commit e62586f415da1bdd2436607ba237aa20be6bf04d Author: ModelHub XC Date: Thu Oct 15 09:00:15 2026 +0800 初始化项目,由ModelHub XC社区提供模型 Model: laion/exp_tas_low_diversity_traces Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..1a561eb --- /dev/null +++ b/.gitattributes @@ -0,0 +1,56 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text + + +*.bz2 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 +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack 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 +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +saved_model/**/* 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 +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zstandard filter=lfs diff=lfs merge=lfs -text +*.tfevents* filter=lfs diff=lfs merge=lfs -text +*.db* filter=lfs diff=lfs merge=lfs -text +*.ark* filter=lfs diff=lfs merge=lfs -text +**/*ckpt*data* filter=lfs diff=lfs merge=lfs -text +**/*ckpt*.meta filter=lfs diff=lfs merge=lfs -text +**/*ckpt*.index filter=lfs diff=lfs merge=lfs -text + +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.gguf* filter=lfs diff=lfs merge=lfs -text +*.ggml filter=lfs diff=lfs merge=lfs -text +*.llamafile* filter=lfs diff=lfs merge=lfs -text +*.pt2 filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text + +model-00004-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text +model-00002-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text +training_args.bin filter=lfs diff=lfs merge=lfs -text +merges.txt filter=lfs diff=lfs merge=lfs -text +model-00003-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text +vocab.json filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text +model-00001-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text \ No newline at end of file diff --git a/README.md b/README.md new file mode 100644 index 0000000..d551d4c --- /dev/null +++ b/README.md @@ -0,0 +1,61 @@ +--- +library_name: transformers +license: apache-2.0 +base_model: Qwen/Qwen3-8B +tags: +- llama-factory +- full +- generated_from_trainer +model-index: +- name: exp_tas_low_diversity_traces + results: [] +--- + + + +# exp_tas_low_diversity_traces + +This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the DCAgent/exp_tas_low_diversity_traces dataset. + +## Model description + +More information needed + +## Intended uses & limitations + +More information needed + +## Training and evaluation data + +More information needed + +## Training procedure + +### Training hyperparameters + +The following hyperparameters were used during training: +- learning_rate: 4e-05 +- train_batch_size: 1 +- eval_batch_size: 8 +- seed: 42 +- distributed_type: multi-GPU +- num_devices: 8 +- gradient_accumulation_steps: 2 +- total_train_batch_size: 16 +- total_eval_batch_size: 64 +- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments +- lr_scheduler_type: cosine +- lr_scheduler_warmup_ratio: 0.1 +- num_epochs: 7.0 + +### Training results + + + +### Framework versions + +- Transformers 4.57.3 +- Pytorch 2.9.0+cu128 +- Datasets 4.4.1 +- Tokenizers 0.22.1 diff --git a/added_tokens.json b/added_tokens.json new file mode 100644 index 0000000..b54f913 --- /dev/null +++ b/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/all_results.json b/all_results.json new file mode 100644 index 0000000..0de1a8d --- /dev/null +++ b/all_results.json @@ -0,0 +1,16 @@ +{ + "achieved_tflops_per_gpu": 3.461307128091577, + "achieved_tflops_per_gpu_theoretical": 420.1001657900728, + "epoch": 7.0, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11300152540206909, + "mfu_percent": 0.3499804982903516, + "mfu_percent_theoretical": 42.47726651062415, + "total_flos": 8.035723720943206e+17, + "train_loss": 0.23149373163243941, + "train_runtime": 29019.8306, + "train_samples_per_second": 1.981, + "train_steps_per_second": 0.124, + "valid_targets_mean": 2509.4, + "valid_targets_min": 1446 +} \ No newline at end of file diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000..01be9b3 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,89 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {{- messages[0].content + '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('') and message.content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} +{%- endfor %} +{%- for message in messages %} + {%- if message.content is string %} + {%- set content = message.content %} + {%- else %} + {%- set content = '' %} + {%- endif %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- if loop.index0 > ns.last_query_index %} + {%- if loop.last or (not loop.last and reasoning_content) %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content.strip('\n') + '\n\n\n' + content.lstrip('\n') }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/config.json b/config.json new file mode 100644 index 0000000..f8a98d4 --- /dev/null +++ b/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3ForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "dtype": "bfloat16", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "layer_types": [ + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention", + "full_attention" + ], + "max_position_embeddings": 40960, + "max_window_layers": 36, + "model_type": "qwen3", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_scaling": null, + "rope_theta": 1000000, + "sliding_window": null, + "tie_word_embeddings": false, + "transformers_version": "4.57.3", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/configuration.json b/configuration.json new file mode 100644 index 0000000..159097f --- /dev/null +++ b/configuration.json @@ -0,0 +1 @@ +{"framework": "pytorch", "task": "others", "allow_remote": true} \ No newline at end of file diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000..9047f32 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,12 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.6, + "top_k": 20, + "top_p": 0.95, + "transformers_version": "4.57.3" +} diff --git a/merges.txt b/merges.txt new file mode 100644 index 0000000..80c1a19 --- /dev/null +++ b/merges.txt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8831e4f1a044471340f7c0a83d7bd71306a5b867e95fd870f74d0c5308a904d5 +size 1671853 diff --git a/model-00001-of-00004.safetensors b/model-00001-of-00004.safetensors new file mode 100644 index 0000000..32f9687 --- /dev/null +++ b/model-00001-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46b7d4dd04d8bc0e90d99b8fa2e336b785ccb540dfb22cd36e11469f3567cefe +size 4902257696 diff --git a/model-00002-of-00004.safetensors b/model-00002-of-00004.safetensors new file mode 100644 index 0000000..4d7d321 --- /dev/null +++ b/model-00002-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:245fc021801836d5c3bac8a3589458d89fcf271c1b6d7f323fce1ec8e2826f38 +size 4915960368 diff --git a/model-00003-of-00004.safetensors b/model-00003-of-00004.safetensors new file mode 100644 index 0000000..bbc3733 --- /dev/null +++ b/model-00003-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de7eb0cec1af776135f6ccd8c0dc6e2d90a55c443c5890abb692c6125895f086 +size 4983068496 diff --git a/model-00004-of-00004.safetensors b/model-00004-of-00004.safetensors new file mode 100644 index 0000000..e6caa63 --- /dev/null +++ b/model-00004-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a849b7df9aa956f8c975ffe4d6aa15b4d70e406450d10ca6712be3499450f488 +size 1580230264 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000..ba886c0 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,407 @@ +{ + "metadata": { + "total_parameters": 308224, + "total_size": 16381470720 + }, + "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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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-00002-of-00004.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.k_norm.weight": "model-00002-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_norm.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-00002-of-00004.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-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_norm.weight": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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-00003-of-00004.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00003-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-00003-of-00004.safetensors", + "model.layers.31.self_attn.k_norm.weight": "model-00003-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_norm.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.32.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.32.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.32.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.32.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.32.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.32.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.layers.32.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.32.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.32.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.layers.32.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.32.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.33.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.layers.33.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.33.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.34.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.34.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.layers.34.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.layers.34.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.34.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.35.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.35.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.35.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.35.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.35.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.layers.35.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.35.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.layers.35.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.35.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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_norm.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-00001-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_norm.weight": "model-00001-of-00004.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.norm.weight": "model-00004-of-00004.safetensors" + } +} diff --git a/run_summary.json b/run_summary.json new file mode 100644 index 0000000..324a306 --- /dev/null +++ b/run_summary.json @@ -0,0 +1,12 @@ +{ + "agent_name": null, + "training_start": null, + "training_end": null, + "created_by": "DCAgent", + "base_model_name": "Qwen/Qwen3-8B", + "dataset_name": "DCAgent/exp_tas_low_diversity_traces", + "training_type": "SFT", + "training_parameters": "https://huggingface.co/laion/exp_tas_low_diversity_traces/blob/main/config.json", + "wandb_link": "https://wandb.ai/DCAgent/OpenThoughts-Agent/runs/exp_tas_low_diversity_traces_Qwen3-8B", + "traces_location_s3": null +} \ No newline at end of file diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000..ac23c0a --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000..cd71f61 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4 +size 11422654 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..e9dc937 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 32768, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/train_results.json b/train_results.json new file mode 100644 index 0000000..0de1a8d --- /dev/null +++ b/train_results.json @@ -0,0 +1,16 @@ +{ + "achieved_tflops_per_gpu": 3.461307128091577, + "achieved_tflops_per_gpu_theoretical": 420.1001657900728, + "epoch": 7.0, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11300152540206909, + "mfu_percent": 0.3499804982903516, + "mfu_percent_theoretical": 42.47726651062415, + "total_flos": 8.035723720943206e+17, + "train_loss": 0.23149373163243941, + "train_runtime": 29019.8306, + "train_samples_per_second": 1.981, + "train_steps_per_second": 0.124, + "valid_targets_mean": 2509.4, + "valid_targets_min": 1446 +} \ No newline at end of file diff --git a/trainer_log.jsonl b/trainer_log.jsonl new file mode 100644 index 0000000..f546920 --- /dev/null +++ b/trainer_log.jsonl @@ -0,0 +1,720 @@ +{"current_steps": 5, "total_steps": 3598, "loss": 0.708, "lr": 4.444444444444445e-07, "epoch": 0.009737098344693282, "percentage": 0.14, "elapsed_time": "0:00:45", "remaining_time": "9:04:02"} +{"current_steps": 10, "total_steps": 3598, "loss": 0.7467, "lr": 1.0000000000000002e-06, "epoch": 0.019474196689386564, "percentage": 0.28, "elapsed_time": "0:01:29", "remaining_time": "8:55:31"} +{"current_steps": 15, "total_steps": 3598, "loss": 0.723, "lr": 1.5555555555555558e-06, "epoch": 0.029211295034079845, "percentage": 0.42, "elapsed_time": "0:02:10", "remaining_time": "8:38:28"} +{"current_steps": 20, "total_steps": 3598, "loss": 0.7416, "lr": 2.1111111111111114e-06, "epoch": 0.03894839337877313, "percentage": 0.56, "elapsed_time": "0:02:43", "remaining_time": "8:07:00"} +{"current_steps": 25, "total_steps": 3598, "loss": 0.7126, "lr": 2.666666666666667e-06, "epoch": 0.04868549172346641, "percentage": 0.69, "elapsed_time": "0:03:23", "remaining_time": "8:03:46"} +{"current_steps": 30, "total_steps": 3598, "loss": 0.7041, "lr": 3.2222222222222227e-06, "epoch": 0.05842259006815969, "percentage": 0.83, "elapsed_time": "0:03:53", "remaining_time": "7:42:13"} +{"current_steps": 35, "total_steps": 3598, "loss": 0.6225, "lr": 3.777777777777778e-06, "epoch": 0.06815968841285297, "percentage": 0.97, "elapsed_time": "0:04:38", "remaining_time": "7:52:13"} +{"current_steps": 40, "total_steps": 3598, "loss": 0.6163, "lr": 4.333333333333334e-06, "epoch": 0.07789678675754626, "percentage": 1.11, "elapsed_time": "0:05:19", "remaining_time": "7:52:55"} +{"current_steps": 45, "total_steps": 3598, "loss": 0.5893, "lr": 4.888888888888889e-06, "epoch": 0.08763388510223953, "percentage": 1.25, "elapsed_time": "0:05:56", "remaining_time": "7:49:21"} +{"current_steps": 50, "total_steps": 3598, "loss": 0.5577, "lr": 5.444444444444445e-06, "epoch": 0.09737098344693282, "percentage": 1.39, "elapsed_time": "0:06:32", "remaining_time": "7:44:37"} +{"current_steps": 55, "total_steps": 3598, "loss": 0.5395, "lr": 6e-06, "epoch": 0.10710808179162609, "percentage": 1.53, "elapsed_time": "0:07:17", "remaining_time": "7:49:47"} +{"current_steps": 60, "total_steps": 3598, "loss": 0.492, "lr": 6.555555555555556e-06, "epoch": 0.11684518013631938, "percentage": 1.67, "elapsed_time": "0:08:00", "remaining_time": "7:51:51"} +{"current_steps": 65, "total_steps": 3598, "loss": 0.5166, "lr": 7.111111111111112e-06, "epoch": 0.12658227848101267, "percentage": 1.81, "elapsed_time": "0:08:48", "remaining_time": "7:58:34"} +{"current_steps": 70, "total_steps": 3598, "loss": 0.4982, "lr": 7.666666666666667e-06, "epoch": 0.13631937682570594, "percentage": 1.95, "elapsed_time": "0:09:24", "remaining_time": "7:54:13"} +{"current_steps": 75, "total_steps": 3598, "loss": 0.5096, "lr": 8.222222222222222e-06, "epoch": 0.1460564751703992, "percentage": 2.08, "elapsed_time": "0:09:53", "remaining_time": "7:44:38"} +{"current_steps": 80, "total_steps": 3598, "loss": 0.4988, "lr": 8.777777777777778e-06, "epoch": 0.15579357351509251, "percentage": 2.22, "elapsed_time": "0:10:37", "remaining_time": "7:46:58"} +{"current_steps": 85, "total_steps": 3598, "loss": 0.4763, "lr": 9.333333333333334e-06, "epoch": 0.1655306718597858, "percentage": 2.36, "elapsed_time": "0:11:18", "remaining_time": "7:47:42"} +{"current_steps": 90, "total_steps": 3598, "loss": 0.4758, "lr": 9.88888888888889e-06, "epoch": 0.17526777020447906, "percentage": 2.5, "elapsed_time": "0:12:02", "remaining_time": "7:49:38"} +{"current_steps": 95, "total_steps": 3598, "loss": 0.4865, "lr": 1.0444444444444445e-05, "epoch": 0.18500486854917234, "percentage": 2.64, "elapsed_time": "0:12:44", "remaining_time": "7:49:41"} +{"current_steps": 100, "total_steps": 3598, "loss": 0.4653, "lr": 1.1000000000000001e-05, "epoch": 0.19474196689386564, "percentage": 2.78, "elapsed_time": "0:13:22", "remaining_time": "7:48:05"} +{"current_steps": 105, "total_steps": 3598, "loss": 0.4581, "lr": 1.1555555555555556e-05, "epoch": 0.2044790652385589, "percentage": 2.92, "elapsed_time": "0:14:00", "remaining_time": "7:46:16"} +{"current_steps": 110, "total_steps": 3598, "loss": 0.4502, "lr": 1.211111111111111e-05, "epoch": 0.21421616358325218, "percentage": 3.06, "elapsed_time": "0:14:39", "remaining_time": "7:44:50"} +{"current_steps": 115, "total_steps": 3598, "loss": 0.4342, "lr": 1.2666666666666667e-05, "epoch": 0.22395326192794549, "percentage": 3.2, "elapsed_time": "0:15:27", "remaining_time": "7:48:14"} +{"current_steps": 120, "total_steps": 3598, "loss": 0.451, "lr": 1.3222222222222223e-05, "epoch": 0.23369036027263876, "percentage": 3.34, "elapsed_time": "0:16:05", "remaining_time": "7:46:32"} +{"current_steps": 125, "total_steps": 3598, "loss": 0.4221, "lr": 1.377777777777778e-05, "epoch": 0.24342745861733203, "percentage": 3.47, "elapsed_time": "0:16:45", "remaining_time": "7:45:29"} +{"current_steps": 130, "total_steps": 3598, "loss": 0.4111, "lr": 1.4333333333333334e-05, "epoch": 0.25316455696202533, "percentage": 3.61, "elapsed_time": "0:17:36", "remaining_time": "7:49:52"} +{"current_steps": 135, "total_steps": 3598, "loss": 0.4129, "lr": 1.488888888888889e-05, "epoch": 0.2629016553067186, "percentage": 3.75, "elapsed_time": "0:18:15", "remaining_time": "7:48:24"} +{"current_steps": 140, "total_steps": 3598, "loss": 0.4197, "lr": 1.5444444444444446e-05, "epoch": 0.2726387536514119, "percentage": 3.89, "elapsed_time": "0:18:52", "remaining_time": "7:46:05"} +{"current_steps": 145, "total_steps": 3598, "loss": 0.3976, "lr": 1.6000000000000003e-05, "epoch": 0.28237585199610515, "percentage": 4.03, "elapsed_time": "0:19:39", "remaining_time": "7:48:01"} +{"current_steps": 150, "total_steps": 3598, "loss": 0.42, "lr": 1.6555555555555556e-05, "epoch": 0.2921129503407984, "percentage": 4.17, "elapsed_time": "0:20:21", "remaining_time": "7:47:48"} +{"current_steps": 155, "total_steps": 3598, "loss": 0.392, "lr": 1.7111111111111112e-05, "epoch": 0.3018500486854917, "percentage": 4.31, "elapsed_time": "0:21:10", "remaining_time": "7:50:13"} +{"current_steps": 160, "total_steps": 3598, "loss": 0.3874, "lr": 1.7666666666666668e-05, "epoch": 0.31158714703018503, "percentage": 4.45, "elapsed_time": "0:21:58", "remaining_time": "7:52:12"} +{"current_steps": 165, "total_steps": 3598, "loss": 0.3808, "lr": 1.8222222222222224e-05, "epoch": 0.3213242453748783, "percentage": 4.59, "elapsed_time": "0:22:39", "remaining_time": "7:51:23"} +{"current_steps": 170, "total_steps": 3598, "loss": 0.4041, "lr": 1.877777777777778e-05, "epoch": 0.3310613437195716, "percentage": 4.72, "elapsed_time": "0:23:20", "remaining_time": "7:50:45"} +{"current_steps": 175, "total_steps": 3598, "loss": 0.379, "lr": 1.9333333333333333e-05, "epoch": 0.34079844206426485, "percentage": 4.86, "elapsed_time": "0:23:54", "remaining_time": "7:47:47"} +{"current_steps": 180, "total_steps": 3598, "loss": 0.4065, "lr": 1.988888888888889e-05, "epoch": 0.3505355404089581, "percentage": 5.0, "elapsed_time": "0:24:33", "remaining_time": "7:46:29"} +{"current_steps": 185, "total_steps": 3598, "loss": 0.4028, "lr": 2.0444444444444446e-05, "epoch": 0.3602726387536514, "percentage": 5.14, "elapsed_time": "0:25:16", "remaining_time": "7:46:22"} +{"current_steps": 190, "total_steps": 3598, "loss": 0.3993, "lr": 2.1000000000000002e-05, "epoch": 0.37000973709834467, "percentage": 5.28, "elapsed_time": "0:25:58", "remaining_time": "7:46:00"} +{"current_steps": 195, "total_steps": 3598, "loss": 0.4064, "lr": 2.1555555555555555e-05, "epoch": 0.379746835443038, "percentage": 5.42, "elapsed_time": "0:26:35", "remaining_time": "7:44:03"} +{"current_steps": 200, "total_steps": 3598, "loss": 0.3681, "lr": 2.2111111111111115e-05, "epoch": 0.3894839337877313, "percentage": 5.56, "elapsed_time": "0:27:07", "remaining_time": "7:40:45"} +{"current_steps": 205, "total_steps": 3598, "loss": 0.3461, "lr": 2.2666666666666668e-05, "epoch": 0.39922103213242455, "percentage": 5.7, "elapsed_time": "0:27:50", "remaining_time": "7:40:41"} +{"current_steps": 210, "total_steps": 3598, "loss": 0.3701, "lr": 2.3222222222222227e-05, "epoch": 0.4089581304771178, "percentage": 5.84, "elapsed_time": "0:28:29", "remaining_time": "7:39:45"} +{"current_steps": 215, "total_steps": 3598, "loss": 0.3819, "lr": 2.377777777777778e-05, "epoch": 0.4186952288218111, "percentage": 5.98, "elapsed_time": "0:29:07", "remaining_time": "7:38:09"} +{"current_steps": 220, "total_steps": 3598, "loss": 0.403, "lr": 