953901052474976, + "num_tokens": 682473028.0, + "step": 26375 + }, + { + "epoch": 2.896551724137931, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5708231925964355, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6989057660102844, + "num_tokens": 682500082.0, + "step": 26376 + }, + { + "epoch": 2.896661541840545, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3222575187683105, + "learning_rate": 1e-06, + "loss": 1.0651, + "mean_token_accuracy": 0.695087194442749, + "num_tokens": 682530425.0, + "step": 26377 + }, + { + "epoch": 2.896771359543158, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6998021602630615, + "learning_rate": 1e-06, + "loss": 1.0363, + "mean_token_accuracy": 0.7017567753791809, + "num_tokens": 682554091.0, + "step": 26378 + }, + { + "epoch": 2.896881177245772, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.511842966079712, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7024155855178833, + "num_tokens": 682579492.0, + "step": 26379 + }, + { + "epoch": 2.8969909949483856, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.538916826248169, + "learning_rate": 1e-06, + "loss": 0.9377, + "mean_token_accuracy": 0.7304947972297668, + "num_tokens": 682603660.0, + "step": 26380 + }, + { + "epoch": 2.8971008126509994, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6175496578216553, + "learning_rate": 1e-06, + "loss": 1.0632, + "mean_token_accuracy": 0.6974040269851685, + "num_tokens": 682631859.0, + "step": 26381 + }, + { + "epoch": 2.897210630353613, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.641042470932007, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.7217395305633545, + "num_tokens": 682656306.0, + "step": 26382 + }, + { + "epoch": 2.8973204480562265, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9006361961364746, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7109963893890381, + "num_tokens": 682679307.0, + "step": 26383 + }, + { + "epoch": 2.89743026575884, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7385549545288086, + "learning_rate": 1e-06, + "loss": 0.9798, + "mean_token_accuracy": 0.7137054800987244, + "num_tokens": 682704005.0, + "step": 26384 + }, + { + "epoch": 2.897540083461454, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4062881469726562, + "learning_rate": 1e-06, + "loss": 0.9511, + "mean_token_accuracy": 0.7189067602157593, + "num_tokens": 682730369.0, + "step": 26385 + }, + { + "epoch": 2.8976499011640677, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4170706272125244, + "learning_rate": 1e-06, + "loss": 1.0934, + "mean_token_accuracy": 0.6994214653968811, + "num_tokens": 682761701.0, + "step": 26386 + }, + { + "epoch": 2.8977597188666815, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4567434787750244, + "learning_rate": 1e-06, + "loss": 0.9571, + "mean_token_accuracy": 0.7198424339294434, + "num_tokens": 682787398.0, + "step": 26387 + }, + { + "epoch": 2.897869536569295, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5918571949005127, + "learning_rate": 1e-06, + "loss": 1.0181, + "mean_token_accuracy": 0.6995424628257751, + "num_tokens": 682813490.0, + "step": 26388 + }, + { + "epoch": 2.8979793542719086, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.673703193664551, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.717953085899353, + "num_tokens": 682839551.0, + "step": 26389 + }, + { + "epoch": 2.8980891719745223, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7156574726104736, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6979635953903198, + "num_tokens": 682864892.0, + "step": 26390 + }, + { + "epoch": 2.898198989677136, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.563512086868286, + "learning_rate": 1e-06, + "loss": 1.0671, + "mean_token_accuracy": 0.6881003379821777, + "num_tokens": 682892896.0, + "step": 26391 + }, + { + "epoch": 2.89830880737975, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.777174472808838, + "learning_rate": 1e-06, + "loss": 1.0191, + "mean_token_accuracy": 0.700316309928894, + "num_tokens": 682915798.0, + "step": 26392 + }, + { + "epoch": 2.898418625082363, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5180273056030273, + "learning_rate": 1e-06, + "loss": 1.0055, + "mean_token_accuracy": 0.7058987617492676, + "num_tokens": 682940538.0, + "step": 26393 + }, + { + "epoch": 2.898528442784977, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.467984199523926, + "learning_rate": 1e-06, + "loss": 1.0854, + "mean_token_accuracy": 0.6820173263549805, + "num_tokens": 682972109.0, + "step": 26394 + }, + { + "epoch": 2.8986382604875907, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.661261796951294, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.70187908411026, + "num_tokens": 682998786.0, + "step": 26395 + }, + { + "epoch": 2.8987480781902044, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6271839141845703, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.6862339377403259, + "num_tokens": 683025291.0, + "step": 26396 + }, + { + "epoch": 2.898857895892818, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.277848243713379, + "learning_rate": 1e-06, + "loss": 0.966, + "mean_token_accuracy": 0.7131797075271606, + "num_tokens": 683056370.0, + "step": 26397 + }, + { + "epoch": 2.8989677135954315, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.562851667404175, + "learning_rate": 1e-06, + "loss": 1.0287, + "mean_token_accuracy": 0.7077334523200989, + "num_tokens": 683084465.0, + "step": 26398 + }, + { + "epoch": 2.8990775312980452, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.798062562942505, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7094213962554932, + "num_tokens": 683108337.0, + "step": 26399 + }, + { + "epoch": 2.899187349000659, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5871572494506836, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7010009288787842, + "num_tokens": 683134858.0, + "step": 26400 + }, + { + "epoch": 2.8992971667032723, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.753988027572632, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.707284152507782, + "num_tokens": 683158821.0, + "step": 26401 + }, + { + "epoch": 2.899406984405886, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5814881324768066, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.6974364519119263, + "num_tokens": 683185864.0, + "step": 26402 + }, + { + "epoch": 2.8995168021085, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5406525135040283, + "learning_rate": 1e-06, + "loss": 1.0173, + "mean_token_accuracy": 0.700018048286438, + "num_tokens": 683211009.0, + "step": 26403 + }, + { + "epoch": 2.8996266198111136, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4620823860168457, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7223951816558838, + "num_tokens": 683240253.0, + "step": 26404 + }, + { + "epoch": 2.8997364375137273, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.61936092376709, + "learning_rate": 1e-06, + "loss": 1.0373, + "mean_token_accuracy": 0.6993207335472107, + "num_tokens": 683264149.0, + "step": 26405 + }, + { + "epoch": 2.8998462552163407, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7522659301757812, + "learning_rate": 1e-06, + "loss": 1.0576, + "mean_token_accuracy": 0.6966344714164734, + "num_tokens": 683292588.0, + "step": 26406 + }, + { + "epoch": 2.8999560729189544, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.667372941970825, + "learning_rate": 1e-06, + "loss": 0.9374, + "mean_token_accuracy": 0.7205480337142944, + "num_tokens": 683314964.0, + "step": 26407 + }, + { + "epoch": 2.900065890621568, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6473495960235596, + "learning_rate": 1e-06, + "loss": 0.8571, + "mean_token_accuracy": 0.7459055781364441, + "num_tokens": 683340361.0, + "step": 26408 + }, + { + "epoch": 2.900175708324182, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8654706478118896, + "learning_rate": 1e-06, + "loss": 0.9656, + "mean_token_accuracy": 0.7102693915367126, + "num_tokens": 683362288.0, + "step": 26409 + }, + { + "epoch": 2.9002855260267957, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.667055130004883, + "learning_rate": 1e-06, + "loss": 1.0565, + "mean_token_accuracy": 0.7047760486602783, + "num_tokens": 683387158.0, + "step": 26410 + }, + { + "epoch": 2.900395343729409, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0916595458984375, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6981831789016724, + "num_tokens": 683408163.0, + "step": 26411 + }, + { + "epoch": 2.9005051614320227, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6107370853424072, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7006834745407104, + "num_tokens": 683434116.0, + "step": 26412 + }, + { + "epoch": 2.9006149791346365, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4973819255828857, + "learning_rate": 1e-06, + "loss": 1.0241, + "mean_token_accuracy": 0.7083277106285095, + "num_tokens": 683463104.0, + "step": 26413 + }, + { + "epoch": 2.9007247968372503, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6167562007904053, + "learning_rate": 1e-06, + "loss": 1.0245, + "mean_token_accuracy": 0.7043139934539795, + "num_tokens": 683488989.0, + "step": 26414 + }, + { + "epoch": 2.900834614539864, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4509623050689697, + "learning_rate": 1e-06, + "loss": 1.0327, + "mean_token_accuracy": 0.7030094265937805, + "num_tokens": 683520293.0, + "step": 26415 + }, + { + "epoch": 2.9009444322424773, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8557612895965576, + "learning_rate": 1e-06, + "loss": 0.896, + "mean_token_accuracy": 0.7299820780754089, + "num_tokens": 683542115.0, + "step": 26416 + }, + { + "epoch": 2.901054249945091, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4908199310302734, + "learning_rate": 1e-06, + "loss": 0.9969, + "mean_token_accuracy": 0.7073745727539062, + "num_tokens": 683572409.0, + "step": 26417 + }, + { + "epoch": 2.901164067647705, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6605801582336426, + "learning_rate": 1e-06, + "loss": 1.0138, + "mean_token_accuracy": 0.7120715379714966, + "num_tokens": 683597498.0, + "step": 26418 + }, + { + "epoch": 2.9012738853503186, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.2074875831604004, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7193130850791931, + "num_tokens": 683632653.0, + "step": 26419 + }, + { + "epoch": 2.9013837030529324, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.401235818862915, + "learning_rate": 1e-06, + "loss": 0.9934, + "mean_token_accuracy": 0.7084531188011169, + "num_tokens": 683662090.0, + "step": 26420 + }, + { + "epoch": 2.9014935207555457, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.75639009475708, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7087124586105347, + "num_tokens": 683684165.0, + "step": 26421 + }, + { + "epoch": 2.9016033384581594, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.623394250869751, + "learning_rate": 1e-06, + "loss": 1.0695, + "mean_token_accuracy": 0.6932483911514282, + "num_tokens": 683710140.0, + "step": 26422 + }, + { + "epoch": 2.901713156160773, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6912333965301514, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6903719902038574, + "num_tokens": 683738357.0, + "step": 26423 + }, + { + "epoch": 2.9018229738633865, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.717540740966797, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7309577465057373, + "num_tokens": 683765150.0, + "step": 26424 + }, + { + "epoch": 2.9019327915660007, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6805267333984375, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6848822832107544, + "num_tokens": 683791127.0, + "step": 26425 + }, + { + "epoch": 2.902042609268614, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.754917860031128, + "learning_rate": 1e-06, + "loss": 0.9898, + "mean_token_accuracy": 0.7141438722610474, + "num_tokens": 683814724.0, + "step": 26426 + }, + { + "epoch": 2.9021524269712278, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 20.569194793701172, + "learning_rate": 1e-06, + "loss": 0.9026, + "mean_token_accuracy": 0.7375481128692627, + "num_tokens": 683834075.0, + "step": 26427 + }, + { + "epoch": 2.9022622446738415, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5333523750305176, + "learning_rate": 1e-06, + "loss": 0.9439, + "mean_token_accuracy": 0.7222111225128174, + "num_tokens": 683865794.0, + "step": 26428 + }, + { + "epoch": 2.902372062376455, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.800349235534668, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7126768827438354, + "num_tokens": 683891188.0, + "step": 26429 + }, + { + "epoch": 2.9024818800790686, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.575061082839966, + "learning_rate": 1e-06, + "loss": 1.1507, + "mean_token_accuracy": 0.6755062341690063, + "num_tokens": 683926150.0, + "step": 26430 + }, + { + "epoch": 2.9025916977816824, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.911978244781494, + "learning_rate": 1e-06, + "loss": 0.9013, + "mean_token_accuracy": 0.7379361391067505, + "num_tokens": 683946763.0, + "step": 26431 + }, + { + "epoch": 2.902701515484296, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5895097255706787, + "learning_rate": 1e-06, + "loss": 1.0007, + "mean_token_accuracy": 0.7124191522598267, + "num_tokens": 683972539.0, + "step": 26432 + }, + { + "epoch": 2.90281133318691, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8064138889312744, + "learning_rate": 1e-06, + "loss": 0.951, + "mean_token_accuracy": 0.7202757000923157, + "num_tokens": 683995041.0, + "step": 26433 + }, + { + "epoch": 2.902921150889523, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3270671367645264, + "learning_rate": 1e-06, + "loss": 1.0749, + "mean_token_accuracy": 0.6905815601348877, + "num_tokens": 684028733.0, + "step": 26434 + }, + { + "epoch": 2.903030968592137, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.530697822570801, + "learning_rate": 1e-06, + "loss": 1.03, + "mean_token_accuracy": 0.700615406036377, + "num_tokens": 684056832.0, + "step": 26435 + }, + { + "epoch": 2.9031407862947507, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3637850284576416, + "learning_rate": 1e-06, + "loss": 0.8908, + "mean_token_accuracy": 0.7337834239006042, + "num_tokens": 684085015.0, + "step": 26436 + }, + { + "epoch": 2.9032506039973645, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.789813756942749, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7280468940734863, + "num_tokens": 684109006.0, + "step": 26437 + }, + { + "epoch": 2.903360421699978, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6772215366363525, + "learning_rate": 1e-06, + "loss": 0.9228, + "mean_token_accuracy": 0.7236368656158447, + "num_tokens": 684132450.0, + "step": 26438 + }, + { + "epoch": 2.9034702394025915, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.282151699066162, + "learning_rate": 1e-06, + "loss": 1.0126, + "mean_token_accuracy": 0.7076320648193359, + "num_tokens": 684164024.0, + "step": 26439 + }, + { + "epoch": 2.9035800571052053, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4235646724700928, + "learning_rate": 1e-06, + "loss": 1.0704, + "mean_token_accuracy": 0.6989811658859253, + "num_tokens": 684193418.0, + "step": 26440 + }, + { + "epoch": 2.903689874807819, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.735008478164673, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.7019256949424744, + "num_tokens": 684219420.0, + "step": 26441 + }, + { + "epoch": 2.903799692510433, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6372082233428955, + "learning_rate": 1e-06, + "loss": 0.802, + "mean_token_accuracy": 0.7604907155036926, + "num_tokens": 684241687.0, + "step": 26442 + }, + { + "epoch": 2.9039095102130466, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.554633855819702, + "learning_rate": 1e-06, + "loss": 0.9741, + "mean_token_accuracy": 0.7146298289299011, + "num_tokens": 684270672.0, + "step": 26443 + }, + { + "epoch": 2.90401932791566, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.834628105163574, + "learning_rate": 1e-06, + "loss": 1.0522, + "mean_token_accuracy": 0.69746994972229, + "num_tokens": 684296671.0, + "step": 26444 + }, + { + "epoch": 2.9041291456182736, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6373515129089355, + "learning_rate": 1e-06, + "loss": 1.0183, + "mean_token_accuracy": 0.7030050754547119, + "num_tokens": 684322906.0, + "step": 26445 + }, + { + "epoch": 2.9042389633208874, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3994977474212646, + "learning_rate": 1e-06, + "loss": 1.053, + "mean_token_accuracy": 0.702499270439148, + "num_tokens": 684352381.0, + "step": 26446 + }, + { + "epoch": 2.904348781023501, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.800950765609741, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.7094647288322449, + "num_tokens": 684376296.0, + "step": 26447 + }, + { + "epoch": 2.904458598726115, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9675238132476807, + "learning_rate": 1e-06, + "loss": 0.9945, + "mean_token_accuracy": 0.7096234560012817, + "num_tokens": 684400919.0, + "step": 26448 + }, + { + "epoch": 2.904568416428728, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6134724617004395, + "learning_rate": 1e-06, + "loss": 1.0459, + "mean_token_accuracy": 0.6997693777084351, + "num_tokens": 684428314.0, + "step": 26449 + }, + { + "epoch": 2.904678234131342, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7424049377441406, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.693510115146637, + "num_tokens": 684452540.0, + "step": 26450 + }, + { + "epoch": 2.9047880518339557, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5765721797943115, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7013595700263977, + "num_tokens": 684477807.0, + "step": 26451 + }, + { + "epoch": 2.904897869536569, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9405744075775146, + "learning_rate": 1e-06, + "loss": 0.9365, + "mean_token_accuracy": 0.7326419353485107, + "num_tokens": 684496994.0, + "step": 26452 + }, + { + "epoch": 2.905007687239183, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5130088329315186, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.6975077986717224, + "num_tokens": 684529066.0, + "step": 26453 + }, + { + "epoch": 2.9051175049417965, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.664846181869507, + "learning_rate": 1e-06, + "loss": 0.919, + "mean_token_accuracy": 0.727249026298523, + "num_tokens": 684551104.0, + "step": 26454 + }, + { + "epoch": 2.9052273226444103, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5877881050109863, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7103042602539062, + "num_tokens": 684577229.0, + "step": 26455 + }, + { + "epoch": 2.905337140347024, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4745900630950928, + "learning_rate": 1e-06, + "loss": 0.9873, + "mean_token_accuracy": 0.7162157893180847, + "num_tokens": 684605572.0, + "step": 26456 + }, + { + "epoch": 2.9054469580496374, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3233885765075684, + "learning_rate": 1e-06, + "loss": 1.0075, + "mean_token_accuracy": 0.7138265371322632, + "num_tokens": 684636023.0, + "step": 26457 + }, + { + "epoch": 2.905556775752251, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.471874237060547, + "learning_rate": 1e-06, + "loss": 1.0655, + "mean_token_accuracy": 0.7017345428466797, + "num_tokens": 684663762.0, + "step": 26458 + }, + { + "epoch": 2.905666593454865, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7390286922454834, + "learning_rate": 1e-06, + "loss": 0.9389, + "mean_token_accuracy": 0.7216084003448486, + "num_tokens": 684685399.0, + "step": 26459 + }, + { + "epoch": 2.9057764111574786, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7808303833007812, + "learning_rate": 1e-06, + "loss": 0.8974, + "mean_token_accuracy": 0.7333971858024597, + "num_tokens": 684706935.0, + "step": 26460 + }, + { + "epoch": 2.9058862288600924, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6179301738739014, + "learning_rate": 1e-06, + "loss": 0.994, + "mean_token_accuracy": 0.7129206657409668, + "num_tokens": 684735676.0, + "step": 26461 + }, + { + "epoch": 2.9059960465627057, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.425996780395508, + "learning_rate": 1e-06, + "loss": 1.0234, + "mean_token_accuracy": 0.7028999328613281, + "num_tokens": 684767002.0, + "step": 26462 + }, + { + "epoch": 2.9061058642653195, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9957051277160645, + "learning_rate": 1e-06, + "loss": 1.0436, + "mean_token_accuracy": 0.6980485320091248, + "num_tokens": 684794623.0, + "step": 26463 + }, + { + "epoch": 2.9062156819679332, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6861987113952637, + "learning_rate": 1e-06, + "loss": 0.9932, + "mean_token_accuracy": 0.7097309827804565, + "num_tokens": 684820573.0, + "step": 26464 + }, + { + "epoch": 2.906325499670547, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.533116102218628, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7064468860626221, + "num_tokens": 684848905.0, + "step": 26465 + }, + { + "epoch": 2.9064353173731607, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5269594192504883, + "learning_rate": 1e-06, + "loss": 0.8964, + "mean_token_accuracy": 0.7331165671348572, + "num_tokens": 684872597.0, + "step": 26466 + }, + { + "epoch": 2.906545135075774, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6234960556030273, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6811238527297974, + "num_tokens": 684899741.0, + "step": 26467 + }, + { + "epoch": 2.906654952778388, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7176284790039062, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7183231115341187, + "num_tokens": 684922939.0, + "step": 26468 + }, + { + "epoch": 2.9067647704810016, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.825801134109497, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7051835656166077, + "num_tokens": 684948105.0, + "step": 26469 + }, + { + "epoch": 2.9068745881836153, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4968831539154053, + "learning_rate": 1e-06, + "loss": 1.069, + "mean_token_accuracy": 0.6944417953491211, + "num_tokens": 684979234.0, + "step": 26470 + }, + { + "epoch": 2.906984405886229, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.883838415145874, + "learning_rate": 1e-06, + "loss": 0.9677, + "mean_token_accuracy": 0.7174572944641113, + "num_tokens": 685001535.0, + "step": 26471 + }, + { + "epoch": 2.9070942235888424, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.74711537361145, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7123488783836365, + "num_tokens": 685024754.0, + "step": 26472 + }, + { + "epoch": 2.907204041291456, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.714799404144287, + "learning_rate": 1e-06, + "loss": 0.9811, + "mean_token_accuracy": 0.7124149799346924, + "num_tokens": 685054084.0, + "step": 26473 + }, + { + "epoch": 2.90731385899407, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.332723379135132, + "learning_rate": 1e-06, + "loss": 1.057, + "mean_token_accuracy": 0.6986681222915649, + "num_tokens": 685084931.0, + "step": 26474 + }, + { + "epoch": 2.9074236766966837, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.061434745788574, + "learning_rate": 1e-06, + "loss": 1.042, + "mean_token_accuracy": 0.6967619061470032, + "num_tokens": 685112701.0, + "step": 26475 + }, + { + "epoch": 2.9075334943992974, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7141807079315186, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.7034018039703369, + "num_tokens": 685139446.0, + "step": 26476 + }, + { + "epoch": 2.9076433121019107, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8822708129882812, + "learning_rate": 1e-06, + "loss": 0.9505, + "mean_token_accuracy": 0.7263257503509521, + "num_tokens": 685160733.0, + "step": 26477 + }, + { + "epoch": 2.9077531298045245, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3569581508636475, + "learning_rate": 1e-06, + "loss": 1.0232, + "mean_token_accuracy": 0.7018626928329468, + "num_tokens": 685189134.0, + "step": 26478 + }, + { + "epoch": 2.9078629475071383, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9468493461608887, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7358464002609253, + "num_tokens": 685211228.0, + "step": 26479 + }, + { + "epoch": 2.9079727652097516, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6727194786071777, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7020546197891235, + "num_tokens": 685236995.0, + "step": 26480 + }, + { + "epoch": 2.9080825829123653, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8011062145233154, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7058780193328857, + "num_tokens": 685262502.0, + "step": 26481 + }, + { + "epoch": 2.908192400614979, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.889854907989502, + "learning_rate": 1e-06, + "loss": 0.9635, + "mean_token_accuracy": 0.7170920372009277, + "num_tokens": 685283923.0, + "step": 26482 + }, + { + "epoch": 2.908302218317593, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5741121768951416, + "learning_rate": 1e-06, + "loss": 1.031, + "mean_token_accuracy": 0.708637535572052, + "num_tokens": 685311142.0, + "step": 26483 + }, + { + "epoch": 2.9084120360202066, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.770627498626709, + "learning_rate": 1e-06, + "loss": 0.9319, + "mean_token_accuracy": 0.7214676141738892, + "num_tokens": 685334840.0, + "step": 26484 + }, + { + "epoch": 2.90852185372282, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.542452335357666, + "learning_rate": 1e-06, + "loss": 1.0294, + "mean_token_accuracy": 0.7040702104568481, + "num_tokens": 685362203.0, + "step": 26485 + }, + { + "epoch": 2.9086316714254337, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7981271743774414, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7008248567581177, + "num_tokens": 685385678.0, + "step": 26486 + }, + { + "epoch": 2.9087414891280474, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6088743209838867, + "learning_rate": 1e-06, + "loss": 1.0858, + "mean_token_accuracy": 0.6952081918716431, + "num_tokens": 685416591.0, + "step": 26487 + }, + { + "epoch": 2.908851306830661, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.966952085494995, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.7030580043792725, + "num_tokens": 685439488.0, + "step": 26488 + }, + { + "epoch": 2.908961124533275, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5298430919647217, + "learning_rate": 1e-06, + "loss": 1.0008, + "mean_token_accuracy": 0.6987638473510742, + "num_tokens": 685467009.0, + "step": 26489 + }, + { + "epoch": 2.9090709422358882, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5865256786346436, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7131048440933228, + "num_tokens": 685492831.0, + "step": 26490 + }, + { + "epoch": 2.909180759938502, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 16.73227310180664, + "learning_rate": 1e-06, + "loss": 1.0751, + "mean_token_accuracy": 0.6996580362319946, + "num_tokens": 685520269.0, + "step": 26491 + }, + { + "epoch": 2.9092905776411158, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.2497594356536865, + "learning_rate": 1e-06, + "loss": 1.0093, + "mean_token_accuracy": 0.714073657989502, + "num_tokens": 685539928.0, + "step": 26492 + }, + { + "epoch": 2.9094003953437295, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.08256459236145, + "learning_rate": 1e-06, + "loss": 1.0889, + "mean_token_accuracy": 0.6913508176803589, + "num_tokens": 685564126.0, + "step": 26493 + }, + { + "epoch": 2.9095102130463433, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6046085357666016, + "learning_rate": 1e-06, + "loss": 0.996, + "mean_token_accuracy": 0.7085244655609131, + "num_tokens": 685590910.0, + "step": 26494 + }, + { + "epoch": 2.9096200307489566, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5799293518066406, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6998898386955261, + "num_tokens": 685620084.0, + "step": 26495 + }, + { + "epoch": 2.9097298484515703, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5042715072631836, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7269449830055237, + "num_tokens": 685647562.0, + "step": 26496 + }, + { + "epoch": 2.909839666154184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4511148929595947, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.7014329433441162, + "num_tokens": 685675349.0, + "step": 26497 + }, + { + "epoch": 2.909949483856798, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.705578327178955, + "learning_rate": 1e-06, + "loss": 0.9566, + "mean_token_accuracy": 0.7202345132827759, + "num_tokens": 685698449.0, + "step": 26498 + }, + { + "epoch": 2.9100593015594116, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.599269390106201, + "learning_rate": 1e-06, + "loss": 1.0863, + "mean_token_accuracy": 0.6835042834281921, + "num_tokens": 685727170.0, + "step": 26499 + }, + { + "epoch": 2.910169119262025, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4226722717285156, + "learning_rate": 1e-06, + "loss": 1.0818, + "mean_token_accuracy": 0.6939122676849365, + "num_tokens": 685759830.0, + "step": 26500 + }, + { + "epoch": 2.9102789369646387, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3524978160858154, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7094755172729492, + "num_tokens": 685790878.0, + "step": 26501 + }, + { + "epoch": 2.9103887546672524, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4386987686157227, + "learning_rate": 1e-06, + "loss": 0.9648, + "mean_token_accuracy": 0.714461624622345, + "num_tokens": 685822032.0, + "step": 26502 + }, + { + "epoch": 2.9104985723698658, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6154990196228027, + "learning_rate": 1e-06, + "loss": 1.0431, + "mean_token_accuracy": 0.7033970952033997, + "num_tokens": 685849320.0, + "step": 26503 + }, + { + "epoch": 2.91060839007248, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7098751068115234, + "learning_rate": 1e-06, + "loss": 1.0445, + "mean_token_accuracy": 0.6941516995429993, + "num_tokens": 685873771.0, + "step": 26504 + }, + { + "epoch": 2.9107182077750933, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4189319610595703, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6917954087257385, + "num_tokens": 685904227.0, + "step": 26505 + }, + { + "epoch": 2.910828025477707, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.621323347091675, + "learning_rate": 1e-06, + "loss": 0.9737, + "mean_token_accuracy": 0.7150390148162842, + "num_tokens": 685931888.0, + "step": 26506 + }, + { + "epoch": 2.910937843180321, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6875147819519043, + "learning_rate": 1e-06, + "loss": 0.9869, + "mean_token_accuracy": 0.7114661931991577, + "num_tokens": 685955824.0, + "step": 26507 + }, + { + "epoch": 2.911047660882934, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6237151622772217, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7044600248336792, + "num_tokens": 685980405.0, + "step": 26508 + }, + { + "epoch": 2.911157478585548, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.8929147720336914, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7093859314918518, + "num_tokens": 686000283.0, + "step": 26509 + }, + { + "epoch": 2.9112672962881616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7034924030303955, + "learning_rate": 1e-06, + "loss": 0.9498, + "mean_token_accuracy": 0.7249473333358765, + "num_tokens": 686024486.0, + "step": 26510 + }, + { + "epoch": 2.9113771139907754, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3900139331817627, + "learning_rate": 1e-06, + "loss": 0.8904, + "mean_token_accuracy": 0.7376561164855957, + "num_tokens": 686049863.0, + "step": 26511 + }, + { + "epoch": 2.911486931693389, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6320347785949707, + "learning_rate": 1e-06, + "loss": 0.9947, + "mean_token_accuracy": 0.7188677191734314, + "num_tokens": 686074858.0, + "step": 26512 + }, + { + "epoch": 2.9115967493960024, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6807563304901123, + "learning_rate": 1e-06, + "loss": 1.0143, + "mean_token_accuracy": 0.7032672166824341, + "num_tokens": 686098909.0, + "step": 26513 + }, + { + "epoch": 2.911706567098616, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.910839319229126, + "learning_rate": 1e-06, + "loss": 1.0238, + "mean_token_accuracy": 0.7000173330307007, + "num_tokens": 686122999.0, + "step": 26514 + }, + { + "epoch": 2.91181638480123, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.889564037322998, + "learning_rate": 1e-06, + "loss": 0.9822, + "mean_token_accuracy": 0.7168135046958923, + "num_tokens": 686146006.0, + "step": 26515 + }, + { + "epoch": 2.9119262025038437, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4444615840911865, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.6966320276260376, + "num_tokens": 686179381.0, + "step": 26516 + }, + { + "epoch": 2.9120360202064575, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.583120107650757, + "learning_rate": 1e-06, + "loss": 1.0077, + "mean_token_accuracy": 0.7032911777496338, + "num_tokens": 686207753.0, + "step": 26517 + }, + { + "epoch": 2.912145837909071, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.604459285736084, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.6990260481834412, + "num_tokens": 686233578.0, + "step": 26518 + }, + { + "epoch": 2.9122556556116845, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6775190830230713, + "learning_rate": 1e-06, + "loss": 0.9751, + "mean_token_accuracy": 0.7206361889839172, + "num_tokens": 686258649.0, + "step": 26519 + }, + { + "epoch": 2.9123654733142983, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4955062866210938, + "learning_rate": 1e-06, + "loss": 0.8571, + "mean_token_accuracy": 0.7537477016448975, + "num_tokens": 686282824.0, + "step": 26520 + }, + { + "epoch": 2.912475291016912, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.515732765197754, + "learning_rate": 1e-06, + "loss": 1.0901, + "mean_token_accuracy": 0.6963882446289062, + "num_tokens": 686315799.0, + "step": 26521 + }, + { + "epoch": 2.912585108719526, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.605350971221924, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.7216073274612427, + "num_tokens": 686339348.0, + "step": 26522 + }, + { + "epoch": 2.912694926422139, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.896329402923584, + "learning_rate": 1e-06, + "loss": 0.9115, + "mean_token_accuracy": 0.7343266010284424, + "num_tokens": 686360029.0, + "step": 26523 + }, + { + "epoch": 2.912804744124753, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5156736373901367, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6884220838546753, + "num_tokens": 686387739.0, + "step": 26524 + }, + { + "epoch": 2.9129145618273666, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.1147396564483643, + "learning_rate": 1e-06, + "loss": 0.9976, + "mean_token_accuracy": 0.7132443189620972, + "num_tokens": 686410635.0, + "step": 26525 + }, + { + "epoch": 2.9130243795299804, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.695298671722412, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7182846069335938, + "num_tokens": 686436426.0, + "step": 26526 + }, + { + "epoch": 2.913134197232594, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.411966323852539, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.695539653301239, + "num_tokens": 686466815.0, + "step": 26527 + }, + { + "epoch": 2.9132440149352075, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7373034954071045, + "learning_rate": 1e-06, + "loss": 1.0647, + "mean_token_accuracy": 0.6933590173721313, + "num_tokens": 686491711.0, + "step": 26528 + }, + { + "epoch": 2.913353832637821, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.704401969909668, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6958523988723755, + "num_tokens": 686515784.0, + "step": 26529 + }, + { + "epoch": 2.913463650340435, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8253531455993652, + "learning_rate": 1e-06, + "loss": 1.0589, + "mean_token_accuracy": 0.7024253606796265, + "num_tokens": 686538227.0, + "step": 26530 + }, + { + "epoch": 2.9135734680430483, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5158512592315674, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6876305341720581, + "num_tokens": 686568077.0, + "step": 26531 + }, + { + "epoch": 2.913683285745662, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4304556846618652, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7068382501602173, + "num_tokens": 686597018.0, + "step": 26532 + }, + { + "epoch": 2.913793103448276, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.245300531387329, + "learning_rate": 1e-06, + "loss": 0.9703, + "mean_token_accuracy": 0.713815450668335, + "num_tokens": 686618219.0, + "step": 26533 + }, + { + "epoch": 2.9139029211508896, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.659447193145752, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7087531089782715, + "num_tokens": 686642533.0, + "step": 26534 + }, + { + "epoch": 2.9140127388535033, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.796168088912964, + "learning_rate": 1e-06, + "loss": 1.0813, + "mean_token_accuracy": 0.6913984417915344, + "num_tokens": 686665114.0, + "step": 26535 + }, + { + "epoch": 2.9141225565561166, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.462592363357544, + "learning_rate": 1e-06, + "loss": 1.0817, + "mean_token_accuracy": 0.686418354511261, + "num_tokens": 686695160.0, + "step": 26536 + }, + { + "epoch": 2.9142323742587304, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.747227668762207, + "learning_rate": 1e-06, + "loss": 1.0544, + "mean_token_accuracy": 0.6927294731140137, + "num_tokens": 686717975.0, + "step": 26537 + }, + { + "epoch": 2.914342191961344, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.560983419418335, + "learning_rate": 1e-06, + "loss": 1.0059, + "mean_token_accuracy": 0.7116369009017944, + "num_tokens": 686744794.0, + "step": 26538 + }, + { + "epoch": 2.914452009663958, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.65938663482666, + "learning_rate": 1e-06, + "loss": 1.0702, + "mean_token_accuracy": 0.6872187256813049, + "num_tokens": 686770704.0, + "step": 26539 + }, + { + "epoch": 2.9145618273665717, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4647939205169678, + "learning_rate": 1e-06, + "loss": 0.9717, + "mean_token_accuracy": 0.7189459800720215, + "num_tokens": 686800460.0, + "step": 26540 + }, + { + "epoch": 2.914671645069185, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.34745717048645, + "learning_rate": 1e-06, + "loss": 1.1479, + "mean_token_accuracy": 0.6710458993911743, + "num_tokens": 686835550.0, + "step": 26541 + }, + { + "epoch": 2.9147814627717987, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.485128402709961, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7138018608093262, + "num_tokens": 686860874.0, + "step": 26542 + }, + { + "epoch": 2.9148912804744125, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.464921236038208, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6968676447868347, + "num_tokens": 686889331.0, + "step": 26543 + }, + { + "epoch": 2.9150010981770262, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5376904010772705, + "learning_rate": 1e-06, + "loss": 1.0216, + "mean_token_accuracy": 0.7034120559692383, + "num_tokens": 686919202.0, + "step": 26544 + }, + { + "epoch": 2.91511091587964, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.702402353286743, + "learning_rate": 1e-06, + "loss": 1.0087, + "mean_token_accuracy": 0.7024309635162354, + "num_tokens": 686942904.0, + "step": 26545 + }, + { + "epoch": 2.9152207335822533, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.4229352474212646, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7125086784362793, + "num_tokens": 686973126.0, + "step": 26546 + }, + { + "epoch": 2.915330551284867, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6531193256378174, + "learning_rate": 1e-06, + "loss": 0.9919, + "mean_token_accuracy": 0.7103704214096069, + "num_tokens": 686998311.0, + "step": 26547 + }, + { + "epoch": 2.915440368987481, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5175247192382812, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7015522718429565, + "num_tokens": 687025114.0, + "step": 26548 + }, + { + "epoch": 2.9155501866900946, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.943887710571289, + "learning_rate": 1e-06, + "loss": 0.8475, + "mean_token_accuracy": 0.737138032913208, + "num_tokens": 687050089.0, + "step": 26549 + }, + { + "epoch": 2.9156600043927083, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3519678115844727, + "learning_rate": 1e-06, + "loss": 1.0465, + "mean_token_accuracy": 0.6947164535522461, + "num_tokens": 687079672.0, + "step": 26550 + }, + { + "epoch": 2.9157698220953217, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5357744693756104, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.718044102191925, + "num_tokens": 687104830.0, + "step": 26551 + }, + { + "epoch": 2.9158796397979354, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8726701736450195, + "learning_rate": 1e-06, + "loss": 0.9325, + "mean_token_accuracy": 0.7259117364883423, + "num_tokens": 687124592.0, + "step": 26552 + }, + { + "epoch": 2.915989457500549, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 16.71117401123047, + "learning_rate": 1e-06, + "loss": 0.9476, + "mean_token_accuracy": 0.7276003360748291, + "num_tokens": 687149660.0, + "step": 26553 + }, + { + "epoch": 2.9160992752031625, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.909046173095703, + "learning_rate": 1e-06, + "loss": 1.07, + "mean_token_accuracy": 0.6903609037399292, + "num_tokens": 687173940.0, + "step": 26554 + }, + { + "epoch": 2.9162090929057767, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9971179962158203, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7033973932266235, + "num_tokens": 687206727.0, + "step": 26555 + }, + { + "epoch": 2.91631891060839, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.762054920196533, + "learning_rate": 1e-06, + "loss": 0.9815, + "mean_token_accuracy": 0.7151602506637573, + "num_tokens": 687227811.0, + "step": 26556 + }, + { + "epoch": 2.9164287283110037, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.4927961826324463, + "learning_rate": 1e-06, + "loss": 1.0195, + "mean_token_accuracy": 0.7085191607475281, + "num_tokens": 687256730.0, + "step": 26557 + }, + { + "epoch": 2.9165385460136175, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.432227373123169, + "learning_rate": 1e-06, + "loss": 0.9242, + "mean_token_accuracy": 0.7247496843338013, + "num_tokens": 687281628.0, + "step": 26558 + }, + { + "epoch": 2.916648363716231, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3961129188537598, + "learning_rate": 1e-06, + "loss": 0.9944, + "mean_token_accuracy": 0.7191004753112793, + "num_tokens": 687310682.0, + "step": 26559 + }, + { + "epoch": 2.9167581814188446, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.506434679031372, + "learning_rate": 1e-06, + "loss": 1.056, + "mean_token_accuracy": 0.6919875741004944, + "num_tokens": 687340045.0, + "step": 26560 + }, + { + "epoch": 2.9168679991214583, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4790713787078857, + "learning_rate": 1e-06, + "loss": 0.9821, + "mean_token_accuracy": 0.7135705947875977, + "num_tokens": 687365898.0, + "step": 26561 + }, + { + "epoch": 2.916977816824072, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4423811435699463, + "learning_rate": 1e-06, + "loss": 0.9414, + "mean_token_accuracy": 0.7370719909667969, + "num_tokens": 687390425.0, + "step": 26562 + }, + { + "epoch": 2.917087634526686, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.39874529838562, + "learning_rate": 1e-06, + "loss": 0.9767, + "mean_token_accuracy": 0.7145048379898071, + "num_tokens": 687412248.0, + "step": 26563 + }, + { + "epoch": 2.917197452229299, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6515936851501465, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7057467103004456, + "num_tokens": 687438042.0, + "step": 26564 + }, + { + "epoch": 2.917307269931913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.4717819690704346, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.6929932832717896, + "num_tokens": 687464699.0, + "step": 26565 + }, + { + "epoch": 2.9174170876345267, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5284230709075928, + "learning_rate": 1e-06, + "loss": 0.9832, + "mean_token_accuracy": 0.7120746374130249, + "num_tokens": 687490154.0, + "step": 26566 + }, + { + "epoch": 2.9175269053371404, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.793313980102539, + "learning_rate": 1e-06, + "loss": 0.8929, + "mean_token_accuracy": 0.7274884581565857, + "num_tokens": 687510151.0, + "step": 26567 + }, + { + "epoch": 2.917636723039754, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.6345226764678955, + "learning_rate": 1e-06, + "loss": 1.0617, + "mean_token_accuracy": 0.7025487422943115, + "num_tokens": 687534651.0, + "step": 26568 + }, + { + "epoch": 2.9177465407423675, + "ewc_loss": 1.043081283569336e-05, + "grad_norm": 2.702143907546997, + "learning_rate": 1e-06, + "loss": 1.0322, + "mean_token_accuracy": 0.6952219009399414, + "num_tokens": 687557368.0, + "step": 26569 + }, + { + "epoch": 2.9178563584449813, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9135429859161377, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7114230394363403, + "num_tokens": 687581983.0, + "step": 26570 + }, + { + "epoch": 2.917966176147595, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.396888494491577, + "learning_rate": 1e-06, + "loss": 1.0301, + "mean_token_accuracy": 0.6969621181488037, + "num_tokens": 687610019.0, + "step": 26571 + }, + { + "epoch": 2.9180759938502088, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.430682420730591, + "learning_rate": 1e-06, + "loss": 1.0726, + "mean_token_accuracy": 0.6882387399673462, + "num_tokens": 687644128.0, + "step": 26572 + }, + { + "epoch": 2.9181858115528225, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.9653289318084717, + "learning_rate": 1e-06, + "loss": 1.0893, + "mean_token_accuracy": 0.6862391233444214, + "num_tokens": 687664735.0, + "step": 26573 + }, + { + "epoch": 2.918295629255436, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.564425230026245, + "learning_rate": 1e-06, + "loss": 0.8679, + "mean_token_accuracy": 0.7393883466720581, + "num_tokens": 687687794.0, + "step": 26574 + }, + { + "epoch": 2.9184054469580496, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.653372049331665, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7012003064155579, + "num_tokens": 687711311.0, + "step": 26575 + }, + { + "epoch": 2.9185152646606634, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8651106357574463, + "learning_rate": 1e-06, + "loss": 0.964, + "mean_token_accuracy": 0.7187074422836304, + "num_tokens": 687733156.0, + "step": 26576 + }, + { + "epoch": 2.918625082363277, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.779143810272217, + "learning_rate": 1e-06, + "loss": 0.9691, + "mean_token_accuracy": 0.7135229110717773, + "num_tokens": 687755265.0, + "step": 26577 + }, + { + "epoch": 2.918734900065891, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8054211139678955, + "learning_rate": 1e-06, + "loss": 0.9766, + "mean_token_accuracy": 0.7135422229766846, + "num_tokens": 687780821.0, + "step": 26578 + }, + { + "epoch": 2.918844717768504, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7863359451293945, + "learning_rate": 1e-06, + "loss": 0.9687, + "mean_token_accuracy": 0.720732569694519, + "num_tokens": 687803688.0, + "step": 26579 + }, + { + "epoch": 2.918954535471118, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.069309949874878, + "learning_rate": 1e-06, + "loss": 0.9626, + "mean_token_accuracy": 0.7192112803459167, + "num_tokens": 687829995.0, + "step": 26580 + }, + { + "epoch": 2.9190643531737317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.468775510787964, + "learning_rate": 1e-06, + "loss": 1.0272, + "mean_token_accuracy": 0.7008090615272522, + "num_tokens": 687859682.0, + "step": 26581 + }, + { + "epoch": 2.919174170876345, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3822593688964844, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6897765398025513, + "num_tokens": 687893976.0, + "step": 26582 + }, + { + "epoch": 2.9192839885789588, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.403797149658203, + "learning_rate": 1e-06, + "loss": 0.91, + "mean_token_accuracy": 0.7289257049560547, + "num_tokens": 687922005.0, + "step": 26583 + }, + { + "epoch": 2.9193938062815725, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5390737056732178, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7147911787033081, + "num_tokens": 687946333.0, + "step": 26584 + }, + { + "epoch": 2.9195036239841863, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7900490760803223, + "learning_rate": 1e-06, + "loss": 0.914, + "mean_token_accuracy": 0.7353583574295044, + "num_tokens": 687966792.0, + "step": 26585 + }, + { + "epoch": 2.9196134416868, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8539319038391113, + "learning_rate": 1e-06, + "loss": 1.0097, + "mean_token_accuracy": 0.7118216753005981, + "num_tokens": 687993284.0, + "step": 26586 + }, + { + "epoch": 2.9197232593894134, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7511188983917236, + "learning_rate": 1e-06, + "loss": 0.9936, + "mean_token_accuracy": 0.7088872790336609, + "num_tokens": 688016354.0, + "step": 26587 + }, + { + "epoch": 2.919833077092027, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5157151222229004, + "learning_rate": 1e-06, + "loss": 0.9559, + "mean_token_accuracy": 0.7219310998916626, + "num_tokens": 688042962.0, + "step": 26588 + }, + { + "epoch": 2.919942894794641, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8095309734344482, + "learning_rate": 1e-06, + "loss": 1.0523, + "mean_token_accuracy": 0.7015079259872437, + "num_tokens": 688065254.0, + "step": 26589 + }, + { + "epoch": 2.9200527124972546, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4508748054504395, + "learning_rate": 1e-06, + "loss": 1.1078, + "mean_token_accuracy": 0.6792140007019043, + "num_tokens": 688094167.0, + "step": 26590 + }, + { + "epoch": 2.9201625301998684, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5974650382995605, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7138214111328125, + "num_tokens": 688121059.0, + "step": 26591 + }, + { + "epoch": 2.9202723479024817, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4889590740203857, + "learning_rate": 1e-06, + "loss": 1.0345, + "mean_token_accuracy": 0.705643355846405, + "num_tokens": 688148228.0, + "step": 26592 + }, + { + "epoch": 2.9203821656050954, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.054964542388916, + "learning_rate": 1e-06, + "loss": 0.9637, + "mean_token_accuracy": 0.720388650894165, + "num_tokens": 688168188.0, + "step": 26593 + }, + { + "epoch": 2.920491983307709, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.602423667907715, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7067383527755737, + "num_tokens": 688193376.0, + "step": 26594 + }, + { + "epoch": 2.920601801010323, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.432952642440796, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.7015143036842346, + "num_tokens": 688223449.0, + "step": 26595 + }, + { + "epoch": 2.9207116187129367, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.694096088409424, + "learning_rate": 1e-06, + "loss": 0.9861, + "mean_token_accuracy": 0.7160946726799011, + "num_tokens": 688248470.0, + "step": 26596 + }, + { + "epoch": 2.92082143641555, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.505204200744629, + "learning_rate": 1e-06, + "loss": 0.8957, + "mean_token_accuracy": 0.7326211333274841, + "num_tokens": 688275671.0, + "step": 26597 + }, + { + "epoch": 2.920931254118164, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.37960147857666, + "learning_rate": 1e-06, + "loss": 0.9916, + "mean_token_accuracy": 0.7095919251441956, + "num_tokens": 688305394.0, + "step": 26598 + }, + { + "epoch": 2.9210410718207775, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.999286413192749, + "learning_rate": 1e-06, + "loss": 0.9975, + "mean_token_accuracy": 0.7052406072616577, + "num_tokens": 688327261.0, + "step": 26599 + }, + { + "epoch": 2.9211508895233913, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5421743392944336, + "learning_rate": 1e-06, + "loss": 0.9576, + "mean_token_accuracy": 0.727105975151062, + "num_tokens": 688354440.0, + "step": 26600 + }, + { + "epoch": 2.921260707226005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.710099697113037, + "learning_rate": 1e-06, + "loss": 0.9921, + "mean_token_accuracy": 0.7094446420669556, + "num_tokens": 688376412.0, + "step": 26601 + }, + { + "epoch": 2.9213705249286184, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.408641815185547, + "learning_rate": 1e-06, + "loss": 1.1138, + "mean_token_accuracy": 0.6810519099235535, + "num_tokens": 688408083.0, + "step": 26602 + }, + { + "epoch": 2.921480342631232, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.653287172317505, + "learning_rate": 1e-06, + "loss": 0.914, + "mean_token_accuracy": 0.7299822568893433, + "num_tokens": 688430832.0, + "step": 26603 + }, + { + "epoch": 2.921590160333846, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8597686290740967, + "learning_rate": 1e-06, + "loss": 1.0112, + "mean_token_accuracy": 0.7055723667144775, + "num_tokens": 688455208.0, + "step": 26604 + }, + { + "epoch": 2.921699978036459, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6867475509643555, + "learning_rate": 1e-06, + "loss": 0.968, + "mean_token_accuracy": 0.7114614248275757, + "num_tokens": 688480867.0, + "step": 26605 + }, + { + "epoch": 2.9218097957390734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5052897930145264, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.6942241787910461, + "num_tokens": 688508797.0, + "step": 26606 + }, + { + "epoch": 2.9219196134416867, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.561734676361084, + "learning_rate": 1e-06, + "loss": 1.1012, + "mean_token_accuracy": 0.6858681440353394, + "num_tokens": 688539193.0, + "step": 26607 + }, + { + "epoch": 2.9220294311443005, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.3905560970306396, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.70119309425354, + "num_tokens": 688568712.0, + "step": 26608 + }, + { + "epoch": 2.9221392488469142, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7670483589172363, + "learning_rate": 1e-06, + "loss": 1.0127, + "mean_token_accuracy": 0.7043792009353638, + "num_tokens": 688592866.0, + "step": 26609 + }, + { + "epoch": 2.9222490665495275, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6254608631134033, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7152500748634338, + "num_tokens": 688617035.0, + "step": 26610 + }, + { + "epoch": 2.9223588842521413, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.507819175720215, + "learning_rate": 1e-06, + "loss": 0.9617, + "mean_token_accuracy": 0.7175651788711548, + "num_tokens": 688642534.0, + "step": 26611 + }, + { + "epoch": 2.922468701954755, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.605452060699463, + "learning_rate": 1e-06, + "loss": 1.0205, + "mean_token_accuracy": 0.7063644528388977, + "num_tokens": 688666128.0, + "step": 26612 + }, + { + "epoch": 2.922578519657369, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.784670352935791, + "learning_rate": 1e-06, + "loss": 0.8865, + "mean_token_accuracy": 0.7361409664154053, + "num_tokens": 688688505.0, + "step": 26613 + }, + { + "epoch": 