2.4333333333333333e-05, "epoch": 0.42843232716650437, "percentage": 6.11, "elapsed_time": "0:29:50", "remaining_time": "7:38:11"} +{"current_steps": 225, "total_steps": 3598, "loss": 0.3901, "lr": 2.4888888888888893e-05, "epoch": 0.43816942551119764, "percentage": 6.25, "elapsed_time": "0:30:26", "remaining_time": "7:36:20"} +{"current_steps": 230, "total_steps": 3598, "loss": 0.4028, "lr": 2.5444444444444446e-05, "epoch": 0.44790652385589097, "percentage": 6.39, "elapsed_time": "0:31:03", "remaining_time": "7:34:44"} +{"current_steps": 235, "total_steps": 3598, "loss": 0.3932, "lr": 2.6000000000000002e-05, "epoch": 0.45764362220058424, "percentage": 6.53, "elapsed_time": "0:31:40", "remaining_time": "7:33:14"} +{"current_steps": 240, "total_steps": 3598, "loss": 0.396, "lr": 2.6555555555555558e-05, "epoch": 0.4673807205452775, "percentage": 6.67, "elapsed_time": "0:32:13", "remaining_time": "7:30:56"} +{"current_steps": 245, "total_steps": 3598, "loss": 0.3861, "lr": 2.7111111111111114e-05, "epoch": 0.4771178188899708, "percentage": 6.81, "elapsed_time": "0:32:49", "remaining_time": "7:29:14"} +{"current_steps": 250, "total_steps": 3598, "loss": 0.3841, "lr": 2.7666666666666667e-05, "epoch": 0.48685491723466406, "percentage": 6.95, "elapsed_time": "0:33:35", "remaining_time": "7:29:48"} +{"current_steps": 255, "total_steps": 3598, "loss": 0.3897, "lr": 2.8222222222222227e-05, "epoch": 0.49659201557935734, "percentage": 7.09, "elapsed_time": "0:34:14", "remaining_time": "7:28:54"} +{"current_steps": 260, "total_steps": 3598, "loss": 0.3422, "lr": 2.877777777777778e-05, "epoch": 0.5063291139240507, "percentage": 7.23, "elapsed_time": "0:35:04", "remaining_time": "7:30:24"} +{"current_steps": 265, "total_steps": 3598, "loss": 0.3726, "lr": 2.9333333333333333e-05, "epoch": 0.5160662122687439, "percentage": 7.37, "elapsed_time": "0:35:52", "remaining_time": "7:31:14"} +{"current_steps": 270, "total_steps": 3598, "loss": 0.3541, "lr": 2.9888888888888892e-05, "epoch": 0.5258033106134372, "percentage": 7.5, "elapsed_time": "0:36:28", "remaining_time": "7:29:35"} +{"current_steps": 275, "total_steps": 3598, "loss": 0.3688, "lr": 3.0444444444444445e-05, "epoch": 0.5355404089581305, "percentage": 7.64, "elapsed_time": "0:36:58", "remaining_time": "7:26:41"} +{"current_steps": 280, "total_steps": 3598, "loss": 0.3908, "lr": 3.1e-05, "epoch": 0.5452775073028238, "percentage": 7.78, "elapsed_time": "0:37:32", "remaining_time": "7:24:53"} +{"current_steps": 285, "total_steps": 3598, "loss": 0.3506, "lr": 3.155555555555556e-05, "epoch": 0.555014605647517, "percentage": 7.92, "elapsed_time": "0:38:18", "remaining_time": "7:25:17"} +{"current_steps": 290, "total_steps": 3598, "loss": 0.3528, "lr": 3.2111111111111114e-05, "epoch": 0.5647517039922103, "percentage": 8.06, "elapsed_time": "0:38:59", "remaining_time": "7:24:41"} +{"current_steps": 295, "total_steps": 3598, "loss": 0.3548, "lr": 3.266666666666667e-05, "epoch": 0.5744888023369036, "percentage": 8.2, "elapsed_time": "0:39:45", "remaining_time": "7:25:06"} +{"current_steps": 300, "total_steps": 3598, "loss": 0.3643, "lr": 3.3222222222222226e-05, "epoch": 0.5842259006815969, "percentage": 8.34, "elapsed_time": "0:40:16", "remaining_time": "7:22:50"} +{"current_steps": 305, "total_steps": 3598, "loss": 0.3757, "lr": 3.377777777777778e-05, "epoch": 0.5939629990262901, "percentage": 8.48, "elapsed_time": "0:41:05", "remaining_time": "7:23:40"} +{"current_steps": 310, "total_steps": 3598, "loss": 0.3929, "lr": 3.433333333333333e-05, "epoch": 0.6037000973709834, "percentage": 8.62, "elapsed_time": "0:41:45", "remaining_time": "7:22:50"} +{"current_steps": 315, "total_steps": 3598, "loss": 0.3531, "lr": 3.4888888888888895e-05, "epoch": 0.6134371957156767, "percentage": 8.75, "elapsed_time": "0:42:32", "remaining_time": "7:23:23"} +{"current_steps": 320, "total_steps": 3598, "loss": 0.3503, "lr": 3.5444444444444445e-05, "epoch": 0.6231742940603701, "percentage": 8.89, "elapsed_time": "0:43:04", "remaining_time": "7:21:16"} +{"current_steps": 325, "total_steps": 3598, "loss": 0.3651, "lr": 3.6e-05, "epoch": 0.6329113924050633, "percentage": 9.03, "elapsed_time": "0:43:38", "remaining_time": "7:19:30"} +{"current_steps": 330, "total_steps": 3598, "loss": 0.3494, "lr": 3.655555555555556e-05, "epoch": 0.6426484907497566, "percentage": 9.17, "elapsed_time": "0:44:14", "remaining_time": "7:18:09"} +{"current_steps": 335, "total_steps": 3598, "loss": 0.3732, "lr": 3.7111111111111113e-05, "epoch": 0.6523855890944499, "percentage": 9.31, "elapsed_time": "0:44:48", "remaining_time": "7:16:25"} +{"current_steps": 340, "total_steps": 3598, "loss": 0.3491, "lr": 3.766666666666667e-05, "epoch": 0.6621226874391432, "percentage": 9.45, "elapsed_time": "0:45:32", "remaining_time": "7:16:25"} +{"current_steps": 345, "total_steps": 3598, "loss": 0.3508, "lr": 3.8222222222222226e-05, "epoch": 0.6718597857838364, "percentage": 9.59, "elapsed_time": "0:46:20", "remaining_time": "7:17:01"} +{"current_steps": 350, "total_steps": 3598, "loss": 0.3635, "lr": 3.877777777777778e-05, "epoch": 0.6815968841285297, "percentage": 9.73, "elapsed_time": "0:46:59", "remaining_time": "7:16:08"} +{"current_steps": 355, "total_steps": 3598, "loss": 0.3387, "lr": 3.933333333333333e-05, "epoch": 0.691333982473223, "percentage": 9.87, "elapsed_time": "0:47:42", "remaining_time": "7:15:47"} +{"current_steps": 360, "total_steps": 3598, "loss": 0.3615, "lr": 3.9888888888888895e-05, "epoch": 0.7010710808179162, "percentage": 10.01, "elapsed_time": "0:48:15", "remaining_time": "7:13:59"} +{"current_steps": 365, "total_steps": 3598, "loss": 0.359, "lr": 3.9999849385947476e-05, "epoch": 0.7108081791626095, "percentage": 10.14, "elapsed_time": "0:48:45", "remaining_time": "7:11:48"} +{"current_steps": 370, "total_steps": 3598, "loss": 0.3444, "lr": 3.999923752024691e-05, "epoch": 0.7205452775073028, "percentage": 10.28, "elapsed_time": "0:49:30", "remaining_time": "7:11:55"} +{"current_steps": 375, "total_steps": 3598, "loss": 0.3404, "lr": 3.999815500390824e-05, "epoch": 0.7302823758519961, "percentage": 10.42, "elapsed_time": "0:50:04", "remaining_time": "7:10:22"} +{"current_steps": 380, "total_steps": 3598, "loss": 0.3359, "lr": 3.9996601862406804e-05, "epoch": 0.7400194741966893, "percentage": 10.56, "elapsed_time": "0:50:53", "remaining_time": "7:10:54"} +{"current_steps": 385, "total_steps": 3598, "loss": 0.3512, "lr": 3.999457813229333e-05, "epoch": 0.7497565725413826, "percentage": 10.7, "elapsed_time": "0:51:29", "remaining_time": "7:09:45"} +{"current_steps": 390, "total_steps": 3598, "loss": 0.336, "lr": 3.999208386119314e-05, "epoch": 0.759493670886076, "percentage": 10.84, "elapsed_time": "0:52:13", "remaining_time": "7:09:35"} +{"current_steps": 395, "total_steps": 3598, "loss": 0.3252, "lr": 3.998911910780497e-05, "epoch": 0.7692307692307693, "percentage": 10.98, "elapsed_time": "0:52:51", "remaining_time": "7:08:38"} +{"current_steps": 400, "total_steps": 3598, "loss": 0.322, "lr": 3.9985683941899654e-05, "epoch": 0.7789678675754625, "percentage": 11.12, "elapsed_time": "0:53:35", "remaining_time": "7:08:24"} +{"current_steps": 405, "total_steps": 3598, "loss": 0.3468, "lr": 3.9981778444318414e-05, "epoch": 0.7887049659201558, "percentage": 11.26, "elapsed_time": "0:54:14", "remaining_time": "7:07:41"} +{"current_steps": 410, "total_steps": 3598, "loss": 0.3257, "lr": 3.9977402706971e-05, "epoch": 0.7984420642648491, "percentage": 11.4, "elapsed_time": "0:55:01", "remaining_time": "7:07:50"} +{"current_steps": 415, "total_steps": 3598, "loss": 0.3419, "lr": 3.997255683283352e-05, "epoch": 0.8081791626095424, "percentage": 11.53, "elapsed_time": "0:55:44", "remaining_time": "7:07:33"} +{"current_steps": 420, "total_steps": 3598, "loss": 0.3449, "lr": 3.9967240935946004e-05, "epoch": 0.8179162609542356, "percentage": 11.67, "elapsed_time": "0:56:18", "remaining_time": "7:06:03"} +{"current_steps": 425, "total_steps": 3598, "loss": 0.3324, "lr": 3.996145514140974e-05, "epoch": 0.8276533592989289, "percentage": 11.81, "elapsed_time": "0:56:55", "remaining_time": "7:04:59"} +{"current_steps": 430, "total_steps": 3598, "loss": 0.3372, "lr": 3.995519958538431e-05, "epoch": 0.8373904576436222, "percentage": 11.95, "elapsed_time": "0:57:27", "remaining_time": "7:03:21"} +{"current_steps": 435, "total_steps": 3598, "loss": 0.3503, "lr": 3.99484744150844e-05, "epoch": 0.8471275559883155, "percentage": 12.09, "elapsed_time": "0:58:00", "remaining_time": "7:01:44"} +{"current_steps": 440, "total_steps": 3598, "loss": 0.3271, "lr": 3.9941279788776324e-05, "epoch": 0.8568646543330087, "percentage": 12.23, "elapsed_time": "0:58:38", "remaining_time": "7:00:52"} +{"current_steps": 445, "total_steps": 3598, "loss": 0.3223, "lr": 3.9933615875774314e-05, "epoch": 0.866601752677702, "percentage": 12.37, "elapsed_time": "0:59:16", "remaining_time": "6:59:57"} +{"current_steps": 450, "total_steps": 3598, "loss": 0.318, "lr": 3.992548285643653e-05, "epoch": 0.8763388510223953, "percentage": 12.51, "elapsed_time": "0:59:57", "remaining_time": "6:59:28"} +{"current_steps": 455, "total_steps": 3598, "loss": 0.3219, "lr": 3.99168809221608e-05, "epoch": 0.8860759493670886, "percentage": 12.65, "elapsed_time": "1:00:34", "remaining_time": "6:58:26"} +{"current_steps": 460, "total_steps": 3598, "loss": 0.3377, "lr": 3.9907810275380133e-05, "epoch": 0.8958130477117819, "percentage": 12.78, "elapsed_time": "1:01:08", "remaining_time": "6:57:05"} +{"current_steps": 465, "total_steps": 3598, "loss": 0.3134, "lr": 3.989827112955794e-05, "epoch": 0.9055501460564752, "percentage": 12.92, "elapsed_time": "1:01:54", "remaining_time": "6:57:05"} +{"current_steps": 470, "total_steps": 3598, "loss": 0.3374, "lr": 3.988826370918306e-05, "epoch": 0.9152872444011685, "percentage": 13.06, "elapsed_time": "1:02:36", "remaining_time": "6:56:42"} +{"current_steps": 475, "total_steps": 3598, "loss": 0.3327, "lr": 3.987778824976439e-05, "epoch": 0.9250243427458618, "percentage": 13.2, "elapsed_time": "1:03:16", "remaining_time": "6:56:00"} +{"current_steps": 480, "total_steps": 3598, "loss": 0.3029, "lr": 3.9866844997825415e-05, "epoch": 0.934761441090555, "percentage": 13.34, "elapsed_time": "1:03:51", "remaining_time": "6:54:49"} +{"current_steps": 485, "total_steps": 3598, "loss": 0.3513, "lr": 3.985543421089839e-05, "epoch": 0.9444985394352483, "percentage": 13.48, "elapsed_time": "1:04:23", "remaining_time": "6:53:18"} +{"current_steps": 490, "total_steps": 3598, "loss": 0.33, "lr": 3.984355615751824e-05, "epoch": 0.9542356377799416, "percentage": 13.62, "elapsed_time": "1:04:55", "remaining_time": "6:51:47"} +{"current_steps": 495, "total_steps": 3598, "loss": 0.3458, "lr": 3.983121111721631e-05, "epoch": 0.9639727361246349, "percentage": 13.76, "elapsed_time": "1:05:30", "remaining_time": "6:50:42"} +{"current_steps": 500, "total_steps": 3598, "loss": 0.3138, "lr": 3.981839938051372e-05, "epoch": 0.9737098344693281, "percentage": 13.9, "elapsed_time": "1:06:10", "remaining_time": "6:50:03"} +{"current_steps": 505, "total_steps": 3598, "loss": 0.3228, "lr": 3.980512124891458e-05, "epoch": 0.9834469328140214, "percentage": 14.04, "elapsed_time": "1:06:50", "remaining_time": "6:49:22"} +{"current_steps": 510, "total_steps": 3598, "loss": 0.3199, "lr": 3.979137703489883e-05, "epoch": 0.9931840311587147, "percentage": 14.17, "elapsed_time": "1:07:30", "remaining_time": "6:48:48"} +{"current_steps": 515, "total_steps": 3598, "loss": 0.3175, "lr": 3.9777167061915e-05, "epoch": 1.0019474196689386, "percentage": 14.31, "elapsed_time": "1:08:15", "remaining_time": "6:48:38"} +{"current_steps": 520, "total_steps": 3598, "loss": 0.2926, "lr": 3.976249166437243e-05, "epoch": 1.011684518013632, "percentage": 14.45, "elapsed_time": "1:09:06", "remaining_time": "6:49:05"} +{"current_steps": 525, "total_steps": 3598, "loss": 0.301, "lr": 3.974735118763359e-05, "epoch": 1.0214216163583252, "percentage": 14.59, "elapsed_time": "1:09:53", "remaining_time": "6:49:07"} +{"current_steps": 530, "total_steps": 3598, "loss": 0.302, "lr": 3.97317459880058e-05, "epoch": 1.0311587147030186, "percentage": 14.73, "elapsed_time": "1:10:32", "remaining_time": "6:48:18"} +{"current_steps": 535, "total_steps": 3598, "loss": 0.2973, "lr": 3.971567643273292e-05, "epoch": 1.0408958130477117, "percentage": 14.87, "elapsed_time": "1:11:10", "remaining_time": "6:47:30"} +{"current_steps": 540, "total_steps": 3598, "loss": 0.2905, "lr": 3.9699142899986685e-05, "epoch": 1.0506329113924051, "percentage": 15.01, "elapsed_time": "1:11:55", "remaining_time": "6:47:17"} +{"current_steps": 545, "total_steps": 3598, "loss": 0.2761, "lr": 3.968214577885782e-05, "epoch": 1.0603700097370983, "percentage": 15.15, "elapsed_time": "1:12:35", "remaining_time": "6:46:37"} +{"current_steps": 550, "total_steps": 3598, "loss": 0.3011, "lr": 3.966468546934687e-05, "epoch": 1.0701071080817917, "percentage": 15.29, "elapsed_time": "1:13:18", "remaining_time": "6:46:14"} +{"current_steps": 555, "total_steps": 3598, "loss": 0.2989, "lr": 3.964676238235479e-05, "epoch": 1.0798442064264848, "percentage": 15.43, "elapsed_time": "1:13:52", "remaining_time": "6:45:02"} +{"current_steps": 560, "total_steps": 3598, "loss": 0.2786, "lr": 3.9628376939673275e-05, "epoch": 1.0895813047711782, "percentage": 15.56, "elapsed_time": "1:14:37", "remaining_time": "6:44:50"} +{"current_steps": 565, "total_steps": 3598, "loss": 0.2856, "lr": 3.960952957397484e-05, "epoch": 1.0993184031158714, "percentage": 15.7, "elapsed_time": "1:15:16", "remaining_time": "6:44:03"} +{"current_steps": 570, "total_steps": 3598, "loss": 0.3081, "lr": 3.9590220728802625e-05, "epoch": 1.1090555014605648, "percentage": 15.84, "elapsed_time": "1:15:51", "remaining_time": "6:43:01"} +{"current_steps": 575, "total_steps": 3598, "loss": 0.2826, "lr": 3.9570450858559975e-05, "epoch": 1.1187925998052581, "percentage": 15.98, "elapsed_time": "1:16:33", "remaining_time": "6:42:30"} +{"current_steps": 580, "total_steps": 3598, "loss": 0.2952, "lr": 3.9550220428499725e-05, "epoch": 1.1285296981499513, "percentage": 16.12, "elapsed_time": "1:17:03", "remaining_time": "6:40:56"} +{"current_steps": 585, "total_steps": 3598, "loss": 0.3074, "lr": 3.9529529914713266e-05, "epoch": 1.1382667964946447, "percentage": 16.26, "elapsed_time": "1:17:45", "remaining_time": "6:40:30"} +{"current_steps": 590, "total_steps": 3598, "loss": 0.2979, "lr": 3.950837980411934e-05, "epoch": 1.1480038948393378, "percentage": 16.4, "elapsed_time": "1:18:32", "remaining_time": "6:40:26"} +{"current_steps": 595, "total_steps": 3598, "loss": 0.2733, "lr": 3.948677059445257e-05, "epoch": 1.1577409931840312, "percentage": 16.54, "elapsed_time": "1:19:17", "remaining_time": "6:40:13"} +{"current_steps": 600, "total_steps": 3598, "loss": 0.2954, "lr": 3.946470279425177e-05, "epoch": 1.1674780915287244, "percentage": 16.68, "elapsed_time": "1:19:50", "remaining_time": "6:38:57"} +{"current_steps": 605, "total_steps": 3598, "loss": 0.2866, "lr": 3.944217692284794e-05, "epoch": 1.1772151898734178, "percentage": 16.81, "elapsed_time": "1:21:09", "remaining_time": "6:41:31"} +{"current_steps": 610, "total_steps": 3598, "loss": 0.2902, "lr": 3.941919351035208e-05, "epoch": 1.186952288218111, "percentage": 16.95, "elapsed_time": "1:21:59", "remaining_time": "6:41:37"} +{"current_steps": 615, "total_steps": 3598, "loss": 0.307, "lr": 3.9395753097642703e-05, "epoch": 1.1966893865628043, "percentage": 17.09, "elapsed_time": "1:22:34", "remaining_time": "6:40:31"} +{"current_steps": 620, "total_steps": 3598, "loss": 0.3072, "lr": 3.937185623635309e-05, "epoch": 1.2064264849074975, "percentage": 17.23, "elapsed_time": "1:23:07", "remaining_time": "6:39:14"} +{"current_steps": 625, "total_steps": 3598, "loss": 0.2815, "lr": 3.934750348885835e-05, "epoch": 1.2161635832521909, "percentage": 17.37, "elapsed_time": "1:23:43", "remaining_time": "6:38:16"} +{"current_steps": 630, "total_steps": 3598, "loss": 0.2897, "lr": 3.932269542826211e-05, "epoch": 1.225900681596884, "percentage": 17.51, "elapsed_time": "1:24:32", "remaining_time": "6:38:15"} +{"current_steps": 635, "total_steps": 3598, "loss": 0.2998, "lr": 3.9297432638383133e-05, "epoch": 1.2356377799415774, "percentage": 17.65, "elapsed_time": "1:25:16", "remaining_time": "6:37:53"} +{"current_steps": 640, "total_steps": 3598, "loss": 0.2892, "lr": 3.9271715713741484e-05, "epoch": 1.2453748782862708, "percentage": 17.79, "elapsed_time": "1:26:06", "remaining_time": "6:37:57"} +{"current_steps": 645, "total_steps": 3598, "loss": 0.2914, "lr": 3.9245545259544575e-05, "epoch": 1.255111976630964, "percentage": 17.93, "elapsed_time": "1:26:45", "remaining_time": "6:37:11"} +{"current_steps": 650, "total_steps": 3598, "loss": 0.2937, "lr": 3.9218921891672955e-05, "epoch": 1.2648490749756571, "percentage": 18.07, "elapsed_time": "1:27:26", "remaining_time": "6:36:36"} +{"current_steps": 655, "total_steps": 3598, "loss": 0.2806, "lr": 3.919184623666575e-05, "epoch": 1.2745861733203505, "percentage": 18.2, "elapsed_time": "1:28:12", "remaining_time": "6:36:21"} +{"current_steps": 660, "total_steps": 3598, "loss": 0.2996, "lr": 3.916431893170596e-05, "epoch": 1.284323271665044, "percentage": 18.34, "elapsed_time": "1:28:47", "remaining_time": "6:35:13"} +{"current_steps": 665, "total_steps": 3598, "loss": 0.2867, "lr": 3.913634062460548e-05, "epoch": 1.294060370009737, "percentage": 18.48, "elapsed_time": "1:29:30", "remaining_time": "6:34:45"} +{"current_steps": 670, "total_steps": 3598, "loss": 0.3031, "lr": 3.910791197378982e-05, "epoch": 1.3037974683544304, "percentage": 18.62, "elapsed_time": "1:30:14", "remaining_time": "6:34:22"} +{"current_steps": 675, "total_steps": 3598, "loss": 0.3224, "lr": 3.9079033648282595e-05, "epoch": 1.3135345666991236, "percentage": 18.76, "elapsed_time": "1:30:53", "remaining_time": "6:33:35"} +{"current_steps": 680, "total_steps": 3598, "loss": 0.2723, "lr": 3.9049706327689844e-05, "epoch": 1.323271665043817, "percentage": 18.9, "elapsed_time": "1:31:41", "remaining_time": "6:33:26"} +{"current_steps": 685, "total_steps": 3598, "loss": 0.2744, "lr": 3.9019930702183995e-05, "epoch": 1.3330087633885102, "percentage": 19.04, "elapsed_time": "1:32:21", "remaining_time": "6:32:47"} +{"current_steps": 690, "total_steps": 3598, "loss": 0.2835, "lr": 3.8989707472487605e-05, "epoch": 1.3427458617332035, "percentage": 19.18, "elapsed_time": "1:32:56", "remaining_time": "6:31:40"} +{"current_steps": 695, "total_steps": 3598, "loss": 0.2898, "lr": 3.8959037349856915e-05, "epoch": 1.352482960077897, "percentage": 19.32, "elapsed_time": "1:33:36", "remaining_time": "6:30:58"} +{"current_steps": 700, "total_steps": 3598, "loss": 0.2916, "lr": 3.8927921056065085e-05, "epoch": 1.36222005842259, "percentage": 19.46, "elapsed_time": "1:34:11", "remaining_time": "6:29:58"} +{"current_steps": 705, "total_steps": 3598, "loss": 0.296, "lr": 3.8896359323385196e-05, "epoch": 1.3719571567672832, "percentage": 19.59, "elapsed_time": "1:34:45", "remaining_time": "6:28:48"} +{"current_steps": 710, "total_steps": 3598, "loss": 0.2983, "lr": 3.886435289457306e-05, "epoch": 1.3816942551119766, "percentage": 19.73, "elapsed_time": "1:35:24", "remaining_time": "6:28:05"} +{"current_steps": 715, "total_steps": 3598, "loss": 0.2662, "lr": 3.883190252284969e-05, "epoch": 1.39143135345667, "percentage": 19.87, "elapsed_time": "1:36:10", "remaining_time": "6:27:46"} +{"current_steps": 720, "total_steps": 3598, "loss": 0.2824, "lr": 3.879900897188361e-05, "epoch": 1.4011684518013632, "percentage": 20.01, "elapsed_time": "1:36:51", "remaining_time": "6:27:11"} +{"current_steps": 725, "total_steps": 3598, "loss": 0.267, "lr": 3.876567301577287e-05, "epoch": 1.4109055501460563, "percentage": 20.15, "elapsed_time": "1:37:28", "remaining_time": "6:26:15"} +{"current_steps": 730, "total_steps": 3598, "loss": 0.2857, "lr": 3.873189543902683e-05, "epoch": 1.4206426484907497, "percentage": 20.29, "elapsed_time": "1:38:10", "remaining_time": "6:25:40"} +{"current_steps": 735, "total_steps": 3598, "loss": 0.2859, "lr": 3.86976770365477e-05, "epoch": 1.4303797468354431, "percentage": 20.43, "elapsed_time": "1:38:51", "remaining_time": "6:25:05"} +{"current_steps": 740, "total_steps": 3598, "loss": 0.2832, "lr": 3.8663018613611824e-05, "epoch": 1.4401168451801363, "percentage": 20.57, "elapsed_time": "1:39:29", "remaining_time": "6:24:14"} +{"current_steps": 745, "total_steps": 3598, "loss": 0.2751, "lr": 3.862792098585074e-05, "epoch": 1.4498539435248297, "percentage": 20.71, "elapsed_time": "1:40:07", "remaining_time": "6:23:26"} +{"current_steps": 750, "total_steps": 3598, "loss": 0.2623, "lr": 3.8592384979231984e-05, "epoch": 1.4595910418695228, "percentage": 