2.9226883373599826, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4521234035491943, + "learning_rate": 1e-06, + "loss": 1.0299, + "mean_token_accuracy": 0.6952618360519409, + "num_tokens": 688719595.0, + "step": 26614 + }, + { + "epoch": 2.922798155062596, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.869701623916626, + "learning_rate": 1e-06, + "loss": 1.0961, + "mean_token_accuracy": 0.6923513412475586, + "num_tokens": 688741262.0, + "step": 26615 + }, + { + "epoch": 2.9229079727652096, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.605926513671875, + "learning_rate": 1e-06, + "loss": 0.9763, + "mean_token_accuracy": 0.711763858795166, + "num_tokens": 688764920.0, + "step": 26616 + }, + { + "epoch": 2.9230177904678234, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4097917079925537, + "learning_rate": 1e-06, + "loss": 1.121, + "mean_token_accuracy": 0.6844339370727539, + "num_tokens": 688796330.0, + "step": 26617 + }, + { + "epoch": 2.923127608170437, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7570114135742188, + "learning_rate": 1e-06, + "loss": 0.9202, + "mean_token_accuracy": 0.7325083017349243, + "num_tokens": 688817560.0, + "step": 26618 + }, + { + "epoch": 2.923237425873051, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6859958171844482, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7161020040512085, + "num_tokens": 688840841.0, + "step": 26619 + }, + { + "epoch": 2.9233472435756642, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.69809627532959, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7115089893341064, + "num_tokens": 688863396.0, + "step": 26620 + }, + { + "epoch": 2.923457061278278, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.825855016708374, + "learning_rate": 1e-06, + "loss": 0.945, + "mean_token_accuracy": 0.7256288528442383, + "num_tokens": 688884373.0, + "step": 26621 + }, + { + "epoch": 2.9235668789808917, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8042609691619873, + "learning_rate": 1e-06, + "loss": 0.907, + "mean_token_accuracy": 0.7381378412246704, + "num_tokens": 688906250.0, + "step": 26622 + }, + { + "epoch": 2.9236766966835055, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.904430627822876, + "learning_rate": 1e-06, + "loss": 0.9693, + "mean_token_accuracy": 0.7160640954971313, + "num_tokens": 688927619.0, + "step": 26623 + }, + { + "epoch": 2.9237865143861193, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.433056116104126, + "learning_rate": 1e-06, + "loss": 0.9568, + "mean_token_accuracy": 0.7300443649291992, + "num_tokens": 688955302.0, + "step": 26624 + }, + { + "epoch": 2.9238963320887326, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3642306327819824, + "learning_rate": 1e-06, + "loss": 1.0867, + "mean_token_accuracy": 0.6923738121986389, + "num_tokens": 688985863.0, + "step": 26625 + }, + { + "epoch": 2.9240061497913463, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.594099998474121, + "learning_rate": 1e-06, + "loss": 0.9467, + "mean_token_accuracy": 0.7215207815170288, + "num_tokens": 689013871.0, + "step": 26626 + }, + { + "epoch": 2.92411596749396, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6876413822174072, + "learning_rate": 1e-06, + "loss": 0.9962, + "mean_token_accuracy": 0.7086269855499268, + "num_tokens": 689036331.0, + "step": 26627 + }, + { + "epoch": 2.924225785196574, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4433810710906982, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7069016695022583, + "num_tokens": 689064330.0, + "step": 26628 + }, + { + "epoch": 2.9243356028991876, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7219130992889404, + "learning_rate": 1e-06, + "loss": 1.0489, + "mean_token_accuracy": 0.6967486143112183, + "num_tokens": 689089193.0, + "step": 26629 + }, + { + "epoch": 2.924445420601801, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4132769107818604, + "learning_rate": 1e-06, + "loss": 1.0131, + "mean_token_accuracy": 0.706714391708374, + "num_tokens": 689117563.0, + "step": 26630 + }, + { + "epoch": 2.9245552383044147, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8728132247924805, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7144569754600525, + "num_tokens": 689141597.0, + "step": 26631 + }, + { + "epoch": 2.9246650560070284, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5468409061431885, + "learning_rate": 1e-06, + "loss": 0.9345, + "mean_token_accuracy": 0.725660502910614, + "num_tokens": 689167430.0, + "step": 26632 + }, + { + "epoch": 2.9247748737096417, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.336486339569092, + "learning_rate": 1e-06, + "loss": 0.9595, + "mean_token_accuracy": 0.7231088876724243, + "num_tokens": 689199668.0, + "step": 26633 + }, + { + "epoch": 2.924884691412256, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.391434907913208, + "learning_rate": 1e-06, + "loss": 1.0017, + "mean_token_accuracy": 0.7070491313934326, + "num_tokens": 689232059.0, + "step": 26634 + }, + { + "epoch": 2.9249945091148692, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.566511392593384, + "learning_rate": 1e-06, + "loss": 1.0739, + "mean_token_accuracy": 0.692180871963501, + "num_tokens": 689259070.0, + "step": 26635 + }, + { + "epoch": 2.925104326817483, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.565186023712158, + "learning_rate": 1e-06, + "loss": 1.1047, + "mean_token_accuracy": 0.6849654316902161, + "num_tokens": 689284076.0, + "step": 26636 + }, + { + "epoch": 2.9252141445200968, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7314798831939697, + "learning_rate": 1e-06, + "loss": 1.0597, + "mean_token_accuracy": 0.6970930695533752, + "num_tokens": 689308431.0, + "step": 26637 + }, + { + "epoch": 2.92532396222271, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.225341796875, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.7098774909973145, + "num_tokens": 689340239.0, + "step": 26638 + }, + { + "epoch": 2.925433779925324, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 7.954233169555664, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7139593958854675, + "num_tokens": 689369272.0, + "step": 26639 + }, + { + "epoch": 2.9255435976279376, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.439687967300415, + "learning_rate": 1e-06, + "loss": 1.1009, + "mean_token_accuracy": 0.6814362406730652, + "num_tokens": 689400323.0, + "step": 26640 + }, + { + "epoch": 2.9256534153305513, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8778610229492188, + "learning_rate": 1e-06, + "loss": 0.983, + "mean_token_accuracy": 0.7133967876434326, + "num_tokens": 689422422.0, + "step": 26641 + }, + { + "epoch": 2.925763233033165, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4975409507751465, + "learning_rate": 1e-06, + "loss": 1.0823, + "mean_token_accuracy": 0.6869684457778931, + "num_tokens": 689451044.0, + "step": 26642 + }, + { + "epoch": 2.9258730507357784, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7206811904907227, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.723354160785675, + "num_tokens": 689473795.0, + "step": 26643 + }, + { + "epoch": 2.925982868438392, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.481417179107666, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.700788676738739, + "num_tokens": 689503380.0, + "step": 26644 + }, + { + "epoch": 2.926092686141006, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8160157203674316, + "learning_rate": 1e-06, + "loss": 1.0198, + "mean_token_accuracy": 0.704586386680603, + "num_tokens": 689526467.0, + "step": 26645 + }, + { + "epoch": 2.9262025038436197, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5460243225097656, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.6939904093742371, + "num_tokens": 689555347.0, + "step": 26646 + }, + { + "epoch": 2.9263123215462334, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2748863697052, + "learning_rate": 1e-06, + "loss": 1.1435, + "mean_token_accuracy": 0.6733294725418091, + "num_tokens": 689592022.0, + "step": 26647 + }, + { + "epoch": 2.9264221392488468, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8910274505615234, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7107648849487305, + "num_tokens": 689612055.0, + "step": 26648 + }, + { + "epoch": 2.9265319569514605, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.51173734664917, + "learning_rate": 1e-06, + "loss": 1.059, + "mean_token_accuracy": 0.6906063556671143, + "num_tokens": 689638227.0, + "step": 26649 + }, + { + "epoch": 2.9266417746540743, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4563777446746826, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7078344225883484, + "num_tokens": 689669711.0, + "step": 26650 + }, + { + "epoch": 2.926751592356688, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5644278526306152, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.693623423576355, + "num_tokens": 689699272.0, + "step": 26651 + }, + { + "epoch": 2.926861410059302, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.431591272354126, + "learning_rate": 1e-06, + "loss": 1.0592, + "mean_token_accuracy": 0.6935842037200928, + "num_tokens": 689729717.0, + "step": 26652 + }, + { + "epoch": 2.926971227761915, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.55226993560791, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7056623697280884, + "num_tokens": 689755552.0, + "step": 26653 + }, + { + "epoch": 2.927081045464529, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.7320926189422607, + "learning_rate": 1e-06, + "loss": 1.1172, + "mean_token_accuracy": 0.6811082363128662, + "num_tokens": 689780275.0, + "step": 26654 + }, + { + "epoch": 2.9271908631671426, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.5971195697784424, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7036378383636475, + "num_tokens": 689806057.0, + "step": 26655 + }, + { + "epoch": 2.9273006808697564, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.730518102645874, + "learning_rate": 1e-06, + "loss": 1.0025, + "mean_token_accuracy": 0.7130569815635681, + "num_tokens": 689831959.0, + "step": 26656 + }, + { + "epoch": 2.92741049857237, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.509428024291992, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7113995552062988, + "num_tokens": 689863640.0, + "step": 26657 + }, + { + "epoch": 2.9275203162749834, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5298595428466797, + "learning_rate": 1e-06, + "loss": 1.0376, + "mean_token_accuracy": 0.6991628408432007, + "num_tokens": 689892249.0, + "step": 26658 + }, + { + "epoch": 2.927630133977597, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5958216190338135, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7102030515670776, + "num_tokens": 689916482.0, + "step": 26659 + }, + { + "epoch": 2.927739951680211, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.624674081802368, + "learning_rate": 1e-06, + "loss": 0.9644, + "mean_token_accuracy": 0.7215882539749146, + "num_tokens": 689941566.0, + "step": 26660 + }, + { + "epoch": 2.9278497693828243, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.581173896789551, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7122690677642822, + "num_tokens": 689968884.0, + "step": 26661 + }, + { + "epoch": 2.927959587085438, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 7.908717632293701, + "learning_rate": 1e-06, + "loss": 1.0508, + "mean_token_accuracy": 0.6972729563713074, + "num_tokens": 689997689.0, + "step": 26662 + }, + { + "epoch": 2.9280694047880518, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8346190452575684, + "learning_rate": 1e-06, + "loss": 0.9659, + "mean_token_accuracy": 0.721070408821106, + "num_tokens": 690020953.0, + "step": 26663 + }, + { + "epoch": 2.9281792224906655, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.808236837387085, + "learning_rate": 1e-06, + "loss": 1.0513, + "mean_token_accuracy": 0.7008179426193237, + "num_tokens": 690049019.0, + "step": 26664 + }, + { + "epoch": 2.9282890401932793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.591543436050415, + "learning_rate": 1e-06, + "loss": 0.9668, + "mean_token_accuracy": 0.7160777449607849, + "num_tokens": 690075642.0, + "step": 26665 + }, + { + "epoch": 2.9283988578958926, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.331336259841919, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7045764923095703, + "num_tokens": 690107687.0, + "step": 26666 + }, + { + "epoch": 2.9285086755985064, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 3.018882989883423, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7028225660324097, + "num_tokens": 690128154.0, + "step": 26667 + }, + { + "epoch": 2.92861849330112, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.848522663116455, + "learning_rate": 1e-06, + "loss": 0.9126, + "mean_token_accuracy": 0.7372084259986877, + "num_tokens": 690149424.0, + "step": 26668 + }, + { + "epoch": 2.928728311003734, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.596912145614624, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7196975946426392, + "num_tokens": 690174474.0, + "step": 26669 + }, + { + "epoch": 2.9288381287063476, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.658531665802002, + "learning_rate": 1e-06, + "loss": 0.9927, + "mean_token_accuracy": 0.7147496938705444, + "num_tokens": 690198756.0, + "step": 26670 + }, + { + "epoch": 2.928947946408961, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.901893377304077, + "learning_rate": 1e-06, + "loss": 0.8712, + "mean_token_accuracy": 0.7435877323150635, + "num_tokens": 690219581.0, + "step": 26671 + }, + { + "epoch": 2.9290577641115747, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.483020067214966, + "learning_rate": 1e-06, + "loss": 1.0535, + "mean_token_accuracy": 0.6958299875259399, + "num_tokens": 690249638.0, + "step": 26672 + }, + { + "epoch": 2.9291675818141885, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.558522939682007, + "learning_rate": 1e-06, + "loss": 1.0911, + "mean_token_accuracy": 0.6914666891098022, + "num_tokens": 690275987.0, + "step": 26673 + }, + { + "epoch": 2.929277399516802, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.8284671306610107, + "learning_rate": 1e-06, + "loss": 0.913, + "mean_token_accuracy": 0.7331490516662598, + "num_tokens": 690295474.0, + "step": 26674 + }, + { + "epoch": 2.929387217219416, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9466731548309326, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7169009447097778, + "num_tokens": 690315944.0, + "step": 26675 + }, + { + "epoch": 2.9294970349220293, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8398284912109375, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7164457440376282, + "num_tokens": 690340715.0, + "step": 26676 + }, + { + "epoch": 2.929606852624643, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.748764991760254, + "learning_rate": 1e-06, + "loss": 0.9816, + "mean_token_accuracy": 0.7136963605880737, + "num_tokens": 690361929.0, + "step": 26677 + }, + { + "epoch": 2.929716670327257, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4366846084594727, + "learning_rate": 1e-06, + "loss": 1.0537, + "mean_token_accuracy": 0.695934534072876, + "num_tokens": 690391351.0, + "step": 26678 + }, + { + "epoch": 2.9298264880298706, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.485553741455078, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.6979033350944519, + "num_tokens": 690421413.0, + "step": 26679 + }, + { + "epoch": 2.9299363057324843, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7615466117858887, + "learning_rate": 1e-06, + "loss": 1.1093, + "mean_token_accuracy": 0.685513436794281, + "num_tokens": 690446312.0, + "step": 26680 + }, + { + "epoch": 2.9300461234350976, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.730309009552002, + "learning_rate": 1e-06, + "loss": 1.1058, + "mean_token_accuracy": 0.68113112449646, + "num_tokens": 690471830.0, + "step": 26681 + }, + { + "epoch": 2.9301559411377114, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.521881103515625, + "learning_rate": 1e-06, + "loss": 1.0737, + "mean_token_accuracy": 0.6915503740310669, + "num_tokens": 690501404.0, + "step": 26682 + }, + { + "epoch": 2.930265758840325, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8116891384124756, + "learning_rate": 1e-06, + "loss": 1.0203, + "mean_token_accuracy": 0.7074620127677917, + "num_tokens": 690524742.0, + "step": 26683 + }, + { + "epoch": 2.9303755765429385, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.795253276824951, + "learning_rate": 1e-06, + "loss": 0.8772, + "mean_token_accuracy": 0.7492983341217041, + "num_tokens": 690546333.0, + "step": 26684 + }, + { + "epoch": 2.9304853942455527, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6808342933654785, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7244172096252441, + "num_tokens": 690569697.0, + "step": 26685 + }, + { + "epoch": 2.930595211948166, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.6276934146881104, + "learning_rate": 1e-06, + "loss": 0.8427, + "mean_token_accuracy": 0.7494649887084961, + "num_tokens": 690593988.0, + "step": 26686 + }, + { + "epoch": 2.9307050296507797, + "ewc_loss": 1.049041748046875e-05, + "grad_norm": 2.686917304992676, + "learning_rate": 1e-06, + "loss": 1.0437, + "mean_token_accuracy": 0.7004185914993286, + "num_tokens": 690618230.0, + "step": 26687 + }, + { + "epoch": 2.9308148473533935, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8506860733032227, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7183433175086975, + "num_tokens": 690639264.0, + "step": 26688 + }, + { + "epoch": 2.930924665056007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.646878957748413, + "learning_rate": 1e-06, + "loss": 0.9269, + "mean_token_accuracy": 0.7298945784568787, + "num_tokens": 690661980.0, + "step": 26689 + }, + { + "epoch": 2.9310344827586206, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.158278465270996, + "learning_rate": 1e-06, + "loss": 0.9819, + "mean_token_accuracy": 0.714292049407959, + "num_tokens": 690685725.0, + "step": 26690 + }, + { + "epoch": 2.9311443004612343, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6977314949035645, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7064249515533447, + "num_tokens": 690710122.0, + "step": 26691 + }, + { + "epoch": 2.931254118163848, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.592202663421631, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7109590172767639, + "num_tokens": 690737360.0, + "step": 26692 + }, + { + "epoch": 2.931363935866462, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7758805751800537, + "learning_rate": 1e-06, + "loss": 1.0118, + "mean_token_accuracy": 0.7057336568832397, + "num_tokens": 690758151.0, + "step": 26693 + }, + { + "epoch": 2.931473753569075, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5909478664398193, + "learning_rate": 1e-06, + "loss": 0.8638, + "mean_token_accuracy": 0.742202639579773, + "num_tokens": 690782626.0, + "step": 26694 + }, + { + "epoch": 2.931583571271689, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7825236320495605, + "learning_rate": 1e-06, + "loss": 0.9402, + "mean_token_accuracy": 0.722576379776001, + "num_tokens": 690804268.0, + "step": 26695 + }, + { + "epoch": 2.9316933889743026, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3997886180877686, + "learning_rate": 1e-06, + "loss": 1.0323, + "mean_token_accuracy": 0.7019073963165283, + "num_tokens": 690833758.0, + "step": 26696 + }, + { + "epoch": 2.9318032066769164, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3113062381744385, + "learning_rate": 1e-06, + "loss": 1.0909, + "mean_token_accuracy": 0.6870613098144531, + "num_tokens": 690867639.0, + "step": 26697 + }, + { + "epoch": 2.93191302437953, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.2501494884490967, + "learning_rate": 1e-06, + "loss": 1.0607, + "mean_token_accuracy": 0.6923555135726929, + "num_tokens": 690902197.0, + "step": 26698 + }, + { + "epoch": 2.9320228420821435, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5049686431884766, + "learning_rate": 1e-06, + "loss": 1.077, + "mean_token_accuracy": 0.690526008605957, + "num_tokens": 690932496.0, + "step": 26699 + }, + { + "epoch": 2.9321326597847572, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.788397789001465, + "learning_rate": 1e-06, + "loss": 0.9966, + "mean_token_accuracy": 0.7122842073440552, + "num_tokens": 690956064.0, + "step": 26700 + }, + { + "epoch": 2.932242477487371, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.654071092605591, + "learning_rate": 1e-06, + "loss": 1.0279, + "mean_token_accuracy": 0.6995593309402466, + "num_tokens": 690980200.0, + "step": 26701 + }, + { + "epoch": 2.9323522951899847, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6405279636383057, + "learning_rate": 1e-06, + "loss": 1.0226, + "mean_token_accuracy": 0.7132058143615723, + "num_tokens": 691003199.0, + "step": 26702 + }, + { + "epoch": 2.9324621128925985, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.727654457092285, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.7092961072921753, + "num_tokens": 691025235.0, + "step": 26703 + }, + { + "epoch": 2.932571930595212, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6245265007019043, + "learning_rate": 1e-06, + "loss": 0.9011, + "mean_token_accuracy": 0.7357525825500488, + "num_tokens": 691049426.0, + "step": 26704 + }, + { + "epoch": 2.9326817482978256, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.385070562362671, + "learning_rate": 1e-06, + "loss": 1.0581, + "mean_token_accuracy": 0.70540452003479, + "num_tokens": 691080489.0, + "step": 26705 + }, + { + "epoch": 2.9327915660004393, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3629555702209473, + "learning_rate": 1e-06, + "loss": 1.0601, + "mean_token_accuracy": 0.7002691030502319, + "num_tokens": 691109836.0, + "step": 26706 + }, + { + "epoch": 2.932901383703053, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7035410404205322, + "learning_rate": 1e-06, + "loss": 1.0172, + "mean_token_accuracy": 0.7106581926345825, + "num_tokens": 691133883.0, + "step": 26707 + }, + { + "epoch": 2.933011201405667, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5498085021972656, + "learning_rate": 1e-06, + "loss": 0.993, + "mean_token_accuracy": 0.7089577913284302, + "num_tokens": 691159481.0, + "step": 26708 + }, + { + "epoch": 2.93312101910828, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0123682022094727, + "learning_rate": 1e-06, + "loss": 0.9591, + "mean_token_accuracy": 0.7127386331558228, + "num_tokens": 691178012.0, + "step": 26709 + }, + { + "epoch": 2.933230836810894, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8097317218780518, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7047755718231201, + "num_tokens": 691200489.0, + "step": 26710 + }, + { + "epoch": 2.9333406545135077, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9731032848358154, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7207045555114746, + "num_tokens": 691220284.0, + "step": 26711 + }, + { + "epoch": 2.933450472216121, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6028831005096436, + "learning_rate": 1e-06, + "loss": 0.9236, + "mean_token_accuracy": 0.7348068952560425, + "num_tokens": 691247468.0, + "step": 26712 + }, + { + "epoch": 2.9335602899187347, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6717336177825928, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7039260268211365, + "num_tokens": 691274301.0, + "step": 26713 + }, + { + "epoch": 2.9336701076213485, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8567657470703125, + "learning_rate": 1e-06, + "loss": 0.9397, + "mean_token_accuracy": 0.7171143293380737, + "num_tokens": 691294714.0, + "step": 26714 + }, + { + "epoch": 2.9337799253239623, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.596407175064087, + "learning_rate": 1e-06, + "loss": 1.0498, + "mean_token_accuracy": 0.6923177242279053, + "num_tokens": 691320042.0, + "step": 26715 + }, + { + "epoch": 2.933889743026576, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.140134811401367, + "learning_rate": 1e-06, + "loss": 0.9234, + "mean_token_accuracy": 0.7309427261352539, + "num_tokens": 691336829.0, + "step": 26716 + }, + { + "epoch": 2.9339995607291893, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.42571759223938, + "learning_rate": 1e-06, + "loss": 0.95, + "mean_token_accuracy": 0.7302380800247192, + "num_tokens": 691366283.0, + "step": 26717 + }, + { + "epoch": 2.934109378431803, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.768186569213867, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7211580276489258, + "num_tokens": 691390445.0, + "step": 26718 + }, + { + "epoch": 2.934219196134417, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4973416328430176, + "learning_rate": 1e-06, + "loss": 0.9633, + "mean_token_accuracy": 0.717011034488678, + "num_tokens": 691416398.0, + "step": 26719 + }, + { + "epoch": 2.9343290138370306, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6277661323547363, + "learning_rate": 1e-06, + "loss": 1.0777, + "mean_token_accuracy": 0.6881729364395142, + "num_tokens": 691451529.0, + "step": 26720 + }, + { + "epoch": 2.9344388315396444, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.561657667160034, + "learning_rate": 1e-06, + "loss": 0.9681, + "mean_token_accuracy": 0.7170240879058838, + "num_tokens": 691476310.0, + "step": 26721 + }, + { + "epoch": 2.9345486492422577, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7895898818969727, + "learning_rate": 1e-06, + "loss": 0.9583, + "mean_token_accuracy": 0.7250659465789795, + "num_tokens": 691499514.0, + "step": 26722 + }, + { + "epoch": 2.9346584669448714, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7708652019500732, + "learning_rate": 1e-06, + "loss": 1.0026, + "mean_token_accuracy": 0.7025219202041626, + "num_tokens": 691521563.0, + "step": 26723 + }, + { + "epoch": 2.934768284647485, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7335171699523926, + "learning_rate": 1e-06, + "loss": 1.0275, + "mean_token_accuracy": 0.7052178382873535, + "num_tokens": 691545425.0, + "step": 26724 + }, + { + "epoch": 2.934878102350099, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5623080730438232, + "learning_rate": 1e-06, + "loss": 1.0885, + "mean_token_accuracy": 0.6803478598594666, + "num_tokens": 691571772.0, + "step": 26725 + }, + { + "epoch": 2.9349879200527127, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.645334005355835, + "learning_rate": 1e-06, + "loss": 1.0616, + "mean_token_accuracy": 0.6910254955291748, + "num_tokens": 691597828.0, + "step": 26726 + }, + { + "epoch": 2.935097737755326, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.444990634918213, + "learning_rate": 1e-06, + "loss": 1.1145, + "mean_token_accuracy": 0.6806262135505676, + "num_tokens": 691627954.0, + "step": 26727 + }, + { + "epoch": 2.9352075554579398, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6172585487365723, + "learning_rate": 1e-06, + "loss": 1.1072, + "mean_token_accuracy": 0.677600622177124, + "num_tokens": 691655678.0, + "step": 26728 + }, + { + "epoch": 2.9353173731605535, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.568230390548706, + "learning_rate": 1e-06, + "loss": 1.0628, + "mean_token_accuracy": 0.7030279636383057, + "num_tokens": 691681956.0, + "step": 26729 + }, + { + "epoch": 2.9354271908631673, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.501394271850586, + "learning_rate": 1e-06, + "loss": 1.0164, + "mean_token_accuracy": 0.712816059589386, + "num_tokens": 691708417.0, + "step": 26730 + }, + { + "epoch": 2.935537008565781, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9064674377441406, + "learning_rate": 1e-06, + "loss": 1.035, + "mean_token_accuracy": 0.7061237692832947, + "num_tokens": 691728792.0, + "step": 26731 + }, + { + "epoch": 2.9356468262683943, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8106627464294434, + "learning_rate": 1e-06, + "loss": 0.9329, + "mean_token_accuracy": 0.7230293154716492, + "num_tokens": 691751611.0, + "step": 26732 + }, + { + "epoch": 2.935756643971008, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.1292896270751953, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.7039304375648499, + "num_tokens": 691773539.0, + "step": 26733 + }, + { + "epoch": 2.935866461673622, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6745364665985107, + "learning_rate": 1e-06, + "loss": 0.9722, + "mean_token_accuracy": 0.7156614065170288, + "num_tokens": 691796543.0, + "step": 26734 + }, + { + "epoch": 2.935976279376235, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4959933757781982, + "learning_rate": 1e-06, + "loss": 0.9892, + "mean_token_accuracy": 0.7157590389251709, + "num_tokens": 691824616.0, + "step": 26735 + }, + { + "epoch": 2.9360860970788494, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7118008136749268, + "learning_rate": 1e-06, + "loss": 1.0609, + "mean_token_accuracy": 0.6903679370880127, + "num_tokens": 691847901.0, + "step": 26736 + }, + { + "epoch": 2.9361959147814627, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4476141929626465, + "learning_rate": 1e-06, + "loss": 0.9718, + "mean_token_accuracy": 0.722412645816803, + "num_tokens": 691875459.0, + "step": 26737 + }, + { + "epoch": 2.9363057324840764, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.587062120437622, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.7007490992546082, + "num_tokens": 691905817.0, + "step": 26738 + }, + { + "epoch": 2.93641555018669, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.407715082168579, + "learning_rate": 1e-06, + "loss": 0.9948, + "mean_token_accuracy": 0.7193295359611511, + "num_tokens": 691932060.0, + "step": 26739 + }, + { + "epoch": 2.9365253678893035, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.677452325820923, + "learning_rate": 1e-06, + "loss": 1.0593, + "mean_token_accuracy": 0.6945289373397827, + "num_tokens": 691959423.0, + "step": 26740 + }, + { + "epoch": 2.9366351855919173, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.433056116104126, + "learning_rate": 1e-06, + "loss": 1.0194, + "mean_token_accuracy": 0.7111376523971558, + "num_tokens": 691989589.0, + "step": 26741 + }, + { + "epoch": 2.936745003294531, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4458577632904053, + "learning_rate": 1e-06, + "loss": 0.9829, + "mean_token_accuracy": 0.7107447385787964, + "num_tokens": 692018265.0, + "step": 26742 + }, + { + "epoch": 2.936854820997145, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6376819610595703, + "learning_rate": 1e-06, + "loss": 1.0229, + "mean_token_accuracy": 0.7069774866104126, + "num_tokens": 692044060.0, + "step": 26743 + }, + { + "epoch": 2.9369646386997585, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4820871353149414, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7130463719367981, + "num_tokens": 692071309.0, + "step": 26744 + }, + { + "epoch": 2.937074456402372, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.75114107131958, + "learning_rate": 1e-06, + "loss": 0.9489, + "mean_token_accuracy": 0.727042019367218, + "num_tokens": 692095403.0, + "step": 26745 + }, + { + "epoch": 2.9371842741049856, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6539251804351807, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.6964620351791382, + "num_tokens": 692120477.0, + "step": 26746 + }, + { + "epoch": 2.9372940918075994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5036685466766357, + "learning_rate": 1e-06, + "loss": 1.007, + "mean_token_accuracy": 0.7022799253463745, + "num_tokens": 692147261.0, + "step": 26747 + }, + { + "epoch": 2.937403909510213, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7398345470428467, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6961637139320374, + "num_tokens": 692170175.0, + "step": 26748 + }, + { + "epoch": 2.937513727212827, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.197977304458618, + "learning_rate": 1e-06, + "loss": 1.0389, + "mean_token_accuracy": 0.6951103806495667, + "num_tokens": 692207159.0, + "step": 26749 + }, + { + "epoch": 2.93762354491544, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8212432861328125, + "learning_rate": 1e-06, + "loss": 0.9621, + "mean_token_accuracy": 0.7241588830947876, + "num_tokens": 692228771.0, + "step": 26750 + }, + { + "epoch": 2.937733362618054, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6963798999786377, + "learning_rate": 1e-06, + "loss": 1.0089, + "mean_token_accuracy": 0.7012259364128113, + "num_tokens": 692252062.0, + "step": 26751 + }, + { + "epoch": 2.9378431803206677, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.811495065689087, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7178584337234497, + "num_tokens": 692276146.0, + "step": 26752 + }, + { + "epoch": 2.9379529980232815, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5942063331604004, + "learning_rate": 1e-06, + "loss": 0.9884, + "mean_token_accuracy": 0.7137244939804077, + "num_tokens": 692302206.0, + "step": 26753 + }, + { + "epoch": 2.9380628157258952, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7119381427764893, + "learning_rate": 1e-06, + "loss": 0.9611, + "mean_token_accuracy": 0.7115814685821533, + "num_tokens": 692325652.0, + "step": 26754 + }, + { + "epoch": 2.9381726334285085, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4871838092803955, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.6972331404685974, + "num_tokens": 692355280.0, + "step": 26755 + }, + { + "epoch": 2.9382824511311223, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6364870071411133, + "learning_rate": 1e-06, + "loss": 1.0385, + "mean_token_accuracy": 0.703126072883606, + "num_tokens": 692379783.0, + "step": 26756 + }, + { + "epoch": 2.938392268833736, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7580478191375732, + "learning_rate": 1e-06, + "loss": 0.9923, + "mean_token_accuracy": 0.7127677798271179, + "num_tokens": 692402919.0, + "step": 26757 + }, + { + "epoch": 2.93850208653635, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.501528739929199, + "learning_rate": 1e-06, + "loss": 0.9733, + "mean_token_accuracy": 0.7166225910186768, + "num_tokens": 692429074.0, + "step": 26758 + }, + { + "epoch": 2.9386119042389636, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.767179250717163, + "learning_rate": 1e-06, + "loss": 1.0051, + "mean_token_accuracy": 0.7075097560882568, + "num_tokens": 692452534.0, + "step": 26759 + }, + { + "epoch": 2.938721721941577, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.676605463027954, + "learning_rate": 1e-06, + "loss": 0.9792, + "mean_token_accuracy": 0.7146612405776978, + "num_tokens": 692477274.0, + "step": 26760 + }, + { + "epoch": 2.9388315396441906, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.956915855407715, + "learning_rate": 1e-06, + "loss": 0.9749, + "mean_token_accuracy": 0.7164894342422485, + "num_tokens": 692500243.0, + "step": 26761 + }, + { + "epoch": 2.9389413573468044, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.656461715698242, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7056871056556702, + "num_tokens": 692523281.0, + "step": 26762 + }, + { + "epoch": 2.9390511750494177, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.728827953338623, + "learning_rate": 1e-06, + "loss": 0.9771, + "mean_token_accuracy": 0.7130910158157349, + "num_tokens": 692550813.0, + "step": 26763 + }, + { + "epoch": 2.9391609927520315, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9640634059906006, + "learning_rate": 1e-06, + "loss": 1.0037, + "mean_token_accuracy": 0.7136085629463196, + "num_tokens": 692571744.0, + "step": 26764 + }, + { + "epoch": 2.939270810454645, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.699228048324585, + "learning_rate": 1e-06, + "loss": 0.99, + "mean_token_accuracy": 0.7119413018226624, + "num_tokens": 692594231.0, + "step": 26765 + }, + { + "epoch": 2.939380628157259, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.956920862197876, + "learning_rate": 1e-06, + "loss": 0.9631, + "mean_token_accuracy": 0.7153996229171753, + "num_tokens": 692614966.0, + "step": 26766 + }, + { + "epoch": 2.9394904458598727, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7446649074554443, + "learning_rate": 1e-06, + "loss": 0.9938, + "mean_token_accuracy": 0.7105662822723389, + "num_tokens": 692638923.0, + "step": 26767 + }, + { + "epoch": 2.939600263562486, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.619127035140991, + "learning_rate": 1e-06, + "loss": 1.0668, + "mean_token_accuracy": 0.6969394087791443, + "num_tokens": 692664177.0, + "step": 26768 + }, + { + "epoch": 2.9397100812651, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6582539081573486, + "learning_rate": 1e-06, + "loss": 0.9808, + "mean_token_accuracy": 0.7160660028457642, + "num_tokens": 692686427.0, + "step": 26769 + }, + { + "epoch": 2.9398198989677136, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.497633934020996, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.7016463279724121, + "num_tokens": 692714060.0, + "step": 26770 + }, + { + "epoch": 2.9399297166703273, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.855865716934204, + "learning_rate": 1e-06, + "loss": 0.8266, + "mean_token_accuracy": 0.7524138689041138, + "num_tokens": 692735925.0, + "step": 26771 + }, + { + "epoch": 2.940039534372941, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9259448051452637, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7061426639556885, + "num_tokens": 692755877.0, + "step": 26772 + }, + { + "epoch": 2.9401493520755544, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7393460273742676, + "learning_rate": 1e-06, + "loss": 1.0828, + "mean_token_accuracy": 0.6903315782546997, + "num_tokens": 692779742.0, + "step": 26773 + }, + { + "epoch": 2.940259169778168, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5830986499786377, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6976467370986938, + "num_tokens": 692807891.0, + "step": 26774 + }, + { + "epoch": 2.940368987480782, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 3.0098211765289307, + "learning_rate": 1e-06, + "loss": 0.9953, + "mean_token_accuracy": 0.7084916234016418, + "num_tokens": 692827979.0, + "step": 26775 + }, + { + "epoch": 2.9404788051833957, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6594960689544678, + "learning_rate": 1e-06, + "loss": 1.0965, + "mean_token_accuracy": 0.682911217212677, + "num_tokens": 692857900.0, + "step": 26776 + }, + { + "epoch": 2.9405886228860094, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.671705722808838, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.69179368019104, + "num_tokens": 692887200.0, + "step": 26777 + }, + { + "epoch": 2.9406984405886227, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.493269443511963, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.714077889919281, + "num_tokens": 692916475.0, + "step": 26778 + }, + { + "epoch": 2.9408082582912365, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.706899404525757, + "learning_rate": 1e-06, + "loss": 1.0412, + "mean_token_accuracy": 0.7015058994293213, + "num_tokens": 692941391.0, + "step": 26779 + }, + { + "epoch": 2.9409180759938502, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5188355445861816, + "learning_rate": 1e-06, + "loss": 0.9609, + "mean_token_accuracy": 0.7149474620819092, + "num_tokens": 692965948.0, + "step": 26780 + }, + { + "epoch": 2.941027893696464, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6674652099609375, + "learning_rate": 1e-06, + "loss": 0.9846, + "mean_token_accuracy": 0.710848867893219, + "num_tokens": 692988724.0, + "step": 26781 + }, + { + "epoch": 2.9411377113990778, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7805423736572266, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7055628895759583, + "num_tokens": 693010367.0, + "step": 26782 + }, + { + "epoch": 2.941247529101691, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.536212205886841, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7166275382041931, + "num_tokens": 693036425.0, + "step": 26783 + }, + { + "epoch": 2.941357346804305, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 4.710778713226318, + "learning_rate": 1e-06, + "loss": 0.9469, + "mean_token_accuracy": 0.7160677313804626, + "num_tokens": 693055801.0, + "step": 26784 + }, + { + "epoch": 2.9414671645069186, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.090487241744995, + "learning_rate": 1e-06, + "loss": 1.0149, + "mean_token_accuracy": 0.7107260227203369, + "num_tokens": 693076977.0, + "step": 26785 + }, + { + "epoch": 2.9415769822095323, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5163347721099854, + "learning_rate": 1e-06, + "loss": 1.0675, + "mean_token_accuracy": 0.6908317804336548, + "num_tokens": 693105603.0, + "step": 26786 + }, + { + "epoch": 2.941686799912146, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5057568550109863, + "learning_rate": 1e-06, + "loss": 1.0021, + "mean_token_accuracy": 0.7149802446365356, + "num_tokens": 693132176.0, + "step": 26787 + }, + { + "epoch": 2.9417966176147594, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.509753942489624, + "learning_rate": 1e-06, + "loss": 0.9524, + "mean_token_accuracy": 0.7199848890304565, + "num_tokens": 693158898.0, + "step": 26788 + }, + { + "epoch": 2.941906435317373, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6247148513793945, + "learning_rate": 1e-06, + "loss": 0.9776, + "mean_token_accuracy": 0.7217273712158203, + "num_tokens": 693181428.0, + "step": 26789 + }, + { + "epoch": 2.942016253019987, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5186336040496826, + "learning_rate": 1e-06, + "loss": 1.0263, + "mean_token_accuracy": 0.7067428827285767, + "num_tokens": 693209373.0, + "step": 26790 + }, + { + "epoch": 2.9421260707226002, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6783010959625244, + "learning_rate": 1e-06, + "loss": 0.9647, + "mean_token_accuracy": 0.7246925830841064, + "num_tokens": 693233344.0, + "step": 26791 + }, + { + "epoch": 2.942235888425214, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.735158681869507, + "learning_rate": 1e-06, + "loss": 0.9757, + "mean_token_accuracy": 0.7114982604980469, + "num_tokens": 693256521.0, + "step": 26792 + }, + { + "epoch": 2.9423457061278278, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7395877838134766, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.7123859524726868, + "num_tokens": 693282044.0, + "step": 26793 + }, + { + "epoch": 2.9424555238304415, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.568676471710205, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7102925777435303, + "num_tokens": 693306197.0, + "step": 26794 + }, + { + "epoch": 2.9425653415330553, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5175158977508545, + "learning_rate": 1e-06, + "loss": 1.0605, + "mean_token_accuracy": 0.6895272731781006, + "num_tokens": 693333976.0, + "step": 26795 + }, + { + "epoch": 2.9426751592356686, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.729778528213501, + "learning_rate": 1e-06, + "loss": 0.974, + "mean_token_accuracy": 0.7172890305519104, + "num_tokens": 693355923.0, + "step": 26796 + }, + { + "epoch": 2.9427849769382823, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8158762454986572, + "learning_rate": 1e-06, + "loss": 1.0346, + "mean_token_accuracy": 0.699131429195404, + "num_tokens": 693383526.0, + "step": 26797 + }, + { + "epoch": 2.942894794640896, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.642024040222168, + "learning_rate": 1e-06, + "loss": 1.0509, + "mean_token_accuracy": 0.7011740207672119, + "num_tokens": 693408827.0, + "step": 26798 + }, + { + "epoch": 2.94300461234351, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5389926433563232, + "learning_rate": 1e-06, + "loss": 0.9987, + "mean_token_accuracy": 0.7128790616989136, + "num_tokens": 693434950.0, + "step": 26799 + }, + { + "epoch": 2.9431144300461236, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5366179943084717, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7094335556030273, + "num_tokens": 693461097.0, + "step": 26800 + }, + { + "epoch": 2.943224247748737, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5749967098236084, + "learning_rate": 1e-06, + "loss": 1.0523, + "mean_token_accuracy": 0.698341965675354, + "num_tokens": 693486772.0, + "step": 26801 + }, + { + "epoch": 2.9433340654513507, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.423318862915039, + "learning_rate": 1e-06, + "loss": 0.9979, + "mean_token_accuracy": 0.7001835703849792, + "num_tokens": 693514765.0, + "step": 26802 + }, + { + "epoch": 2.9434438831539644, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7174153327941895, + "learning_rate": 1e-06, + "loss": 1.0784, + "mean_token_accuracy": 0.6966944932937622, + "num_tokens": 693538919.0, + "step": 26803 + }, + { + "epoch": 2.943553700856578, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.621408700942993, + "learning_rate": 1e-06, + "loss": 0.9083, + "mean_token_accuracy": 0.7291703224182129, + "num_tokens": 693560016.0, + "step": 26804 + }, + { + "epoch": 2.943663518559192, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.528362274169922, + "learning_rate": 1e-06, + "loss": 1.0451, + "mean_token_accuracy": 0.6978127956390381, + "num_tokens": 693590642.0, + "step": 26805 + }, + { + "epoch": 2.9437733362618053, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9672813415527344, + "learning_rate": 1e-06, + "loss": 1.0114, + "mean_token_accuracy": 0.7039567232131958, + "num_tokens": 693614369.0, + "step": 26806 + }, + { + "epoch": 2.943883153964419, + "ewc_loss": 1.1086463928222656e-05, + "grad_norm": 80.53755950927734, + "learning_rate": 1e-06, + "loss": 1.026, + "mean_token_accuracy": 0.7056799530982971, + "num_tokens": 693643798.0, + "step": 26807 + }, + { + "epoch": 2.9439929716670328, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.41805362701416, + "learning_rate": 1e-06, + "loss": 1.0896, + "mean_token_accuracy": 0.6878988146781921, + "num_tokens": 693676593.0, + "step": 26808 + }, + { + "epoch": 2.9441027893696465, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9268569946289062, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7195420265197754, + "num_tokens": 693698578.0, + "step": 26809 + }, + { + "epoch": 2.9442126070722603, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3276426792144775, + "learning_rate": 1e-06, + "loss": 0.9941, + "mean_token_accuracy": 0.709163248538971, + "num_tokens": 693728855.0, + "step": 26810 + }, + { + "epoch": 2.9443224247748736, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3982791900634766, + "learning_rate": 1e-06, + "loss": 1.1098, + "mean_token_accuracy": 0.6797206401824951, + "num_tokens": 693757497.0, + "step": 26811 + }, + { + "epoch": 2.9444322424774874, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7050516605377197, + "learning_rate": 1e-06, + "loss": 0.9683, + "mean_token_accuracy": 0.7155418992042542, + "num_tokens": 693780563.0, + "step": 26812 + }, + { + "epoch": 2.944542060180101, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.467113971710205, + "learning_rate": 1e-06, + "loss": 1.0314, + "mean_token_accuracy": 0.697756290435791, + "num_tokens": 693807523.0, + "step": 26813 + }, + { + "epoch": 2.9446518778827144, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.732151508331299, + "learning_rate": 1e-06, + "loss": 1.0449, + "mean_token_accuracy": 0.6969482898712158, + "num_tokens": 693831195.0, + "step": 26814 + }, + { + "epoch": 2.9447616955853286, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.552492141723633, + "learning_rate": 1e-06, + "loss": 1.1273, + "mean_token_accuracy": 0.6867214441299438, + "num_tokens": 693860382.0, + "step": 26815 + }, + { + "epoch": 2.944871513287942, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4934566020965576, + "learning_rate": 1e-06, + "loss": 1.0482, + "mean_token_accuracy": 0.7010467052459717, + "num_tokens": 693887384.0, + "step": 26816 + }, + { + "epoch": 2.9449813309905557, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.624739170074463, + "learning_rate": 1e-06, + "loss": 0.9529, + "mean_token_accuracy": 0.7167642116546631, + "num_tokens": 693910703.0, + "step": 26817 + }, + { + "epoch": 2.9450911486931695, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.66123628616333, + "learning_rate": 1e-06, + "loss": 1.0747, + "mean_token_accuracy": 0.6915447115898132, + "num_tokens": 693935462.0, + "step": 26818 + }, + { + "epoch": 2.9452009663957828, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7262179851531982, + "learning_rate": 1e-06, + "loss": 0.9247, + "mean_token_accuracy": 0.7227233648300171, + "num_tokens": 693956186.0, + "step": 26819 + }, + { + "epoch": 2.9453107840983965, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4795761108398438, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7054353952407837, + "num_tokens": 693985511.0, + "step": 26820 + }, + { + "epoch": 2.9454206018010103, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7366530895233154, + "learning_rate": 1e-06, + "loss": 0.8996, + "mean_token_accuracy": 0.7286032438278198, + "num_tokens": 694008428.0, + "step": 26821 + }, + { + "epoch": 2.945530419503624, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.549163341522217, + "learning_rate": 1e-06, + "loss": 1.0174, + "mean_token_accuracy": 0.7123136520385742, + "num_tokens": 694035928.0, + "step": 26822 + }, + { + "epoch": 2.945640237206238, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7475950717926025, + "learning_rate": 1e-06, + "loss": 1.058, + "mean_token_accuracy": 0.6912536025047302, + "num_tokens": 694061531.0, + "step": 26823 + }, + { + "epoch": 2.945750054908851, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7937779426574707, + "learning_rate": 1e-06, + "loss": 1.0207, + "mean_token_accuracy": 0.7060284614562988, + "num_tokens": 694084998.0, + "step": 26824 + }, + { + "epoch": 2.945859872611465, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8255343437194824, + "learning_rate": 1e-06, + "loss": 0.9724, + "mean_token_accuracy": 0.7113008499145508, + "num_tokens": 694106285.0, + "step": 26825 + }, + { + "epoch": 2.9459696903140786, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.420457124710083, + "learning_rate": 1e-06, + "loss": 1.0271, + "mean_token_accuracy": 0.7015395164489746, + "num_tokens": 694135134.0, + "step": 26826 + }, + { + "epoch": 2.9460795080166924, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.680103063583374, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7232235074043274, + "num_tokens": 694160220.0, + "step": 26827 + }, + { + "epoch": 2.946189325719306, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.908193826675415, + "learning_rate": 1e-06, + "loss": 1.0224, + "mean_token_accuracy": 0.7027522921562195, + "num_tokens": 694181966.0, + "step": 26828 + }, + { + "epoch": 2.9462991434219195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.572453022003174, + "learning_rate": 1e-06, + "loss": 0.9925, + "mean_token_accuracy": 0.7096079587936401, + "num_tokens": 694207810.0, + "step": 26829 + }, + { + "epoch": 2.946408961124533, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.338275671005249, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.7013624310493469, + "num_tokens": 694238163.0, + "step": 26830 + }, + { + "epoch": 2.946518778827147, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.950033187866211, + "learning_rate": 1e-06, + "loss": 0.94, + "mean_token_accuracy": 0.7263307571411133, + "num_tokens": 694260657.0, + "step": 26831 + }, + { + "epoch": 2.9466285965297607, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.394041061401367, + "learning_rate": 1e-06, + "loss": 1.0814, + "mean_token_accuracy": 0.6844398975372314, + "num_tokens": 694290003.0, + "step": 26832 + }, + { + "epoch": 2.9467384142323745, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5743350982666016, + "learning_rate": 1e-06, + "loss": 1.038, + "mean_token_accuracy": 0.7073792815208435, + "num_tokens": 694315162.0, + "step": 26833 + }, + { + "epoch": 2.946848231934988, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4984118938446045, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7088927030563354, + "num_tokens": 694342414.0, + "step": 26834 + }, + { + "epoch": 2.9469580496376016, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.671008825302124, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7260458469390869, + "num_tokens": 694367724.0, + "step": 26835 + }, + { + "epoch": 2.9470678673402153, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8058245182037354, + "learning_rate": 1e-06, + "loss": 0.9322, + "mean_token_accuracy": 0.7281087636947632, + "num_tokens": 694388463.0, + "step": 26836 + }, + { + "epoch": 2.947177685042829, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4485065937042236, + "learning_rate": 1e-06, + "loss": 0.9468, + "mean_token_accuracy": 0.7194018363952637, + "num_tokens": 694415520.0, + "step": 26837 + }, + { + "epoch": 2.947287502745443, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.865738868713379, + "learning_rate": 1e-06, + "loss": 1.0472, + "mean_token_accuracy": 0.6960231065750122, + "num_tokens": 694438833.0, + "step": 26838 + }, + { + "epoch": 2.947397320448056, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6405515670776367, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7226237058639526, + "num_tokens": 694466746.0, + "step": 26839 + }, + { + "epoch": 2.94750713815067, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.919523000717163, + "learning_rate": 1e-06, + "loss": 0.8929, + "mean_token_accuracy": 0.7331907749176025, + "num_tokens": 694488620.0, + "step": 26840 + }, + { + "epoch": 2.9476169558532836, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8093738555908203, + "learning_rate": 1e-06, + "loss": 1.0249, + "mean_token_accuracy": 0.6987753510475159, + "num_tokens": 694513493.0, + "step": 26841 + }, + { + "epoch": 2.947726773555897, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.681931257247925, + "learning_rate": 1e-06, + "loss": 0.9894, + "mean_token_accuracy": 0.7168457508087158, + "num_tokens": 694535953.0, + "step": 26842 + }, + { + "epoch": 2.9478365912585107, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5543899536132812, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7195146083831787, + "num_tokens": 694563545.0, + "step": 26843 + }, + { + "epoch": 2.9479464089611245, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.414039373397827, + "learning_rate": 1e-06, + "loss": 0.9903, + "mean_token_accuracy": 0.7134126424789429, + "num_tokens": 694592406.0, + "step": 26844 + }, + { + "epoch": 2.9480562266637382, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.810014009475708, + "learning_rate": 1e-06, + "loss": 1.0499, + "mean_token_accuracy": 0.6948904991149902, + "num_tokens": 694615286.0, + "step": 26845 + }, + { + "epoch": 2.948166044366352, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.489518165588379, + "learning_rate": 1e-06, + "loss": 1.0983, + "mean_token_accuracy": 0.6857572793960571, + "num_tokens": 694643591.0, + "step": 26846 + }, + { + "epoch": 2.9482758620689653, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.781984567642212, + "learning_rate": 1e-06, + "loss": 0.9425, + "mean_token_accuracy": 0.7170993089675903, + "num_tokens": 694664116.0, + "step": 26847 + }, + { + "epoch": 2.948385679771579, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6925957202911377, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7073870897293091, + "num_tokens": 694688798.0, + "step": 26848 + }, + { + "epoch": 2.948495497474193, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6964571475982666, + "learning_rate": 1e-06, + "loss": 0.8823, + "mean_token_accuracy": 0.7350516319274902, + "num_tokens": 694710847.0, + "step": 26849 + }, + { + "epoch": 2.9486053151768066, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.967090606689453, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7163628339767456, + "num_tokens": 694729400.0, + "step": 26850 + }, + { + "epoch": 2.9487151328794203, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5208353996276855, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6883373260498047, + "num_tokens": 