20.84, "elapsed_time": "1:40:56", "remaining_time": "6:23:19"} +{"current_steps": 755, "total_steps": 3598, "loss": 0.2949, "lr": 3.8556411430039656e-05, "epoch": 1.4693281402142162, "percentage": 20.98, "elapsed_time": "1:41:23", "remaining_time": "6:21:47"} +{"current_steps": 760, "total_steps": 3598, "loss": 0.3034, "lr": 3.852000118485471e-05, "epoch": 1.4790652385589094, "percentage": 21.12, "elapsed_time": "1:41:53", "remaining_time": "6:20:27"} +{"current_steps": 765, "total_steps": 3598, "loss": 0.2768, "lr": 3.8483155100535074e-05, "epoch": 1.4888023369036028, "percentage": 21.26, "elapsed_time": "1:42:44", "remaining_time": "6:20:28"} +{"current_steps": 770, "total_steps": 3598, "loss": 0.2734, "lr": 3.8445874044195475e-05, "epoch": 1.4985394352482961, "percentage": 21.4, "elapsed_time": "1:43:20", "remaining_time": "6:19:33"} +{"current_steps": 775, "total_steps": 3598, "loss": 0.2707, "lr": 3.840815889318699e-05, "epoch": 1.5082765335929893, "percentage": 21.54, "elapsed_time": "1:44:06", "remaining_time": "6:19:14"} +{"current_steps": 780, "total_steps": 3598, "loss": 0.2731, "lr": 3.837001053507645e-05, "epoch": 1.5180136319376825, "percentage": 21.68, "elapsed_time": "1:44:47", "remaining_time": "6:18:35"} +{"current_steps": 785, "total_steps": 3598, "loss": 0.2886, "lr": 3.833142986762556e-05, "epoch": 1.5277507302823758, "percentage": 21.82, "elapsed_time": "1:45:20", "remaining_time": "6:17:30"} +{"current_steps": 790, "total_steps": 3598, "loss": 0.2832, "lr": 3.82924177987697e-05, "epoch": 1.5374878286270692, "percentage": 21.96, "elapsed_time": "1:46:01", "remaining_time": "6:16:52"} +{"current_steps": 795, "total_steps": 3598, "loss": 0.284, "lr": 3.8252975246596634e-05, "epoch": 1.5472249269717624, "percentage": 22.1, "elapsed_time": "1:46:46", "remaining_time": "6:16:27"} +{"current_steps": 800, "total_steps": 3598, "loss": 0.3103, "lr": 3.821310313932486e-05, "epoch": 1.5569620253164556, "percentage": 22.23, "elapsed_time": "1:47:23", "remaining_time": "6:15:36"} +{"current_steps": 805, "total_steps": 3598, "loss": 0.2919, "lr": 3.8172802415281796e-05, "epoch": 1.566699123661149, "percentage": 22.37, "elapsed_time": "1:47:58", "remaining_time": "6:14:38"} +{"current_steps": 810, "total_steps": 3598, "loss": 0.2841, "lr": 3.813207402288167e-05, "epoch": 1.5764362220058423, "percentage": 22.51, "elapsed_time": "1:48:32", "remaining_time": "6:13:36"} +{"current_steps": 815, "total_steps": 3598, "loss": 0.2652, "lr": 3.8090918920603214e-05, "epoch": 1.5861733203505355, "percentage": 22.65, "elapsed_time": "1:49:16", "remaining_time": "6:13:08"} +{"current_steps": 820, "total_steps": 3598, "loss": 0.2753, "lr": 3.804933807696712e-05, "epoch": 1.5959104186952289, "percentage": 22.79, "elapsed_time": "1:49:54", "remaining_time": "6:12:19"} +{"current_steps": 825, "total_steps": 3598, "loss": 0.295, "lr": 3.80073324705132e-05, "epoch": 1.6056475170399223, "percentage": 22.93, "elapsed_time": "1:50:29", "remaining_time": "6:11:22"} +{"current_steps": 830, "total_steps": 3598, "loss": 0.2655, "lr": 3.7964903089777443e-05, "epoch": 1.6153846153846154, "percentage": 23.07, "elapsed_time": "1:51:12", "remaining_time": "6:10:51"} +{"current_steps": 835, "total_steps": 3598, "loss": 0.276, "lr": 3.792205093326865e-05, "epoch": 1.6251217137293086, "percentage": 23.21, "elapsed_time": "1:51:57", "remaining_time": "6:10:29"} +{"current_steps": 840, "total_steps": 3598, "loss": 0.284, "lr": 3.7878777009445025e-05, "epoch": 1.634858812074002, "percentage": 23.35, "elapsed_time": "1:52:39", "remaining_time": "6:09:52"} +{"current_steps": 845, "total_steps": 3598, "loss": 0.2742, "lr": 3.7835082336690376e-05, "epoch": 1.6445959104186954, "percentage": 23.49, "elapsed_time": "1:53:30", "remaining_time": "6:09:48"} +{"current_steps": 850, "total_steps": 3598, "loss": 0.3167, "lr": 3.7790967943290186e-05, "epoch": 1.6543330087633885, "percentage": 23.62, "elapsed_time": "1:54:02", "remaining_time": "6:08:40"} +{"current_steps": 855, "total_steps": 3598, "loss": 0.2733, "lr": 3.7746434867407405e-05, "epoch": 1.6640701071080817, "percentage": 23.76, "elapsed_time": "1:54:42", "remaining_time": "6:08:00"} +{"current_steps": 860, "total_steps": 3598, "loss": 0.2809, "lr": 3.7701484157058016e-05, "epoch": 1.673807205452775, "percentage": 23.9, "elapsed_time": "1:55:24", "remaining_time": "6:07:25"} +{"current_steps": 865, "total_steps": 3598, "loss": 0.2884, "lr": 3.7656116870086376e-05, "epoch": 1.6835443037974684, "percentage": 24.04, "elapsed_time": "1:55:59", "remaining_time": "6:06:28"} +{"current_steps": 870, "total_steps": 3598, "loss": 0.2819, "lr": 3.761033407414032e-05, "epoch": 1.6932814021421616, "percentage": 24.18, "elapsed_time": "1:56:39", "remaining_time": "6:05:47"} +{"current_steps": 875, "total_steps": 3598, "loss": 0.2881, "lr": 3.756413684664602e-05, "epoch": 1.7030185004868548, "percentage": 24.32, "elapsed_time": "1:57:20", "remaining_time": "6:05:10"} +{"current_steps": 880, "total_steps": 3598, "loss": 0.266, "lr": 3.751752627478266e-05, "epoch": 1.7127555988315482, "percentage": 24.46, "elapsed_time": "1:58:05", "remaining_time": "6:04:45"} +{"current_steps": 885, "total_steps": 3598, "loss": 0.273, "lr": 3.747050345545683e-05, "epoch": 1.7224926971762415, "percentage": 24.6, "elapsed_time": "1:58:40", "remaining_time": "6:03:47"} +{"current_steps": 890, "total_steps": 3598, "loss": 0.283, "lr": 3.7423069495276705e-05, "epoch": 1.7322297955209347, "percentage": 24.74, "elapsed_time": "1:59:26", "remaining_time": "6:03:24"} +{"current_steps": 895, "total_steps": 3598, "loss": 0.2665, "lr": 3.737522551052603e-05, "epoch": 1.741966893865628, "percentage": 24.87, "elapsed_time": "1:59:59", "remaining_time": "6:02:22"} +{"current_steps": 900, "total_steps": 3598, "loss": 0.2681, "lr": 3.732697262713783e-05, "epoch": 1.7517039922103215, "percentage": 25.01, "elapsed_time": "2:00:42", "remaining_time": "6:01:51"} +{"current_steps": 905, "total_steps": 3598, "loss": 0.2891, "lr": 3.727831198066791e-05, "epoch": 1.7614410905550146, "percentage": 25.15, "elapsed_time": "2:01:28", "remaining_time": "6:01:28"} +{"current_steps": 910, "total_steps": 3598, "loss": 0.2733, "lr": 3.722924471626815e-05, "epoch": 1.7711781888997078, "percentage": 25.29, "elapsed_time": "2:02:11", "remaining_time": "6:00:56"} +{"current_steps": 915, "total_steps": 3598, "loss": 0.2766, "lr": 3.7179771988659526e-05, "epoch": 1.7809152872444012, "percentage": 25.43, "elapsed_time": "2:02:48", "remaining_time": "6:00:06"} +{"current_steps": 920, "total_steps": 3598, "loss": 0.2827, "lr": 3.712989496210497e-05, "epoch": 1.7906523855890946, "percentage": 25.57, "elapsed_time": "2:03:21", "remaining_time": "5:59:05"} +{"current_steps": 925, "total_steps": 3598, "loss": 0.2852, "lr": 3.707961481038196e-05, "epoch": 1.8003894839337877, "percentage": 25.71, "elapsed_time": "2:04:01", "remaining_time": "5:58:24"} +{"current_steps": 930, "total_steps": 3598, "loss": 0.2849, "lr": 3.702893271675487e-05, "epoch": 1.810126582278481, "percentage": 25.85, "elapsed_time": "2:04:29", "remaining_time": "5:57:09"} +{"current_steps": 935, "total_steps": 3598, "loss": 0.2681, "lr": 3.697784987394716e-05, "epoch": 1.8198636806231743, "percentage": 25.99, "elapsed_time": "2:05:08", "remaining_time": "5:56:25"} +{"current_steps": 940, "total_steps": 3598, "loss": 0.2617, "lr": 3.692636748411329e-05, "epoch": 1.8296007789678677, "percentage": 26.13, "elapsed_time": "2:05:45", "remaining_time": "5:55:37"} +{"current_steps": 945, "total_steps": 3598, "loss": 0.2695, "lr": 3.687448675881043e-05, "epoch": 1.8393378773125608, "percentage": 26.26, "elapsed_time": "2:06:32", "remaining_time": "5:55:14"} +{"current_steps": 950, "total_steps": 3598, "loss": 0.276, "lr": 3.682220891896995e-05, "epoch": 1.849074975657254, "percentage": 26.4, "elapsed_time": "2:07:08", "remaining_time": "5:54:22"} +{"current_steps": 955, "total_steps": 3598, "loss": 0.2941, "lr": 3.676953519486868e-05, "epoch": 1.8588120740019474, "percentage": 26.54, "elapsed_time": "2:07:46", "remaining_time": "5:53:37"} +{"current_steps": 960, "total_steps": 3598, "loss": 0.2916, "lr": 3.671646682609999e-05, "epoch": 1.8685491723466408, "percentage": 26.68, "elapsed_time": "2:08:33", "remaining_time": "5:53:17"} +{"current_steps": 965, "total_steps": 3598, "loss": 0.2929, "lr": 3.666300506154455e-05, "epoch": 1.878286270691334, "percentage": 26.82, "elapsed_time": "2:09:08", "remaining_time": "5:52:20"} +{"current_steps": 970, "total_steps": 3598, "loss": 0.2989, "lr": 3.660915115934103e-05, "epoch": 1.8880233690360273, "percentage": 26.96, "elapsed_time": "2:09:52", "remaining_time": "5:51:52"} +{"current_steps": 975, "total_steps": 3598, "loss": 0.2749, "lr": 3.6554906386856394e-05, "epoch": 1.8977604673807207, "percentage": 27.1, "elapsed_time": "2:10:26", "remaining_time": "5:50:54"} +{"current_steps": 980, "total_steps": 3598, "loss": 0.2905, "lr": 3.650027202065617e-05, "epoch": 1.9074975657254138, "percentage": 27.24, "elapsed_time": "2:11:02", "remaining_time": "5:50:04"} +{"current_steps": 985, "total_steps": 3598, "loss": 0.2736, "lr": 3.644524934647432e-05, "epoch": 1.917234664070107, "percentage": 27.38, "elapsed_time": "2:11:51", "remaining_time": "5:49:48"} +{"current_steps": 990, "total_steps": 3598, "loss": 0.2846, "lr": 3.6389839659183064e-05, "epoch": 1.9269717624148004, "percentage": 27.52, "elapsed_time": "2:12:20", "remaining_time": "5:48:39"} +{"current_steps": 995, "total_steps": 3598, "loss": 0.2777, "lr": 3.633404426276234e-05, "epoch": 1.9367088607594938, "percentage": 27.65, "elapsed_time": "2:13:07", "remaining_time": "5:48:16"} +{"current_steps": 1000, "total_steps": 3598, "loss": 0.2712, "lr": 3.6277864470269164e-05, "epoch": 1.946445959104187, "percentage": 27.79, "elapsed_time": "2:13:43", "remaining_time": "5:47:24"} +{"current_steps": 1005, "total_steps": 3598, "loss": 0.2662, "lr": 3.622130160380668e-05, "epoch": 1.95618305744888, "percentage": 27.93, "elapsed_time": "2:14:17", "remaining_time": "5:46:28"} +{"current_steps": 1010, "total_steps": 3598, "loss": 0.2711, "lr": 3.616435699449311e-05, "epoch": 1.9659201557935735, "percentage": 28.07, "elapsed_time": "2:14:57", "remaining_time": "5:45:49"} +{"current_steps": 1015, "total_steps": 3598, "loss": 0.2974, "lr": 3.6107031982430374e-05, "epoch": 1.9756572541382669, "percentage": 28.21, "elapsed_time": "2:15:33", "remaining_time": "5:44:59"} +{"current_steps": 1020, "total_steps": 3598, "loss": 0.2775, "lr": 3.604932791667258e-05, "epoch": 1.98539435248296, "percentage": 28.35, "elapsed_time": "2:16:22", "remaining_time": "5:44:40"} +{"current_steps": 1025, "total_steps": 3598, "loss": 0.2423, "lr": 3.599124615519427e-05, "epoch": 1.9951314508276532, "percentage": 28.49, "elapsed_time": "2:17:12", "remaining_time": "5:44:25"} +{"current_steps": 1030, "total_steps": 3598, "loss": 0.2514, "lr": 3.5932788064858454e-05, "epoch": 2.0038948393378773, "percentage": 28.63, "elapsed_time": "2:17:51", "remaining_time": "5:43:42"} +{"current_steps": 1035, "total_steps": 3598, "loss": 0.2485, "lr": 3.5873955021384464e-05, "epoch": 2.0136319376825704, "percentage": 28.77, "elapsed_time": "2:18:22", "remaining_time": "5:42:38"} +{"current_steps": 1040, "total_steps": 3598, "loss": 0.2476, "lr": 3.5814748409315545e-05, "epoch": 2.023369036027264, "percentage": 28.9, "elapsed_time": "2:19:04", "remaining_time": "5:42:04"} +{"current_steps": 1045, "total_steps": 3598, "loss": 0.242, "lr": 3.575516962198632e-05, "epoch": 2.033106134371957, "percentage": 29.04, "elapsed_time": "2:19:39", "remaining_time": "5:41:11"} +{"current_steps": 1050, "total_steps": 3598, "loss": 0.2429, "lr": 3.5695220061489935e-05, "epoch": 2.0428432327166504, "percentage": 29.18, "elapsed_time": "2:20:19", "remaining_time": "5:40:31"} +{"current_steps": 1055, "total_steps": 3598, "loss": 0.2464, "lr": 3.563490113864514e-05, "epoch": 2.0525803310613435, "percentage": 29.32, "elapsed_time": "2:20:53", "remaining_time": "5:39:35"} +{"current_steps": 1060, "total_steps": 3598, "loss": 0.2339, "lr": 3.557421427296302e-05, "epoch": 2.062317429406037, "percentage": 29.46, "elapsed_time": "2:21:32", "remaining_time": "5:38:53"} +{"current_steps": 1065, "total_steps": 3598, "loss": 0.2394, "lr": 3.551316089261363e-05, "epoch": 2.0720545277507303, "percentage": 29.6, "elapsed_time": "2:22:13", "remaining_time": "5:38:15"} +{"current_steps": 1070, "total_steps": 3598, "loss": 0.228, "lr": 3.545174243439236e-05, "epoch": 2.0817916260954235, "percentage": 29.74, "elapsed_time": "2:22:57", "remaining_time": "5:37:45"} +{"current_steps": 1075, "total_steps": 3598, "loss": 0.2415, "lr": 3.538996034368615e-05, "epoch": 2.091528724440117, "percentage": 29.88, "elapsed_time": "2:23:36", "remaining_time": "5:37:03"} +{"current_steps": 1080, "total_steps": 3598, "loss": 0.243, "lr": 3.5327816074439445e-05, "epoch": 2.1012658227848102, "percentage": 30.02, "elapsed_time": "2:24:22", "remaining_time": "5:36:36"} +{"current_steps": 1085, "total_steps": 3598, "loss": 0.2331, "lr": 3.526531108912e-05, "epoch": 2.1110029211295034, "percentage": 30.16, "elapsed_time": "2:25:00", "remaining_time": "5:35:51"} +{"current_steps": 1090, "total_steps": 3598, "loss": 0.2261, "lr": 3.520244685868446e-05, "epoch": 2.1207400194741965, "percentage": 30.29, "elapsed_time": "2:25:41", "remaining_time": "5:35:12"} +{"current_steps": 1095, "total_steps": 3598, "loss": 0.237, "lr": 3.513922486254373e-05, "epoch": 2.13047711781889, "percentage": 30.43, "elapsed_time": "2:26:16", "remaining_time": "5:34:22"} +{"current_steps": 1100, "total_steps": 3598, "loss": 0.2408, "lr": 3.507564658852817e-05, "epoch": 2.1402142161635833, "percentage": 30.57, "elapsed_time": "2:27:11", "remaining_time": "5:34:15"} +{"current_steps": 1105, "total_steps": 3598, "loss": 0.2488, "lr": 3.501171353285258e-05, "epoch": 2.1499513145082765, "percentage": 30.71, "elapsed_time": "2:27:47", "remaining_time": "5:33:26"} +{"current_steps": 1110, "total_steps": 3598, "loss": 0.24, "lr": 3.494742720008102e-05, "epoch": 2.1596884128529696, "percentage": 30.85, "elapsed_time": "2:28:24", "remaining_time": "5:32:39"} +{"current_steps": 1115, "total_steps": 3598, "loss": 0.2348, "lr": 3.488278910309131e-05, "epoch": 2.1694255111976632, "percentage": 30.99, "elapsed_time": "2:28:57", "remaining_time": "5:31:42"} +{"current_steps": 1120, "total_steps": 3598, "loss": 0.2477, "lr": 3.481780076303953e-05, "epoch": 2.1791626095423564, "percentage": 31.13, "elapsed_time": "2:29:42", "remaining_time": "5:31:14"} +{"current_steps": 1125, "total_steps": 3598, "loss": 0.2484, "lr": 3.475246370932419e-05, "epoch": 2.1888997078870496, "percentage": 31.27, "elapsed_time": "2:30:17", "remaining_time": "5:30:22"} +{"current_steps": 1130, "total_steps": 3598, "loss": 0.2347, "lr": 3.468677947955019e-05, "epoch": 2.1986368062317427, "percentage": 31.41, "elapsed_time": "2:30:58", "remaining_time": "5:29:45"} +{"current_steps": 1135, "total_steps": 3598, "loss": 0.2236, "lr": 3.4620749619492704e-05, "epoch": 2.2083739045764363, "percentage": 31.55, "elapsed_time": "2:31:40", "remaining_time": "5:29:08"} +{"current_steps": 1140, "total_steps": 3598, "loss": 0.2428, "lr": 3.455437568306076e-05, "epoch": 2.2181110029211295, "percentage": 31.68, "elapsed_time": "2:32:28", "remaining_time": "5:28:44"} +{"current_steps": 1145, "total_steps": 3598, "loss": 0.2152, "lr": 3.448765923226069e-05, "epoch": 2.2278481012658227, "percentage": 31.82, "elapsed_time": "2:33:13", "remaining_time": "5:28:16"} +{"current_steps": 1150, "total_steps": 3598, "loss": 0.2371, "lr": 3.442060183715935e-05, "epoch": 2.2375851996105163, "percentage": 31.96, "elapsed_time": "2:33:53", "remaining_time": "5:27:35"} +{"current_steps": 1155, "total_steps": 3598, "loss": 0.2553, "lr": 3.4353205075847195e-05, "epoch": 2.2473222979552094, "percentage": 32.1, "elapsed_time": "2:34:24", "remaining_time": "5:26:36"} +{"current_steps": 1160, "total_steps": 3598, "loss": 0.2327, "lr": 3.428547053440113e-05, "epoch": 2.2570593962999026, "percentage": 32.24, "elapsed_time": "2:35:05", "remaining_time": "5:25:57"} +{"current_steps": 1165, "total_steps": 3598, "loss": 0.2368, "lr": 3.421739980684718e-05, "epoch": 2.2667964946445958, "percentage": 32.38, "elapsed_time": "2:35:50", "remaining_time": "5:25:27"} +{"current_steps": 1170, "total_steps": 3598, "loss": 0.2442, "lr": 3.4148994495123e-05, "epoch": 2.2765335929892894, "percentage": 32.52, "elapsed_time": "2:36:25", "remaining_time": "5:24:37"} +{"current_steps": 1175, "total_steps": 3598, "loss": 0.2294, "lr": 3.408025620904012e-05, "epoch": 2.2862706913339825, "percentage": 32.66, "elapsed_time": "2:37:12", "remaining_time": "5:24:11"} +{"current_steps": 1180, "total_steps": 3598, "loss": 0.2448, "lr": 3.401118656624613e-05, "epoch": 2.2960077896786757, "percentage": 32.8, "elapsed_time": "2:37:45", "remaining_time": "5:23:15"} +{"current_steps": 1185, "total_steps": 3598, "loss": 0.2507, "lr": 3.3941787192186584e-05, "epoch": 2.305744888023369, "percentage": 32.93, "elapsed_time": "2:38:22", "remaining_time": "5:22:30"} +{"current_steps": 1190, "total_steps": 3598, "loss": 0.2282, "lr": 3.387205972006671e-05, "epoch": 2.3154819863680625, "percentage": 33.07, "elapsed_time": "2:39:00", "remaining_time": "5:21:45"} +{"current_steps": 1195, "total_steps": 3598, "loss": 0.2157, "lr": 3.3802005790813046e-05, "epoch": 2.3252190847127556, "percentage": 33.21, "elapsed_time": "2:39:32", "remaining_time": "5:20:48"} +{"current_steps": 1200, "total_steps": 3598, "loss": 0.2426, "lr": 3.373162705303476e-05, "epoch": 2.334956183057449, "percentage": 33.35, "elapsed_time": "2:40:13", "remaining_time": "5:20:11"} +{"current_steps": 1205, "total_steps": 3598, "loss": 0.2414, "lr": 3.366092516298489e-05, "epoch": 2.344693281402142, "percentage": 33.49, "elapsed_time": "2:41:33", "remaining_time": "5:20:49"} +{"current_steps": 1210, "total_steps": 3598, "loss": 0.2357, "lr": 3.358990178452135e-05, "epoch": 2.3544303797468356, "percentage": 33.63, "elapsed_time": "2:42:14", "remaining_time": "5:20:11"} +{"current_steps": 1215, "total_steps": 3598, "loss": 0.2617, "lr": 3.351855858906778e-05, "epoch": 2.3641674780915287, "percentage": 33.77, "elapsed_time": "2:42:55", "remaining_time": "5:19:32"} +{"current_steps": 1220, "total_steps": 3598, "loss": 0.2463, "lr": 3.344689725557422e-05, "epoch": 2.373904576436222, "percentage": 33.91, "elapsed_time": "2:43:35", "remaining_time": "5:18:52"} +{"current_steps": 1225, "total_steps": 3598, "loss": 0.2442, "lr": 3.337491947047757e-05, "epoch": 2.3836416747809155, "percentage": 34.05, "elapsed_time": "2:44:11", "remaining_time": "5:18:02"} +{"current_steps": 1230, "total_steps": 3598, "loss": 0.2421, "lr": 3.330262692766193e-05, "epoch": 2.3933787731256086, "percentage": 34.19, "elapsed_time": "2:44:48", "remaining_time": "5:17:16"} +{"current_steps": 1235, "total_steps": 3598, "loss": 0.2603, "lr": 3.323002132841875e-05, "epoch": 2.403115871470302, "percentage": 34.32, "elapsed_time": "2:45:25", "remaining_time": "5:16:31"} +{"current_steps": 1240, "total_steps": 3598, "loss": 0.2351, "lr": 3.315710438140674e-05, "epoch": 2.412852969814995, "percentage": 34.46, "elapsed_time": "2:46:00", "remaining_time": "5:15:41"} +{"current_steps": 1245, "total_steps": 3598, "loss": 0.2439, "lr": 3.308387780261169e-05, "epoch": 2.4225900681596886, "percentage": 34.6, "elapsed_time": "2:46:40", "remaining_time": "5:15:00"} +{"current_steps": 1250, "total_steps": 3598, "loss": 0.2336, "lr": 3.3010343315306106e-05, "epoch": 2.4323271665043817, "percentage": 34.74, "elapsed_time": "2:47:24", "remaining_time": "5:14:26"} +{"current_steps": 1255, "total_steps": 3598, "loss": 0.252, "lr": 3.293650265000864e-05, "epoch": 2.442064264849075, "percentage": 34.88, "elapsed_time": "2:48:00", "remaining_time": "5:13:39"} +{"current_steps": 1260, "total_steps": 3598, "loss": 0.2395, "lr": 3.286235754444335e-05, "epoch": 2.451801363193768, "percentage": 35.02, "elapsed_time": "2:48:36", "remaining_time": "5:12:51"} +{"current_steps": 1265, "total_steps": 3598, "loss": 0.2413, "lr": 3.278790974349882e-05, "epoch": 2.4615384615384617, "percentage": 35.16, "elapsed_time": "2:49:11", "remaining_time": "5:12:01"} +{"current_steps": 1270, "total_steps": 3598, "loss": 0.2318, "lr": 3.271316099918709e-05, "epoch": 2.471275559883155, "percentage": 35.3, "elapsed_time": "2:50:02", "remaining_time": "5:11:41"} +{"current_steps": 1275, "total_steps": 3598, "loss": 0.237, "lr": 3.2638113070602436e-05, "epoch": 2.481012658227848, "percentage": 35.44, "elapsed_time": "2:50:49", "remaining_time": "5:11:14"} +{"current_steps": 1280, "total_steps": 3598, "loss": 0.2367, "lr": 3.256276772387997e-05, "epoch": 2.4907497565725416, "percentage": 35.58, "elapsed_time": "2:51:38", "remaining_time": "5:10:49"} +{"current_steps": 