694760140.0, + "step": 26851 + }, + { + "epoch": 2.9488249505820336, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7601702213287354, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6969465017318726, + "num_tokens": 694782721.0, + "step": 26852 + }, + { + "epoch": 2.9489347682846474, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.553504705429077, + "learning_rate": 1e-06, + "loss": 1.0331, + "mean_token_accuracy": 0.711480438709259, + "num_tokens": 694807509.0, + "step": 26853 + }, + { + "epoch": 2.949044585987261, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6636016368865967, + "learning_rate": 1e-06, + "loss": 0.9878, + "mean_token_accuracy": 0.7101445198059082, + "num_tokens": 694833524.0, + "step": 26854 + }, + { + "epoch": 2.949154403689875, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.666090726852417, + "learning_rate": 1e-06, + "loss": 0.9172, + "mean_token_accuracy": 0.7398189902305603, + "num_tokens": 694855875.0, + "step": 26855 + }, + { + "epoch": 2.9492642213924887, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8139920234680176, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.7071301937103271, + "num_tokens": 694877608.0, + "step": 26856 + }, + { + "epoch": 2.949374039095102, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.6872622966766357, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7113583087921143, + "num_tokens": 694901942.0, + "step": 26857 + }, + { + "epoch": 2.9494838567977157, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6618359088897705, + "learning_rate": 1e-06, + "loss": 0.9706, + "mean_token_accuracy": 0.7193988561630249, + "num_tokens": 694925693.0, + "step": 26858 + }, + { + "epoch": 2.9495936745003295, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4770424365997314, + "learning_rate": 1e-06, + "loss": 1.0019, + "mean_token_accuracy": 0.708295464515686, + "num_tokens": 694957187.0, + "step": 26859 + }, + { + "epoch": 2.9497034922029433, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6528913974761963, + "learning_rate": 1e-06, + "loss": 1.0147, + "mean_token_accuracy": 0.7038168907165527, + "num_tokens": 694982495.0, + "step": 26860 + }, + { + "epoch": 2.949813309905557, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.578619956970215, + "learning_rate": 1e-06, + "loss": 0.9731, + "mean_token_accuracy": 0.7165374755859375, + "num_tokens": 695006353.0, + "step": 26861 + }, + { + "epoch": 2.9499231276081703, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.925327777862549, + "learning_rate": 1e-06, + "loss": 0.9782, + "mean_token_accuracy": 0.7179895639419556, + "num_tokens": 695030589.0, + "step": 26862 + }, + { + "epoch": 2.950032945310784, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7794079780578613, + "learning_rate": 1e-06, + "loss": 1.0878, + "mean_token_accuracy": 0.6975843906402588, + "num_tokens": 695055152.0, + "step": 26863 + }, + { + "epoch": 2.950142763013398, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7000114917755127, + "learning_rate": 1e-06, + "loss": 1.0045, + "mean_token_accuracy": 0.7151360511779785, + "num_tokens": 695077887.0, + "step": 26864 + }, + { + "epoch": 2.950252580716011, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3264787197113037, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6941678524017334, + "num_tokens": 695108433.0, + "step": 26865 + }, + { + "epoch": 2.9503623984186254, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.628009080886841, + "learning_rate": 1e-06, + "loss": 1.0332, + "mean_token_accuracy": 0.7026042938232422, + "num_tokens": 695134430.0, + "step": 26866 + }, + { + "epoch": 2.9504722161212387, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.700343132019043, + "learning_rate": 1e-06, + "loss": 1.032, + "mean_token_accuracy": 0.7055034041404724, + "num_tokens": 695157626.0, + "step": 26867 + }, + { + "epoch": 2.9505820338238524, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5544402599334717, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.694249153137207, + "num_tokens": 695185852.0, + "step": 26868 + }, + { + "epoch": 2.950691851526466, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 7.8901238441467285, + "learning_rate": 1e-06, + "loss": 1.0358, + "mean_token_accuracy": 0.7020405530929565, + "num_tokens": 695213099.0, + "step": 26869 + }, + { + "epoch": 2.9508016692290795, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6620213985443115, + "learning_rate": 1e-06, + "loss": 1.0652, + "mean_token_accuracy": 0.6919692158699036, + "num_tokens": 695239514.0, + "step": 26870 + }, + { + "epoch": 2.9509114869316933, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.625840425491333, + "learning_rate": 1e-06, + "loss": 1.0277, + "mean_token_accuracy": 0.6977221965789795, + "num_tokens": 695267943.0, + "step": 26871 + }, + { + "epoch": 2.951021304634307, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.2180657386779785, + "learning_rate": 1e-06, + "loss": 0.8823, + "mean_token_accuracy": 0.7421833872795105, + "num_tokens": 695287393.0, + "step": 26872 + }, + { + "epoch": 2.9511311223369208, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6263926029205322, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.708547055721283, + "num_tokens": 695310829.0, + "step": 26873 + }, + { + "epoch": 2.9512409400395345, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4609484672546387, + "learning_rate": 1e-06, + "loss": 0.9695, + "mean_token_accuracy": 0.72284996509552, + "num_tokens": 695335851.0, + "step": 26874 + }, + { + "epoch": 2.951350757742148, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.801483392715454, + "learning_rate": 1e-06, + "loss": 0.9951, + "mean_token_accuracy": 0.7122199535369873, + "num_tokens": 695359724.0, + "step": 26875 + }, + { + "epoch": 2.9514605754447616, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5606982707977295, + "learning_rate": 1e-06, + "loss": 1.0259, + "mean_token_accuracy": 0.7002822756767273, + "num_tokens": 695383927.0, + "step": 26876 + }, + { + "epoch": 2.9515703931473753, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4132049083709717, + "learning_rate": 1e-06, + "loss": 1.0072, + "mean_token_accuracy": 0.7105116248130798, + "num_tokens": 695415377.0, + "step": 26877 + }, + { + "epoch": 2.951680210849989, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6382479667663574, + "learning_rate": 1e-06, + "loss": 1.0923, + "mean_token_accuracy": 0.6859531402587891, + "num_tokens": 695444063.0, + "step": 26878 + }, + { + "epoch": 2.951790028552603, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.652496576309204, + "learning_rate": 1e-06, + "loss": 0.9707, + "mean_token_accuracy": 0.7142180800437927, + "num_tokens": 695468197.0, + "step": 26879 + }, + { + "epoch": 2.951899846255216, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4383389949798584, + "learning_rate": 1e-06, + "loss": 1.1051, + "mean_token_accuracy": 0.6827055215835571, + "num_tokens": 695498017.0, + "step": 26880 + }, + { + "epoch": 2.95200966395783, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5004336833953857, + "learning_rate": 1e-06, + "loss": 1.0839, + "mean_token_accuracy": 0.6809104084968567, + "num_tokens": 695523901.0, + "step": 26881 + }, + { + "epoch": 2.9521194816604437, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.656055212020874, + "learning_rate": 1e-06, + "loss": 1.0106, + "mean_token_accuracy": 0.7092792987823486, + "num_tokens": 695552061.0, + "step": 26882 + }, + { + "epoch": 2.9522292993630574, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.439687967300415, + "learning_rate": 1e-06, + "loss": 1.1212, + "mean_token_accuracy": 0.6805044412612915, + "num_tokens": 695582133.0, + "step": 26883 + }, + { + "epoch": 2.952339117065671, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4290430545806885, + "learning_rate": 1e-06, + "loss": 1.0052, + "mean_token_accuracy": 0.7067550420761108, + "num_tokens": 695609439.0, + "step": 26884 + }, + { + "epoch": 2.9524489347682845, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7511537075042725, + "learning_rate": 1e-06, + "loss": 1.0635, + "mean_token_accuracy": 0.7062060236930847, + "num_tokens": 695634479.0, + "step": 26885 + }, + { + "epoch": 2.9525587524708983, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.668833017349243, + "learning_rate": 1e-06, + "loss": 1.0046, + "mean_token_accuracy": 0.6989286541938782, + "num_tokens": 695659018.0, + "step": 26886 + }, + { + "epoch": 2.952668570173512, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4708354473114014, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.7059047818183899, + "num_tokens": 695685941.0, + "step": 26887 + }, + { + "epoch": 2.952778387876126, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.871624708175659, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.703987717628479, + "num_tokens": 695706360.0, + "step": 26888 + }, + { + "epoch": 2.9528882055787395, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4041316509246826, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.7205908298492432, + "num_tokens": 695736570.0, + "step": 26889 + }, + { + "epoch": 2.952998023281353, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.844346761703491, + "learning_rate": 1e-06, + "loss": 0.985, + "mean_token_accuracy": 0.7143476009368896, + "num_tokens": 695758780.0, + "step": 26890 + }, + { + "epoch": 2.9531078409839666, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8874130249023438, + "learning_rate": 1e-06, + "loss": 1.0156, + "mean_token_accuracy": 0.7045968174934387, + "num_tokens": 695781814.0, + "step": 26891 + }, + { + "epoch": 2.9532176586865804, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3171627521514893, + "learning_rate": 1e-06, + "loss": 0.975, + "mean_token_accuracy": 0.7149466872215271, + "num_tokens": 695816254.0, + "step": 26892 + }, + { + "epoch": 2.9533274763891937, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.785696029663086, + "learning_rate": 1e-06, + "loss": 1.0375, + "mean_token_accuracy": 0.7009315490722656, + "num_tokens": 695841352.0, + "step": 26893 + }, + { + "epoch": 2.9534372940918074, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5512428283691406, + "learning_rate": 1e-06, + "loss": 1.0324, + "mean_token_accuracy": 0.6994385719299316, + "num_tokens": 695869778.0, + "step": 26894 + }, + { + "epoch": 2.953547111794421, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.574500799179077, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.7006996273994446, + "num_tokens": 695895169.0, + "step": 26895 + }, + { + "epoch": 2.953656929497035, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.778301954269409, + "learning_rate": 1e-06, + "loss": 0.9783, + "mean_token_accuracy": 0.7208149433135986, + "num_tokens": 695929771.0, + "step": 26896 + }, + { + "epoch": 2.9537667471996487, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7737982273101807, + "learning_rate": 1e-06, + "loss": 0.9748, + "mean_token_accuracy": 0.7125867605209351, + "num_tokens": 695952223.0, + "step": 26897 + }, + { + "epoch": 2.953876564902262, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5131843090057373, + "learning_rate": 1e-06, + "loss": 1.0791, + "mean_token_accuracy": 0.6952570676803589, + "num_tokens": 695983239.0, + "step": 26898 + }, + { + "epoch": 2.953986382604876, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6568081378936768, + "learning_rate": 1e-06, + "loss": 1.005, + "mean_token_accuracy": 0.7061119675636292, + "num_tokens": 696007787.0, + "step": 26899 + }, + { + "epoch": 2.9540962003074895, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6343653202056885, + "learning_rate": 1e-06, + "loss": 1.0062, + "mean_token_accuracy": 0.7093762755393982, + "num_tokens": 696032528.0, + "step": 26900 + }, + { + "epoch": 2.9542060180101033, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5227761268615723, + "learning_rate": 1e-06, + "loss": 1.0642, + "mean_token_accuracy": 0.6887363195419312, + "num_tokens": 696061706.0, + "step": 26901 + }, + { + "epoch": 2.954315835712717, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5131311416625977, + "learning_rate": 1e-06, + "loss": 1.0159, + "mean_token_accuracy": 0.7086696624755859, + "num_tokens": 696090386.0, + "step": 26902 + }, + { + "epoch": 2.9544256534153304, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6204681396484375, + "learning_rate": 1e-06, + "loss": 0.962, + "mean_token_accuracy": 0.7164140939712524, + "num_tokens": 696114610.0, + "step": 26903 + }, + { + "epoch": 2.954535471117944, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.23258900642395, + "learning_rate": 1e-06, + "loss": 0.988, + "mean_token_accuracy": 0.7152280807495117, + "num_tokens": 696147123.0, + "step": 26904 + }, + { + "epoch": 2.954645288820558, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4636306762695312, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.712838351726532, + "num_tokens": 696176382.0, + "step": 26905 + }, + { + "epoch": 2.9547551065231716, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3430604934692383, + "learning_rate": 1e-06, + "loss": 1.0158, + "mean_token_accuracy": 0.705793023109436, + "num_tokens": 696207007.0, + "step": 26906 + }, + { + "epoch": 2.9548649242257854, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5286338329315186, + "learning_rate": 1e-06, + "loss": 1.0179, + "mean_token_accuracy": 0.70802903175354, + "num_tokens": 696233648.0, + "step": 26907 + }, + { + "epoch": 2.9549747419283987, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.1111221313476562, + "learning_rate": 1e-06, + "loss": 1.0016, + "mean_token_accuracy": 0.7076903581619263, + "num_tokens": 696261850.0, + "step": 26908 + }, + { + "epoch": 2.9550845596310125, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.1616320610046387, + "learning_rate": 1e-06, + "loss": 0.9641, + "mean_token_accuracy": 0.7245826721191406, + "num_tokens": 696283083.0, + "step": 26909 + }, + { + "epoch": 2.955194377333626, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5337939262390137, + "learning_rate": 1e-06, + "loss": 1.1023, + "mean_token_accuracy": 0.678414523601532, + "num_tokens": 696318250.0, + "step": 26910 + }, + { + "epoch": 2.95530419503624, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.943072557449341, + "learning_rate": 1e-06, + "loss": 0.9725, + "mean_token_accuracy": 0.7161446809768677, + "num_tokens": 696338696.0, + "step": 26911 + }, + { + "epoch": 2.9554140127388537, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9089627265930176, + "learning_rate": 1e-06, + "loss": 1.0015, + "mean_token_accuracy": 0.7158955335617065, + "num_tokens": 696359347.0, + "step": 26912 + }, + { + "epoch": 2.955523830441467, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9811816215515137, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7227809429168701, + "num_tokens": 696379979.0, + "step": 26913 + }, + { + "epoch": 2.955633648144081, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4204416275024414, + "learning_rate": 1e-06, + "loss": 1.0778, + "mean_token_accuracy": 0.6938631534576416, + "num_tokens": 696410744.0, + "step": 26914 + }, + { + "epoch": 2.9557434658466946, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.453813314437866, + "learning_rate": 1e-06, + "loss": 1.044, + "mean_token_accuracy": 0.703039288520813, + "num_tokens": 696442078.0, + "step": 26915 + }, + { + "epoch": 2.955853283549308, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.873044013977051, + "learning_rate": 1e-06, + "loss": 0.9692, + "mean_token_accuracy": 0.7122757434844971, + "num_tokens": 696464422.0, + "step": 26916 + }, + { + "epoch": 2.955963101251922, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.737358808517456, + "learning_rate": 1e-06, + "loss": 0.9779, + "mean_token_accuracy": 0.7169475555419922, + "num_tokens": 696488774.0, + "step": 26917 + }, + { + "epoch": 2.9560729189545354, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.590219020843506, + "learning_rate": 1e-06, + "loss": 1.0793, + "mean_token_accuracy": 0.6936873197555542, + "num_tokens": 696516657.0, + "step": 26918 + }, + { + "epoch": 2.956182736657149, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.984867811203003, + "learning_rate": 1e-06, + "loss": 1.0155, + "mean_token_accuracy": 0.7074508666992188, + "num_tokens": 696537618.0, + "step": 26919 + }, + { + "epoch": 2.956292554359763, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7407386302948, + "learning_rate": 1e-06, + "loss": 0.9434, + "mean_token_accuracy": 0.7267077565193176, + "num_tokens": 696562197.0, + "step": 26920 + }, + { + "epoch": 2.956402372062376, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7375755310058594, + "learning_rate": 1e-06, + "loss": 1.0401, + "mean_token_accuracy": 0.7007287740707397, + "num_tokens": 696588380.0, + "step": 26921 + }, + { + "epoch": 2.95651218976499, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5782082080841064, + "learning_rate": 1e-06, + "loss": 1.0395, + "mean_token_accuracy": 0.7050182223320007, + "num_tokens": 696612837.0, + "step": 26922 + }, + { + "epoch": 2.9566220074676037, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8029773235321045, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.7105008959770203, + "num_tokens": 696633842.0, + "step": 26923 + }, + { + "epoch": 2.9567318251702175, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6700546741485596, + "learning_rate": 1e-06, + "loss": 1.006, + "mean_token_accuracy": 0.7035233974456787, + "num_tokens": 696660609.0, + "step": 26924 + }, + { + "epoch": 2.9568416428728312, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.623154401779175, + "learning_rate": 1e-06, + "loss": 1.0809, + "mean_token_accuracy": 0.6955269575119019, + "num_tokens": 696688257.0, + "step": 26925 + }, + { + "epoch": 2.9569514605754446, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.574324607849121, + "learning_rate": 1e-06, + "loss": 1.0113, + "mean_token_accuracy": 0.7209325432777405, + "num_tokens": 696712726.0, + "step": 26926 + }, + { + "epoch": 2.9570612782780583, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.374191999435425, + "learning_rate": 1e-06, + "loss": 1.0024, + "mean_token_accuracy": 0.7104417681694031, + "num_tokens": 696743551.0, + "step": 26927 + }, + { + "epoch": 2.957171095980672, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5042760372161865, + "learning_rate": 1e-06, + "loss": 0.9437, + "mean_token_accuracy": 0.7221324443817139, + "num_tokens": 696772162.0, + "step": 26928 + }, + { + "epoch": 2.957280913683286, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4795196056365967, + "learning_rate": 1e-06, + "loss": 0.9781, + "mean_token_accuracy": 0.7204958200454712, + "num_tokens": 696796890.0, + "step": 26929 + }, + { + "epoch": 2.9573907313858996, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7040867805480957, + "learning_rate": 1e-06, + "loss": 1.0377, + "mean_token_accuracy": 0.6998692750930786, + "num_tokens": 696821425.0, + "step": 26930 + }, + { + "epoch": 2.957500549088513, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5171759128570557, + "learning_rate": 1e-06, + "loss": 1.0379, + "mean_token_accuracy": 0.7045336961746216, + "num_tokens": 696849118.0, + "step": 26931 + }, + { + "epoch": 2.9576103667911267, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.601057767868042, + "learning_rate": 1e-06, + "loss": 1.0344, + "mean_token_accuracy": 0.6996755599975586, + "num_tokens": 696875342.0, + "step": 26932 + }, + { + "epoch": 2.9577201844937404, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.613765239715576, + "learning_rate": 1e-06, + "loss": 1.0068, + "mean_token_accuracy": 0.7046045064926147, + "num_tokens": 696900139.0, + "step": 26933 + }, + { + "epoch": 2.957830002196354, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5947606563568115, + "learning_rate": 1e-06, + "loss": 1.0382, + "mean_token_accuracy": 0.6954014897346497, + "num_tokens": 696925544.0, + "step": 26934 + }, + { + "epoch": 2.957939819898968, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.492703914642334, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7043426036834717, + "num_tokens": 696957200.0, + "step": 26935 + }, + { + "epoch": 2.9580496376015812, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5802407264709473, + "learning_rate": 1e-06, + "loss": 1.0519, + "mean_token_accuracy": 0.6959153413772583, + "num_tokens": 696984092.0, + "step": 26936 + }, + { + "epoch": 2.958159455304195, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.874690055847168, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.720676600933075, + "num_tokens": 697009435.0, + "step": 26937 + }, + { + "epoch": 2.9582692730068088, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4773387908935547, + "learning_rate": 1e-06, + "loss": 1.0103, + "mean_token_accuracy": 0.716323971748352, + "num_tokens": 697040127.0, + "step": 26938 + }, + { + "epoch": 2.9583790907094225, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.536459445953369, + "learning_rate": 1e-06, + "loss": 1.0235, + "mean_token_accuracy": 0.7076139450073242, + "num_tokens": 697068990.0, + "step": 26939 + }, + { + "epoch": 2.9584889084120363, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6551520824432373, + "learning_rate": 1e-06, + "loss": 1.0168, + "mean_token_accuracy": 0.7053407430648804, + "num_tokens": 697093253.0, + "step": 26940 + }, + { + "epoch": 2.9585987261146496, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2768704891204834, + "learning_rate": 1e-06, + "loss": 1.0638, + "mean_token_accuracy": 0.6940436959266663, + "num_tokens": 697127762.0, + "step": 26941 + }, + { + "epoch": 2.9587085438172633, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.160236120223999, + "learning_rate": 1e-06, + "loss": 1.0388, + "mean_token_accuracy": 0.6949344873428345, + "num_tokens": 697152355.0, + "step": 26942 + }, + { + "epoch": 2.958818361519877, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.578601598739624, + "learning_rate": 1e-06, + "loss": 0.929, + "mean_token_accuracy": 0.7273832559585571, + "num_tokens": 697177570.0, + "step": 26943 + }, + { + "epoch": 2.9589281792224904, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.508592128753662, + "learning_rate": 1e-06, + "loss": 0.881, + "mean_token_accuracy": 0.7370387315750122, + "num_tokens": 697203614.0, + "step": 26944 + }, + { + "epoch": 2.959037996925104, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6031911373138428, + "learning_rate": 1e-06, + "loss": 0.9657, + "mean_token_accuracy": 0.7227280735969543, + "num_tokens": 697227800.0, + "step": 26945 + }, + { + "epoch": 2.959147814627718, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.735383987426758, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7062219381332397, + "num_tokens": 697249865.0, + "step": 26946 + }, + { + "epoch": 2.9592576323303317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.645418643951416, + "learning_rate": 1e-06, + "loss": 0.9897, + "mean_token_accuracy": 0.7079766392707825, + "num_tokens": 697272909.0, + "step": 26947 + }, + { + "epoch": 2.9593674500329454, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.369617223739624, + "learning_rate": 1e-06, + "loss": 0.9459, + "mean_token_accuracy": 0.7287546992301941, + "num_tokens": 697300674.0, + "step": 26948 + }, + { + "epoch": 2.9594772677355587, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.521400213241577, + "learning_rate": 1e-06, + "loss": 0.9047, + "mean_token_accuracy": 0.741877555847168, + "num_tokens": 697327265.0, + "step": 26949 + }, + { + "epoch": 2.9595870854381725, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.45166277885437, + "learning_rate": 1e-06, + "loss": 0.9023, + "mean_token_accuracy": 0.7347781658172607, + "num_tokens": 697352568.0, + "step": 26950 + }, + { + "epoch": 2.9596969031407863, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3351263999938965, + "learning_rate": 1e-06, + "loss": 1.0225, + "mean_token_accuracy": 0.7015335559844971, + "num_tokens": 697383626.0, + "step": 26951 + }, + { + "epoch": 2.9598067208434, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7682414054870605, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.7102247476577759, + "num_tokens": 697406230.0, + "step": 26952 + }, + { + "epoch": 2.9599165385460138, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7834718227386475, + "learning_rate": 1e-06, + "loss": 0.9862, + "mean_token_accuracy": 0.7183911800384521, + "num_tokens": 697428583.0, + "step": 26953 + }, + { + "epoch": 2.960026356248627, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8382527828216553, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7049790620803833, + "num_tokens": 697450490.0, + "step": 26954 + }, + { + "epoch": 2.960136173951241, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.915492057800293, + "learning_rate": 1e-06, + "loss": 1.0497, + "mean_token_accuracy": 0.6953949928283691, + "num_tokens": 697474413.0, + "step": 26955 + }, + { + "epoch": 2.9602459916538546, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7180423736572266, + "learning_rate": 1e-06, + "loss": 1.0826, + "mean_token_accuracy": 0.6842515468597412, + "num_tokens": 697500328.0, + "step": 26956 + }, + { + "epoch": 2.9603558093564684, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.480064630508423, + "learning_rate": 1e-06, + "loss": 0.9678, + "mean_token_accuracy": 0.7230463624000549, + "num_tokens": 697531539.0, + "step": 26957 + }, + { + "epoch": 2.960465627059082, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5972843170166016, + "learning_rate": 1e-06, + "loss": 1.0868, + "mean_token_accuracy": 0.6902540922164917, + "num_tokens": 697557611.0, + "step": 26958 + }, + { + "epoch": 2.9605754447616954, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5907580852508545, + "learning_rate": 1e-06, + "loss": 1.0208, + "mean_token_accuracy": 0.6993087530136108, + "num_tokens": 697584171.0, + "step": 26959 + }, + { + "epoch": 2.960685262464309, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4665427207946777, + "learning_rate": 1e-06, + "loss": 1.0276, + "mean_token_accuracy": 0.706239640712738, + "num_tokens": 697610562.0, + "step": 26960 + }, + { + "epoch": 2.960795080166923, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7527894973754883, + "learning_rate": 1e-06, + "loss": 0.999, + "mean_token_accuracy": 0.7151573896408081, + "num_tokens": 697632479.0, + "step": 26961 + }, + { + "epoch": 2.9609048978695367, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.592212200164795, + "learning_rate": 1e-06, + "loss": 1.0304, + "mean_token_accuracy": 0.702364444732666, + "num_tokens": 697659337.0, + "step": 26962 + }, + { + "epoch": 2.9610147155721505, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.318262815475464, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7126220464706421, + "num_tokens": 697692917.0, + "step": 26963 + }, + { + "epoch": 2.9611245332747638, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3522422313690186, + "learning_rate": 1e-06, + "loss": 1.0578, + "mean_token_accuracy": 0.6981013417243958, + "num_tokens": 697722169.0, + "step": 26964 + }, + { + "epoch": 2.9612343509773775, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5257697105407715, + "learning_rate": 1e-06, + "loss": 1.0157, + "mean_token_accuracy": 0.7064942717552185, + "num_tokens": 697750091.0, + "step": 26965 + }, + { + "epoch": 2.9613441686799913, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5017030239105225, + "learning_rate": 1e-06, + "loss": 1.0674, + "mean_token_accuracy": 0.6972519159317017, + "num_tokens": 697776446.0, + "step": 26966 + }, + { + "epoch": 2.961453986382605, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 16.778778076171875, + "learning_rate": 1e-06, + "loss": 1.0252, + "mean_token_accuracy": 0.6998940706253052, + "num_tokens": 697796895.0, + "step": 26967 + }, + { + "epoch": 2.961563804085219, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5922980308532715, + "learning_rate": 1e-06, + "loss": 1.1111, + "mean_token_accuracy": 0.6877191066741943, + "num_tokens": 697827015.0, + "step": 26968 + }, + { + "epoch": 2.961673621787832, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.932286500930786, + "learning_rate": 1e-06, + "loss": 0.9548, + "mean_token_accuracy": 0.7326358556747437, + "num_tokens": 697849556.0, + "step": 26969 + }, + { + "epoch": 2.961783439490446, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.792941093444824, + "learning_rate": 1e-06, + "loss": 0.9486, + "mean_token_accuracy": 0.7245285511016846, + "num_tokens": 697873031.0, + "step": 26970 + }, + { + "epoch": 2.9618932571930596, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4952051639556885, + "learning_rate": 1e-06, + "loss": 1.0161, + "mean_token_accuracy": 0.7022852897644043, + "num_tokens": 697902202.0, + "step": 26971 + }, + { + "epoch": 2.962003074895673, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.541722536087036, + "learning_rate": 1e-06, + "loss": 0.9837, + "mean_token_accuracy": 0.7167999744415283, + "num_tokens": 697929995.0, + "step": 26972 + }, + { + "epoch": 2.9621128925982867, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.663543701171875, + "learning_rate": 1e-06, + "loss": 0.9672, + "mean_token_accuracy": 0.7168433666229248, + "num_tokens": 697954809.0, + "step": 26973 + }, + { + "epoch": 2.9622227103009005, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5733842849731445, + "learning_rate": 1e-06, + "loss": 1.0892, + "mean_token_accuracy": 0.6827909350395203, + "num_tokens": 697982219.0, + "step": 26974 + }, + { + "epoch": 2.962332528003514, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5578579902648926, + "learning_rate": 1e-06, + "loss": 1.0369, + "mean_token_accuracy": 0.6998865008354187, + "num_tokens": 698010780.0, + "step": 26975 + }, + { + "epoch": 2.962442345706128, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4668657779693604, + "learning_rate": 1e-06, + "loss": 0.9868, + "mean_token_accuracy": 0.7169809341430664, + "num_tokens": 698036413.0, + "step": 26976 + }, + { + "epoch": 2.9625521634087413, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.717925786972046, + "learning_rate": 1e-06, + "loss": 0.9713, + "mean_token_accuracy": 0.7187968492507935, + "num_tokens": 698062900.0, + "step": 26977 + }, + { + "epoch": 2.962661981111355, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8241448402404785, + "learning_rate": 1e-06, + "loss": 1.0438, + "mean_token_accuracy": 0.6975688934326172, + "num_tokens": 698087121.0, + "step": 26978 + }, + { + "epoch": 2.962771798813969, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8309435844421387, + "learning_rate": 1e-06, + "loss": 1.0579, + "mean_token_accuracy": 0.6892543435096741, + "num_tokens": 698113106.0, + "step": 26979 + }, + { + "epoch": 2.9628816165165826, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6989595890045166, + "learning_rate": 1e-06, + "loss": 0.9762, + "mean_token_accuracy": 0.7201113700866699, + "num_tokens": 698136953.0, + "step": 26980 + }, + { + "epoch": 2.9629914342191963, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6951351165771484, + "learning_rate": 1e-06, + "loss": 1.0733, + "mean_token_accuracy": 0.6851776838302612, + "num_tokens": 698160688.0, + "step": 26981 + }, + { + "epoch": 2.9631012519218096, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5915417671203613, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.715961217880249, + "num_tokens": 698186106.0, + "step": 26982 + }, + { + "epoch": 2.9632110696244234, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5913493633270264, + "learning_rate": 1e-06, + "loss": 0.9855, + "mean_token_accuracy": 0.713426947593689, + "num_tokens": 698213086.0, + "step": 26983 + }, + { + "epoch": 2.963320887327037, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3004562854766846, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7242505550384521, + "num_tokens": 698243181.0, + "step": 26984 + }, + { + "epoch": 2.963430705029651, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.638655185699463, + "learning_rate": 1e-06, + "loss": 0.9743, + "mean_token_accuracy": 0.716537594795227, + "num_tokens": 698267595.0, + "step": 26985 + }, + { + "epoch": 2.9635405227322646, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6500165462493896, + "learning_rate": 1e-06, + "loss": 0.9431, + "mean_token_accuracy": 0.7221798300743103, + "num_tokens": 698291108.0, + "step": 26986 + }, + { + "epoch": 2.963650340434878, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4778740406036377, + "learning_rate": 1e-06, + "loss": 0.9918, + "mean_token_accuracy": 0.7114506959915161, + "num_tokens": 698321436.0, + "step": 26987 + }, + { + "epoch": 2.9637601581374917, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8503780364990234, + "learning_rate": 1e-06, + "loss": 1.0268, + "mean_token_accuracy": 0.7072932720184326, + "num_tokens": 698348173.0, + "step": 26988 + }, + { + "epoch": 2.9638699758401055, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3520689010620117, + "learning_rate": 1e-06, + "loss": 1.1114, + "mean_token_accuracy": 0.6857423782348633, + "num_tokens": 698382618.0, + "step": 26989 + }, + { + "epoch": 2.9639797935427192, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.124911308288574, + "learning_rate": 1e-06, + "loss": 1.0525, + "mean_token_accuracy": 0.7006527185440063, + "num_tokens": 698419430.0, + "step": 26990 + }, + { + "epoch": 2.964089611245333, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.742375612258911, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7247635126113892, + "num_tokens": 698440289.0, + "step": 26991 + }, + { + "epoch": 2.9641994289479463, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6180930137634277, + "learning_rate": 1e-06, + "loss": 1.0031, + "mean_token_accuracy": 0.7081831097602844, + "num_tokens": 698468329.0, + "step": 26992 + }, + { + "epoch": 2.96430924665056, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7093780040740967, + "learning_rate": 1e-06, + "loss": 0.9765, + "mean_token_accuracy": 0.712417483329773, + "num_tokens": 698493506.0, + "step": 26993 + }, + { + "epoch": 2.964419064353174, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9451193809509277, + "learning_rate": 1e-06, + "loss": 0.9351, + "mean_token_accuracy": 0.723891019821167, + "num_tokens": 698512949.0, + "step": 26994 + }, + { + "epoch": 2.964528882055787, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5263826847076416, + "learning_rate": 1e-06, + "loss": 1.0707, + "mean_token_accuracy": 0.6976500749588013, + "num_tokens": 698540485.0, + "step": 26995 + }, + { + "epoch": 2.9646386997584013, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5332932472229004, + "learning_rate": 1e-06, + "loss": 1.1004, + "mean_token_accuracy": 0.6836861968040466, + "num_tokens": 698568590.0, + "step": 26996 + }, + { + "epoch": 2.9647485174610146, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5736024379730225, + "learning_rate": 1e-06, + "loss": 0.9882, + "mean_token_accuracy": 0.7093116641044617, + "num_tokens": 698595857.0, + "step": 26997 + }, + { + "epoch": 2.9648583351636284, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.282348871231079, + "learning_rate": 1e-06, + "loss": 1.087, + "mean_token_accuracy": 0.6885951161384583, + "num_tokens": 698630604.0, + "step": 26998 + }, + { + "epoch": 2.964968152866242, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5419280529022217, + "learning_rate": 1e-06, + "loss": 0.9744, + "mean_token_accuracy": 0.7247538566589355, + "num_tokens": 698655461.0, + "step": 26999 + }, + { + "epoch": 2.9650779705688555, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.290541172027588, + "learning_rate": 1e-06, + "loss": 1.084, + "mean_token_accuracy": 0.6870285868644714, + "num_tokens": 698690583.0, + "step": 27000 + }, + { + "epoch": 2.9651877882714692, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.67700457572937, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6897490620613098, + "num_tokens": 698717539.0, + "step": 27001 + }, + { + "epoch": 2.965297605974083, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7158420085906982, + "learning_rate": 1e-06, + "loss": 0.9522, + "mean_token_accuracy": 0.7250416278839111, + "num_tokens": 698740933.0, + "step": 27002 + }, + { + "epoch": 2.9654074236766967, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.648134469985962, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7118488550186157, + "num_tokens": 698767324.0, + "step": 27003 + }, + { + "epoch": 2.9655172413793105, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.537705183029175, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.7020308971405029, + "num_tokens": 698795914.0, + "step": 27004 + }, + { + "epoch": 2.965627059081924, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.832731008529663, + "learning_rate": 1e-06, + "loss": 1.0315, + "mean_token_accuracy": 0.7026388645172119, + "num_tokens": 698821613.0, + "step": 27005 + }, + { + "epoch": 2.9657368767845376, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8529629707336426, + "learning_rate": 1e-06, + "loss": 1.039, + "mean_token_accuracy": 0.7020232677459717, + "num_tokens": 698842556.0, + "step": 27006 + }, + { + "epoch": 2.9658466944871513, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.559034824371338, + "learning_rate": 1e-06, + "loss": 1.0398, + "mean_token_accuracy": 0.707137942314148, + "num_tokens": 698871271.0, + "step": 27007 + }, + { + "epoch": 2.965956512189765, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.551281452178955, + "learning_rate": 1e-06, + "loss": 1.075, + "mean_token_accuracy": 0.6888883709907532, + "num_tokens": 698897929.0, + "step": 27008 + }, + { + "epoch": 2.966066329892379, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.708827018737793, + "learning_rate": 1e-06, + "loss": 1.0622, + "mean_token_accuracy": 0.6894817352294922, + "num_tokens": 698922956.0, + "step": 27009 + }, + { + "epoch": 2.966176147594992, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3937339782714844, + "learning_rate": 1e-06, + "loss": 0.9564, + "mean_token_accuracy": 0.7198804616928101, + "num_tokens": 698952848.0, + "step": 27010 + }, + { + "epoch": 2.966285965297606, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5813193321228027, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7130229473114014, + "num_tokens": 698977153.0, + "step": 27011 + }, + { + "epoch": 2.9663957830002197, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8676095008850098, + "learning_rate": 1e-06, + "loss": 0.8808, + "mean_token_accuracy": 0.7426644563674927, + "num_tokens": 698996246.0, + "step": 27012 + }, + { + "epoch": 2.9665056007028334, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.744330883026123, + "learning_rate": 1e-06, + "loss": 0.9968, + "mean_token_accuracy": 0.7088619470596313, + "num_tokens": 699018202.0, + "step": 27013 + }, + { + "epoch": 2.966615418405447, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8427834510803223, + "learning_rate": 1e-06, + "loss": 0.9881, + "mean_token_accuracy": 0.7036443948745728, + "num_tokens": 699039269.0, + "step": 27014 + }, + { + "epoch": 2.9667252361080605, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.57452654838562, + "learning_rate": 1e-06, + "loss": 1.0318, + "mean_token_accuracy": 0.700576901435852, + "num_tokens": 699067947.0, + "step": 27015 + }, + { + "epoch": 2.9668350538106742, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7805540561676025, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7030617594718933, + "num_tokens": 699093164.0, + "step": 27016 + }, + { + "epoch": 2.966944871513288, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.757632255554199, + "learning_rate": 1e-06, + "loss": 0.9103, + "mean_token_accuracy": 0.7298521995544434, + "num_tokens": 699116466.0, + "step": 27017 + }, + { + "epoch": 2.9670546892159018, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8108551502227783, + "learning_rate": 1e-06, + "loss": 0.918, + "mean_token_accuracy": 0.7331362962722778, + "num_tokens": 699138646.0, + "step": 27018 + }, + { + "epoch": 2.9671645069185155, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7476913928985596, + "learning_rate": 1e-06, + "loss": 0.9671, + "mean_token_accuracy": 0.7129652500152588, + "num_tokens": 699163571.0, + "step": 27019 + }, + { + "epoch": 2.967274324621129, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4143903255462646, + "learning_rate": 1e-06, + "loss": 1.0835, + "mean_token_accuracy": 0.6960911750793457, + "num_tokens": 699194222.0, + "step": 27020 + }, + { + "epoch": 2.9673841423237426, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5018980503082275, + "learning_rate": 1e-06, + "loss": 1.0121, + "mean_token_accuracy": 0.703516960144043, + "num_tokens": 699225044.0, + "step": 27021 + }, + { + "epoch": 2.9674939600263563, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5307278633117676, + "learning_rate": 1e-06, + "loss": 1.0536, + "mean_token_accuracy": 0.6916748285293579, + "num_tokens": 699254793.0, + "step": 27022 + }, + { + "epoch": 2.9676037777289697, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0836048126220703, + "learning_rate": 1e-06, + "loss": 1.0066, + "mean_token_accuracy": 0.7077223062515259, + "num_tokens": 699276903.0, + "step": 27023 + }, + { + "epoch": 2.9677135954315834, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.426337718963623, + "learning_rate": 1e-06, + "loss": 0.9526, + "mean_token_accuracy": 0.7215641736984253, + "num_tokens": 699306065.0, + "step": 27024 + }, + { + "epoch": 2.967823413134197, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8926196098327637, + "learning_rate": 1e-06, + "loss": 1.0653, + "mean_token_accuracy": 0.7014898061752319, + "num_tokens": 699329481.0, + "step": 27025 + }, + { + "epoch": 2.967933230836811, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3108181953430176, + "learning_rate": 1e-06, + "loss": 1.0554, + "mean_token_accuracy": 0.6913350224494934, + "num_tokens": 699361293.0, + "step": 27026 + }, + { + "epoch": 2.9680430485394247, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.519568920135498, + "learning_rate": 1e-06, + "loss": 0.9632, + "mean_token_accuracy": 0.7190255522727966, + "num_tokens": 699387539.0, + "step": 27027 + }, + { + "epoch": 2.968152866242038, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8939762115478516, + "learning_rate": 1e-06, + "loss": 0.9978, + "mean_token_accuracy": 0.7158697843551636, + "num_tokens": 699407609.0, + "step": 27028 + }, + { + "epoch": 2.9682626839446518, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5029666423797607, + "learning_rate": 1e-06, + "loss": 1.0425, + "mean_token_accuracy": 0.7014221549034119, + "num_tokens": 699435534.0, + "step": 27029 + }, + { + "epoch": 2.9683725016472655, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5837619304656982, + "learning_rate": 1e-06, + "loss": 1.0041, + "mean_token_accuracy": 0.705783486366272, + "num_tokens": 699460880.0, + "step": 27030 + }, + { + "epoch": 2.9684823193498793, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0447463989257812, + "learning_rate": 1e-06, + "loss": 0.9024, + "mean_token_accuracy": 0.7294968366622925, + "num_tokens": 699481884.0, + "step": 27031 + }, + { + "epoch": 2.968592137052493, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0982744693756104, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7101781368255615, + "num_tokens": 699506015.0, + "step": 27032 + }, + { + "epoch": 2.9687019547551063, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.759302854537964, + "learning_rate": 1e-06, + "loss": 1.0502, + "mean_token_accuracy": 0.6972751617431641, + "num_tokens": 699532009.0, + "step": 27033 + }, + { + "epoch": 2.96881177245772, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8211891651153564, + "learning_rate": 1e-06, + "loss": 1.0073, + "mean_token_accuracy": 0.7142212390899658, + "num_tokens": 699553030.0, + "step": 27034 + }, + { + "epoch": 2.968921590160334, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.498096227645874, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7149733304977417, + "num_tokens": 699579827.0, + "step": 27035 + }, + { + "epoch": 2.9690314078629476, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.531122922897339, + "learning_rate": 1e-06, + "loss": 1.0132, + "mean_token_accuracy": 0.7065873146057129, + "num_tokens": 699606097.0, + "step": 27036 + }, + { + "epoch": 2.9691412255655614, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.65242338180542, + "learning_rate": 1e-06, + "loss": 0.959, + "mean_token_accuracy": 0.7201507091522217, + "num_tokens": 699628139.0, + "step": 27037 + }, + { + "epoch": 2.9692510432681747, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.445793628692627, + "learning_rate": 1e-06, + "loss": 1.0291, + "mean_token_accuracy": 0.7038860321044922, + "num_tokens": 699656894.0, + "step": 27038 + }, + { + "epoch": 2.9693608609707884, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.688246965408325, + "learning_rate": 1e-06, + "loss": 1.0305, + "mean_token_accuracy": 0.7027953863143921, + "num_tokens": 699682476.0, + "step": 27039 + }, + { + "epoch": 2.969470678673402, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7634284496307373, + "learning_rate": 1e-06, + "loss": 1.0284, + "mean_token_accuracy": 0.6992825269699097, + "num_tokens": 699708794.0, + "step": 27040 + }, + { + "epoch": 2.969580496376016, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8262178897857666, + "learning_rate": 1e-06, + "loss": 0.9441, + "mean_token_accuracy": 0.7166787385940552, + "num_tokens": 699731205.0, + "step": 27041 + }, + { + "epoch": 2.9696903140786297, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9959540367126465, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7146848440170288, + "num_tokens": 699757450.0, + "step": 27042 + }, + { + "epoch": 2.969800131781243, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.717700958251953, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.724647045135498, + "num_tokens": 699778921.0, + "step": 27043 + }, + { + "epoch": 2.969909949483857, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.528823137283325, + "learning_rate": 1e-06, + "loss": 0.9942, + "mean_token_accuracy": 0.7097680568695068, + "num_tokens": 699808168.0, + "step": 27044 + }, + { + "epoch": 2.9700197671864705, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4507830142974854, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.7066951990127563, + "num_tokens": 699838825.0, + "step": 27045 + }, + { + "epoch": 2.970129584889084, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4419174194335938, + "learning_rate": 1e-06, + "loss": 1.0463, + "mean_token_accuracy": 0.6999018788337708, + "num_tokens": 699866348.0, + "step": 27046 + }, + { + "epoch": 2.970239402591698, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.644895076751709, + "learning_rate": 1e-06, + "loss": 0.9245, + "mean_token_accuracy": 0.7221956253051758, + "num_tokens": 699889755.0, + "step": 27047 + }, + { + "epoch": 2.9703492202943114, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4042041301727295, + "learning_rate": 1e-06, + "loss": 1.0979, + "mean_token_accuracy": 0.6878657341003418, + "num_tokens": 699922208.0, + "step": 27048 + }, + { + "epoch": 2.970459037996925, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0701067447662354, + "learning_rate": 1e-06, + "loss": 1.0337, + "mean_token_accuracy": 0.7040426731109619, + "num_tokens": 699941048.0, + "step": 27049 + }, + { + "epoch": 2.970568855699539, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4772226810455322, + "learning_rate": 1e-06, + "loss": 1.0478, + "mean_token_accuracy": 0.693206787109375, + "num_tokens": 699969241.0, + "step": 27050 + }, + { + "epoch": 2.970678673402152, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6858246326446533, + "learning_rate": 1e-06, + "loss": 0.9623, + "mean_token_accuracy": 0.7112798690795898, + "num_tokens": 699993630.0, + "step": 27051 + }, + { + "epoch": 2.970788491104766, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5501291751861572, + "learning_rate": 1e-06, + "loss": 0.9662, + "mean_token_accuracy": 0.7132022380828857, + "num_tokens": 700020713.0, + "step": 27052 + }, + { + "epoch": 2.9708983088073797, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7037205696105957, + "learning_rate": 1e-06, + "loss": 0.9902, + "mean_token_accuracy": 0.7197980880737305, + "num_tokens": 700043333.0, + "step": 27053 + }, + { + "epoch": 2.9710081265099935, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.809072256088257, + "learning_rate": 1e-06, + "loss": 1.0595, + "mean_token_accuracy": 0.6968526840209961, + "num_tokens": 700071824.0, + "step": 27054 + }, + { + "epoch": 2.971117944212607, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8461859226226807, + "learning_rate": 1e-06, + "loss": 0.9339, + "mean_token_accuracy": 0.7260463237762451, + "num_tokens": 700096285.0, + "step": 27055 + }, + { + "epoch": 2.9712277619152205, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.694356679916382, + "learning_rate": 1e-06, + "loss": 1.024, + "mean_token_accuracy": 0.7015252709388733, + "num_tokens": 700122349.0, + "step": 27056 + }, + { + "epoch": 2.9713375796178343, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.311790943145752, + "learning_rate": 1e-06, + "loss": 1.0485, + "mean_token_accuracy": 0.7090237140655518, + "num_tokens": 700152605.0, + "step": 27057 + }, + { + "epoch": 2.971447397320448, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.624910831451416, + "learning_rate": 1e-06, + "loss": 0.9315, + "mean_token_accuracy": 0.7301214337348938, + "num_tokens": 700173611.0, + "step": 27058 + }, + { + "epoch": 2.971557215023062, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.36649227142334, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.7003962993621826, + "num_tokens": 700205386.0, + "step": 27059 + }, + { + "epoch": 2.9716670327256756, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4066483974456787, + "learning_rate": 1e-06, + "loss": 0.9567, + "mean_token_accuracy": 0.7098885774612427, + "num_tokens": 700235046.0, + "step": 27060 + }, + { + "epoch": 2.971776850428289, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.380427360534668, + "learning_rate": 1e-06, + "loss": 1.022, + "mean_token_accuracy": 0.7019617557525635, + "num_tokens": 700265594.0, + "step": 27061 + }, + { + "epoch": 2.9718866681309026, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.2405354976654053, + "learning_rate": 1e-06, + "loss": 0.9493, + "mean_token_accuracy": 0.7214792370796204, + "num_tokens": 700284713.0, + "step": 27062 + }, + { + "epoch": 2.9719964858335164, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.004103660583496, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.706005871295929, + "num_tokens": 700311304.0, + "step": 27063 + }, + { + "epoch": 2.97210630353613, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.769378423690796, + "learning_rate": 1e-06, + "loss": 0.9958, + "mean_token_accuracy": 0.7113548517227173, + "num_tokens": 700334528.0, + "step": 27064 + }, + { + "epoch": 2.972216121238744, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.509028911590576, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.697792649269104, + "num_tokens": 700365145.0, + "step": 27065 + }, + { + "epoch": 2.972325938941357, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4914276599884033, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7166166305541992, + "num_tokens": 700394174.0, + "step": 27066 + }, + { + "epoch": 2.972435756643971, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4982240200042725, + "learning_rate": 1e-06, + "loss": 0.9879, + "mean_token_accuracy": 0.7145934700965881, + "num_tokens": 700422187.0, + "step": 27067 + }, + { + "epoch": 2.9725455743465847, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5853817462921143, + "learning_rate": 1e-06, + "loss": 1.1145, + "mean_token_accuracy": 0.6779000759124756, + "num_tokens": 700449559.0, + "step": 27068 + }, + { + "epoch": 2.9726553920491985, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4412035942077637, + "learning_rate": 1e-06, + "loss": 1.0566, + "mean_token_accuracy": 0.6977648138999939, + "num_tokens": 700479833.0, + "step": 27069 + }, + { + "epoch": 2.9727652097518122, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6714107990264893, + "learning_rate": 1e-06, + "loss": 0.9373, + "mean_token_accuracy": 0.7214133739471436, + "num_tokens": 700501218.0, + "step": 27070 + }, + { + "epoch": 2.9728750274544256, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5908164978027344, + "learning_rate": 1e-06, + "loss": 1.094, + "mean_token_accuracy": 0.6808292269706726, + "num_tokens": 700529618.0, + "step": 27071 + }, + { + "epoch": 2.9729848451570393, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 8.802597045898438, + "learning_rate": 1e-06, + "loss": 0.986, + "mean_token_accuracy": 0.7099125385284424, + "num_tokens": 700555811.0, + "step": 27072 + }, + { + "epoch": 2.973094662859653, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.5365662574768066, + "learning_rate": 1e-06, + "loss": 0.8757, + "mean_token_accuracy": 0.7435985803604126, + "num_tokens": 700574821.0, + "step": 27073 + }, + { + "epoch": 2.9732044805622664, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.790774345397949, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.6984794735908508, + "num_tokens": 700599791.0, + "step": 27074 + }, + { + "epoch": 2.97331429826488, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.572977304458618, + "learning_rate": 1e-06, + "loss": 1.0001, + "mean_token_accuracy": 0.7123711109161377, + "num_tokens": 700627048.0, + "step": 27075 + }, + { + "epoch": 2.973424115967494, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.994157552719116, + "learning_rate": 1e-06, + "loss": 1.0585, + "mean_token_accuracy": 0.6949342489242554, + "num_tokens": 700656950.0, + "step": 27076 + }, + { + "epoch": 2.9735339336701077, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.457994222640991, + "learning_rate": 1e-06, + "loss": 1.0353, + "mean_token_accuracy": 0.6959269046783447, + "num_tokens": 700686909.0, + "step": 27077 + }, + { + "epoch": 2.9736437513727214, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.604055881500244, + "learning_rate": 1e-06, + "loss": 0.9192, + "mean_token_accuracy": 0.7251019477844238, + "num_tokens": 700711077.0, + "step": 27078 + }, + { + "epoch": 2.9737535690753347, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5349133014678955, + "learning_rate": 1e-06, + "loss": 1.0133, + "mean_token_accuracy": 0.7081547379493713, + "num_tokens": 700737323.0, + "step": 27079 + }, + { + "epoch": 2.9738633867779485, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.957036018371582, + "learning_rate": 1e-06, + "loss": 1.025, + "mean_token_accuracy": 0.702094316482544, + "num_tokens": 700763459.0, + "step": 27080 + }, + { + "epoch": 2.9739732044805622, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5162739753723145, + "learning_rate": 1e-06, + "loss": 0.8942, + "mean_token_accuracy": 0.7380493879318237, + "num_tokens": 700790020.0, + "step": 27081 + }, + { + "epoch": 2.974083022183176, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7696282863616943, + "learning_rate": 1e-06, + "loss": 0.8888, + "mean_token_accuracy": 0.7431458830833435, + "num_tokens": 700810259.0, + "step": 27082 + }, + { + "epoch": 2.9741928398857898, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4696524143218994, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6964775323867798, + "num_tokens": 700837021.0, + "step": 27083 + }, + { + "epoch": 2.974302657588403, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.38204026222229, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7126736640930176, + "num_tokens": 700864837.0, + "step": 27084 + }, + { + "epoch": 2.974412475291017, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.392300844192505, + "learning_rate": 1e-06, + "loss": 1.014, + "mean_token_accuracy": 0.7069442868232727, + "num_tokens": 700893517.0, + "step": 27085 + }, + { + "epoch": 2.9745222929936306, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.657641649246216, + "learning_rate": 1e-06, + "loss": 1.0586, + "mean_token_accuracy": 0.6945976614952087, + "num_tokens": 700918292.0, + "step": 27086 + }, + { + "epoch": 2.9746321106962443, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9603116512298584, + "learning_rate": 