1285, "total_steps": 3598, "loss": 0.2395, "lr": 3.248712673215405e-05, "epoch": 2.5004868549172348, "percentage": 35.71, "elapsed_time": "2:52:18", "remaining_time": "5:10:09"} +{"current_steps": 1290, "total_steps": 3598, "loss": 0.2463, "lr": 3.241119187551661e-05, "epoch": 2.510223953261928, "percentage": 35.85, "elapsed_time": "2:53:08", "remaining_time": "5:09:47"} +{"current_steps": 1295, "total_steps": 3598, "loss": 0.2416, "lr": 3.233496494097521e-05, "epoch": 2.519961051606621, "percentage": 35.99, "elapsed_time": "2:53:51", "remaining_time": "5:09:10"} +{"current_steps": 1300, "total_steps": 3598, "loss": 0.2465, "lr": 3.2258447722411e-05, "epoch": 2.5296981499513143, "percentage": 36.13, "elapsed_time": "2:54:23", "remaining_time": "5:08:15"} +{"current_steps": 1305, "total_steps": 3598, "loss": 0.2383, "lr": 3.2181642020536536e-05, "epoch": 2.539435248296008, "percentage": 36.27, "elapsed_time": "2:54:59", "remaining_time": "5:07:27"} +{"current_steps": 1310, "total_steps": 3598, "loss": 0.2257, "lr": 3.210454964285333e-05, "epoch": 2.549172346640701, "percentage": 36.41, "elapsed_time": "2:55:35", "remaining_time": "5:06:41"} +{"current_steps": 1315, "total_steps": 3598, "loss": 0.2169, "lr": 3.202717240360941e-05, "epoch": 2.558909444985394, "percentage": 36.55, "elapsed_time": "2:56:17", "remaining_time": "5:06:03"} +{"current_steps": 1320, "total_steps": 3598, "loss": 0.2327, "lr": 3.194951212375654e-05, "epoch": 2.568646543330088, "percentage": 36.69, "elapsed_time": "2:56:57", "remaining_time": "5:05:22"} +{"current_steps": 1325, "total_steps": 3598, "loss": 0.2357, "lr": 3.18715706309074e-05, "epoch": 2.578383641674781, "percentage": 36.83, "elapsed_time": "2:57:41", "remaining_time": "5:04:50"} +{"current_steps": 1330, "total_steps": 3598, "loss": 0.2502, "lr": 3.17933497592926e-05, "epoch": 2.588120740019474, "percentage": 36.96, "elapsed_time": "2:58:12", "remaining_time": "5:03:53"} +{"current_steps": 1335, "total_steps": 3598, "loss": 0.2534, "lr": 3.171485134971748e-05, "epoch": 2.5978578383641677, "percentage": 37.1, "elapsed_time": "2:58:49", "remaining_time": "5:03:07"} +{"current_steps": 1340, "total_steps": 3598, "loss": 0.2361, "lr": 3.163607724951877e-05, "epoch": 2.607594936708861, "percentage": 37.24, "elapsed_time": "2:59:24", "remaining_time": "5:02:18"} +{"current_steps": 1345, "total_steps": 3598, "loss": 0.2381, "lr": 3.155702931252119e-05, "epoch": 2.617332035053554, "percentage": 37.38, "elapsed_time": "3:00:08", "remaining_time": "5:01:44"} +{"current_steps": 1350, "total_steps": 3598, "loss": 0.2203, "lr": 3.147770939899376e-05, "epoch": 2.627069133398247, "percentage": 37.52, "elapsed_time": "3:00:42", "remaining_time": "5:00:54"} +{"current_steps": 1355, "total_steps": 3598, "loss": 0.2286, "lr": 3.1398119375606046e-05, "epoch": 2.6368062317429404, "percentage": 37.66, "elapsed_time": "3:01:23", "remaining_time": "5:00:16"} +{"current_steps": 1360, "total_steps": 3598, "loss": 0.2414, "lr": 3.13182611153842e-05, "epoch": 2.646543330087634, "percentage": 37.8, "elapsed_time": "3:02:02", "remaining_time": "4:59:34"} +{"current_steps": 1365, "total_steps": 3598, "loss": 0.2344, "lr": 3.1238136497666944e-05, "epoch": 2.656280428432327, "percentage": 37.94, "elapsed_time": "3:02:42", "remaining_time": "4:58:53"} +{"current_steps": 1370, "total_steps": 3598, "loss": 0.2388, "lr": 3.115774740806128e-05, "epoch": 2.6660175267770203, "percentage": 38.08, "elapsed_time": "3:03:35", "remaining_time": "4:58:33"} +{"current_steps": 1375, "total_steps": 3598, "loss": 0.2362, "lr": 3.1077095738398153e-05, "epoch": 2.675754625121714, "percentage": 38.22, "elapsed_time": "3:04:08", "remaining_time": "4:57:43"} +{"current_steps": 1380, "total_steps": 3598, "loss": 0.2473, "lr": 3.099618338668791e-05, "epoch": 2.685491723466407, "percentage": 38.35, "elapsed_time": "3:04:50", "remaining_time": "4:57:04"} +{"current_steps": 1385, "total_steps": 3598, "loss": 0.2214, "lr": 3.0915012257075635e-05, "epoch": 2.6952288218111002, "percentage": 38.49, "elapsed_time": "3:05:29", "remaining_time": "4:56:23"} +{"current_steps": 1390, "total_steps": 3598, "loss": 0.2472, "lr": 3.083358425979637e-05, "epoch": 2.704965920155794, "percentage": 38.63, "elapsed_time": "3:06:15", "remaining_time": "4:55:52"} +{"current_steps": 1395, "total_steps": 3598, "loss": 0.2428, "lr": 3.075190131113009e-05, "epoch": 2.714703018500487, "percentage": 38.77, "elapsed_time": "3:06:51", "remaining_time": "4:55:05"} +{"current_steps": 1400, "total_steps": 3598, "loss": 0.2436, "lr": 3.0669965333356695e-05, "epoch": 2.72444011684518, "percentage": 38.91, "elapsed_time": "3:07:43", "remaining_time": "4:54:44"} +{"current_steps": 1405, "total_steps": 3598, "loss": 0.2423, "lr": 3.058777825471071e-05, "epoch": 2.7341772151898733, "percentage": 39.05, "elapsed_time": "3:08:26", "remaining_time": "4:54:08"} +{"current_steps": 1410, "total_steps": 3598, "loss": 0.2221, "lr": 3.0505342009335898e-05, "epoch": 2.7439143135345665, "percentage": 39.19, "elapsed_time": "3:09:06", "remaining_time": "4:53:27"} +{"current_steps": 1415, "total_steps": 3598, "loss": 0.2509, "lr": 3.0422658537239823e-05, "epoch": 2.75365141187926, "percentage": 39.33, "elapsed_time": "3:09:50", "remaining_time": "4:52:53"} +{"current_steps": 1420, "total_steps": 3598, "loss": 0.2306, "lr": 3.0339729784248103e-05, "epoch": 2.7633885102239533, "percentage": 39.47, "elapsed_time": "3:10:29", "remaining_time": "4:52:10"} +{"current_steps": 1425, "total_steps": 3598, "loss": 0.246, "lr": 3.025655770195866e-05, "epoch": 2.7731256085686464, "percentage": 39.61, "elapsed_time": "3:11:07", "remaining_time": "4:51:26"} +{"current_steps": 1430, "total_steps": 3598, "loss": 0.2365, "lr": 3.0173144247695796e-05, "epoch": 2.78286270691334, "percentage": 39.74, "elapsed_time": "3:11:50", "remaining_time": "4:50:50"} +{"current_steps": 1435, "total_steps": 3598, "loss": 0.2161, "lr": 3.008949138446413e-05, "epoch": 2.792599805258033, "percentage": 39.88, "elapsed_time": "3:12:30", "remaining_time": "4:50:09"} +{"current_steps": 1440, "total_steps": 3598, "loss": 0.2319, "lr": 3.0005601080902376e-05, "epoch": 2.8023369036027264, "percentage": 40.02, "elapsed_time": "3:13:17", "remaining_time": "4:49:39"} +{"current_steps": 1445, "total_steps": 3598, "loss": 0.2283, "lr": 2.9921475311237055e-05, "epoch": 2.8120740019474195, "percentage": 40.16, "elapsed_time": "3:14:08", "remaining_time": "4:49:15"} +{"current_steps": 1450, "total_steps": 3598, "loss": 0.2414, "lr": 2.9837116055236e-05, "epoch": 2.8218111002921127, "percentage": 40.3, "elapsed_time": "3:14:43", "remaining_time": "4:48:27"} +{"current_steps": 1455, "total_steps": 3598, "loss": 0.2334, "lr": 2.975252529816178e-05, "epoch": 2.8315481986368063, "percentage": 40.44, "elapsed_time": "3:15:24", "remaining_time": "4:47:48"} +{"current_steps": 1460, "total_steps": 3598, "loss": 0.2257, "lr": 2.9667705030724976e-05, "epoch": 2.8412852969814995, "percentage": 40.58, "elapsed_time": "3:15:58", "remaining_time": "4:46:59"} +{"current_steps": 1465, "total_steps": 3598, "loss": 0.225, "lr": 2.9582657249037335e-05, "epoch": 2.8510223953261926, "percentage": 40.72, "elapsed_time": "3:16:44", "remaining_time": "4:46:26"} +{"current_steps": 1470, "total_steps": 3598, "loss": 0.2405, "lr": 2.9497383954564807e-05, "epoch": 2.8607594936708862, "percentage": 40.86, "elapsed_time": "3:17:23", "remaining_time": "4:45:44"} +{"current_steps": 1475, "total_steps": 3598, "loss": 0.2342, "lr": 2.941188715408041e-05, "epoch": 2.8704965920155794, "percentage": 40.99, "elapsed_time": "3:18:09", "remaining_time": "4:45:12"} +{"current_steps": 1480, "total_steps": 3598, "loss": 0.2276, "lr": 2.9326168859617044e-05, "epoch": 2.8802336903602725, "percentage": 41.13, "elapsed_time": "3:18:48", "remaining_time": "4:44:30"} +{"current_steps": 1485, "total_steps": 3598, "loss": 0.2319, "lr": 2.9240231088420112e-05, "epoch": 2.889970788704966, "percentage": 41.27, "elapsed_time": "3:19:38", "remaining_time": "4:44:04"} +{"current_steps": 1490, "total_steps": 3598, "loss": 0.2321, "lr": 2.9154075862900047e-05, "epoch": 2.8997078870496593, "percentage": 41.41, "elapsed_time": "3:20:23", "remaining_time": "4:43:29"} +{"current_steps": 1495, "total_steps": 3598, "loss": 0.2476, "lr": 2.9067705210584764e-05, "epoch": 2.9094449853943525, "percentage": 41.55, "elapsed_time": "3:21:04", "remaining_time": "4:42:50"} +{"current_steps": 1500, "total_steps": 3598, "loss": 0.2307, "lr": 2.898112116407186e-05, "epoch": 2.9191820837390456, "percentage": 41.69, "elapsed_time": "3:21:44", "remaining_time": "4:42:10"} +{"current_steps": 1505, "total_steps": 3598, "loss": 0.2436, "lr": 2.889432576098087e-05, "epoch": 2.928919182083739, "percentage": 41.83, "elapsed_time": "3:22:25", "remaining_time": "4:41:30"} +{"current_steps": 1510, "total_steps": 3598, "loss": 0.2196, "lr": 2.8807321043905253e-05, "epoch": 2.9386562804284324, "percentage": 41.97, "elapsed_time": "3:23:10", "remaining_time": "4:40:56"} +{"current_steps": 1515, "total_steps": 3598, "loss": 0.2466, "lr": 2.872010906036435e-05, "epoch": 2.9483933787731256, "percentage": 42.11, "elapsed_time": "3:23:57", "remaining_time": "4:40:25"} +{"current_steps": 1520, "total_steps": 3598, "loss": 0.2383, "lr": 2.863269186275519e-05, "epoch": 2.9581304771178187, "percentage": 42.25, "elapsed_time": "3:24:28", "remaining_time": "4:39:32"} +{"current_steps": 1525, "total_steps": 3598, "loss": 0.2358, "lr": 2.8545071508304193e-05, "epoch": 2.9678675754625123, "percentage": 42.38, "elapsed_time": "3:25:08", "remaining_time": "4:38:52"} +{"current_steps": 1530, "total_steps": 3598, "loss": 0.2333, "lr": 2.8457250059018765e-05, "epoch": 2.9776046738072055, "percentage": 42.52, "elapsed_time": "3:25:52", "remaining_time": "4:38:16"} +{"current_steps": 1535, "total_steps": 3598, "loss": 0.2399, "lr": 2.836922958163875e-05, "epoch": 2.9873417721518987, "percentage": 42.66, "elapsed_time": "3:26:44", "remaining_time": "4:37:50"} +{"current_steps": 1540, "total_steps": 3598, "loss": 0.2403, "lr": 2.8281012147587816e-05, "epoch": 2.9970788704965923, "percentage": 42.8, "elapsed_time": "3:27:22", "remaining_time": "4:37:07"} +{"current_steps": 1545, "total_steps": 3598, "loss": 0.2116, "lr": 2.8192599832924677e-05, "epoch": 3.005842259006816, "percentage": 42.94, "elapsed_time": "3:28:00", "remaining_time": "4:36:24"} +{"current_steps": 1550, "total_steps": 3598, "loss": 0.2033, "lr": 2.810399471829429e-05, "epoch": 3.015579357351509, "percentage": 43.08, "elapsed_time": "3:28:31", "remaining_time": "4:35:30"} +{"current_steps": 1555, "total_steps": 3598, "loss": 0.2044, "lr": 2.8015198888878835e-05, "epoch": 3.0253164556962027, "percentage": 43.22, "elapsed_time": "3:29:20", "remaining_time": "4:35:01"} +{"current_steps": 1560, "total_steps": 3598, "loss": 0.2216, "lr": 2.7926214434348645e-05, "epoch": 3.035053554040896, "percentage": 43.36, "elapsed_time": "3:29:49", "remaining_time": "4:34:07"} +{"current_steps": 1565, "total_steps": 3598, "loss": 0.2148, "lr": 2.78370434488131e-05, "epoch": 3.044790652385589, "percentage": 43.5, "elapsed_time": "3:30:31", "remaining_time": "4:33:28"} +{"current_steps": 1570, "total_steps": 3598, "loss": 0.2039, "lr": 2.7747688030771253e-05, "epoch": 3.0545277507302826, "percentage": 43.64, "elapsed_time": "3:31:12", "remaining_time": "4:32:49"} +{"current_steps": 1575, "total_steps": 3598, "loss": 0.2026, "lr": 2.765815028306249e-05, "epoch": 3.0642648490749758, "percentage": 43.77, "elapsed_time": "3:31:46", "remaining_time": "4:32:00"} +{"current_steps": 1580, "total_steps": 3598, "loss": 0.2062, "lr": 2.756843231281707e-05, "epoch": 3.074001947419669, "percentage": 43.91, "elapsed_time": "3:32:20", "remaining_time": "4:31:12"} +{"current_steps": 1585, "total_steps": 3598, "loss": 0.2066, "lr": 2.747853623140646e-05, "epoch": 3.083739045764362, "percentage": 44.05, "elapsed_time": "3:32:48", "remaining_time": "4:30:15"} +{"current_steps": 1590, "total_steps": 3598, "loss": 0.2087, "lr": 2.7388464154393735e-05, "epoch": 3.0934761441090557, "percentage": 44.19, "elapsed_time": "3:33:21", "remaining_time": "4:29:26"} +{"current_steps": 1595, "total_steps": 3598, "loss": 0.2075, "lr": 2.7298218201483725e-05, "epoch": 3.103213242453749, "percentage": 44.33, "elapsed_time": "3:34:03", "remaining_time": "4:28:48"} +{"current_steps": 1600, "total_steps": 3598, "loss": 0.2003, "lr": 2.7207800496473157e-05, "epoch": 3.112950340798442, "percentage": 44.47, "elapsed_time": "3:34:43", "remaining_time": "4:28:08"} +{"current_steps": 1605, "total_steps": 3598, "loss": 0.2042, "lr": 2.7117213167200696e-05, "epoch": 3.122687439143135, "percentage": 44.61, "elapsed_time": "3:35:22", "remaining_time": "4:27:26"} +{"current_steps": 1610, "total_steps": 3598, "loss": 0.2217, "lr": 2.7026458345496826e-05, "epoch": 3.132424537487829, "percentage": 44.75, "elapsed_time": "3:35:57", "remaining_time": "4:26:39"} +{"current_steps": 1615, "total_steps": 3598, "loss": 0.2121, "lr": 2.6935538167133702e-05, "epoch": 3.142161635832522, "percentage": 44.89, "elapsed_time": "3:36:30", "remaining_time": "4:25:50"} +{"current_steps": 1620, "total_steps": 3598, "loss": 0.2083, "lr": 2.6844454771774898e-05, "epoch": 3.151898734177215, "percentage": 45.03, "elapsed_time": "3:37:13", "remaining_time": "4:25:13"} +{"current_steps": 1625, "total_steps": 3598, "loss": 0.201, "lr": 2.6753210302925044e-05, "epoch": 3.1616358325219083, "percentage": 45.16, "elapsed_time": "3:37:50", "remaining_time": "4:24:29"} +{"current_steps": 1630, "total_steps": 3598, "loss": 0.2071, "lr": 2.6661806907879385e-05, "epoch": 3.171372930866602, "percentage": 45.3, "elapsed_time": "3:38:22", "remaining_time": "4:23:39"} +{"current_steps": 1635, "total_steps": 3598, "loss": 0.2029, "lr": 2.6570246737673216e-05, "epoch": 3.181110029211295, "percentage": 45.44, "elapsed_time": "3:39:02", "remaining_time": "4:22:59"} +{"current_steps": 1640, "total_steps": 3598, "loss": 0.2102, "lr": 2.6478531947031325e-05, "epoch": 3.190847127555988, "percentage": 45.58, "elapsed_time": "3:39:45", "remaining_time": "4:22:22"} +{"current_steps": 1645, "total_steps": 3598, "loss": 0.2055, "lr": 2.638666469431723e-05, "epoch": 3.2005842259006814, "percentage": 45.72, "elapsed_time": "3:40:27", "remaining_time": "4:21:44"} +{"current_steps": 1650, "total_steps": 3598, "loss": 0.2032, "lr": 2.6294647141482413e-05, "epoch": 3.210321324245375, "percentage": 45.86, "elapsed_time": "3:41:03", "remaining_time": "4:20:58"} +{"current_steps": 1655, "total_steps": 3598, "loss": 0.1842, "lr": 2.6202481454015426e-05, "epoch": 3.220058422590068, "percentage": 46.0, "elapsed_time": "3:41:36", "remaining_time": "4:20:10"} +{"current_steps": 1660, "total_steps": 3598, "loss": 0.2165, "lr": 2.6110169800890937e-05, "epoch": 3.2297955209347613, "percentage": 46.14, "elapsed_time": "3:42:18", "remaining_time": "4:19:32"} +{"current_steps": 1665, "total_steps": 3598, "loss": 0.2061, "lr": 2.6017714354518696e-05, "epoch": 3.239532619279455, "percentage": 46.28, "elapsed_time": "3:43:00", "remaining_time": "4:18:54"} +{"current_steps": 1670, "total_steps": 3598, "loss": 0.1873, "lr": 2.592511729069239e-05, "epoch": 3.249269717624148, "percentage": 46.41, "elapsed_time": "3:43:44", "remaining_time": "4:18:18"} +{"current_steps": 1675, "total_steps": 3598, "loss": 0.2108, "lr": 2.5832380788538457e-05, "epoch": 3.2590068159688412, "percentage": 46.55, "elapsed_time": "3:44:22", "remaining_time": "4:17:35"} +{"current_steps": 1680, "total_steps": 3598, "loss": 0.2054, "lr": 2.573950703046479e-05, "epoch": 3.2687439143135344, "percentage": 46.69, "elapsed_time": "3:45:05", "remaining_time": "4:16:58"} +{"current_steps": 1685, "total_steps": 3598, "loss": 0.2147, "lr": 2.564649820210939e-05, "epoch": 3.278481012658228, "percentage": 46.83, "elapsed_time": "3:45:44", "remaining_time": "4:16:17"} +{"current_steps": 1690, "total_steps": 3598, "loss": 0.208, "lr": 2.5553356492288933e-05, "epoch": 3.288218111002921, "percentage": 46.97, "elapsed_time": "3:46:14", "remaining_time": "4:15:25"} +{"current_steps": 1695, "total_steps": 3598, "loss": 0.191, "lr": 2.5460084092947233e-05, "epoch": 3.2979552093476143, "percentage": 47.11, "elapsed_time": "3:46:48", "remaining_time": "4:14:38"} +{"current_steps": 1700, "total_steps": 3598, "loss": 0.1957, "lr": 2.5366683199103692e-05, "epoch": 3.3076923076923075, "percentage": 47.25, "elapsed_time": "3:47:33", "remaining_time": "4:14:03"} +{"current_steps": 1705, "total_steps": 3598, "loss": 0.2008, "lr": 2.5273156008801614e-05, "epoch": 3.317429406037001, "percentage": 47.39, "elapsed_time": "3:48:10", "remaining_time": "4:13:19"} +{"current_steps": 1710, "total_steps": 3598, "loss": 0.1971, "lr": 2.517950472305651e-05, "epoch": 3.3271665043816943, "percentage": 47.53, "elapsed_time": "3:48:53", "remaining_time": "4:12:43"} +{"current_steps": 1715, "total_steps": 3598, "loss": 0.206, "lr": 2.508573154580427e-05, "epoch": 3.3369036027263874, "percentage": 47.67, "elapsed_time": "3:49:36", "remaining_time": "4:12:06"} +{"current_steps": 1720, "total_steps": 3598, "loss": 0.2081, "lr": 2.4991838683849306e-05, "epoch": 3.346640701071081, "percentage": 47.8, "elapsed_time": "3:50:18", "remaining_time": "4:11:28"} +{"current_steps": 1725, "total_steps": 3598, "loss": 0.2063, "lr": 2.4897828346812637e-05, "epoch": 3.356377799415774, "percentage": 47.94, "elapsed_time": "3:50:57", "remaining_time": "4:10:46"} +{"current_steps": 1730, "total_steps": 3598, "loss": 0.2123, "lr": 2.4803702747079858e-05, "epoch": 3.3661148977604674, "percentage": 48.08, "elapsed_time": "3:51:32", "remaining_time": "4:10:00"} +{"current_steps": 1735, "total_steps": 3598, "loss": 0.1934, "lr": 2.4709464099749093e-05, "epoch": 3.3758519961051605, "percentage": 48.22, "elapsed_time": "3:52:16", "remaining_time": "4:09:24"} +{"current_steps": 1740, "total_steps": 3598, "loss": 0.2036, "lr": 2.4615114622578875e-05, "epoch": 3.385589094449854, "percentage": 48.36, "elapsed_time": "3:53:02", "remaining_time": "4:08:50"} +{"current_steps": 1745, "total_steps": 3598, "loss": 0.2045, "lr": 2.4520656535935927e-05, "epoch": 3.3953261927945473, "percentage": 48.5, "elapsed_time": "3:53:39", "remaining_time": "4:08:07"} +{"current_steps": 1750, "total_steps": 3598, "loss": 0.1952, "lr": 2.4426092062742933e-05, "epoch": 3.4050632911392404, "percentage": 48.64, "elapsed_time": "3:54:25", "remaining_time": "4:07:33"} +{"current_steps": 1755, "total_steps": 3598, "loss": 0.1975, "lr": 2.4331423428426227e-05, "epoch": 3.4148003894839336, "percentage": 48.78, "elapsed_time": "3:55:11", "remaining_time": "4:06:58"} +{"current_steps": 1760, "total_steps": 3598, "loss": 0.2202, "lr": 2.4236652860863396e-05, "epoch": 3.424537487828627, "percentage": 48.92, "elapsed_time": "3:55:45", "remaining_time": "4:06:12"} +{"current_steps": 1765, "total_steps": 3598, "loss": 0.2102, "lr": 2.414178259033087e-05, "epoch": 3.4342745861733204, "percentage": 49.06, "elapsed_time": "3:56:32", "remaining_time": "4:05:38"} +{"current_steps": 1770, "total_steps": 3598, "loss": 0.2097, "lr": 2.404681484945143e-05, "epoch": 3.4440116845180135, "percentage": 49.19, "elapsed_time": "3:57:02", "remaining_time": "4:04:48"} +{"current_steps": 1775, "total_steps": 3598, "loss": 0.1976, "lr": 2.3951751873141683e-05, "epoch": 3.453748782862707, "percentage": 49.33, "elapsed_time": "3:57:46", "remaining_time": "4:04:12"} +{"current_steps": 1780, "total_steps": 3598, "loss": 0.2127, "lr": 2.385659589855944e-05, "epoch": 3.4634858812074003, "percentage": 49.47, "elapsed_time": "3:58:23", "remaining_time": "4:03:28"} +{"current_steps": 1785, "total_steps": 3598, "loss": 0.1887, "lr": 2.3761349165051075e-05, "epoch": 3.4732229795520935, "percentage": 49.61, "elapsed_time": "3:59:01", "remaining_time": "4:02:46"} +{"current_steps": 1790, "total_steps": 3598, "loss": 0.2166, "lr": 2.366601391409884e-05, "epoch": 3.4829600778967866, "percentage": 49.75, "elapsed_time": "3:59:37", "remaining_time": "4:02:01"} +{"current_steps": 1795, "total_steps": 3598, "loss": 0.2002, "lr": 2.35705923892681e-05, "epoch": 3.49269717624148, "percentage": 49.89, "elapsed_time": "4:00:11", "remaining_time": "4:01:15"} +{"current_steps": 1800, "total_steps": 3598, "loss": 0.2096, "lr": 2.3475086836154547e-05, "epoch": 3.5024342745861734, "percentage": 50.03, "elapsed_time": "4:00:47", "remaining_time": "4:00:31"} +{"current_steps": 1805, "total_steps": 3598, "loss": 0.1772, "lr": 2.3379499502331347e-05, "epoch": 3.5121713729308666, "percentage": 50.17, "elapsed_time": "4:02:09", "remaining_time": "4:00:33"} +{"current_steps": 1810, "total_steps": 3598, "loss": 0.1856, "lr": 2.328383263729624e-05, "epoch": 3.5219084712755597, "percentage": 50.31, "elapsed_time": "4:02:47", "remaining_time": "3:59:50"} +{"current_steps": 1815, "total_steps": 3598, "loss": 0.2185, "lr": 2.318808849241861e-05, "epoch": 3.5316455696202533, "percentage": 50.44, "elapsed_time": "4:03:32", "remaining_time": "3:59:14"} +{"current_steps": 1820, "total_steps": 3598, "loss": 0.2051, "lr": 2.3092269320886497e-05, "epoch": 3.5413826679649465, "percentage": 50.58, "elapsed_time": "4:04:17", "remaining_time": "3:58:38"} +{"current_steps": 1825, "total_steps": 3598, "loss": 0.2096, "lr": 2.299637737765358e-05, "epoch": 3.5511197663096397, "percentage": 50.72, "elapsed_time": "4:05:05", "remaining_time": "3:58:06"} +{"current_steps": 1830, "total_steps": 3598, "loss": 0.1907, "lr": 2.290041491938612e-05, "epoch": 3.5608568646543333, "percentage": 50.86, "elapsed_time": "4:05:46", "remaining_time": "3:57:27"} +{"current_steps": 1835, "total_steps": 3598, "loss": 0.2121, "lr": 2.280438420440981e-05, "epoch": 3.5705939629990264, "percentage": 51.0, "elapsed_time": "4:06:28", "remaining_time": "3:56:48"} +{"current_steps": 1840, "total_steps": 3598, "loss": 0.2186, "lr": 2.2708287492656675e-05, "epoch": 3.5803310613437196, "percentage": 51.14, "elapsed_time": "4:07:13", "remaining_time": "3:56:12"} +{"current_steps": 1845, "total_steps": 3598, "loss": 0.2088, "lr": 2.2612127045611867e-05, "epoch": 3.5900681596884128, "percentage": 51.28, "elapsed_time": "4:07:56", "remaining_time": "3:55:35"} +{"current_steps": 1850, "total_steps": 3598, "loss": 0.2022, "lr": 2.2515905126260455e-05, "epoch": 3.599805258033106, "percentage": 51.42, "elapsed_time": "4:08:29", "remaining_time": "3:54:47"} +{"current_steps": 1855, "total_steps": 3598, "loss": 0.2092, "lr": 2.2419623999034146e-05, "epoch": 3.6095423563777995, "percentage": 51.56, "elapsed_time": "4:09:05", "remaining_time": "3:54:02"} +{"current_steps": 1860, "total_steps": 3598, "loss": 0.1893, "lr": 2.232328592975802e-05, "epoch": 3.6192794547224927, "percentage": 51.7, "elapsed_time": "4:09:51", "remaining_time": "3:53:28"} +{"current_steps": 1865, "total_steps": 3598, "loss": 0.1946, "lr": 2.2226893185597204e-05, "epoch": 3.629016553067186, "percentage": 51.83, "elapsed_time": "4:10:31", "remaining_time": "3:52:47"} +{"current_steps": 1870, "total_steps": 3598, "loss": 0.196, "lr": 2.213044803500351e-05, "epoch": 3.6387536514118795, "percentage": 51.97, "elapsed_time": "4:11:23", "remaining_time": "3:52:18"} +{"current_steps": 1875, "total_steps": 3598, "loss": 0.2029, "lr": 2.2033952747662052e-05, "epoch": 3.6484907497565726, "percentage": 52.11, "elapsed_time": "4:12:03", "remaining_time": "3:51:37"} +{"current_steps": 1880, "total_steps": 3598, "loss": 0.1914, "lr": 2.193740959443782e-05, "epoch": 3.6582278481012658, "percentage": 52.25, "elapsed_time": "4:12:47", "remaining_time": "3:51:00"} +{"current_steps": 1885, "total_steps": 3598, "loss": 0.2055, "lr": 2.1840820847322274e-05, "epoch": 3.667964946445959, "percentage": 52.39, "elapsed_time": "4:13:37", "remaining_time": "3:50:28"} +{"current_steps": 1890, "total_steps": 3598, "loss": 0.2075, "lr": 2.1744188779379844e-05, "epoch": 3.6777020447906525, "percentage": 52.53, "elapsed_time": "4:14:14", "remaining_time": "3:49:45"} +{"current_steps": 1895, "total_steps": 3598, "loss": 0.2103, "lr": 2.1647515664694455e-05, "epoch": 3.6874391431353457, "percentage": 52.67, "elapsed_time": "4:14:46", "remaining_time": "3:48:57"} +{"current_steps": 1900, "total_steps": 3598, "loss": 0.1933, "lr": 2.155080377831599e-05, "epoch": 3.697176241480039, "percentage": 52.81, "elapsed_time": "4:15:23", "remaining_time": "3:48:14"} +{"current_steps": 1905, "total_steps": 3598, "loss": 0.2161, "lr": 2.145405539620677e-05, "epoch": 3.706913339824732, "percentage": 52.95, "elapsed_time": "4:15:57", "remaining_time": "3:47:28"} +{"current_steps": 1910, "total_steps": 3598, "loss": 0.2129, "lr": 2.135727279518798e-05, "epoch": 3.7166504381694256, "percentage": 53.09, "elapsed_time": "4:16:35", "remaining_time": "3:46:45"} +{"current_steps": 1915, "total_steps": 3598, "loss": 0.2094, "lr": 2.1260458252886117e-05, "epoch": 3.726387536514119, "percentage": 53.22, "elapsed_time": "4:17:14", "remaining_time": "3:46:04"} +{"current_steps": 1920, "total_steps": 3598, "loss": 0.2045, "lr": 2.1163614047679322e-05, "epoch": 3.736124634858812, "percentage": 53.36, "elapsed_time": "4:17:58", "remaining_time": "3:45:27"} +{"current_steps": 1925, "total_steps": 3598, "loss": 0.2036, "lr": 2.1066742458643853e-05, "epoch": 3.7458617332035056, "percentage": 53.5, "elapsed_time": "4:18:41", "remaining_time": "3:44:49"} +{"current_steps": 1930, "total_steps": 3598, "loss": 0.2101, "lr": 2.096984576550037e-05, "epoch": 3.7555988315481987, "percentage": 53.64, "elapsed_time": "4:19:24", "remaining_time": "3:44:11"} +{"current_steps": 1935, "total_steps": 3598, "loss": 0.1981, "lr": 2.087292624856033e-05, "epoch": 3.765335929892892, "percentage": 53.78, "elapsed_time": "4:20:08", "remaining_time": "3:43:34"} +{"current_steps": 1940, "total_steps": 3598, "loss": 0.2074, "lr": 2.0775986188672324e-05, "epoch": 3.775073028237585, "percentage": 53.92, "elapsed_time": "4:20:45", "remaining_time": "3:42:51"} +{"current_steps": 1945, "total_steps": 3598, "loss": 0.1962, "lr": 2.067902786716836e-05, "epoch": 3.7848101265822782, "percentage": 54.06, "elapsed_time": "4:21:32", "remaining_time": "3:42:16"} +{"current_steps": 1950, "total_steps": 3598, "loss": 0.2023, "lr": 2.0582053565810242e-05, "epoch": 3.794547224926972, "percentage": 54.2, "elapsed_time": "4:22:09", "remaining_time": "3:41:33"} +{"current_steps": 1955, "total_steps": 3598, "loss": 0.1887, "lr": 2.0485065566735795e-05, "epoch": 3.804284323271665, "percentage": 54.34, "elapsed_time": "4:22:50", "remaining_time": "3:40:53"} +{"current_steps": 1960, "total_steps": 3598, "loss": 0.2155, "lr": 2.038806615240523e-05, "epoch": 3.814021421616358, "percentage": 54.47, "elapsed_time": "4:23:36", "remaining_time": "3:40:17"} +{"current_steps": 1965, "total_steps": 3598, "loss": 0.1964, "lr": 2.0291057605547393e-05, "epoch": 3.8237585199610518, "percentage": 54.61, "elapsed_time": "4:24:19", "remaining_time": "3:39:39"} +{"current_steps": 1970, "total_steps": 3598, "loss": 0.2138, "lr": 2.0194042209106023e-05, "epoch": 3.833495618305745, "percentage": 54.75, "elapsed_time": "4:24:54", "remaining_time": "3:38:55"} +{"current_steps": 1975, "total_steps": 3598, "loss": 0.1954, "lr": 2.0097022246186092e-05, "epoch": 3.843232716650438, "percentage": 54.89, "elapsed_time": "4:25:39", "remaining_time": "3:38:18"} +{"current_steps": 1980, "total_steps": 3598, "loss": 0.2095, "lr": 2e-05, "epoch": 3.8529698149951317, "percentage": 55.03, "elapsed_time": "4:26:17", "remaining_time": "3:37:36"} +{"current_steps": 1985, "total_steps": 3598, "loss": 0.2051, "lr": 1.9902977753813918e-05, "epoch": 3.862706913339825, "percentage": 55.17, "elapsed_time": "4:26:57", "remaining_time": "3:36:56"} +{"current_steps": 1990, "total_steps": 3598, "loss": 0.2073, "lr": 1.980595779089398e-05, "epoch": 3.872444011684518, "percentage": 55.31, "elapsed_time": "4:27:42", "remaining_time": "3:36:19"} +{"current_steps": 1995, "total_steps": 3598, "loss": 0.2031, "lr": 1.9708942394452613e-05, "epoch": 3.882181110029211, "percentage": 55.45, "elapsed_time": "4:28:26", "remaining_time": "3:35:41"} +{"current_steps": 2000, "total_steps": 3598, "loss": 0.1919, "lr": 1.9611933847594775e-05, "epoch": 3.8919182083739043, "percentage": 55.59, "elapsed_time": "4:29:06", "remaining_time": "3:35:01"} +{"current_steps": 2005, "total_steps": 3598, "loss": 0.1994, "lr": 1.951493443326421e-05, "epoch": 3.901655306718598, "percentage": 55.73, "elapsed_time": "4:29:58", "remaining_time": "3:34:29"} +{"current_steps": 2010, "total_steps": 3598, "loss": 0.2083, "lr": 1.9417946434189768e-05, "epoch": 3.911392405063291, "percentage": 55.86, "elapsed_time": "4:30:41", "remaining_time": "3:33:51"} +{"current_steps": 2015, "total_steps": 3598, "loss": 0.2127, "lr": 1.9320972132831643e-05, "epoch": 3.9211295034079843, "percentage": 56.0, "elapsed_time": "4:31:15", "remaining_time": "3:33:06"} +{"current_steps": 2020, "total_steps": 3598, "loss": 0.1982, "lr": 1.9224013811327686e-05, "epoch": 3.930866601752678, "percentage": 56.14, "elapsed_time": "4:32:04", "remaining_time": "3:32:32"} +{"current_steps": 2025, "total_steps": 3598, "loss": 0.1968, "lr": 1.912707375143967e-05, "epoch": 3.940603700097371, "percentage": 56.28, "elapsed_time": "4:32:40", "remaining_time": "3:31:48"} +{"current_steps": 2030, "total_steps": 3598, "loss": 0.1988, "lr": 1.9030154234499635e-05, "epoch": 3.950340798442064, "percentage": 56.42, "elapsed_time": "4:33:15", "remaining_time": "3:31:04"} +{"current_steps": 2035, "total_steps": 3598, "loss": 0.2011, "lr": 1.8933257541356154e-05, "epoch": 3.960077896786758, "percentage": 56.56, "elapsed_time": "4:33:53", "remaining_time": "3:30:21"} +{"current_steps": 2040, "total_steps": 3598, "loss": 0.2216, "lr": 1.8836385952320685e-05, "epoch": 3.969814995131451, "percentage": 56.7, "elapsed_time": "4:34:22", "remaining_time": "3:29:32"} +{"current_steps": 2045, "total_steps": 3598, "loss": 0.1897, "lr": 1.873954174711389e-05, "epoch": 3.979552093476144, "percentage": 56.84, "elapsed_time": "4:34:57", "remaining_time": "3:28:48"} +{"current_steps": 2050, "total_steps": 3598, "loss": 0.1966, "lr": 1.8642727204812025e-05, "epoch": 3.9892891918208373, "percentage": 56.98, "elapsed_time": "4:35:31", "remaining_time": "3:28:03"} +{"current_steps": 2055, "total_steps": 3598, "loss": 0.1898, "lr": 1.854594460379324e-05, "epoch": 3.9990262901655305, "percentage": 57.12, "elapsed_time": "4:36:22", "remaining_time": "3:27:31"} +{"current_steps": 2060, "total_steps": 3598, "loss": 0.1806, "lr": 1.8449196221684015e-05, "epoch": 4.0077896786757545, "percentage": 57.25, "elapsed_time": "4:37:00", "remaining_time": "3:26:48"} +{"current_steps": 2065, "total_steps": 3598, "loss": 0.1745, "lr": 1.835248433530555e-05, "epoch": 4.017526777020448, "percentage": 57.39, "elapsed_time": "4:37:36", "remaining_time": "3:26:05"} +{"current_steps": 2070, "total_steps": 3598, "loss": 0.1803, "lr": 1.825581122062016e-05, "epoch": 4.027263875365141, "percentage": 57.53, "elapsed_time": "4:38:18", "remaining_time": "3:25:26"} +{"current_steps": 2075, "total_steps": 3598, "loss": 0.1797, "lr": 1.815917915267773e-05, "epoch": 4.0370009737098345, "percentage": 57.67, "elapsed_time": "4:39:05", "remaining_time": "3:24:50"} +{"current_steps": 2080, "total_steps": 3598, "loss": 0.175, "lr": 1.8062590405562185e-05, "epoch": 4.046738072054528, "percentage": 57.81, "elapsed_time": "4:39:36", "remaining_time": "3:24:03"} +{"current_steps": 2085, "total_steps": 3598, "loss": 0.1788, "lr": 1.7966047252337958e-05, "epoch": 4.056475170399221, "percentage": 57.95, "elapsed_time": "4:40:13", "remaining_time": "3:23:21"} +{"current_steps": 2090, "total_steps": 3598, "loss": 0.1825, "lr": 1.78695519649965e-05, "epoch": 4.066212268743914, "percentage": 58.09, "elapsed_time": "4:40:59", "remaining_time": "3:22:44"} +{"current_steps": 2095, "total_steps": 3598, "loss": 0.1801, "lr": 1.7773106814402796e-05, "epoch": 4.075949367088608, "percentage": 58.23, "elapsed_time": "4:41:49", "remaining_time": "3:22:11"} +{"current_steps": 2100, "total_steps": 3598, "loss": 0.1579, "lr": 1.7676714070241987e-05, "epoch": 4.085686465433301, "percentage": 58.37, "elapsed_time": "4:42:35", "remaining_time": "3:21:35"} +{"current_steps": 2105, "total_steps": 3598, "loss": 0.1745, "lr": 1.7580376000965868e-05, "epoch": 4.095423563777994, "percentage": 58.5, "elapsed_time": "4:43:11", "remaining_time": "3:20:51"} +{"current_steps": 2110, "total_steps": 3598, "loss": 0.1724, "lr": 1.748409487373955e-05, "epoch": 4.105160662122687, "percentage": 58.64, "elapsed_time": "4:43:46", "remaining_time": "3:20:07"} +{"current_steps": 2115, "total_steps": 3598, "loss": 0.1797, "lr": 1.7387872954388136e-05, "epoch": 4.114897760467381, "percentage": 58.78, "elapsed_time": "4:44:21", "remaining_time": "3:19:23"} +{"current_steps": 2120, "total_steps": 3598, "loss": 0.1781, "lr": 1.729171250734333e-05, "epoch": 4.124634858812074, "percentage": 58.92, "elapsed_time": "4:45:00", "remaining_time": "3:18:42"} +{"current_steps": 2125, "total_steps": 3598, "loss": 0.1902, "lr": 1.71956157955902e-05, "epoch": 4.134371957156767, "percentage": 59.06, "elapsed_time": "4:45:33", "remaining_time": "3:17:56"} +{"current_steps": 2130, "total_steps": 3598, "loss": 0.1846, "lr": 1.7099585080613884e-05, "epoch": 4.144109055501461, "percentage": 59.2, "elapsed_time": "4:46:15", "remaining_time": "3:17:17"} +{"current_steps": 2135, "total_steps": 3598, "loss": 0.166, "lr": 1.7003622622346424e-05, "epoch": 4.153846153846154, "percentage": 59.34, "elapsed_time": "4:46:52", "remaining_time": "3:16:34"} +{"current_steps": 2140, "total_steps": 3598, "loss": 0.1665, "lr": 1.6907730679113513e-05, "epoch": 4.163583252190847, "percentage": 59.48, "elapsed_time": "4:47:35", "remaining_time": "3:15:56"} +{"current_steps": 2145, "total_steps": 3598, "loss": 0.1691, "lr": 1.6811911507581394e-05, "epoch": 4.1733203505355405, "percentage": 59.62, "elapsed_time": "4:48:12", "remaining_time": "3:15:13"} +{"current_steps": 2150, "total_steps": 3598, "loss": 0.1743, "lr": 1.6716167362703764e-05, "epoch": 4.183057448880234, "percentage": 59.76, "elapsed_time": "4:48:52", "remaining_time": "3:14:33"} +{"current_steps": 2155, "total_steps": 3598, "loss": 0.1895, "lr": 1.6620500497668656e-05, "epoch": 4.192794547224927, "percentage": 59.89, "elapsed_time": "4:49:19", "remaining_time": "3:13:44"} +{"current_steps": 2160, "total_steps": 3598, "loss": 0.1656, "lr": 1.6524913163845456e-05, "epoch": 4.2025316455696204, "percentage": 60.03, "elapsed_time": "4:50:01", "remaining_time": "3:13:04"} +{"current_steps": 2165, "total_steps": 3598, "loss": 0.1892, "lr": 1.6429407610731902e-05, "epoch": 4.212268743914313, "percentage": 60.17, "elapsed_time": "4:50:46", "remaining_time": "3:12:27"} +{"current_steps": 2170, "total_steps": 3598, "loss": 0.1775, "lr": 1.6333986085901167e-05, "epoch": 4.222005842259007, "percentage": 60.31, "elapsed_time": "4:51:28", "remaining_time": "3:11:48"} +{"current_steps": 2175, "total_steps": 3598, "loss": 0.1723, "lr": 1.6238650834948935e-05, "epoch": 4.2317429406037, "percentage": 60.45, "elapsed_time": "4:52:10", "remaining_time": "3:11:09"} +{"current_steps": 2180, "total_steps": 3598, "loss": 0.1879, "lr": 1.614340410144056e-05, "epoch": 4.241480038948393, "percentage": 60.59, "elapsed_time": "4:52:54", "remaining_time": "3:10:31"} +{"current_steps": 2185, "total_steps": 3598, "loss": 0.1729, "lr": 1.604824812685832e-05, "epoch": 4.251217137293087, "percentage": 60.73, "elapsed_time": "4:53:33", "remaining_time": "3:09:50"} +{"current_steps": 2190, "total_steps": 3598, "loss": 0.1829, "lr": 1.5953185150548574e-05, "epoch": 4.26095423563778, "percentage": 60.87, "elapsed_time": "4:54:03", "remaining_time": "3:09:03"} +{"current_steps": 2195, "total_steps": 3598, "loss": 0.1753, "lr": 1.5858217409669138e-05, "epoch": 4.270691333982473, "percentage": 61.01, "elapsed_time": "4:54:48", "remaining_time": "3:08:26"} +{"current_steps": 2200, "total_steps": 3598, "loss": 0.1718, "lr": 1.576334713913661e-05, "epoch": 4.280428432327167, "percentage": 61.15, "elapsed_time": "4:55:32", "remaining_time": "3:07:47"} +{"current_steps": 2205, "total_steps": 3598, "loss": 0.1777, "lr": 1.566857657157378e-05, "epoch": 4.29016553067186, "percentage": 61.28, "elapsed_time": "4:56:17", "remaining_time": "3:07:10"} +{"current_steps": 2210, "total_steps": 3598, "loss": 0.1692, "lr": 1.557390793725707e-05, "epoch": 4.299902629016553, "percentage": 61.42, "elapsed_time": "4:56:59", "remaining_time": "3:06:31"} +{"current_steps": 2215, "total_steps": 3598, "loss": 0.1867, "lr": 1.5479343464064077e-05, "epoch": 4.309639727361247, "percentage": 61.56, "elapsed_time": "4:57:38", "remaining_time": "3:05:50"} +{"current_steps": 2220, "total_steps": 3598, "loss": 0.1711, "lr": 1.5384885377421132e-05, "epoch": 4.319376825705939, "percentage": 61.7, "elapsed_time": "4:58:13", "remaining_time": "3:05:06"} +{"current_steps": 2225, "total_steps": 3598, "loss": 0.176, "lr": 1.5290535900250914e-05, "epoch": 4.329113924050633, "percentage": 61.84, "elapsed_time": "4:58:52", "remaining_time": "3:04:25"} +{"current_steps": 2230, "total_steps": 3598, "loss": 0.1898, "lr": 1.5196297252920142e-05, "epoch": 4.3388510223953265, "percentage": 61.98, "elapsed_time": "4:59:28", "remaining_time": "3:03:42"} +{"current_steps": 2235, "total_steps": 3598, "loss": 0.1777, "lr": 1.5102171653187367e-05, "epoch": 4.348588120740019, "percentage": 62.12, "elapsed_time": "4:59:57", "remaining_time": "3:02:55"} +{"current_steps": 2240, "total_steps": 3598, "loss": 0.1856, "lr": 1.5008161316150699e-05, "epoch": 4.358325219084713, "percentage": 62.26, "elapsed_time": "5:00:41", "remaining_time": "3:02:17"} +{"current_steps": 2245, "total_steps": 3598, "loss": 0.1752, "lr": 1.4914268454195737e-05, "epoch": 4.368062317429406, "percentage": 62.4, "elapsed_time": "5:01:24", "remaining_time": "3:01:38"} +{"current_steps": 2250, "total_steps": 3598, "loss": 0.1808, "lr": 1.4820495276943491e-05, "epoch": 4.377799415774099, "percentage": 62.53, "elapsed_time": "5:01:58", "remaining_time": "3:00:55"} +{"current_steps": 2255, "total_steps": 3598, "loss": 0.1743, "lr": 1.4726843991198389e-05, "epoch": 4.387536514118793, "percentage": 62.67, "elapsed_time": "5:02:42", "remaining_time": "3:00:17"} +{"current_steps": 2260, "total_steps": 3598, "loss": 0.1727, "lr": 1.4633316800896316e-05, "epoch": 4.3972736124634855, "percentage": 62.81, "elapsed_time": "5:03:28", "remaining_time": "2:59:39"} +{"current_steps": 2265, "total_steps": 3598, "loss": 0.1779, "lr": 1.453991590705277e-05, "epoch": 4.407010710808179, "percentage": 62.95, "elapsed_time": "5:04:01", "remaining_time": "2:58:55"} +{"current_steps": 2270, "total_steps": 3598, "loss": 0.1667, "lr": 1.4446643507711068e-05, "epoch": 4.416747809152873, "percentage": 63.09, "elapsed_time": "5:04:49", "remaining_time": "2:58:19"} +{"current_steps": 2275, "total_steps": 3598, "loss": 0.1853, "lr": 1.4353501797890617e-05, "epoch": 4.426484907497565, "percentage": 63.23, "elapsed_time": "5:05:26", "remaining_time": "2:57:37"} +{"current_steps": 2280, "total_steps": 3598, "loss": 0.1761, "lr": 1.4260492969535218e-05, "epoch": 4.436222005842259, "percentage": 63.37, "elapsed_time": "5:06:06", "remaining_time": "2:56:56"} +{"current_steps": 2285, "total_steps": 3598, "loss": 0.1866, "lr": 1.416761921146155e-05, "epoch": 4.445959104186953, "percentage": 63.51, "elapsed_time": "5:06:45", "remaining_time": "2:56:16"} +{"current_steps": 2290, "total_steps": 3598, "loss": 0.1825, "lr": 1.4074882709307617e-05, "epoch": 4.455696202531645, "percentage": 63.65, "elapsed_time": "5:07:27", "remaining_time": "2:55:37"} +{"current_steps": 2295, "total_steps": 3598, "loss": 0.1776, "lr": 1.398228564548131e-05, "epoch": 4.465433300876339, "percentage": 63.79, "elapsed_time": "5:08:16", "remaining_time": "2:55:01"} +{"current_steps": 2300, "total_steps": 3598, "loss": 0.1734, "lr": 1.3889830199109064e-05, "epoch": 4.4751703992210325, "percentage": 63.92, "elapsed_time": "5:08:53", "remaining_time": "2:54:19"} +{"current_steps": 2305, "total_steps": 3598, "loss": 0.1775, "lr": 1.3797518545984577e-05, "epoch": 4.484907497565725, "percentage": 64.06, "elapsed_time": "5:09:37", "remaining_time": "2:53:40"} +{"current_steps": 2310, "total_steps": 3598, "loss": 0.1676, "lr": 1.3705352858517596e-05, "epoch": 4.494644595910419, "percentage": 64.2, "elapsed_time": "5:10:21", "remaining_time": "2:53:02"} +{"current_steps": 2315, "total_steps": 3598, "loss": 0.1875, "lr": 1.3613335305682769e-05, "epoch": 4.504381694255112, "percentage": 64.34, "elapsed_time": "5:10:59", "remaining_time": "2:52:21"} +{"current_steps": 2320, "total_steps": 3598, "loss": 0.1664, "lr": 1.3521468052968682e-05, "epoch": 4.514118792599805, "percentage": 64.48, "elapsed_time": "5:11:48", "remaining_time": "2:51:46"} +{"current_steps": 2325, "total_steps": 3598, "loss": 0.1729, "lr": 1.3429753262326793e-05, "epoch": 4.523855890944499, "percentage": 64.62, "elapsed_time": "5:12:30", "remaining_time": "2:51:06"} +{"current_steps": 2330, "total_steps": 3598, "loss": 0.1775, "lr": 1.3338193092120628e-05, "epoch": 4.5335929892891915, "percentage": 64.76, "elapsed_time": "5:13:11", "remaining_time": "2:50:26"} +{"current_steps": 2335, "total_steps": 3598, "loss": 0.1708, "lr": 1.3246789697074958e-05, "epoch": 4.543330087633885, "percentage": 64.9, "elapsed_time": "5:13:55", "remaining_time": "2:49:48"} +{"current_steps": 2340, "total_steps": 3598, "loss": 0.1806, "lr": 1.3155545228225104e-05, "epoch": 4.553067185978579, "percentage": 