1e-06, + "loss": 0.9585, + "mean_token_accuracy": 0.7151263952255249, + "num_tokens": 700942703.0, + "step": 27087 + }, + { + "epoch": 2.974741928398858, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.754427909851074, + "learning_rate": 1e-06, + "loss": 0.9826, + "mean_token_accuracy": 0.7203707695007324, + "num_tokens": 700966432.0, + "step": 27088 + }, + { + "epoch": 2.9748517461014714, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6207528114318848, + "learning_rate": 1e-06, + "loss": 0.998, + "mean_token_accuracy": 0.7078045606613159, + "num_tokens": 700994539.0, + "step": 27089 + }, + { + "epoch": 2.974961563804085, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2661056518554688, + "learning_rate": 1e-06, + "loss": 1.06, + "mean_token_accuracy": 0.6983920335769653, + "num_tokens": 701028518.0, + "step": 27090 + }, + { + "epoch": 2.975071381506699, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.670802593231201, + "learning_rate": 1e-06, + "loss": 0.9661, + "mean_token_accuracy": 0.7159552574157715, + "num_tokens": 701053782.0, + "step": 27091 + }, + { + "epoch": 2.9751811992093127, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.608039379119873, + "learning_rate": 1e-06, + "loss": 0.9343, + "mean_token_accuracy": 0.7264789342880249, + "num_tokens": 701077793.0, + "step": 27092 + }, + { + "epoch": 2.9752910169119264, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.553584575653076, + "learning_rate": 1e-06, + "loss": 1.1457, + "mean_token_accuracy": 0.6877784132957458, + "num_tokens": 701104306.0, + "step": 27093 + }, + { + "epoch": 2.9754008346145397, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5841379165649414, + "learning_rate": 1e-06, + "loss": 0.9592, + "mean_token_accuracy": 0.7168066501617432, + "num_tokens": 701129809.0, + "step": 27094 + }, + { + "epoch": 2.9755106523171535, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5371694564819336, + "learning_rate": 1e-06, + "loss": 1.0029, + "mean_token_accuracy": 0.7185986638069153, + "num_tokens": 701157533.0, + "step": 27095 + }, + { + "epoch": 2.9756204700197673, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4490957260131836, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7077256441116333, + "num_tokens": 701188408.0, + "step": 27096 + }, + { + "epoch": 2.9757302877223806, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6704256534576416, + "learning_rate": 1e-06, + "loss": 1.0712, + "mean_token_accuracy": 0.6964287757873535, + "num_tokens": 701212371.0, + "step": 27097 + }, + { + "epoch": 2.9758401054249948, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.670586109161377, + "learning_rate": 1e-06, + "loss": 0.9252, + "mean_token_accuracy": 0.7249457836151123, + "num_tokens": 701235803.0, + "step": 27098 + }, + { + "epoch": 2.975949923127608, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4636483192443848, + "learning_rate": 1e-06, + "loss": 1.0341, + "mean_token_accuracy": 0.6979773044586182, + "num_tokens": 701263089.0, + "step": 27099 + }, + { + "epoch": 2.976059740830222, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6841323375701904, + "learning_rate": 1e-06, + "loss": 1.0407, + "mean_token_accuracy": 0.6980287432670593, + "num_tokens": 701288344.0, + "step": 27100 + }, + { + "epoch": 2.9761695585328356, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.087351083755493, + "learning_rate": 1e-06, + "loss": 1.0548, + "mean_token_accuracy": 0.7010678052902222, + "num_tokens": 701314693.0, + "step": 27101 + }, + { + "epoch": 2.976279376235449, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9687206745147705, + "learning_rate": 1e-06, + "loss": 0.9154, + "mean_token_accuracy": 0.7314488887786865, + "num_tokens": 701332861.0, + "step": 27102 + }, + { + "epoch": 2.9763891939380627, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5803418159484863, + "learning_rate": 1e-06, + "loss": 1.0836, + "mean_token_accuracy": 0.6864312887191772, + "num_tokens": 701362650.0, + "step": 27103 + }, + { + "epoch": 2.9764990116406764, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6010630130767822, + "learning_rate": 1e-06, + "loss": 1.0393, + "mean_token_accuracy": 0.6987377405166626, + "num_tokens": 701387886.0, + "step": 27104 + }, + { + "epoch": 2.97660882934329, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6926863193511963, + "learning_rate": 1e-06, + "loss": 1.0054, + "mean_token_accuracy": 0.7078142762184143, + "num_tokens": 701411826.0, + "step": 27105 + }, + { + "epoch": 2.976718647045904, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.573505401611328, + "learning_rate": 1e-06, + "loss": 0.9616, + "mean_token_accuracy": 0.7181329727172852, + "num_tokens": 701437492.0, + "step": 27106 + }, + { + "epoch": 2.9768284647485173, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5920872688293457, + "learning_rate": 1e-06, + "loss": 0.9999, + "mean_token_accuracy": 0.7062320113182068, + "num_tokens": 701467832.0, + "step": 27107 + }, + { + "epoch": 2.976938282451131, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9011332988739014, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7044411897659302, + "num_tokens": 701495198.0, + "step": 27108 + }, + { + "epoch": 2.9770481001537448, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4287803173065186, + "learning_rate": 1e-06, + "loss": 1.047, + "mean_token_accuracy": 0.6995208263397217, + "num_tokens": 701523997.0, + "step": 27109 + }, + { + "epoch": 2.9771579178563585, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.719604730606079, + "learning_rate": 1e-06, + "loss": 1.0834, + "mean_token_accuracy": 0.6886304616928101, + "num_tokens": 701549938.0, + "step": 27110 + }, + { + "epoch": 2.9772677355589723, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.469957113265991, + "learning_rate": 1e-06, + "loss": 1.033, + "mean_token_accuracy": 0.7025870084762573, + "num_tokens": 701576929.0, + "step": 27111 + }, + { + "epoch": 2.9773775532615856, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0308351516723633, + "learning_rate": 1e-06, + "loss": 0.9801, + "mean_token_accuracy": 0.7058645486831665, + "num_tokens": 701594938.0, + "step": 27112 + }, + { + "epoch": 2.9774873709641994, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.699435234069824, + "learning_rate": 1e-06, + "loss": 1.0686, + "mean_token_accuracy": 0.6924116611480713, + "num_tokens": 701620664.0, + "step": 27113 + }, + { + "epoch": 2.977597188666813, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7681047916412354, + "learning_rate": 1e-06, + "loss": 0.9219, + "mean_token_accuracy": 0.7263140678405762, + "num_tokens": 701642530.0, + "step": 27114 + }, + { + "epoch": 2.977707006369427, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6303796768188477, + "learning_rate": 1e-06, + "loss": 1.0325, + "mean_token_accuracy": 0.7025567889213562, + "num_tokens": 701667475.0, + "step": 27115 + }, + { + "epoch": 2.9778168240720406, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.186361074447632, + "learning_rate": 1e-06, + "loss": 1.0394, + "mean_token_accuracy": 0.7030845880508423, + "num_tokens": 701689185.0, + "step": 27116 + }, + { + "epoch": 2.977926641774654, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5575685501098633, + "learning_rate": 1e-06, + "loss": 0.8977, + "mean_token_accuracy": 0.7312412858009338, + "num_tokens": 701717697.0, + "step": 27117 + }, + { + "epoch": 2.9780364594772677, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4322457313537598, + "learning_rate": 1e-06, + "loss": 0.9302, + "mean_token_accuracy": 0.7277019023895264, + "num_tokens": 701743206.0, + "step": 27118 + }, + { + "epoch": 2.9781462771798815, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.479877471923828, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.713397204875946, + "num_tokens": 701766654.0, + "step": 27119 + }, + { + "epoch": 2.978256094882495, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.619575262069702, + "learning_rate": 1e-06, + "loss": 1.0296, + "mean_token_accuracy": 0.7026091814041138, + "num_tokens": 701789536.0, + "step": 27120 + }, + { + "epoch": 2.978365912585109, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.618069887161255, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7122886180877686, + "num_tokens": 701812744.0, + "step": 27121 + }, + { + "epoch": 2.9784757302877223, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6397242546081543, + "learning_rate": 1e-06, + "loss": 0.9497, + "mean_token_accuracy": 0.7236661911010742, + "num_tokens": 701836365.0, + "step": 27122 + }, + { + "epoch": 2.978585547990336, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7661163806915283, + "learning_rate": 1e-06, + "loss": 0.9769, + "mean_token_accuracy": 0.7141996026039124, + "num_tokens": 701857860.0, + "step": 27123 + }, + { + "epoch": 2.97869536569295, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5141470432281494, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.6929333209991455, + "num_tokens": 701886383.0, + "step": 27124 + }, + { + "epoch": 2.978805183395563, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.772848606109619, + "learning_rate": 1e-06, + "loss": 0.894, + "mean_token_accuracy": 0.7466891407966614, + "num_tokens": 701907847.0, + "step": 27125 + }, + { + "epoch": 2.978915001098177, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.514446973800659, + "learning_rate": 1e-06, + "loss": 0.9904, + "mean_token_accuracy": 0.7104087471961975, + "num_tokens": 701936129.0, + "step": 27126 + }, + { + "epoch": 2.9790248188007906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7440311908721924, + "learning_rate": 1e-06, + "loss": 1.0348, + "mean_token_accuracy": 0.7057170271873474, + "num_tokens": 701959243.0, + "step": 27127 + }, + { + "epoch": 2.9791346365034044, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9652035236358643, + "learning_rate": 1e-06, + "loss": 0.9912, + "mean_token_accuracy": 0.7152884006500244, + "num_tokens": 701985609.0, + "step": 27128 + }, + { + "epoch": 2.979244454206018, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6526432037353516, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.6993058919906616, + "num_tokens": 702010856.0, + "step": 27129 + }, + { + "epoch": 2.9793542719086314, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6065590381622314, + "learning_rate": 1e-06, + "loss": 1.0673, + "mean_token_accuracy": 0.6931918859481812, + "num_tokens": 702036322.0, + "step": 27130 + }, + { + "epoch": 2.979464089611245, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.480902671813965, + "learning_rate": 1e-06, + "loss": 0.954, + "mean_token_accuracy": 0.7228626608848572, + "num_tokens": 702064103.0, + "step": 27131 + }, + { + "epoch": 2.979573907313859, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5510475635528564, + "learning_rate": 1e-06, + "loss": 0.9893, + "mean_token_accuracy": 0.7140768766403198, + "num_tokens": 702089732.0, + "step": 27132 + }, + { + "epoch": 2.9796837250164727, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.717858076095581, + "learning_rate": 1e-06, + "loss": 0.9046, + "mean_token_accuracy": 0.7342155575752258, + "num_tokens": 702111620.0, + "step": 27133 + }, + { + "epoch": 2.9797935427190865, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5910513401031494, + "learning_rate": 1e-06, + "loss": 1.1014, + "mean_token_accuracy": 0.6803679466247559, + "num_tokens": 702136971.0, + "step": 27134 + }, + { + "epoch": 2.9799033604217, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4030239582061768, + "learning_rate": 1e-06, + "loss": 1.0035, + "mean_token_accuracy": 0.7051241397857666, + "num_tokens": 702166116.0, + "step": 27135 + }, + { + "epoch": 2.9800131781243135, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.306265354156494, + "learning_rate": 1e-06, + "loss": 0.8704, + "mean_token_accuracy": 0.7358288168907166, + "num_tokens": 702180819.0, + "step": 27136 + }, + { + "epoch": 2.9801229958269273, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7874224185943604, + "learning_rate": 1e-06, + "loss": 0.9772, + "mean_token_accuracy": 0.7152894139289856, + "num_tokens": 702201448.0, + "step": 27137 + }, + { + "epoch": 2.980232813529541, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.684464693069458, + "learning_rate": 1e-06, + "loss": 1.0169, + "mean_token_accuracy": 0.7089076042175293, + "num_tokens": 702225696.0, + "step": 27138 + }, + { + "epoch": 2.980342631232155, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.63391375541687, + "learning_rate": 1e-06, + "loss": 0.9981, + "mean_token_accuracy": 0.7064904570579529, + "num_tokens": 702252138.0, + "step": 27139 + }, + { + "epoch": 2.980452448934768, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7002687454223633, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7198017835617065, + "num_tokens": 702275638.0, + "step": 27140 + }, + { + "epoch": 2.980562266637382, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6346654891967773, + "learning_rate": 1e-06, + "loss": 0.9551, + "mean_token_accuracy": 0.7204647064208984, + "num_tokens": 702298503.0, + "step": 27141 + }, + { + "epoch": 2.9806720843399956, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.68044114112854, + "learning_rate": 1e-06, + "loss": 0.9612, + "mean_token_accuracy": 0.7262787818908691, + "num_tokens": 702321521.0, + "step": 27142 + }, + { + "epoch": 2.9807819020426094, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5943658351898193, + "learning_rate": 1e-06, + "loss": 1.0265, + "mean_token_accuracy": 0.6957359910011292, + "num_tokens": 702343966.0, + "step": 27143 + }, + { + "epoch": 2.980891719745223, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7811408042907715, + "learning_rate": 1e-06, + "loss": 1.0415, + "mean_token_accuracy": 0.697363018989563, + "num_tokens": 702366429.0, + "step": 27144 + }, + { + "epoch": 2.9810015374478365, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9271316528320312, + "learning_rate": 1e-06, + "loss": 0.9847, + "mean_token_accuracy": 0.7107023000717163, + "num_tokens": 702388325.0, + "step": 27145 + }, + { + "epoch": 2.9811113551504502, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4519221782684326, + "learning_rate": 1e-06, + "loss": 1.0913, + "mean_token_accuracy": 0.6939992308616638, + "num_tokens": 702417213.0, + "step": 27146 + }, + { + "epoch": 2.981221172853064, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.637925863265991, + "learning_rate": 1e-06, + "loss": 1.0397, + "mean_token_accuracy": 0.6969321966171265, + "num_tokens": 702441237.0, + "step": 27147 + }, + { + "epoch": 2.9813309905556777, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.392756700515747, + "learning_rate": 1e-06, + "loss": 1.1017, + "mean_token_accuracy": 0.6841012239456177, + "num_tokens": 702474699.0, + "step": 27148 + }, + { + "epoch": 2.9814408082582915, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.37966251373291, + "learning_rate": 1e-06, + "loss": 1.0178, + "mean_token_accuracy": 0.7046794295310974, + "num_tokens": 702504887.0, + "step": 27149 + }, + { + "epoch": 2.981550625960905, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7582552433013916, + "learning_rate": 1e-06, + "loss": 1.0701, + "mean_token_accuracy": 0.6900047063827515, + "num_tokens": 702528723.0, + "step": 27150 + }, + { + "epoch": 2.9816604436635186, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.863588571548462, + "learning_rate": 1e-06, + "loss": 0.9344, + "mean_token_accuracy": 0.724947452545166, + "num_tokens": 702551150.0, + "step": 27151 + }, + { + "epoch": 2.9817702613661323, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.70231294631958, + "learning_rate": 1e-06, + "loss": 0.9463, + "mean_token_accuracy": 0.7283879518508911, + "num_tokens": 702574087.0, + "step": 27152 + }, + { + "epoch": 2.9818800790687456, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.569023847579956, + "learning_rate": 1e-06, + "loss": 1.0264, + "mean_token_accuracy": 0.7000356316566467, + "num_tokens": 702601130.0, + "step": 27153 + }, + { + "epoch": 2.9819898967713594, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.345799207687378, + "learning_rate": 1e-06, + "loss": 1.0297, + "mean_token_accuracy": 0.7092291116714478, + "num_tokens": 702633442.0, + "step": 27154 + }, + { + "epoch": 2.982099714473973, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.613436698913574, + "learning_rate": 1e-06, + "loss": 0.9465, + "mean_token_accuracy": 0.7179868221282959, + "num_tokens": 702659621.0, + "step": 27155 + }, + { + "epoch": 2.982209532176587, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.444707155227661, + "learning_rate": 1e-06, + "loss": 0.9961, + "mean_token_accuracy": 0.7142435312271118, + "num_tokens": 702687776.0, + "step": 27156 + }, + { + "epoch": 2.9823193498792007, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6870076656341553, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7122045755386353, + "num_tokens": 702712274.0, + "step": 27157 + }, + { + "epoch": 2.982429167581814, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5995590686798096, + "learning_rate": 1e-06, + "loss": 0.9877, + "mean_token_accuracy": 0.712409257888794, + "num_tokens": 702740462.0, + "step": 27158 + }, + { + "epoch": 2.9825389852844277, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.650634288787842, + "learning_rate": 1e-06, + "loss": 0.9547, + "mean_token_accuracy": 0.7169377207756042, + "num_tokens": 702764661.0, + "step": 27159 + }, + { + "epoch": 2.9826488029870415, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.549980401992798, + "learning_rate": 1e-06, + "loss": 1.0088, + "mean_token_accuracy": 0.7032207250595093, + "num_tokens": 702792044.0, + "step": 27160 + }, + { + "epoch": 2.9827586206896552, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7591333389282227, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7127856612205505, + "num_tokens": 702813905.0, + "step": 27161 + }, + { + "epoch": 2.982868438392269, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8314967155456543, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7097256779670715, + "num_tokens": 702834398.0, + "step": 27162 + }, + { + "epoch": 2.9829782560948823, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.618140459060669, + "learning_rate": 1e-06, + "loss": 1.0584, + "mean_token_accuracy": 0.6964563131332397, + "num_tokens": 702860805.0, + "step": 27163 + }, + { + "epoch": 2.983088073797496, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.88642954826355, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7116116285324097, + "num_tokens": 702887110.0, + "step": 27164 + }, + { + "epoch": 2.98319789150011, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 3.154268264770508, + "learning_rate": 1e-06, + "loss": 1.0372, + "mean_token_accuracy": 0.6990504264831543, + "num_tokens": 702914694.0, + "step": 27165 + }, + { + "epoch": 2.9833077092027236, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.53227162361145, + "learning_rate": 1e-06, + "loss": 1.0433, + "mean_token_accuracy": 0.6934486627578735, + "num_tokens": 702941666.0, + "step": 27166 + }, + { + "epoch": 2.9834175269053373, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.8810222148895264, + "learning_rate": 1e-06, + "loss": 0.9582, + "mean_token_accuracy": 0.7179574966430664, + "num_tokens": 702962865.0, + "step": 27167 + }, + { + "epoch": 2.9835273446079507, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.511709213256836, + "learning_rate": 1e-06, + "loss": 0.9243, + "mean_token_accuracy": 0.7213976979255676, + "num_tokens": 702991618.0, + "step": 27168 + }, + { + "epoch": 2.9836371623105644, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.911275625228882, + "learning_rate": 1e-06, + "loss": 1.0792, + "mean_token_accuracy": 0.6890289783477783, + "num_tokens": 703023298.0, + "step": 27169 + }, + { + "epoch": 2.983746980013178, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.436262845993042, + "learning_rate": 1e-06, + "loss": 1.0512, + "mean_token_accuracy": 0.6983431577682495, + "num_tokens": 703053813.0, + "step": 27170 + }, + { + "epoch": 2.983856797715792, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0401110649108887, + "learning_rate": 1e-06, + "loss": 0.9269, + "mean_token_accuracy": 0.7261791229248047, + "num_tokens": 703072059.0, + "step": 27171 + }, + { + "epoch": 2.9839666154184057, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8235528469085693, + "learning_rate": 1e-06, + "loss": 1.043, + "mean_token_accuracy": 0.6975482702255249, + "num_tokens": 703096897.0, + "step": 27172 + }, + { + "epoch": 2.984076433121019, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0787956714630127, + "learning_rate": 1e-06, + "loss": 0.9985, + "mean_token_accuracy": 0.7049281597137451, + "num_tokens": 703117645.0, + "step": 27173 + }, + { + "epoch": 2.9841862508236328, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.527846336364746, + "learning_rate": 1e-06, + "loss": 0.9239, + "mean_token_accuracy": 0.7375881671905518, + "num_tokens": 703139908.0, + "step": 27174 + }, + { + "epoch": 2.9842960685262465, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5540566444396973, + "learning_rate": 1e-06, + "loss": 0.925, + "mean_token_accuracy": 0.7300244569778442, + "num_tokens": 703163662.0, + "step": 27175 + }, + { + "epoch": 2.98440588622886, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5002095699310303, + "learning_rate": 1e-06, + "loss": 0.9965, + "mean_token_accuracy": 0.71703040599823, + "num_tokens": 703188222.0, + "step": 27176 + }, + { + "epoch": 2.984515703931474, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7506675720214844, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7124158143997192, + "num_tokens": 703213448.0, + "step": 27177 + }, + { + "epoch": 2.9846255216340873, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.489396810531616, + "learning_rate": 1e-06, + "loss": 0.9758, + "mean_token_accuracy": 0.7134362459182739, + "num_tokens": 703239870.0, + "step": 27178 + }, + { + "epoch": 2.984735339336701, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6346068382263184, + "learning_rate": 1e-06, + "loss": 0.9857, + "mean_token_accuracy": 0.709018349647522, + "num_tokens": 703262075.0, + "step": 27179 + }, + { + "epoch": 2.984845157039315, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 3.0094573497772217, + "learning_rate": 1e-06, + "loss": 1.0071, + "mean_token_accuracy": 0.7044739127159119, + "num_tokens": 703286704.0, + "step": 27180 + }, + { + "epoch": 2.984954974741928, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.471386671066284, + "learning_rate": 1e-06, + "loss": 1.0734, + "mean_token_accuracy": 0.6862420439720154, + "num_tokens": 703315082.0, + "step": 27181 + }, + { + "epoch": 2.985064792444542, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.549163579940796, + "learning_rate": 1e-06, + "loss": 0.9415, + "mean_token_accuracy": 0.7199584245681763, + "num_tokens": 703342324.0, + "step": 27182 + }, + { + "epoch": 2.9851746101471557, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4096739292144775, + "learning_rate": 1e-06, + "loss": 1.01, + "mean_token_accuracy": 0.707129955291748, + "num_tokens": 703373898.0, + "step": 27183 + }, + { + "epoch": 2.9852844278497694, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9015421867370605, + "learning_rate": 1e-06, + "loss": 0.9579, + "mean_token_accuracy": 0.7215420007705688, + "num_tokens": 703394399.0, + "step": 27184 + }, + { + "epoch": 2.985394245552383, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.674236297607422, + "learning_rate": 1e-06, + "loss": 1.0424, + "mean_token_accuracy": 0.7013102173805237, + "num_tokens": 703418105.0, + "step": 27185 + }, + { + "epoch": 2.9855040632549965, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4723286628723145, + "learning_rate": 1e-06, + "loss": 1.0446, + "mean_token_accuracy": 0.6965576410293579, + "num_tokens": 703445569.0, + "step": 27186 + }, + { + "epoch": 2.9856138809576103, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.836690902709961, + "learning_rate": 1e-06, + "loss": 1.0413, + "mean_token_accuracy": 0.702671229839325, + "num_tokens": 703466434.0, + "step": 27187 + }, + { + "epoch": 2.985723698660224, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6436996459960938, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7119319438934326, + "num_tokens": 703493661.0, + "step": 27188 + }, + { + "epoch": 2.985833516362838, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.871708869934082, + "learning_rate": 1e-06, + "loss": 1.0171, + "mean_token_accuracy": 0.7104068994522095, + "num_tokens": 703515129.0, + "step": 27189 + }, + { + "epoch": 2.9859433340654515, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0353434085845947, + "learning_rate": 1e-06, + "loss": 0.9587, + "mean_token_accuracy": 0.7206248641014099, + "num_tokens": 703531781.0, + "step": 27190 + }, + { + "epoch": 2.986053151768065, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.3196511268615723, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6943697929382324, + "num_tokens": 703564640.0, + "step": 27191 + }, + { + "epoch": 2.9861629694706786, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5378987789154053, + "learning_rate": 1e-06, + "loss": 0.9982, + "mean_token_accuracy": 0.7086971998214722, + "num_tokens": 703590692.0, + "step": 27192 + }, + { + "epoch": 2.9862727871732924, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.743989944458008, + "learning_rate": 1e-06, + "loss": 1.0056, + "mean_token_accuracy": 0.7044501304626465, + "num_tokens": 703614104.0, + "step": 27193 + }, + { + "epoch": 2.986382604875906, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.490600824356079, + "learning_rate": 1e-06, + "loss": 1.0117, + "mean_token_accuracy": 0.7120349407196045, + "num_tokens": 703643107.0, + "step": 27194 + }, + { + "epoch": 2.98649242257852, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.355808973312378, + "learning_rate": 1e-06, + "loss": 1.0745, + "mean_token_accuracy": 0.684994101524353, + "num_tokens": 703675675.0, + "step": 27195 + }, + { + "epoch": 2.986602240281133, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4684088230133057, + "learning_rate": 1e-06, + "loss": 1.0145, + "mean_token_accuracy": 0.7033388614654541, + "num_tokens": 703705056.0, + "step": 27196 + }, + { + "epoch": 2.986712057983747, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.651515483856201, + "learning_rate": 1e-06, + "loss": 0.9619, + "mean_token_accuracy": 0.732591450214386, + "num_tokens": 703728771.0, + "step": 27197 + }, + { + "epoch": 2.9868218756863607, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.398442268371582, + "learning_rate": 1e-06, + "loss": 1.0339, + "mean_token_accuracy": 0.7104451656341553, + "num_tokens": 703757949.0, + "step": 27198 + }, + { + "epoch": 2.9869316933889745, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.690608263015747, + "learning_rate": 1e-06, + "loss": 0.961, + "mean_token_accuracy": 0.717480480670929, + "num_tokens": 703781668.0, + "step": 27199 + }, + { + "epoch": 2.987041511091588, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.541450262069702, + "learning_rate": 1e-06, + "loss": 0.9673, + "mean_token_accuracy": 0.7188557386398315, + "num_tokens": 703810232.0, + "step": 27200 + }, + { + "epoch": 2.9871513287942015, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4599459171295166, + "learning_rate": 1e-06, + "loss": 1.0039, + "mean_token_accuracy": 0.7121289968490601, + "num_tokens": 703837505.0, + "step": 27201 + }, + { + "epoch": 2.9872611464968153, + "ewc_loss": 1.0669231414794922e-05, + "grad_norm": 3.084125518798828, + "learning_rate": 1e-06, + "loss": 1.0434, + "mean_token_accuracy": 0.702002763748169, + "num_tokens": 703864601.0, + "step": 27202 + }, + { + "epoch": 2.987370964199429, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7953195571899414, + "learning_rate": 1e-06, + "loss": 1.1298, + "mean_token_accuracy": 0.675614595413208, + "num_tokens": 703896134.0, + "step": 27203 + }, + { + "epoch": 2.9874807819020424, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9578118324279785, + "learning_rate": 1e-06, + "loss": 0.978, + "mean_token_accuracy": 0.7165544033050537, + "num_tokens": 703916331.0, + "step": 27204 + }, + { + "epoch": 2.987590599604656, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4992449283599854, + "learning_rate": 1e-06, + "loss": 1.0107, + "mean_token_accuracy": 0.7078179121017456, + "num_tokens": 703944317.0, + "step": 27205 + }, + { + "epoch": 2.98770041730727, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9382922649383545, + "learning_rate": 1e-06, + "loss": 0.9956, + "mean_token_accuracy": 0.7133884429931641, + "num_tokens": 703965788.0, + "step": 27206 + }, + { + "epoch": 2.9878102350098836, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6206114292144775, + "learning_rate": 1e-06, + "loss": 1.0003, + "mean_token_accuracy": 0.7114014625549316, + "num_tokens": 703991689.0, + "step": 27207 + }, + { + "epoch": 2.9879200527124974, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6217916011810303, + "learning_rate": 1e-06, + "loss": 0.973, + "mean_token_accuracy": 0.7189639210700989, + "num_tokens": 704018020.0, + "step": 27208 + }, + { + "epoch": 2.9880298704151107, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7710838317871094, + "learning_rate": 1e-06, + "loss": 1.0102, + "mean_token_accuracy": 0.7031506299972534, + "num_tokens": 704040366.0, + "step": 27209 + }, + { + "epoch": 2.9881396881177245, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.526768207550049, + "learning_rate": 1e-06, + "loss": 1.0295, + "mean_token_accuracy": 0.7092423439025879, + "num_tokens": 704064833.0, + "step": 27210 + }, + { + "epoch": 2.988249505820338, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.833735466003418, + "learning_rate": 1e-06, + "loss": 0.9864, + "mean_token_accuracy": 0.7184352278709412, + "num_tokens": 704087321.0, + "step": 27211 + }, + { + "epoch": 2.988359323522952, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.543201208114624, + "learning_rate": 1e-06, + "loss": 1.1071, + "mean_token_accuracy": 0.6843331456184387, + "num_tokens": 704115895.0, + "step": 27212 + }, + { + "epoch": 2.9884691412255657, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.239260196685791, + "learning_rate": 1e-06, + "loss": 1.0063, + "mean_token_accuracy": 0.7187854051589966, + "num_tokens": 704148074.0, + "step": 27213 + }, + { + "epoch": 2.988578958928179, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.640371561050415, + "learning_rate": 1e-06, + "loss": 1.0349, + "mean_token_accuracy": 0.701758861541748, + "num_tokens": 704175384.0, + "step": 27214 + }, + { + "epoch": 2.988688776630793, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5741617679595947, + "learning_rate": 1e-06, + "loss": 0.9753, + "mean_token_accuracy": 0.7187963724136353, + "num_tokens": 704198495.0, + "step": 27215 + }, + { + "epoch": 2.9887985943334066, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.53570294380188, + "learning_rate": 1e-06, + "loss": 1.0456, + "mean_token_accuracy": 0.6939100027084351, + "num_tokens": 704226538.0, + "step": 27216 + }, + { + "epoch": 2.9889084120360203, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.779533863067627, + "learning_rate": 1e-06, + "loss": 0.9204, + "mean_token_accuracy": 0.7266455888748169, + "num_tokens": 704248083.0, + "step": 27217 + }, + { + "epoch": 2.989018229738634, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.884972333908081, + "learning_rate": 1e-06, + "loss": 0.8954, + "mean_token_accuracy": 0.7301756143569946, + "num_tokens": 704267508.0, + "step": 27218 + }, + { + "epoch": 2.9891280474412474, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4582231044769287, + "learning_rate": 1e-06, + "loss": 1.0891, + "mean_token_accuracy": 0.687706708908081, + "num_tokens": 704298993.0, + "step": 27219 + }, + { + "epoch": 2.989237865143861, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8482255935668945, + "learning_rate": 1e-06, + "loss": 0.9625, + "mean_token_accuracy": 0.7204856872558594, + "num_tokens": 704323108.0, + "step": 27220 + }, + { + "epoch": 2.989347682846475, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9596309661865234, + "learning_rate": 1e-06, + "loss": 1.0094, + "mean_token_accuracy": 0.7106765508651733, + "num_tokens": 704343373.0, + "step": 27221 + }, + { + "epoch": 2.9894575005490887, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.682267904281616, + "learning_rate": 1e-06, + "loss": 1.0687, + "mean_token_accuracy": 0.7000961303710938, + "num_tokens": 704366538.0, + "step": 27222 + }, + { + "epoch": 2.9895673182517024, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.9735236167907715, + "learning_rate": 1e-06, + "loss": 1.0199, + "mean_token_accuracy": 0.700742781162262, + "num_tokens": 704390386.0, + "step": 27223 + }, + { + "epoch": 2.9896771359543157, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6706109046936035, + "learning_rate": 1e-06, + "loss": 1.085, + "mean_token_accuracy": 0.6933699250221252, + "num_tokens": 704415032.0, + "step": 27224 + }, + { + "epoch": 2.9897869536569295, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4755451679229736, + "learning_rate": 1e-06, + "loss": 0.9479, + "mean_token_accuracy": 0.7279822826385498, + "num_tokens": 704442716.0, + "step": 27225 + }, + { + "epoch": 2.9898967713595432, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.634247064590454, + "learning_rate": 1e-06, + "loss": 0.9646, + "mean_token_accuracy": 0.7195887565612793, + "num_tokens": 704467676.0, + "step": 27226 + }, + { + "epoch": 2.9900065890621566, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7798941135406494, + "learning_rate": 1e-06, + "loss": 0.9484, + "mean_token_accuracy": 0.7170089483261108, + "num_tokens": 704488473.0, + "step": 27227 + }, + { + "epoch": 2.9901164067647708, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2345986366271973, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7043018341064453, + "num_tokens": 704520933.0, + "step": 27228 + }, + { + "epoch": 2.990226224467384, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.302037239074707, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7168346643447876, + "num_tokens": 704550513.0, + "step": 27229 + }, + { + "epoch": 2.990336042169998, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5710554122924805, + "learning_rate": 1e-06, + "loss": 1.1119, + "mean_token_accuracy": 0.6873677372932434, + "num_tokens": 704577609.0, + "step": 27230 + }, + { + "epoch": 2.9904458598726116, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7399682998657227, + "learning_rate": 1e-06, + "loss": 0.9872, + "mean_token_accuracy": 0.7120176553726196, + "num_tokens": 704601933.0, + "step": 27231 + }, + { + "epoch": 2.990555677575225, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.716433048248291, + "learning_rate": 1e-06, + "loss": 1.0141, + "mean_token_accuracy": 0.7091737985610962, + "num_tokens": 704624988.0, + "step": 27232 + }, + { + "epoch": 2.9906654952778386, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4752583503723145, + "learning_rate": 1e-06, + "loss": 0.9866, + "mean_token_accuracy": 0.7080932855606079, + "num_tokens": 704653433.0, + "step": 27233 + }, + { + "epoch": 2.9907753129804524, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.520132541656494, + "learning_rate": 1e-06, + "loss": 0.9813, + "mean_token_accuracy": 0.7106191515922546, + "num_tokens": 704681291.0, + "step": 27234 + }, + { + "epoch": 2.990885130683066, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.990550994873047, + "learning_rate": 1e-06, + "loss": 0.9324, + "mean_token_accuracy": 0.7225741744041443, + "num_tokens": 704704250.0, + "step": 27235 + }, + { + "epoch": 2.99099494838568, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7084059715270996, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7256326675415039, + "num_tokens": 704729508.0, + "step": 27236 + }, + { + "epoch": 2.9911047660882932, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7706775665283203, + "learning_rate": 1e-06, + "loss": 0.9834, + "mean_token_accuracy": 0.7115432024002075, + "num_tokens": 704753815.0, + "step": 27237 + }, + { + "epoch": 2.991214583790907, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4966185092926025, + "learning_rate": 1e-06, + "loss": 1.0955, + "mean_token_accuracy": 0.6814168691635132, + "num_tokens": 704785221.0, + "step": 27238 + }, + { + "epoch": 2.9913244014935207, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3842992782592773, + "learning_rate": 1e-06, + "loss": 1.0649, + "mean_token_accuracy": 0.7034873366355896, + "num_tokens": 704815571.0, + "step": 27239 + }, + { + "epoch": 2.9914342191961345, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.452277660369873, + "learning_rate": 1e-06, + "loss": 1.0069, + "mean_token_accuracy": 0.7085666656494141, + "num_tokens": 704841008.0, + "step": 27240 + }, + { + "epoch": 2.9915440368987483, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4927666187286377, + "learning_rate": 1e-06, + "loss": 0.9963, + "mean_token_accuracy": 0.7137142419815063, + "num_tokens": 704867545.0, + "step": 27241 + }, + { + "epoch": 2.9916538546013616, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7700469493865967, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7092880010604858, + "num_tokens": 704891252.0, + "step": 27242 + }, + { + "epoch": 2.9917636723039753, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.935438394546509, + "learning_rate": 1e-06, + "loss": 0.9988, + "mean_token_accuracy": 0.7132242918014526, + "num_tokens": 704910961.0, + "step": 27243 + }, + { + "epoch": 2.991873490006589, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.39056396484375, + "learning_rate": 1e-06, + "loss": 1.0759, + "mean_token_accuracy": 0.6903312802314758, + "num_tokens": 704943947.0, + "step": 27244 + }, + { + "epoch": 2.991983307709203, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.439586877822876, + "learning_rate": 1e-06, + "loss": 1.0034, + "mean_token_accuracy": 0.7067645192146301, + "num_tokens": 704974192.0, + "step": 27245 + }, + { + "epoch": 2.9920931254118166, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.900559186935425, + "learning_rate": 1e-06, + "loss": 0.9773, + "mean_token_accuracy": 0.7232249975204468, + "num_tokens": 704995669.0, + "step": 27246 + }, + { + "epoch": 2.99220294311443, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6515607833862305, + "learning_rate": 1e-06, + "loss": 0.9516, + "mean_token_accuracy": 0.7156533002853394, + "num_tokens": 705020032.0, + "step": 27247 + }, + { + "epoch": 2.9923127608170437, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.647221088409424, + "learning_rate": 1e-06, + "loss": 1.0864, + "mean_token_accuracy": 0.6901647448539734, + "num_tokens": 705048849.0, + "step": 27248 + }, + { + "epoch": 2.9924225785196574, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7965500354766846, + "learning_rate": 1e-06, + "loss": 0.9812, + "mean_token_accuracy": 0.7145498991012573, + "num_tokens": 705077489.0, + "step": 27249 + }, + { + "epoch": 2.992532396222271, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.432969570159912, + "learning_rate": 1e-06, + "loss": 0.9838, + "mean_token_accuracy": 0.7181591987609863, + "num_tokens": 705105991.0, + "step": 27250 + }, + { + "epoch": 2.992642213924885, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.939943790435791, + "learning_rate": 1e-06, + "loss": 0.9865, + "mean_token_accuracy": 0.7143413424491882, + "num_tokens": 705126974.0, + "step": 27251 + }, + { + "epoch": 2.9927520316274983, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5958895683288574, + "learning_rate": 1e-06, + "loss": 1.0822, + "mean_token_accuracy": 0.685967206954956, + "num_tokens": 705151444.0, + "step": 27252 + }, + { + "epoch": 2.992861849330112, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.792132616043091, + "learning_rate": 1e-06, + "loss": 0.9685, + "mean_token_accuracy": 0.7230597138404846, + "num_tokens": 705172830.0, + "step": 27253 + }, + { + "epoch": 2.9929716670327258, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5341858863830566, + "learning_rate": 1e-06, + "loss": 1.0404, + "mean_token_accuracy": 0.7052576541900635, + "num_tokens": 705202731.0, + "step": 27254 + }, + { + "epoch": 2.993081484735339, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9687392711639404, + "learning_rate": 1e-06, + "loss": 0.9825, + "mean_token_accuracy": 0.7094364166259766, + "num_tokens": 705223406.0, + "step": 27255 + }, + { + "epoch": 2.993191302437953, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4892184734344482, + "learning_rate": 1e-06, + "loss": 1.0448, + "mean_token_accuracy": 0.6965134143829346, + "num_tokens": 705252641.0, + "step": 27256 + }, + { + "epoch": 2.9933011201405666, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4656739234924316, + "learning_rate": 1e-06, + "loss": 1.0255, + "mean_token_accuracy": 0.7032736539840698, + "num_tokens": 705283871.0, + "step": 27257 + }, + { + "epoch": 2.9934109378431804, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4989848136901855, + "learning_rate": 1e-06, + "loss": 0.9736, + "mean_token_accuracy": 0.7155113816261292, + "num_tokens": 705308264.0, + "step": 27258 + }, + { + "epoch": 2.993520755545794, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8257877826690674, + "learning_rate": 1e-06, + "loss": 0.9381, + "mean_token_accuracy": 0.7210155129432678, + "num_tokens": 705331376.0, + "step": 27259 + }, + { + "epoch": 2.9936305732484074, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.478646755218506, + "learning_rate": 1e-06, + "loss": 1.0886, + "mean_token_accuracy": 0.6827265024185181, + "num_tokens": 705359289.0, + "step": 27260 + }, + { + "epoch": 2.993740390951021, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5827507972717285, + "learning_rate": 1e-06, + "loss": 1.034, + "mean_token_accuracy": 0.6950865983963013, + "num_tokens": 705386707.0, + "step": 27261 + }, + { + "epoch": 2.993850208653635, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.4543099403381348, + "learning_rate": 1e-06, + "loss": 0.949, + "mean_token_accuracy": 0.7190960645675659, + "num_tokens": 705418513.0, + "step": 27262 + }, + { + "epoch": 2.9939600263562487, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.2840163707733154, + "learning_rate": 1e-06, + "loss": 0.937, + "mean_token_accuracy": 0.7240873575210571, + "num_tokens": 705447497.0, + "step": 27263 + }, + { + "epoch": 2.9940698440588625, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.455721855163574, + "learning_rate": 1e-06, + "loss": 1.0009, + "mean_token_accuracy": 0.7075196504592896, + "num_tokens": 705474844.0, + "step": 27264 + }, + { + "epoch": 2.9941796617614758, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.715803861618042, + "learning_rate": 1e-06, + "loss": 0.9133, + "mean_token_accuracy": 0.7356013059616089, + "num_tokens": 705496732.0, + "step": 27265 + }, + { + "epoch": 2.9942894794640895, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.696368932723999, + "learning_rate": 1e-06, + "loss": 0.9658, + "mean_token_accuracy": 0.7150681614875793, + "num_tokens": 705519700.0, + "step": 27266 + }, + { + "epoch": 2.9943992971667033, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.510160207748413, + "learning_rate": 1e-06, + "loss": 0.9645, + "mean_token_accuracy": 0.7184260487556458, + "num_tokens": 705546443.0, + "step": 27267 + }, + { + "epoch": 2.994509114869317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7109405994415283, + "learning_rate": 1e-06, + "loss": 1.004, + "mean_token_accuracy": 0.7153365015983582, + "num_tokens": 705569904.0, + "step": 27268 + }, + { + "epoch": 2.994618932571931, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.902775764465332, + "learning_rate": 1e-06, + "loss": 0.9984, + "mean_token_accuracy": 0.7167902588844299, + "num_tokens": 705593737.0, + "step": 27269 + }, + { + "epoch": 2.994728750274544, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5464940071105957, + "learning_rate": 1e-06, + "loss": 0.9896, + "mean_token_accuracy": 0.7084099650382996, + "num_tokens": 705620490.0, + "step": 27270 + }, + { + "epoch": 2.994838567977158, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8225228786468506, + "learning_rate": 1e-06, + "loss": 0.9627, + "mean_token_accuracy": 0.7265048027038574, + "num_tokens": 705641312.0, + "step": 27271 + }, + { + "epoch": 2.9949483856797716, + "ewc_loss": 1.0907649993896484e-05, + "grad_norm": 4.221529960632324, + "learning_rate": 1e-06, + "loss": 0.9581, + "mean_token_accuracy": 0.7224767208099365, + "num_tokens": 705668248.0, + "step": 27272 + }, + { + "epoch": 2.9950582033823854, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.729308605194092, + "learning_rate": 1e-06, + "loss": 1.0076, + "mean_token_accuracy": 0.7021812200546265, + "num_tokens": 705696773.0, + "step": 27273 + }, + { + "epoch": 2.995168021084999, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.456271171569824, + "learning_rate": 1e-06, + "loss": 1.0577, + "mean_token_accuracy": 0.6965153217315674, + "num_tokens": 705728765.0, + "step": 27274 + }, + { + "epoch": 2.9952778387876124, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5014731884002686, + "learning_rate": 1e-06, + "loss": 1.0546, + "mean_token_accuracy": 0.6913372874259949, + "num_tokens": 705755618.0, + "step": 27275 + }, + { + "epoch": 2.995387656490226, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4947850704193115, + "learning_rate": 1e-06, + "loss": 1.0246, + "mean_token_accuracy": 0.7027971148490906, + "num_tokens": 705784998.0, + "step": 27276 + }, + { + "epoch": 2.99549747419284, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.697516918182373, + "learning_rate": 1e-06, + "loss": 0.9011, + "mean_token_accuracy": 0.7415229082107544, + "num_tokens": 705807142.0, + "step": 27277 + }, + { + "epoch": 2.9956072918954533, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.560194969177246, + "learning_rate": 1e-06, + "loss": 1.0027, + "mean_token_accuracy": 0.7065476179122925, + "num_tokens": 705834156.0, + "step": 27278 + }, + { + "epoch": 2.9957171095980675, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5171589851379395, + "learning_rate": 1e-06, + "loss": 0.9989, + "mean_token_accuracy": 0.7080339193344116, + "num_tokens": 705858564.0, + "step": 27279 + }, + { + "epoch": 2.995826927300681, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.409846067428589, + "learning_rate": 1e-06, + "loss": 1.0488, + "mean_token_accuracy": 0.6988322138786316, + "num_tokens": 705888824.0, + "step": 27280 + }, + { + "epoch": 2.9959367450032945, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.79834246635437, + "learning_rate": 1e-06, + "loss": 1.0386, + "mean_token_accuracy": 0.7124370336532593, + "num_tokens": 705912030.0, + "step": 27281 + }, + { + "epoch": 2.9960465627059083, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5987026691436768, + "learning_rate": 1e-06, + "loss": 1.0266, + "mean_token_accuracy": 0.7039890885353088, + "num_tokens": 705937277.0, + "step": 27282 + }, + { + "epoch": 2.9961563804085216, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.7927353382110596, + "learning_rate": 1e-06, + "loss": 0.9387, + "mean_token_accuracy": 0.7220735549926758, + "num_tokens": 705959095.0, + "step": 27283 + }, + { + "epoch": 2.9962661981111354, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.969111919403076, + "learning_rate": 1e-06, + "loss": 0.9435, + "mean_token_accuracy": 0.7232558727264404, + "num_tokens": 705980298.0, + "step": 27284 + }, + { + "epoch": 2.996376015813749, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.6913838386535645, + "learning_rate": 1e-06, + "loss": 0.9092, + "mean_token_accuracy": 0.7347408533096313, + "num_tokens": 706002573.0, + "step": 27285 + }, + { + "epoch": 2.996485833516363, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8153746128082275, + "learning_rate": 1e-06, + "loss": 1.0361, + "mean_token_accuracy": 0.7023268938064575, + "num_tokens": 706024834.0, + "step": 27286 + }, + { + "epoch": 2.9965956512189766, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.715672492980957, + "learning_rate": 1e-06, + "loss": 1.0447, + "mean_token_accuracy": 0.6904383897781372, + "num_tokens": 706050593.0, + "step": 27287 + }, + { + "epoch": 2.99670546892159, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5358164310455322, + "learning_rate": 1e-06, + "loss": 1.0752, + "mean_token_accuracy": 0.6945854425430298, + "num_tokens": 706076878.0, + "step": 27288 + }, + { + "epoch": 2.9968152866242037, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.672266721725464, + "learning_rate": 1e-06, + "loss": 0.8945, + "mean_token_accuracy": 0.7342802882194519, + "num_tokens": 706098592.0, + "step": 27289 + }, + { + "epoch": 2.9969251043268175, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6230740547180176, + "learning_rate": 1e-06, + "loss": 0.9202, + "mean_token_accuracy": 0.7302238941192627, + "num_tokens": 706122101.0, + "step": 27290 + }, + { + "epoch": 2.9970349220294312, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.699035167694092, + "learning_rate": 1e-06, + "loss": 0.9914, + "mean_token_accuracy": 0.7099590301513672, + "num_tokens": 706150293.0, + "step": 27291 + }, + { + "epoch": 2.997144739732045, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7146337032318115, + "learning_rate": 1e-06, + "loss": 0.9528, + "mean_token_accuracy": 0.7329476475715637, + "num_tokens": 706172552.0, + "step": 27292 + }, + { + "epoch": 2.9972545574346583, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4646990299224854, + "learning_rate": 1e-06, + "loss": 1.0163, + "mean_token_accuracy": 0.7059531807899475, + "num_tokens": 706202391.0, + "step": 27293 + }, + { + "epoch": 2.997364375137272, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.6548421382904053, + "learning_rate": 1e-06, + "loss": 1.0517, + "mean_token_accuracy": 0.696317732334137, + "num_tokens": 706226644.0, + "step": 27294 + }, + { + "epoch": 2.997474192839886, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.88509464263916, + "learning_rate": 1e-06, + "loss": 0.9389, + "mean_token_accuracy": 0.7252182960510254, + "num_tokens": 706245912.0, + "step": 27295 + }, + { + "epoch": 2.9975840105424996, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.8133163452148438, + "learning_rate": 1e-06, + "loss": 0.9939, + "mean_token_accuracy": 0.7100213766098022, + "num_tokens": 706268033.0, + "step": 27296 + }, + { + "epoch": 2.9976938282451133, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.7185866832733154, + "learning_rate": 1e-06, + "loss": 0.8937, + "mean_token_accuracy": 0.7341158986091614, + "num_tokens": 706290252.0, + "step": 27297 + }, + { + "epoch": 2.9978036459477266, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4559731483459473, + "learning_rate": 1e-06, + "loss": 1.104, + "mean_token_accuracy": 0.6885204911231995, + "num_tokens": 706320660.0, + "step": 27298 + }, + { + "epoch": 2.9979134636503404, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.766633987426758, + "learning_rate": 1e-06, + "loss": 1.0082, + "mean_token_accuracy": 0.7042466402053833, + "num_tokens": 706342544.0, + "step": 27299 + }, + { + "epoch": 2.998023281352954, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5143508911132812, + "learning_rate": 1e-06, + "loss": 0.9578, + "mean_token_accuracy": 0.7210665941238403, + "num_tokens": 706369294.0, + "step": 27300 + }, + { + "epoch": 2.998133099055568, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.582245349884033, + "learning_rate": 1e-06, + "loss": 0.8661, + "mean_token_accuracy": 0.7476222515106201, + "num_tokens": 706392264.0, + "step": 27301 + }, + { + "epoch": 2.9982429167581817, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.479041814804077, + "learning_rate": 1e-06, + "loss": 1.0452, + "mean_token_accuracy": 0.6988614797592163, + "num_tokens": 706423052.0, + "step": 27302 + }, + { + "epoch": 2.998352734460795, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.611030340194702, + "learning_rate": 1e-06, + "loss": 0.9184, + "mean_token_accuracy": 0.7293212413787842, + "num_tokens": 706448674.0, + "step": 27303 + }, + { + "epoch": 2.9984625521634087, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.556629180908203, + "learning_rate": 1e-06, + "loss": 0.9818, + "mean_token_accuracy": 0.7126346230506897, + "num_tokens": 706475897.0, + "step": 27304 + }, + { + "epoch": 2.9985723698660225, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.4613184928894043, + "learning_rate": 1e-06, + "loss": 0.9546, + "mean_token_accuracy": 0.7216278314590454, + "num_tokens": 706508259.0, + "step": 27305 + }, + { + "epoch": 2.998682187568636, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.9888346195220947, + "learning_rate": 1e-06, + "loss": 1.0129, + "mean_token_accuracy": 0.7106131911277771, + "num_tokens": 706531279.0, + "step": 27306 + }, + { + "epoch": 2.9987920052712496, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.535064697265625, + "learning_rate": 1e-06, + "loss": 1.0381, + "mean_token_accuracy": 0.7003545761108398, + "num_tokens": 706557018.0, + "step": 27307 + }, + { + "epoch": 2.9989018229738633, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.5143258571624756, + "learning_rate": 1e-06, + "loss": 1.0551, + "mean_token_accuracy": 0.7020407915115356, + "num_tokens": 706587542.0, + "step": 27308 + }, + { + "epoch": 2.999011640676477, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 3.0389373302459717, + "learning_rate": 1e-06, + "loss": 1.0175, + "mean_token_accuracy": 0.7127450704574585, + "num_tokens": 706606653.0, + "step": 27309 + }, + { + "epoch": 2.999121458379091, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.867339611053467, + "learning_rate": 1e-06, + "loss": 0.9791, + "mean_token_accuracy": 0.7207731008529663, + "num_tokens": 706631700.0, + "step": 27310 + }, + { + "epoch": 2.999231276081704, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.522280216217041, + "learning_rate": 1e-06, + "loss": 0.963, + "mean_token_accuracy": 0.7166911959648132, + "num_tokens": 706656969.0, + "step": 27311 + }, + { + "epoch": 2.999341093784318, + "ewc_loss": 1.0609626770019531e-05, + "grad_norm": 2.710869073867798, + "learning_rate": 1e-06, + "loss": 1.001, + "mean_token_accuracy": 0.7048547267913818, + "num_tokens": 706681011.0, + "step": 27312 + }, + { + "epoch": 2.9994509114869317, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.3989932537078857, + "learning_rate": 1e-06, + "loss": 0.9804, + "mean_token_accuracy": 0.7142921090126038, + "num_tokens": 706707342.0, + "step": 27313 + }, + { + "epoch": 2.9995607291895454, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.8864362239837646, + "learning_rate": 1e-06, + "loss": 0.9696, + "mean_token_accuracy": 0.7158312797546387, + "num_tokens": 706731699.0, + "step": 27314 + }, + { + "epoch": 2.999670546892159, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.5619146823883057, + "learning_rate": 1e-06, + "loss": 1.0043, + "mean_token_accuracy": 0.7060459852218628, + "num_tokens": 706757635.0, + "step": 27315 + }, + { + "epoch": 2.9997803645947725, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.41471791267395, + "learning_rate": 1e-06, + "loss": 0.9495, + "mean_token_accuracy": 0.7245469689369202, + "num_tokens": 706787813.0, + "step": 27316 + }, + { + "epoch": 2.9998901822973862, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.57116961479187, + "learning_rate": 1e-06, + "loss": 1.0908, + "mean_token_accuracy": 0.6812887191772461, + "num_tokens": 706821902.0, + "step": 27317 + }, + { + "epoch": 3.0, + "ewc_loss": 1.055002212524414e-05, + "grad_norm": 2.511716842651367, + "learning_rate": 1e-06, + "loss": 0.9802, + "mean_token_accuracy": 0.712732195854187, + "num_tokens": 706854136.0, + "step": 27318 + }, + { + "epoch": 3.0, + "ewc_loss": 1.055002212524414e-05, + "step": 27318, + "total_flos": 3.1829322264618205e+19, + "train_loss": 1.0248253059881185, + "train_runtime": 37191.1487, + "train_samples_per_second": 11.752, + "train_steps_per_second": 0.735 + } + ], + "logging_steps": 1, + "max_steps": 27318, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 13659, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 3.1829322264618205e+19, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000..2bf4d54 --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfbc37849dd543d66b83869bf1d40183dd0a02e0257a255dad3a9dbf6254b92e +size 13393