65.04, "elapsed_time": "5:14:29", "remaining_time": "2:49:04"} +{"current_steps": 2345, "total_steps": 3598, "loss": 0.1765, "lr": 1.3064461832866304e-05, "epoch": 4.5628042843232715, "percentage": 65.18, "elapsed_time": "5:15:06", "remaining_time": "2:48:22"} +{"current_steps": 2350, "total_steps": 3598, "loss": 0.1689, "lr": 1.297354165450318e-05, "epoch": 4.572541382667965, "percentage": 65.31, "elapsed_time": "5:15:50", "remaining_time": "2:47:43"} +{"current_steps": 2355, "total_steps": 3598, "loss": 0.1845, "lr": 1.2882786832799305e-05, "epoch": 4.582278481012658, "percentage": 65.45, "elapsed_time": "5:16:30", "remaining_time": "2:47:03"} +{"current_steps": 2360, "total_steps": 3598, "loss": 0.1814, "lr": 1.2792199503526848e-05, "epoch": 4.592015579357351, "percentage": 65.59, "elapsed_time": "5:17:20", "remaining_time": "2:46:28"} +{"current_steps": 2365, "total_steps": 3598, "loss": 0.1835, "lr": 1.2701781798516288e-05, "epoch": 4.601752677702045, "percentage": 65.73, "elapsed_time": "5:17:54", "remaining_time": "2:45:44"} +{"current_steps": 2370, "total_steps": 3598, "loss": 0.1777, "lr": 1.2611535845606273e-05, "epoch": 4.611489776046738, "percentage": 65.87, "elapsed_time": "5:18:33", "remaining_time": "2:45:03"} +{"current_steps": 2375, "total_steps": 3598, "loss": 0.1719, "lr": 1.2521463768593545e-05, "epoch": 4.621226874391431, "percentage": 66.01, "elapsed_time": "5:19:26", "remaining_time": "2:44:29"} +{"current_steps": 2380, "total_steps": 3598, "loss": 0.1782, "lr": 1.2431567687182939e-05, "epoch": 4.630963972736125, "percentage": 66.15, "elapsed_time": "5:20:00", "remaining_time": "2:43:46"} +{"current_steps": 2385, "total_steps": 3598, "loss": 0.1763, "lr": 1.234184971693751e-05, "epoch": 4.640701071080818, "percentage": 66.29, "elapsed_time": "5:20:41", "remaining_time": "2:43:06"} +{"current_steps": 2390, "total_steps": 3598, "loss": 0.184, "lr": 1.2252311969228754e-05, "epoch": 4.650438169425511, "percentage": 66.43, "elapsed_time": "5:21:15", "remaining_time": "2:42:22"} +{"current_steps": 2395, "total_steps": 3598, "loss": 0.1777, "lr": 1.2162956551186904e-05, "epoch": 4.660175267770205, "percentage": 66.56, "elapsed_time": "5:22:00", "remaining_time": "2:41:44"} +{"current_steps": 2400, "total_steps": 3598, "loss": 0.1732, "lr": 1.2073785565651362e-05, "epoch": 4.669912366114898, "percentage": 66.7, "elapsed_time": "5:22:35", "remaining_time": "2:41:01"} +{"current_steps": 2405, "total_steps": 3598, "loss": 0.1698, "lr": 1.1984801111121173e-05, "epoch": 4.679649464459591, "percentage": 66.84, "elapsed_time": "5:23:58", "remaining_time": "2:40:42"} +{"current_steps": 2410, "total_steps": 3598, "loss": 0.1814, "lr": 1.1896005281705717e-05, "epoch": 4.689386562804284, "percentage": 66.98, "elapsed_time": "5:24:48", "remaining_time": "2:40:06"} +{"current_steps": 2415, "total_steps": 3598, "loss": 0.1842, "lr": 1.180740016707533e-05, "epoch": 4.6991236611489775, "percentage": 67.12, "elapsed_time": "5:25:23", "remaining_time": "2:39:23"} +{"current_steps": 2420, "total_steps": 3598, "loss": 0.1739, "lr": 1.1718987852412193e-05, "epoch": 4.708860759493671, "percentage": 67.26, "elapsed_time": "5:26:00", "remaining_time": "2:38:41"} +{"current_steps": 2425, "total_steps": 3598, "loss": 0.1767, "lr": 1.1630770418361255e-05, "epoch": 4.718597857838364, "percentage": 67.4, "elapsed_time": "5:26:40", "remaining_time": "2:38:01"} +{"current_steps": 2430, "total_steps": 3598, "loss": 0.1828, "lr": 1.154274994098124e-05, "epoch": 4.728334956183057, "percentage": 67.54, "elapsed_time": "5:27:19", "remaining_time": "2:37:20"} +{"current_steps": 2435, "total_steps": 3598, "loss": 0.1603, "lr": 1.1454928491695803e-05, "epoch": 4.738072054527751, "percentage": 67.68, "elapsed_time": "5:27:54", "remaining_time": "2:36:36"} +{"current_steps": 2440, "total_steps": 3598, "loss": 0.1797, "lr": 1.1367308137244817e-05, "epoch": 4.747809152872444, "percentage": 67.82, "elapsed_time": "5:28:43", "remaining_time": "2:36:00"} +{"current_steps": 2445, "total_steps": 3598, "loss": 0.1758, "lr": 1.1279890939635659e-05, "epoch": 4.757546251217137, "percentage": 67.95, "elapsed_time": "5:29:23", "remaining_time": "2:35:19"} +{"current_steps": 2450, "total_steps": 3598, "loss": 0.1702, "lr": 1.1192678956094753e-05, "epoch": 4.767283349561831, "percentage": 68.09, "elapsed_time": "5:30:15", "remaining_time": "2:34:44"} +{"current_steps": 2455, "total_steps": 3598, "loss": 0.193, "lr": 1.1105674239019132e-05, "epoch": 4.777020447906524, "percentage": 68.23, "elapsed_time": "5:30:52", "remaining_time": "2:34:02"} +{"current_steps": 2460, "total_steps": 3598, "loss": 0.1864, "lr": 1.1018878835928142e-05, "epoch": 4.786757546251217, "percentage": 68.37, "elapsed_time": "5:31:26", "remaining_time": "2:33:19"} +{"current_steps": 2465, "total_steps": 3598, "loss": 0.169, "lr": 1.093229478941525e-05, "epoch": 4.79649464459591, "percentage": 68.51, "elapsed_time": "5:32:06", "remaining_time": "2:32:38"} +{"current_steps": 2470, "total_steps": 3598, "loss": 0.1568, "lr": 1.0845924137099953e-05, "epoch": 4.806231742940604, "percentage": 68.65, "elapsed_time": "5:32:58", "remaining_time": "2:32:03"} +{"current_steps": 2475, "total_steps": 3598, "loss": 0.1896, "lr": 1.0759768911579896e-05, "epoch": 4.815968841285297, "percentage": 68.79, "elapsed_time": "5:33:34", "remaining_time": "2:31:21"} +{"current_steps": 2480, "total_steps": 3598, "loss": 0.1786, "lr": 1.0673831140382962e-05, "epoch": 4.82570593962999, "percentage": 68.93, "elapsed_time": "5:34:07", "remaining_time": "2:30:37"} +{"current_steps": 2485, "total_steps": 3598, "loss": 0.1823, "lr": 1.0588112845919594e-05, "epoch": 4.8354430379746836, "percentage": 69.07, "elapsed_time": "5:34:44", "remaining_time": "2:29:55"} +{"current_steps": 2490, "total_steps": 3598, "loss": 0.1756, "lr": 1.0502616045435202e-05, "epoch": 4.845180136319377, "percentage": 69.21, "elapsed_time": "5:35:20", "remaining_time": "2:29:13"} +{"current_steps": 2495, "total_steps": 3598, "loss": 0.1774, "lr": 1.0417342750962671e-05, "epoch": 4.85491723466407, "percentage": 69.34, "elapsed_time": "5:36:01", "remaining_time": "2:28:33"} +{"current_steps": 2500, "total_steps": 3598, "loss": 0.1812, "lr": 1.0332294969275032e-05, "epoch": 4.8646543330087635, "percentage": 69.48, "elapsed_time": "5:36:44", "remaining_time": "2:27:53"} +{"current_steps": 2505, "total_steps": 3598, "loss": 0.1753, "lr": 1.0247474701838222e-05, "epoch": 4.874391431353457, "percentage": 69.62, "elapsed_time": "5:37:20", "remaining_time": "2:27:11"} +{"current_steps": 2510, "total_steps": 3598, "loss": 0.1799, "lr": 1.0162883944763999e-05, "epoch": 4.88412852969815, "percentage": 69.76, "elapsed_time": "5:37:50", "remaining_time": "2:26:26"} +{"current_steps": 2515, "total_steps": 3598, "loss": 0.1725, "lr": 1.0078524688762955e-05, "epoch": 4.893865628042843, "percentage": 69.9, "elapsed_time": "5:38:33", "remaining_time": "2:25:47"} +{"current_steps": 2520, "total_steps": 3598, "loss": 0.1811, "lr": 9.994398919097634e-06, "epoch": 4.903602726387536, "percentage": 70.04, "elapsed_time": "5:39:14", "remaining_time": "2:25:07"} +{"current_steps": 2525, "total_steps": 3598, "loss": 0.1808, "lr": 9.910508615535876e-06, "epoch": 4.91333982473223, "percentage": 70.18, "elapsed_time": "5:39:51", "remaining_time": "2:24:25"} +{"current_steps": 2530, "total_steps": 3598, "loss": 0.1763, "lr": 9.826855752304209e-06, "epoch": 4.923076923076923, "percentage": 70.32, "elapsed_time": "5:40:28", "remaining_time": "2:23:43"} +{"current_steps": 2535, "total_steps": 3598, "loss": 0.1804, "lr": 9.743442298041348e-06, "epoch": 4.932814021421616, "percentage": 70.46, "elapsed_time": "5:41:09", "remaining_time": "2:23:03"} +{"current_steps": 2540, "total_steps": 3598, "loss": 0.1838, "lr": 9.660270215751905e-06, "epoch": 4.94255111976631, "percentage": 70.59, "elapsed_time": "5:41:46", "remaining_time": "2:22:21"} +{"current_steps": 2545, "total_steps": 3598, "loss": 0.1819, "lr": 9.577341462760182e-06, "epoch": 4.952288218111003, "percentage": 70.73, "elapsed_time": "5:42:21", "remaining_time": "2:21:39"} +{"current_steps": 2550, "total_steps": 3598, "loss": 0.1795, "lr": 9.494657990664105e-06, "epoch": 4.962025316455696, "percentage": 70.87, "elapsed_time": "5:42:59", "remaining_time": "2:20:57"} +{"current_steps": 2555, "total_steps": 3598, "loss": 0.1837, "lr": 9.412221745289296e-06, "epoch": 4.97176241480039, "percentage": 71.01, "elapsed_time": "5:43:44", "remaining_time": "2:20:19"} +{"current_steps": 2560, "total_steps": 3598, "loss": 0.1748, "lr": 9.330034666643304e-06, "epoch": 4.981499513145083, "percentage": 71.15, "elapsed_time": "5:44:25", "remaining_time": "2:19:39"} +{"current_steps": 2565, "total_steps": 3598, "loss": 0.1809, "lr": 9.248098688869917e-06, "epoch": 4.991236611489776, "percentage": 71.29, "elapsed_time": "5:45:04", "remaining_time": "2:18:58"} +{"current_steps": 2570, "total_steps": 3598, "loss": 0.1823, "lr": 9.166415740203644e-06, "epoch": 5.0, "percentage": 71.43, "elapsed_time": "5:45:41", "remaining_time": "2:18:16"} +{"current_steps": 2575, "total_steps": 3598, "loss": 0.1469, "lr": 9.084987742924365e-06, "epoch": 5.009737098344694, "percentage": 71.57, "elapsed_time": "5:46:16", "remaining_time": "2:17:34"} +{"current_steps": 2580, "total_steps": 3598, "loss": 0.1381, "lr": 9.003816613312097e-06, "epoch": 5.019474196689386, "percentage": 71.71, "elapsed_time": "5:46:59", "remaining_time": "2:16:55"} +{"current_steps": 2585, "total_steps": 3598, "loss": 0.1565, "lr": 8.922904261601853e-06, "epoch": 5.02921129503408, "percentage": 71.85, "elapsed_time": "5:47:31", "remaining_time": "2:16:11"} +{"current_steps": 2590, "total_steps": 3598, "loss": 0.1567, "lr": 8.842252591938725e-06, "epoch": 5.0389483933787735, "percentage": 71.98, "elapsed_time": "5:48:13", "remaining_time": "2:15:31"} +{"current_steps": 2595, "total_steps": 3598, "loss": 0.1518, "lr": 8.76186350233306e-06, "epoch": 5.048685491723466, "percentage": 72.12, "elapsed_time": "5:48:53", "remaining_time": "2:14:51"} +{"current_steps": 2600, "total_steps": 3598, "loss": 0.1638, "lr": 8.681738884615805e-06, "epoch": 5.05842259006816, "percentage": 72.26, "elapsed_time": "5:49:38", "remaining_time": "2:14:12"} +{"current_steps": 2605, "total_steps": 3598, "loss": 0.1596, "lr": 8.601880624393964e-06, "epoch": 5.068159688412853, "percentage": 72.4, "elapsed_time": "5:50:19", "remaining_time": "2:13:32"} +{"current_steps": 2610, "total_steps": 3598, "loss": 0.164, "lr": 8.522290601006245e-06, "epoch": 5.077896786757546, "percentage": 72.54, "elapsed_time": "5:50:59", "remaining_time": "2:12:51"} +{"current_steps": 2615, "total_steps": 3598, "loss": 0.1575, "lr": 8.442970687478815e-06, "epoch": 5.08763388510224, "percentage": 72.68, "elapsed_time": "5:51:38", "remaining_time": "2:12:11"} +{"current_steps": 2620, "total_steps": 3598, "loss": 0.1572, "lr": 8.363922750481239e-06, "epoch": 5.0973709834469325, "percentage": 72.82, "elapsed_time": "5:52:19", "remaining_time": "2:11:31"} +{"current_steps": 2625, "total_steps": 3598, "loss": 0.1532, "lr": 8.28514865028253e-06, "epoch": 5.107108081791626, "percentage": 72.96, "elapsed_time": "5:52:58", "remaining_time": "2:10:50"} +{"current_steps": 2630, "total_steps": 3598, "loss": 0.1631, "lr": 8.2066502407074e-06, "epoch": 5.11684518013632, "percentage": 73.1, "elapsed_time": "5:53:36", "remaining_time": "2:10:08"} +{"current_steps": 2635, "total_steps": 3598, "loss": 0.1534, "lr": 8.128429369092599e-06, "epoch": 5.1265822784810124, "percentage": 73.24, "elapsed_time": "5:54:15", "remaining_time": "2:09:27"} +{"current_steps": 2640, "total_steps": 3598, "loss": 0.1597, "lr": 8.050487876243473e-06, "epoch": 5.136319376825706, "percentage": 73.37, "elapsed_time": "5:54:54", "remaining_time": "2:08:47"} +{"current_steps": 2645, "total_steps": 3598, "loss": 0.161, "lr": 7.972827596390594e-06, "epoch": 5.146056475170399, "percentage": 73.51, "elapsed_time": "5:55:33", "remaining_time": "2:08:06"} +{"current_steps": 2650, "total_steps": 3598, "loss": 0.163, "lr": 7.89545035714667e-06, "epoch": 5.155793573515092, "percentage": 73.65, "elapsed_time": "5:56:04", "remaining_time": "2:07:22"} +{"current_steps": 2655, "total_steps": 3598, "loss": 0.1562, "lr": 7.81835797946347e-06, "epoch": 5.165530671859786, "percentage": 73.79, "elapsed_time": "5:56:50", "remaining_time": "2:06:44"} +{"current_steps": 2660, "total_steps": 3598, "loss": 0.1532, "lr": 7.741552277589e-06, "epoch": 5.175267770204479, "percentage": 73.93, "elapsed_time": "5:57:22", "remaining_time": "2:06:01"} +{"current_steps": 2665, "total_steps": 3598, "loss": 0.1555, "lr": 7.665035059024794e-06, "epoch": 5.185004868549172, "percentage": 74.07, "elapsed_time": "5:58:07", "remaining_time": "2:05:22"} +{"current_steps": 2670, "total_steps": 3598, "loss": 0.1467, "lr": 7.5888081244833936e-06, "epoch": 5.194741966893866, "percentage": 74.21, "elapsed_time": "5:58:47", "remaining_time": "2:04:42"} +{"current_steps": 2675, "total_steps": 3598, "loss": 0.1535, "lr": 7.512873267845948e-06, "epoch": 5.204479065238559, "percentage": 74.35, "elapsed_time": "5:59:19", "remaining_time": "2:03:59"} +{"current_steps": 2680, "total_steps": 3598, "loss": 0.1598, "lr": 7.437232276120037e-06, "epoch": 5.214216163583252, "percentage": 74.49, "elapsed_time": "6:00:04", "remaining_time": "2:03:20"} +{"current_steps": 2685, "total_steps": 3598, "loss": 0.1569, "lr": 7.361886929397568e-06, "epoch": 5.223953261927946, "percentage": 74.62, "elapsed_time": "6:00:44", "remaining_time": "2:02:39"} +{"current_steps": 2690, "total_steps": 3598, "loss": 0.1572, "lr": 7.286839000812922e-06, "epoch": 5.233690360272639, "percentage": 74.76, "elapsed_time": "6:01:22", "remaining_time": "2:01:58"} +{"current_steps": 2695, "total_steps": 3598, "loss": 0.1582, "lr": 7.212090256501185e-06, "epoch": 5.243427458617332, "percentage": 74.9, "elapsed_time": "6:01:53", "remaining_time": "2:01:15"} +{"current_steps": 2700, "total_steps": 3598, "loss": 0.1507, "lr": 7.137642455556654e-06, "epoch": 5.253164556962025, "percentage": 75.04, "elapsed_time": "6:02:38", "remaining_time": "2:00:36"} +{"current_steps": 2705, "total_steps": 3598, "loss": 0.1593, "lr": 7.063497349991364e-06, "epoch": 5.2629016553067185, "percentage": 75.18, "elapsed_time": "6:03:20", "remaining_time": "1:59:56"} +{"current_steps": 2710, "total_steps": 3598, "loss": 0.1556, "lr": 6.9896566846939e-06, "epoch": 5.272638753651412, "percentage": 75.32, "elapsed_time": "6:04:03", "remaining_time": "1:59:17"} +{"current_steps": 2715, "total_steps": 3598, "loss": 0.1641, "lr": 6.916122197388322e-06, "epoch": 5.282375851996105, "percentage": 75.46, "elapsed_time": "6:04:43", "remaining_time": "1:58:37"} +{"current_steps": 2720, "total_steps": 3598, "loss": 0.1651, "lr": 6.842895618593275e-06, "epoch": 5.292112950340798, "percentage": 75.6, "elapsed_time": "6:05:28", "remaining_time": "1:57:58"} +{"current_steps": 2725, "total_steps": 3598, "loss": 0.1555, "lr": 6.769978671581257e-06, "epoch": 5.301850048685492, "percentage": 75.74, "elapsed_time": "6:06:10", "remaining_time": "1:57:18"} +{"current_steps": 2730, "total_steps": 3598, "loss": 0.16, "lr": 6.697373072338065e-06, "epoch": 5.311587147030185, "percentage": 75.88, "elapsed_time": "6:06:49", "remaining_time": "1:56:37"} +{"current_steps": 2735, "total_steps": 3598, "loss": 0.1569, "lr": 6.625080529522432e-06, "epoch": 5.321324245374878, "percentage": 76.01, "elapsed_time": "6:07:33", "remaining_time": "1:55:58"} +{"current_steps": 2740, "total_steps": 3598, "loss": 0.1636, "lr": 6.55310274442579e-06, "epoch": 5.331061343719572, "percentage": 76.15, "elapsed_time": "6:08:15", "remaining_time": "1:55:18"} +{"current_steps": 2745, "total_steps": 3598, "loss": 0.1714, "lr": 6.481441410932221e-06, "epoch": 5.340798442064265, "percentage": 76.29, "elapsed_time": "6:08:52", "remaining_time": "1:54:37"} +{"current_steps": 2750, "total_steps": 3598, "loss": 0.1583, "lr": 6.410098215478655e-06, "epoch": 5.350535540408958, "percentage": 76.43, "elapsed_time": "6:09:30", "remaining_time": "1:53:56"} +{"current_steps": 2755, "total_steps": 3598, "loss": 0.1537, "lr": 6.339074837015115e-06, "epoch": 5.360272638753651, "percentage": 76.57, "elapsed_time": "6:10:13", "remaining_time": "1:53:17"} +{"current_steps": 2760, "total_steps": 3598, "loss": 0.1645, "lr": 6.268372946965245e-06, "epoch": 5.370009737098345, "percentage": 76.71, "elapsed_time": "6:10:54", "remaining_time": "1:52:36"} +{"current_steps": 2765, "total_steps": 3598, "loss": 0.157, "lr": 6.197994209186959e-06, "epoch": 5.379746835443038, "percentage": 76.85, "elapsed_time": "6:11:27", "remaining_time": "1:51:54"} +{"current_steps": 2770, "total_steps": 3598, "loss": 0.1522, "lr": 6.1279402799332934e-06, "epoch": 5.389483933787731, "percentage": 76.99, "elapsed_time": "6:12:05", "remaining_time": "1:51:13"} +{"current_steps": 2775, "total_steps": 3598, "loss": 0.1668, "lr": 6.058212807813426e-06, "epoch": 5.3992210321324245, "percentage": 77.13, "elapsed_time": "6:12:41", "remaining_time": "1:50:31"} +{"current_steps": 2780, "total_steps": 3598, "loss": 0.1567, "lr": 5.988813433753869e-06, "epoch": 5.408958130477118, "percentage": 77.27, "elapsed_time": "6:13:26", "remaining_time": "1:49:53"} +{"current_steps": 2785, "total_steps": 3598, "loss": 0.1661, "lr": 5.919743790959888e-06, "epoch": 5.418695228821811, "percentage": 77.4, "elapsed_time": "6:14:11", "remaining_time": "1:49:14"} +{"current_steps": 2790, "total_steps": 3598, "loss": 0.1558, "lr": 5.851005504877012e-06, "epoch": 5.4284323271665045, "percentage": 77.54, "elapsed_time": "6:14:56", "remaining_time": "1:48:35"} +{"current_steps": 2795, "total_steps": 3598, "loss": 0.1597, "lr": 5.782600193152819e-06, "epoch": 5.438169425511198, "percentage": 77.68, "elapsed_time": "6:15:31", "remaining_time": "1:47:53"} +{"current_steps": 2800, "total_steps": 3598, "loss": 0.1663, "lr": 5.714529465598872e-06, "epoch": 5.447906523855891, "percentage": 77.82, "elapsed_time": "6:16:05", "remaining_time": "1:47:11"} +{"current_steps": 2805, "total_steps": 3598, "loss": 0.1666, "lr": 5.646794924152808e-06, "epoch": 5.457643622200584, "percentage": 77.96, "elapsed_time": "6:16:49", "remaining_time": "1:46:31"} +{"current_steps": 2810, "total_steps": 3598, "loss": 0.1544, "lr": 5.579398162840655e-06, "epoch": 5.467380720545277, "percentage": 78.1, "elapsed_time": "6:17:31", "remaining_time": "1:45:52"} +{"current_steps": 2815, "total_steps": 3598, "loss": 0.1563, "lr": 5.512340767739315e-06, "epoch": 5.477117818889971, "percentage": 78.24, "elapsed_time": "6:18:04", "remaining_time": "1:45:09"} +{"current_steps": 2820, "total_steps": 3598, "loss": 0.1641, "lr": 5.445624316939244e-06, "epoch": 5.486854917234664, "percentage": 78.38, "elapsed_time": "6:18:59", "remaining_time": "1:44:33"} +{"current_steps": 2825, "total_steps": 3598, "loss": 0.1538, "lr": 5.379250380507302e-06, "epoch": 5.496592015579357, "percentage": 78.52, "elapsed_time": "6:19:49", "remaining_time": "1:43:55"} +{"current_steps": 2830, "total_steps": 3598, "loss": 0.1647, "lr": 5.313220520449811e-06, "epoch": 5.506329113924051, "percentage": 78.65, "elapsed_time": "6:20:33", "remaining_time": "1:43:16"} +{"current_steps": 2835, "total_steps": 3598, "loss": 0.1672, "lr": 5.247536290675821e-06, "epoch": 5.516066212268744, "percentage": 78.79, "elapsed_time": "6:21:11", "remaining_time": "1:42:35"} +{"current_steps": 2840, "total_steps": 3598, "loss": 0.1661, "lr": 5.182199236960476e-06, "epoch": 5.525803310613437, "percentage": 78.93, "elapsed_time": "6:21:46", "remaining_time": "1:41:53"} +{"current_steps": 2845, "total_steps": 3598, "loss": 0.1655, "lr": 5.117210896908702e-06, "epoch": 5.535540408958131, "percentage": 79.07, "elapsed_time": "6:22:27", "remaining_time": "1:41:13"} +{"current_steps": 2850, "total_steps": 3598, "loss": 0.158, "lr": 5.0525727999189865e-06, "epoch": 5.545277507302824, "percentage": 79.21, "elapsed_time": "6:23:06", "remaining_time": "1:40:32"} +{"current_steps": 2855, "total_steps": 3598, "loss": 0.1617, "lr": 4.988286467147416e-06, "epoch": 5.555014605647517, "percentage": 79.35, "elapsed_time": "6:23:44", "remaining_time": "1:39:51"} +{"current_steps": 2860, "total_steps": 3598, "loss": 0.1618, "lr": 4.924353411471834e-06, "epoch": 5.5647517039922105, "percentage": 79.49, "elapsed_time": "6:24:18", "remaining_time": "1:39:10"} +{"current_steps": 2865, "total_steps": 3598, "loss": 0.1668, "lr": 4.860775137456275e-06, "epoch": 5.574488802336903, "percentage": 79.63, "elapsed_time": "6:25:03", "remaining_time": "1:38:30"} +{"current_steps": 2870, "total_steps": 3598, "loss": 0.1642, "lr": 4.797553141315543e-06, "epoch": 5.584225900681597, "percentage": 79.77, "elapsed_time": "6:25:43", "remaining_time": "1:37:50"} +{"current_steps": 2875, "total_steps": 3598, "loss": 0.164, "lr": 4.734688910880001e-06, "epoch": 5.5939629990262905, "percentage": 79.91, "elapsed_time": "6:26:22", "remaining_time": "1:37:09"} +{"current_steps": 2880, "total_steps": 3598, "loss": 0.1572, "lr": 4.672183925560559e-06, "epoch": 5.603700097370983, "percentage": 80.04, "elapsed_time": "6:26:58", "remaining_time": "1:36:28"} +{"current_steps": 2885, "total_steps": 3598, "loss": 0.1563, "lr": 4.610039656313854e-06, "epoch": 5.613437195715677, "percentage": 80.18, "elapsed_time": "6:27:41", "remaining_time": "1:35:48"} +{"current_steps": 2890, "total_steps": 3598, "loss": 0.1541, "lr": 4.548257565607643e-06, "epoch": 5.62317429406037, "percentage": 80.32, "elapsed_time": "6:28:22", "remaining_time": "1:35:08"} +{"current_steps": 2895, "total_steps": 3598, "loss": 0.1533, "lr": 4.486839107386378e-06, "epoch": 5.632911392405063, "percentage": 80.46, "elapsed_time": "6:28:59", "remaining_time": "1:34:27"} +{"current_steps": 2900, "total_steps": 3598, "loss": 0.1683, "lr": 4.425785727036979e-06, "epoch": 5.642648490749757, "percentage": 80.6, "elapsed_time": "6:29:46", "remaining_time": "1:33:48"} +{"current_steps": 2905, "total_steps": 3598, "loss": 0.1696, "lr": 4.365098861354862e-06, "epoch": 5.65238558909445, "percentage": 80.74, "elapsed_time": "6:30:24", "remaining_time": "1:33:08"} +{"current_steps": 2910, "total_steps": 3598, "loss": 0.1768, "lr": 4.304779938510073e-06, "epoch": 5.662122687439143, "percentage": 80.88, "elapsed_time": "6:31:11", "remaining_time": "1:32:29"} +{"current_steps": 2915, "total_steps": 3598, "loss": 0.1629, "lr": 4.244830378013689e-06, "epoch": 5.671859785783837, "percentage": 81.02, "elapsed_time": "6:31:57", "remaining_time": "1:31:50"} +{"current_steps": 2920, "total_steps": 3598, "loss": 0.1516, "lr": 4.185251590684458e-06, "epoch": 5.681596884128529, "percentage": 81.16, "elapsed_time": "6:32:37", "remaining_time": "1:31:09"} +{"current_steps": 2925, "total_steps": 3598, "loss": 0.1523, "lr": 4.126044978615544e-06, "epoch": 5.691333982473223, "percentage": 81.3, "elapsed_time": "6:33:20", "remaining_time": "1:30:30"} +{"current_steps": 2930, "total_steps": 3598, "loss": 0.1552, "lr": 4.067211935141551e-06, "epoch": 5.701071080817917, "percentage": 81.43, "elapsed_time": "6:33:57", "remaining_time": "1:29:49"} +{"current_steps": 2935, "total_steps": 3598, "loss": 0.1727, "lr": 4.008753844805737e-06, "epoch": 5.710808179162609, "percentage": 81.57, "elapsed_time": "6:34:32", "remaining_time": "1:29:07"} +{"current_steps": 2940, "total_steps": 3598, "loss": 0.1658, "lr": 3.950672083327425e-06, "epoch": 5.720545277507303, "percentage": 81.71, "elapsed_time": "6:35:15", "remaining_time": "1:28:27"} +{"current_steps": 2945, "total_steps": 3598, "loss": 0.1479, "lr": 3.8929680175696315e-06, "epoch": 5.730282375851996, "percentage": 81.85, "elapsed_time": "6:35:53", "remaining_time": "1:27:46"} +{"current_steps": 2950, "total_steps": 3598, "loss": 0.1554, "lr": 3.835643005506893e-06, "epoch": 5.740019474196689, "percentage": 81.99, "elapsed_time": "6:36:34", "remaining_time": "1:27:06"} +{"current_steps": 2955, "total_steps": 3598, "loss": 0.1507, "lr": 3.7786983961933234e-06, "epoch": 5.749756572541383, "percentage": 82.13, "elapsed_time": "6:37:22", "remaining_time": "1:26:28"} +{"current_steps": 2960, "total_steps": 3598, "loss": 0.1645, "lr": 3.7221355297308483e-06, "epoch": 5.759493670886076, "percentage": 82.27, "elapsed_time": "6:37:57", "remaining_time": "1:25:46"} +{"current_steps": 2965, "total_steps": 3598, "loss": 0.176, "lr": 3.665955737237665e-06, "epoch": 5.769230769230769, "percentage": 82.41, "elapsed_time": "6:38:32", "remaining_time": "1:25:05"} +{"current_steps": 2970, "total_steps": 3598, "loss": 0.1526, "lr": 3.6101603408169373e-06, "epoch": 5.778967867575463, "percentage": 82.55, "elapsed_time": "6:39:11", "remaining_time": "1:24:24"} +{"current_steps": 2975, "total_steps": 3598, "loss": 0.1481, "lr": 3.5547506535256825e-06, "epoch": 5.7887049659201555, "percentage": 82.68, "elapsed_time": "6:39:58", "remaining_time": "1:23:45"} +{"current_steps": 2980, "total_steps": 3598, "loss": 0.16, "lr": 3.4997279793438387e-06, "epoch": 5.798442064264849, "percentage": 82.82, "elapsed_time": "6:40:25", "remaining_time": "1:23:02"} +{"current_steps": 2985, "total_steps": 3598, "loss": 0.1663, "lr": 3.445093613143611e-06, "epoch": 5.808179162609543, "percentage": 82.96, "elapsed_time": "6:41:06", "remaining_time": "1:22:22"} +{"current_steps": 2990, "total_steps": 3598, "loss": 0.1517, "lr": 3.390848840658978e-06, "epoch": 5.817916260954235, "percentage": 83.1, "elapsed_time": "6:41:55", "remaining_time": "1:21:43"} +{"current_steps": 2995, "total_steps": 3598, "loss": 0.1547, "lr": 3.3369949384554514e-06, "epoch": 5.827653359298929, "percentage": 83.24, "elapsed_time": "6:42:34", "remaining_time": "1:21:03"} +{"current_steps": 3000, "total_steps": 3598, "loss": 0.1445, "lr": 3.2835331739000178e-06, "epoch": 5.837390457643622, "percentage": 83.38, "elapsed_time": "6:43:21", "remaining_time": "1:20:24"} +{"current_steps": 3005, "total_steps": 3598, "loss": 0.1614, "lr": 3.2304648051313146e-06, "epoch": 5.847127555988315, "percentage": 83.52, "elapsed_time": "6:44:57", "remaining_time": "1:19:54"} +{"current_steps": 3010, "total_steps": 3598, "loss": 0.1567, "lr": 3.177791081030057e-06, "epoch": 5.856864654333009, "percentage": 83.66, "elapsed_time": "6:45:32", "remaining_time": "1:19:13"} +{"current_steps": 3015, "total_steps": 3598, "loss": 0.1591, "lr": 3.125513241189575e-06, "epoch": 5.866601752677702, "percentage": 83.8, "elapsed_time": "6:46:00", "remaining_time": "1:18:30"} +{"current_steps": 3020, "total_steps": 3598, "loss": 0.1526, "lr": 3.073632515886711e-06, "epoch": 5.876338851022395, "percentage": 83.94, "elapsed_time": "6:46:49", "remaining_time": "1:17:51"} +{"current_steps": 3025, "total_steps": 3598, "loss": 0.1519, "lr": 3.0221501260528406e-06, "epoch": 5.886075949367089, "percentage": 84.07, "elapsed_time": "6:47:33", "remaining_time": "1:17:11"} +{"current_steps": 3030, "total_steps": 3598, "loss": 0.1532, "lr": 2.9710672832451303e-06, "epoch": 5.895813047711782, "percentage": 84.21, "elapsed_time": "6:48:20", "remaining_time": "1:16:32"} +{"current_steps": 3035, "total_steps": 3598, "loss": 0.1693, "lr": 2.9203851896180422e-06, "epoch": 5.905550146056475, "percentage": 84.35, "elapsed_time": "6:48:55", "remaining_time": "1:15:51"} +{"current_steps": 3040, "total_steps": 3598, "loss": 0.1614, "lr": 2.8701050378950303e-06, "epoch": 5.915287244401169, "percentage": 84.49, "elapsed_time": "6:49:28", "remaining_time": "1:15:09"} +{"current_steps": 3045, "total_steps": 3598, "loss": 0.1605, "lr": 2.820228011340478e-06, "epoch": 5.9250243427458615, "percentage": 84.63, "elapsed_time": "6:50:01", "remaining_time": "1:14:27"} +{"current_steps": 3050, "total_steps": 3598, "loss": 0.1589, "lr": 2.7707552837318584e-06, "epoch": 5.934761441090555, "percentage": 84.77, "elapsed_time": "6:50:48", "remaining_time": "1:13:48"} +{"current_steps": 3055, "total_steps": 3598, "loss": 0.1628, "lr": 2.7216880193320915e-06, "epoch": 5.944498539435248, "percentage": 84.91, "elapsed_time": "6:51:32", "remaining_time": "1:13:08"} +{"current_steps": 3060, "total_steps": 3598, "loss": 0.1666, "lr": 2.673027372862178e-06, "epoch": 5.9542356377799415, "percentage": 85.05, "elapsed_time": "6:52:10", "remaining_time": "1:12:27"} +{"current_steps": 3065, "total_steps": 3598, "loss": 0.1535, "lr": 2.624774489473978e-06, "epoch": 5.963972736124635, "percentage": 85.19, "elapsed_time": "6:52:51", "remaining_time": "1:11:47"} +{"current_steps": 3070, "total_steps": 3598, "loss": 0.1594, "lr": 2.5769305047233004e-06, "epoch": 5.973709834469328, "percentage": 85.33, "elapsed_time": "6:53:42", "remaining_time": "1:11:09"} +{"current_steps": 3075, "total_steps": 3598, "loss": 0.1599, "lr": 2.5294965445431775e-06, "epoch": 5.983446932814021, "percentage": 85.46, "elapsed_time": "6:54:19", "remaining_time": "1:10:28"} +{"current_steps": 3080, "total_steps": 3598, "loss": 0.1573, "lr": 2.4824737252173447e-06, "epoch": 5.993184031158715, "percentage": 85.6, "elapsed_time": "6:54:58", "remaining_time": "1:09:47"} +{"current_steps": 3085, "total_steps": 3598, "loss": 0.1501, "lr": 2.4358631533539902e-06, "epoch": 6.001947419668939, "percentage": 85.74, "elapsed_time": "6:55:31", "remaining_time": "1:09:05"} +{"current_steps": 3090, "total_steps": 3598, "loss": 0.1446, "lr": 2.3896659258596877e-06, "epoch": 6.011684518013632, "percentage": 85.88, "elapsed_time": "6:56:16", "remaining_time": "1:08:26"} +{"current_steps": 3095, "total_steps": 3598, "loss": 0.1506, "lr": 2.3438831299136265e-06, "epoch": 6.021421616358325, "percentage": 86.02, "elapsed_time": "6:56:55", "remaining_time": "1:07:45"} +{"current_steps": 3100, "total_steps": 3598, "loss": 0.15, "lr": 2.2985158429419885e-06, "epoch": 6.031158714703018, "percentage": 86.16, "elapsed_time": "6:57:25", "remaining_time": "1:07:03"} +{"current_steps": 3105, "total_steps": 3598, "loss": 0.1479, "lr": 2.2535651325926012e-06, "epoch": 6.040895813047712, "percentage": 86.3, "elapsed_time": "6:58:10", "remaining_time": "1:06:23"} +{"current_steps": 3110, "total_steps": 3598, "loss": 0.1459, "lr": 2.209032056709821e-06, "epoch": 6.050632911392405, "percentage": 86.44, "elapsed_time": "6:58:54", "remaining_time": "1:05:43"} +{"current_steps": 3115, "total_steps": 3598, "loss": 0.1688, "lr": 2.1649176633096312e-06, "epoch": 6.060370009737098, "percentage": 86.58, "elapsed_time": "6:59:27", "remaining_time": "1:05:02"} +{"current_steps": 3120, "total_steps": 3598, "loss": 0.1532, "lr": 2.121222990554981e-06, "epoch": 6.070107108081792, "percentage": 86.71, "elapsed_time": "7:00:07", "remaining_time": "1:04:21"} +{"current_steps": 3125, "total_steps": 3598, "loss": 0.1558, "lr": 2.0779490667313485e-06, "epoch": 6.079844206426485, "percentage": 86.85, "elapsed_time": "7:00:42", "remaining_time": "1:03:40"} +{"current_steps": 3130, "total_steps": 3598, "loss": 0.1494, "lr": 2.0350969102225603e-06, "epoch": 6.089581304771178, "percentage": 86.99, "elapsed_time": "7:01:21", "remaining_time": "1:03:00"} +{"current_steps": 3135, "total_steps": 3598, "loss": 0.1524, "lr": 1.9926675294868024e-06, "epoch": 6.099318403115872, "percentage": 87.13, "elapsed_time": "7:02:09", "remaining_time": "1:02:20"} +{"current_steps": 3140, "total_steps": 3598, "loss": 0.1457, "lr": 1.9506619230328884e-06, "epoch": 6.109055501460565, "percentage": 87.27, "elapsed_time": "7:02:52", "remaining_time": "1:01:40"} +{"current_steps": 3145, "total_steps": 3598, "loss": 0.1451, "lr": 1.9090810793967885e-06, "epoch": 6.118792599805258, "percentage": 87.41, "elapsed_time": "7:03:31", "remaining_time": "1:01:00"} +{"current_steps": 3150, "total_steps": 3598, "loss": 0.1555, "lr": 1.8679259771183322e-06, "epoch": 6.1285296981499515, "percentage": 87.55, "elapsed_time": "7:04:05", "remaining_time": "1:00:18"} +{"current_steps": 3155, "total_steps": 3598, "loss": 0.1507, "lr": 1.8271975847182056e-06, "epoch": 6.138266796494644, "percentage": 87.69, "elapsed_time": "7:04:47", "remaining_time": "0:59:38"} +{"current_steps": 3160, "total_steps": 3598, "loss": 0.1436, "lr": 1.7868968606751425e-06, "epoch": 6.148003894839338, "percentage": 87.83, "elapsed_time": "7:05:28", "remaining_time": "0:58:58"} +{"current_steps": 3165, "total_steps": 3598, "loss": 0.1445, "lr": 1.7470247534033703e-06, "epoch": 6.1577409931840315, "percentage": 87.97, "elapsed_time": "7:06:16", "remaining_time": "0:58:19"} +{"current_steps": 3170, "total_steps": 3598, "loss": 0.1661, "lr": 1.7075822012303022e-06, "epoch": 6.167478091528724, "percentage": 88.1, "elapsed_time": "7:06:56", "remaining_time": "0:57:38"} +{"current_steps": 3175, "total_steps": 3598, "loss": 0.1512, "lr": 1.6685701323744386e-06, "epoch": 6.177215189873418, "percentage": 88.24, "elapsed_time": "7:07:34", "remaining_time": "0:56:57"} +{"current_steps": 3180, "total_steps": 3598, "loss": 0.1548, "lr": 1.6299894649235448e-06, "epoch": 6.186952288218111, "percentage": 88.38, "elapsed_time": "7:08:19", "remaining_time": "0:56:18"} +{"current_steps": 3185, "total_steps": 3598, "loss": 0.1457, "lr": 1.591841106813017e-06, "epoch": 6.196689386562804, "percentage": 88.52, "elapsed_time": "7:08:52", "remaining_time": "0:55:36"} +{"current_steps": 3190, "total_steps": 3598, "loss": 0.1446, "lr": 1.5541259558045308e-06, "epoch": 6.206426484907498, "percentage": 88.66, "elapsed_time": "7:09:34", "remaining_time": "0:54:56"} +{"current_steps": 3195, "total_steps": 3598, "loss": 0.1526, "lr": 1.5168448994649266e-06, "epoch": 6.21616358325219, "percentage": 88.8, "elapsed_time": "7:10:12", "remaining_time": "0:54:15"} +{"current_steps": 3200, "total_steps": 3598, "loss": 0.1609, "lr": 1.479998815145296e-06, "epoch": 6.225900681596884, "percentage": 88.94, "elapsed_time": "7:10:39", "remaining_time": "0:53:33"} +{"current_steps": 3205, "total_steps": 3598, "loss": 0.1498, "lr": 1.4435885699603524e-06, "epoch": 6.235637779941578, "percentage": 89.08, "elapsed_time": "7:11:17", "remaining_time": "0:52:53"} +{"current_steps": 3210, "total_steps": 3598, "loss": 0.1466, "lr": 1.4076150207680185e-06, "epoch": 6.24537487828627, "percentage": 89.22, "elapsed_time": "7:11:55", "remaining_time": "0:52:12"} +{"current_steps": 3215, "total_steps": 3598, "loss": 0.1455, "lr": 1.3720790141492657e-06, "epoch": 6.255111976630964, "percentage": 89.36, "elapsed_time": "7:12:39", "remaining_time": "0:51:32"} +{"current_steps": 3220, "total_steps": 3598, "loss": 0.141, "lr": 1.336981386388183e-06, "epoch": 6.264849074975658, "percentage": 89.49, "elapsed_time": "7:13:18", "remaining_time": "0:50:51"} +{"current_steps": 3225, "total_steps": 3598, "loss": 0.1516, "lr": 1.3023229634523027e-06, "epoch": 6.27458617332035, "percentage": 89.63, "elapsed_time": "7:13:54", "remaining_time": "0:50:11"} +{"current_steps": 3230, "total_steps": 3598, "loss": 0.1476, "lr": 1.2681045609731756e-06, "epoch": 6.284323271665044, "percentage": 89.77, "elapsed_time": "7:14:35", "remaining_time": "0:49:30"} +{"current_steps": 3235, "total_steps": 3598, "loss": 0.143, "lr": 1.2343269842271366e-06, "epoch": 6.2940603700097375, "percentage": 89.91, "elapsed_time": "7:15:15", "remaining_time": "0:48:50"} +{"current_steps": 3240, "total_steps": 3598, "loss": 0.1602, "lr": 1.2009910281163962e-06, "epoch": 6.30379746835443, "percentage": 90.05, "elapsed_time": "7:15:55", "remaining_time": "0:48:10"} +{"current_steps": 3245, "total_steps": 3598, "loss": 0.1447, "lr": 1.168097477150314e-06, "epoch": 6.313534566699124, "percentage": 90.19, "elapsed_time": "7:16:29", "remaining_time": "0:47:28"} +{"current_steps": 3250, "total_steps": 3598, "loss": 0.147, "lr": 1.1356471054269447e-06, "epoch": 6.3232716650438165, "percentage": 90.33, "elapsed_time": "7:17:16", "remaining_time": "0:46:49"} +{"current_steps": 3255, "total_steps": 3598, "loss": 0.1452, "lr": 1.1036406766148033e-06, "epoch": 6.33300876338851, "percentage": 90.47, "elapsed_time": "7:17:57", "remaining_time": "0:46:09"} +{"current_steps": 3260, "total_steps": 3598, "loss": 0.1418, "lr": 1.0720789439349177e-06, "epoch": 6.342745861733204, "percentage": 90.61, "elapsed_time": "7:18:41", "remaining_time": "0:45:29"} +{"current_steps": 3265, "total_steps": 3598, "loss": 0.1502, "lr": 1.0409626501430847e-06, "epoch": 6.3524829600778965, "percentage": 90.74, "elapsed_time": "7:19:11", "remaining_time": "0:44:47"} +{"current_steps": 3270, "total_steps": 3598, "loss": 0.1411, "lr": 1.0102925275123954e-06, "epoch": 6.36222005842259, "percentage": 90.88, "elapsed_time": "7:19:47", "remaining_time": "0:44:06"} +{"current_steps": 3275, "total_steps": 3598, "loss": 0.1528, "lr": 9.800692978160064e-07, "epoch": 6.371957156767284, "percentage": 91.02, "elapsed_time": "7:20:20", "remaining_time": "0:43:25"} +{"current_steps": 3280, "total_steps": 3598, "loss": 0.1591, "lr": 9.502936723101563e-07, "epoch": 6.381694255111976, "percentage": 91.16, "elapsed_time": "7:21:03", "remaining_time": "0:42:45"} +{"current_steps": 3285, "total_steps": 3598, "loss": 0.1399, "lr": 9.209663517174094e-07, "epoch": 6.39143135345667, "percentage": 91.3, "elapsed_time": "7:21:47", "remaining_time": "0:42:05"} +{"current_steps": 3290, "total_steps": 3598, "loss": 0.154, "lr": 8.920880262101894e-07, "epoch": 6.401168451801363, "percentage": 91.44, "elapsed_time": "7:22:29", "remaining_time": "0:41:25"} +{"current_steps": 3295, "total_steps": 3598, "loss": 0.1563, "lr": 8.636593753945188e-07, "epoch": 6.410905550146056, "percentage": 91.58, "elapsed_time": "7:23:04", "remaining_time": "0:40:44"} +{"current_steps": 3300, "total_steps": 3598, "loss": 0.1575, "lr": 8.356810682940386e-07, "epoch": 6.42064264849075, "percentage": 91.72, "elapsed_time": "7:23:49", "remaining_time": "0:40:04"} +{"current_steps": 3305, "total_steps": 3598, "loss": 0.1518, "lr": 8.081537633342562e-07, "epoch": 6.430379746835443, "percentage": 91.86, "elapsed_time": "7:24:34", "remaining_time": "0:39:24"} +{"current_steps": 3310, "total_steps": 3598, "loss": 0.1487, "lr": 7.810781083270491e-07, "epoch": 6.440116845180136, "percentage": 92.0, "elapsed_time": "7:25:17", "remaining_time": "0:38:44"} +{"current_steps": 3315, "total_steps": 3598, "loss": 0.1542, "lr": 7.544547404554236e-07, "epoch": 6.44985394352483, "percentage": 92.13, "elapsed_time": "7:25:55", "remaining_time": "0:38:04"} +{"current_steps": 3320, "total_steps": 3598, "loss": 0.1519, "lr": 7.282842862585204e-07, "epoch": 6.459591041869523, "percentage": 92.27, "elapsed_time": "7:26:36", "remaining_time": "0:37:23"} +{"current_steps": 3325, "total_steps": 3598, "loss": 0.1493, "lr": 7.025673616168704e-07, "epoch": 6.469328140214216, "percentage": 92.41, "elapsed_time": "7:27:06", "remaining_time": "0:36:42"} +{"current_steps": 3330, "total_steps": 3598, "loss": 0.1613, "lr": 6.773045717378912e-07, "epoch": 6.47906523855891, "percentage": 92.55, "elapsed_time": "7:27:45", "remaining_time": "0:36:02"} +{"current_steps": 3335, "total_steps": 3598, "loss": 0.1466, "lr": 6.524965111416604e-07, "epoch": 6.4888023369036025, "percentage": 92.69, "elapsed_time": "7:28:11", "remaining_time": "0:35:20"} +{"current_steps": 3340, "total_steps": 3598, "loss": 0.1598, "lr": 6.281437636469135e-07, "epoch": 6.498539435248296, "percentage": 92.83, "elapsed_time": "7:28:56", "remaining_time": "0:34:40"} +{"current_steps": 3345, "total_steps": 3598, "loss": 0.1437, "lr": 6.042469023573016e-07, "epoch": 6.508276533592989, "percentage": 92.97, "elapsed_time": "7:29:24", "remaining_time": "0:33:59"} +{"current_steps": 3350, "total_steps": 3598, "loss": 0.1531, "lr": 5.808064896479227e-07, "epoch": 6.5180136319376825, "percentage": 93.11, "elapsed_time": "7:29:51", "remaining_time": "0:33:18"} +{"current_steps": 3355, "total_steps": 3598, "loss": 0.1486, "lr": 5.578230771520665e-07, "epoch": 6.527750730282376, "percentage": 93.25, "elapsed_time": "7:30:28", "remaining_time": "0:32:37"} +{"current_steps": 3360, "total_steps": 3598, "loss": 0.1419, "lr": 5.352972057482375e-07, "epoch": 6.537487828627069, "percentage": 93.39, "elapsed_time": "7:31:03", "remaining_time": "0:31:57"} +{"current_steps": 3365, "total_steps": 3598, "loss": 0.1358, "lr": 5.132294055474307e-07, "epoch": 6.547224926971762, "percentage": 93.52, "elapsed_time": "7:31:42", "remaining_time": "0:31:16"} +{"current_steps": 3370, "total_steps": 3598, "loss": 0.1354, "lr": 4.916201958806621e-07, "epoch": 6.556962025316456, "percentage": 93.66, "elapsed_time": "7:32:25", "remaining_time": "0:30:36"} +{"current_steps": 3375, "total_steps": 3598, "loss": 0.1494, "lr": 4.704700852867361e-07, "epoch": 6.566699123661149, "percentage": 93.8, "elapsed_time": "7:33:04", "remaining_time": "0:29:56"} +{"current_steps": 3380, "total_steps": 3598, "loss": 0.1508, "lr": 4.497795715002795e-07, "epoch": 6.576436222005842, "percentage": 93.94, "elapsed_time": "7:33:44", "remaining_time": "0:29:15"} +{"current_steps": 3385, "total_steps": 3598, "loss": 0.1468, "lr": 4.295491414400288e-07, "epoch": 6.586173320350536, "percentage": 94.08, "elapsed_time": "7:34:19", "remaining_time": "0:28:35"} +{"current_steps": 3390, "total_steps": 3598, "loss": 0.1451, "lr": 4.0977927119737695e-07, "epoch": 6.595910418695229, "percentage": 94.22, "elapsed_time": "7:34:54", "remaining_time": "0:27:54"} +{"current_steps": 3395, "total_steps": 3598, "loss": 0.153, "lr": 3.9047042602516237e-07, "epoch": 6.605647517039922, "percentage": 94.36, "elapsed_time": "7:35:39", "remaining_time": "0:27:14"} +{"current_steps": 3400, "total_steps": 3598, "loss": 0.1397, "lr": 3.716230603267268e-07, "epoch": 6.615384615384615, "percentage": 94.5, "elapsed_time": "7:36:13", "remaining_time": "0:26:34"} +{"current_steps": 3405, "total_steps": 3598, "loss": 0.1611, "lr": 3.532376176452146e-07, "epoch": 6.625121713729309, "percentage": 94.64, "elapsed_time": "7:36:59", "remaining_time": "0:25:54"} +{"current_steps": 3410, "total_steps": 3598, "loss": 0.1452, "lr": 3.353145306531347e-07, "epoch": 6.634858812074002, "percentage": 94.77, "elapsed_time": "7:37:37", "remaining_time": "0:25:13"} +{"current_steps": 3415, "total_steps": 3598, "loss": 0.1577, "lr": 3.17854221142182e-07, "epoch": 6.644595910418695, "percentage": 94.91, "elapsed_time": "7:38:23", "remaining_time": "0:24:33"} +{"current_steps": 3420, "total_steps": 3598, "loss": 0.1526, "lr": 3.008571000133187e-07, "epoch": 6.6543330087633885, "percentage": 95.05, "elapsed_time": "7:39:11", "remaining_time": "0:23:53"} +{"current_steps": 3425, "total_steps": 3598, "loss": 0.1619, "lr": 2.843235672670841e-07, "epoch": 6.664070107108082, "percentage": 95.19, "elapsed_time": "7:39:53", "remaining_time": "0:23:13"} +{"current_steps": 3430, "total_steps": 3598, "loss": 0.148, "lr": 2.682540119942023e-07, "epoch": 6.673807205452775, "percentage": 95.33, "elapsed_time": "7:40:33", "remaining_time": "0:22:33"} +{"current_steps": 3435, "total_steps": 3598, "loss": 0.1451, "lr": 2.5264881236640947e-07, "epoch": 6.6835443037974684, "percentage": 95.47, "elapsed_time": "7:41:16", "remaining_time": "0:21:53"} +{"current_steps": 3440, "total_steps": 3598, "loss": 0.1542, "lr": 2.375083356275676e-07, "epoch": 6.693281402142162, "percentage": 95.61, "elapsed_time": "7:41:56", "remaining_time": "0:21:13"} +{"current_steps": 3445, "total_steps": 3598, "loss": 0.1493, "lr": 2.2283293808501138e-07, "epoch": 6.703018500486855, "percentage": 95.75, "elapsed_time": "7:42:35", "remaining_time": "0:20:32"} +{"current_steps": 3450, "total_steps": 3598, "loss": 0.1605, "lr": 2.086229651011684e-07, "epoch": 6.712755598831548, "percentage": 95.89, "elapsed_time": "7:43:17", "remaining_time": "0:19:52"} +{"current_steps": 3455, "total_steps": 3598, "loss": 0.145, "lr": 1.9487875108542997e-07, "epoch": 6.722492697176241, "percentage": 96.03, "elapsed_time": "7:43:56", "remaining_time": "0:19:12"} +{"current_steps": 3460, "total_steps": 3598, "loss": 0.1497, "lr": 1.8160061948628406e-07, "epoch": 6.732229795520935, "percentage": 96.16, "elapsed_time": "7:44:42", "remaining_time": "0:18:32"} +{"current_steps": 3465, "total_steps": 3598, "loss": 0.1492, "lr": 1.6878888278369255e-07, "epoch": 6.741966893865628, "percentage": 96.3, "elapsed_time": "7:45:26", "remaining_time": "0:17:51"} +{"current_steps": 3470, "total_steps": 3598, "loss": 0.1472, "lr": 1.5644384248176159e-07, "epoch": 6.751703992210321, "percentage": 96.44, "elapsed_time": "7:46:09", "remaining_time": "0:17:11"} +{"current_steps": 3475, "total_steps": 3598, "loss": 0.1511, "lr": 1.445657891016161e-07, "epoch": 6.761441090555015, "percentage": 96.58, "elapsed_time": "7:46:51", "remaining_time": "0:16:31"} +{"current_steps": 3480, "total_steps": 3598, "loss": 0.1486, "lr": 1.3315500217458533e-07, "epoch": 6.771178188899708, "percentage": 96.72, "elapsed_time": "7:47:32", "remaining_time": "0:15:51"} +{"current_steps": 3485, "total_steps": 3598, "loss": 0.1483, "lr": 1.2221175023561238e-07, "epoch": 6.780915287244401, "percentage": 96.86, "elapsed_time": "7:48:12", "remaining_time": "0:15:10"} +{"current_steps": 3490, "total_steps": 3598, "loss": 0.1412, "lr": 1.1173629081694392e-07, "epoch": 6.790652385589095, "percentage": 97.0, "elapsed_time": "7:48:56", "remaining_time": "0:14:30"} +{"current_steps": 3495, "total_steps": 3598, "loss": 0.1635, "lr": 1.0172887044205936e-07, "epoch": 6.800389483933788, "percentage": 97.14, "elapsed_time": "7:49:33", "remaining_time": "0:13:50"} +{"current_steps": 3500, "total_steps": 3598, "loss": 0.1495, "lr": 9.21897246198733e-08, "epoch": 6.810126582278481, "percentage": 97.28, "elapsed_time": "7:50:14", "remaining_time": "0:13:10"} +{"current_steps": 3505, "total_steps": 3598, "loss": 0.1501, "lr": 8.311907783920437e-08, "epoch": 6.8198636806231745, "percentage": 97.42, "elapsed_time": "7:50:57", "remaining_time": "0:12:29"} +{"current_steps": 3510, "total_steps": 3598, "loss": 0.1541, "lr": 7.451714356347062e-08, "epoch": 6.829600778967867, "percentage": 97.55, "elapsed_time": "7:51:41", "remaining_time": "0:11:49"} +{"current_steps": 3515, "total_steps": 3598, "loss": 0.1445, "lr": 6.638412422568463e-08, "epoch": 6.839337877312561, "percentage": 97.69, "elapsed_time": "7:52:16", "remaining_time": "0:11:09"} +{"current_steps": 3520, "total_steps": 3598, "loss": 0.1609, "lr": 5.8720211223677324e-08, "epoch": 6.849074975657254, "percentage": 97.83, "elapsed_time": "7:52:47", "remaining_time": "0:10:28"} +{"current_steps": 3525, "total_steps": 3598, "loss": 0.1438, "lr": 5.1525584915605996e-08, "epoch": 6.858812074001947, "percentage": 97.97, "elapsed_time": "7:53:27", "remaining_time": "0:09:48"} +{"current_steps": 3530, "total_steps": 3598, "loss": 0.1444, "lr": 4.480041461569773e-08, "epoch": 6.868549172346641, "percentage": 98.11, "elapsed_time": "7:54:08", "remaining_time": "0:09:08"} +{"current_steps": 3535, "total_steps": 3598, "loss": 0.1439, "lr": 3.854485859026591e-08, "epoch": 6.878286270691334, "percentage": 98.25, "elapsed_time": "7:54:50", "remaining_time": "0:08:27"} +{"current_steps": 3540, "total_steps": 3598, "loss": 0.1471, "lr": 3.2759064054002086e-08, "epoch": 6.888023369036027, "percentage": 98.39, "elapsed_time": "7:55:37", "remaining_time": "0:07:47"} +{"current_steps": 3545, "total_steps": 3598, "loss": 0.1493, "lr": 2.744316716648543e-08, "epoch": 6.897760467380721, "percentage": 98.53, "elapsed_time": "7:56:17", "remaining_time": "0:07:07"} +{"current_steps": 3550, "total_steps": 3598, "loss": 0.1481, "lr": 2.259729302900304e-08, "epoch": 6.907497565725414, "percentage": 98.67, "elapsed_time": "7:56:50", "remaining_time": "0:06:26"} +{"current_steps": 3555, "total_steps": 3598, "loss": 0.1542, "lr": 1.822155568159012e-08, "epoch": 6.917234664070107, "percentage": 98.8, "elapsed_time": "7:57:26", "remaining_time": "0:05:46"} +{"current_steps": 3560, "total_steps": 3598, "loss": 0.1528, "lr": 1.431605810034764e-08, "epoch": 6.926971762414801, "percentage": 98.94, "elapsed_time": "7:58:03", "remaining_time": "0:05:06"} +{"current_steps": 3565, "total_steps": 3598, "loss": 0.135, "lr": 1.0880892195028747e-08, "epoch": 6.936708860759493, "percentage": 99.08, "elapsed_time": "7:58:46", "remaining_time": "0:04:25"} +{"current_steps": 3570, "total_steps": 3598, "loss": 0.1487, "lr": 7.916138806864927e-09, "epoch": 6.946445959104187, "percentage": 99.22, "elapsed_time": "7:59:32", "remaining_time": "0:03:45"} +{"current_steps": 3575, "total_steps": 3598, "loss": 0.1454, "lr": 5.421867706671968e-09, "epoch": 6.9561830574488805, "percentage": 99.36, "elapsed_time": "8:00:05", "remaining_time": "0:03:05"} +{"current_steps": 3580, "total_steps": 3598, "loss": 0.157, "lr": 3.3981375932001792e-09, "epoch": 6.965920155793573, "percentage": 99.5, "elapsed_time": "8:00:36", "remaining_time": "0:02:24"} +{"current_steps": 3585, "total_steps": 3598, "loss": 0.1479, "lr": 1.8449960917599207e-09, "epoch": 6.975657254138267, "percentage": 99.64, "elapsed_time": "8:01:15", "remaining_time": "0:01:44"} +{"current_steps": 3590, "total_steps": 3598, "loss": 0.1611, "lr": 7.624797530958461e-10, "epoch": 6.98539435248296, "percentage": 99.78, "elapsed_time": "8:01:50", "remaining_time": "0:01:04"} +{"current_steps": 3595, "total_steps": 3598, "loss": 0.1441, "lr": 1.5061405252980721e-10, "epoch": 6.995131450827653, "percentage": 99.92, "elapsed_time": "8:02:27", "remaining_time": "0:00:24"} +{"current_steps": 3598, "total_steps": 3598, "epoch": 7.0, "percentage": 100.0, "elapsed_time": "8:03:36", "remaining_time": "0:00:00"} diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000..809e2ad --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,7956 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.0, + "eval_steps": 500, + "global_step": 3598, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.009737098344693282, + "grad_norm": 6.659023843744882, + "learning_rate": 4.444444444444445e-07, + "loss": 0.708, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.3341186046600342, + "step": 5, + "valid_targets_mean": 3131.6, + "valid_targets_min": 1131 + }, + { + "epoch": 0.019474196689386564, + "grad_norm": 7.096902024555233, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.7467, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.3632035255432129, + "step": 10, + "valid_targets_mean": 3640.4, + "valid_targets_min": 1212 + }, + { + "epoch": 0.029211295034079845, + "grad_norm": 8.390948057011439, + "learning_rate": 1.5555555555555558e-06, + "loss": 0.723, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.3690454959869385, + "step": 15, + "valid_targets_mean": 2841.9, + "valid_targets_min": 1460 + }, + { + "epoch": 0.03894839337877313, + "grad_norm": 4.941585314582268, + "learning_rate": 2.1111111111111114e-06, + "loss": 0.7416, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.34727782011032104, + "step": 20, + "valid_targets_mean": 2777.1, + "valid_targets_min": 816 + }, + { + "epoch": 0.04868549172346641, + "grad_norm": 3.910822741404409, + "learning_rate": 2.666666666666667e-06, + "loss": 0.7126, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.3296957314014435, + "step": 25, + "valid_targets_mean": 2570.0, + "valid_targets_min": 819 + }, + { + "epoch": 0.05842259006815969, + "grad_norm": 3.6743426354854005, + "learning_rate": 3.2222222222222227e-06, + "loss": 0.7041, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.30102282762527466, + "step": 30, + "valid_targets_mean": 1553.8, + "valid_targets_min": 919 + }, + { + "epoch": 0.06815968841285297, + "grad_norm": 2.6901440323068724, + "learning_rate": 3.777777777777778e-06, + "loss": 0.6225, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.3363445997238159, + "step": 35, + "valid_targets_mean": 2740.1, + "valid_targets_min": 956 + }, + { + "epoch": 0.07789678675754626, + "grad_norm": 1.5292949966980411, + "learning_rate": 4.333333333333334e-06, + "loss": 0.6163, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2927810847759247, + "step": 40, + "valid_targets_mean": 3039.4, + "valid_targets_min": 1511 + }, + { + "epoch": 0.08763388510223953, + "grad_norm": 1.3434308671523765, + "learning_rate": 4.888888888888889e-06, + "loss": 0.5893, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.3337903916835785, + "step": 45, + "valid_targets_mean": 3930.4, + "valid_targets_min": 1675 + }, + { + "epoch": 0.09737098344693282, + "grad_norm": 0.9373741509167212, + "learning_rate": 5.444444444444445e-06, + "loss": 0.5577, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18909615278244019, + "step": 50, + "valid_targets_mean": 3237.6, + "valid_targets_min": 1383 + }, + { + "epoch": 0.10710808179162609, + "grad_norm": 1.3352877614003678, + "learning_rate": 6e-06, + "loss": 0.5395, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2800602316856384, + "step": 55, + "valid_targets_mean": 1897.5, + "valid_targets_min": 989 + }, + { + "epoch": 0.11684518013631938, + "grad_norm": 0.7395266220632808, + "learning_rate": 6.555555555555556e-06, + "loss": 0.492, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2359597086906433, + "step": 60, + "valid_targets_mean": 4924.5, + "valid_targets_min": 880 + }, + { + "epoch": 0.12658227848101267, + "grad_norm": 0.9102371594210115, + "learning_rate": 7.111111111111112e-06, + "loss": 0.5166, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2459840178489685, + "step": 65, + "valid_targets_mean": 2506.4, + "valid_targets_min": 590 + }, + { + "epoch": 0.13631937682570594, + "grad_norm": 0.8136377348280819, + "learning_rate": 7.666666666666667e-06, + "loss": 0.4982, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.24337783455848694, + "step": 70, + "valid_targets_mean": 3078.9, + "valid_targets_min": 1886 + }, + { + "epoch": 0.1460564751703992, + "grad_norm": 0.989281570628736, + "learning_rate": 8.222222222222222e-06, + "loss": 0.5096, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2819697856903076, + "step": 75, + "valid_targets_mean": 2441.6, + "valid_targets_min": 1560 + }, + { + "epoch": 0.15579357351509251, + "grad_norm": 0.9036044445773288, + "learning_rate": 8.777777777777778e-06, + "loss": 0.4988, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.25402408838272095, + "step": 80, + "valid_targets_mean": 3471.0, + "valid_targets_min": 1593 + }, + { + "epoch": 0.1655306718597858, + "grad_norm": 0.7988252984615115, + "learning_rate": 9.333333333333334e-06, + "loss": 0.4763, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23960669338703156, + "step": 85, + "valid_targets_mean": 2990.8, + "valid_targets_min": 609 + }, + { + "epoch": 0.17526777020447906, + "grad_norm": 0.7571243365246795, + "learning_rate": 9.88888888888889e-06, + "loss": 0.4758, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.28586897253990173, + "step": 90, + "valid_targets_mean": 3950.6, + "valid_targets_min": 843 + }, + { + "epoch": 0.18500486854917234, + "grad_norm": 0.7339048901803402, + "learning_rate": 1.0444444444444445e-05, + "loss": 0.4865, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2568185329437256, + "step": 95, + "valid_targets_mean": 3648.8, + "valid_targets_min": 2224 + }, + { + "epoch": 0.19474196689386564, + "grad_norm": 0.7948184895802777, + "learning_rate": 1.1000000000000001e-05, + "loss": 0.4653, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.25145047903060913, + "step": 100, + "valid_targets_mean": 3199.6, + "valid_targets_min": 1424 + }, + { + "epoch": 0.2044790652385589, + "grad_norm": 0.7168429039256338, + "learning_rate": 1.1555555555555556e-05, + "loss": 0.4581, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.26357921957969666, + "step": 105, + "valid_targets_mean": 3663.4, + "valid_targets_min": 1249 + }, + { + "epoch": 0.21421616358325218, + "grad_norm": 0.7547313993147083, + "learning_rate": 1.211111111111111e-05, + "loss": 0.4502, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23485320806503296, + "step": 110, + "valid_targets_mean": 4071.5, + "valid_targets_min": 328 + }, + { + "epoch": 0.22395326192794549, + "grad_norm": 0.5836593751970055, + "learning_rate": 1.2666666666666667e-05, + "loss": 0.4342, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17011061310768127, + "step": 115, + "valid_targets_mean": 3115.8, + "valid_targets_min": 1423 + }, + { + "epoch": 0.23369036027263876, + "grad_norm": 0.7449951697430623, + "learning_rate": 1.3222222222222223e-05, + "loss": 0.451, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.21737040579319, + "step": 120, + "valid_targets_mean": 2913.0, + "valid_targets_min": 662 + }, + { + "epoch": 0.24342745861733203, + "grad_norm": 0.8242919386330488, + "learning_rate": 1.377777777777778e-05, + "loss": 0.4221, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2685893177986145, + "step": 125, + "valid_targets_mean": 3352.2, + "valid_targets_min": 1471 + }, + { + "epoch": 0.25316455696202533, + "grad_norm": 0.6620888315425654, + "learning_rate": 1.4333333333333334e-05, + "loss": 0.4111, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2265009582042694, + "step": 130, + "valid_targets_mean": 3928.9, + "valid_targets_min": 1108 + }, + { + "epoch": 0.2629016553067186, + "grad_norm": 0.7782866385397257, + "learning_rate": 1.488888888888889e-05, + "loss": 0.4129, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19729581475257874, + "step": 135, + "valid_targets_mean": 2165.6, + "valid_targets_min": 952 + }, + { + "epoch": 0.2726387536514119, + "grad_norm": 0.9252799291147483, + "learning_rate": 1.5444444444444446e-05, + "loss": 0.4197, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.22347350418567657, + "step": 140, + "valid_targets_mean": 1940.4, + "valid_targets_min": 949 + }, + { + "epoch": 0.28237585199610515, + "grad_norm": 0.7315249415292264, + "learning_rate": 1.6000000000000003e-05, + "loss": 0.3976, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2102898508310318, + "step": 145, + "valid_targets_mean": 3084.5, + "valid_targets_min": 1159 + }, + { + "epoch": 0.2921129503407984, + "grad_norm": 1.02712502502217, + "learning_rate": 1.6555555555555556e-05, + "loss": 0.42, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2322973608970642, + "step": 150, + "valid_targets_mean": 1999.4, + "valid_targets_min": 507 + }, + { + "epoch": 0.3018500486854917, + "grad_norm": 0.6139566424152655, + "learning_rate": 1.7111111111111112e-05, + "loss": 0.392, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1920212060213089, + "step": 155, + "valid_targets_mean": 4469.4, + "valid_targets_min": 1199 + }, + { + "epoch": 0.31158714703018503, + "grad_norm": 0.8626688830616602, + "learning_rate": 1.7666666666666668e-05, + "loss": 0.3874, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15205690264701843, + "step": 160, + "valid_targets_mean": 2269.1, + "valid_targets_min": 1221 + }, + { + "epoch": 0.3213242453748783, + "grad_norm": 0.7369325509923896, + "learning_rate": 1.8222222222222224e-05, + "loss": 0.3808, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1633809208869934, + "step": 165, + "valid_targets_mean": 3191.4, + "valid_targets_min": 693 + }, + { + "epoch": 0.3310613437195716, + "grad_norm": 0.9261185610744254, + "learning_rate": 1.877777777777778e-05, + "loss": 0.4041, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2576698362827301, + "step": 170, + "valid_targets_mean": 2714.6, + "valid_targets_min": 793 + }, + { + "epoch": 0.34079844206426485, + "grad_norm": 0.6904663234626722, + "learning_rate": 1.9333333333333333e-05, + "loss": 0.379, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19274544715881348, + "step": 175, + "valid_targets_mean": 3690.5, + "valid_targets_min": 1294 + }, + { + "epoch": 0.3505355404089581, + "grad_norm": 0.8172102038691861, + "learning_rate": 1.988888888888889e-05, + "loss": 0.4065, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.20005688071250916, + "step": 180, + "valid_targets_mean": 3082.0, + "valid_targets_min": 1208 + }, + { + "epoch": 0.3602726387536514, + "grad_norm": 0.7380067514394181, + "learning_rate": 2.0444444444444446e-05, + "loss": 0.4028, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.171134814620018, + "step": 185, + "valid_targets_mean": 2461.5, + "valid_targets_min": 1228 + }, + { + "epoch": 0.37000973709834467, + "grad_norm": 0.7600421793317651, + "learning_rate": 2.1000000000000002e-05, + "loss": 0.3993, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23319822549819946, + "step": 190, + "valid_targets_mean": 3521.8, + "valid_targets_min": 993 + }, + { + "epoch": 0.379746835443038, + "grad_norm": 0.6656904832221395, + "learning_rate": 2.1555555555555555e-05, + "loss": 0.4064, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.21867111325263977, + "step": 195, + "valid_targets_mean": 4448.8, + "valid_targets_min": 1679 + }, + { + "epoch": 0.3894839337877313, + "grad_norm": 0.8878862701932538, + "learning_rate": 2.2111111111111115e-05, + "loss": 0.3681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.20832327008247375, + "step": 200, + "valid_targets_mean": 2900.9, + "valid_targets_min": 967 + }, + { + "epoch": 0.39922103213242455, + "grad_norm": 0.5989319061194426, + "learning_rate": 2.2666666666666668e-05, + "loss": 0.3461, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14120624959468842, + "step": 205, + "valid_targets_mean": 3533.9, + "valid_targets_min": 808 + }, + { + "epoch": 0.4089581304771178, + "grad_norm": 0.6157267635559596, + "learning_rate": 2.3222222222222227e-05, + "loss": 0.3701, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1939328908920288, + "step": 210, + "valid_targets_mean": 4087.9, + "valid_targets_min": 1256 + }, + { + "epoch": 0.4186952288218111, + "grad_norm": 0.6990114686440374, + "learning_rate": 2.377777777777778e-05, + "loss": 0.3819, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18667009472846985, + "step": 215, + "valid_targets_mean": 2944.9, + "valid_targets_min": 709 + }, + { + "epoch": 0.42843232716650437, + "grad_norm": 0.8210463913714243, + "learning_rate": 2.4333333333333333e-05, + "loss": 0.403, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2342182993888855, + "step": 220, + "valid_targets_mean": 3265.4, + "valid_targets_min": 940 + }, + { + "epoch": 0.43816942551119764, + "grad_norm": 0.8024376262887541, + "learning_rate": 2.4888888888888893e-05, + "loss": 0.3901, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.20045126974582672, + "step": 225, + "valid_targets_mean": 3143.4, + "valid_targets_min": 751 + }, + { + "epoch": 0.44790652385589097, + "grad_norm": 0.824782738636105, + "learning_rate": 2.5444444444444446e-05, + "loss": 0.4028, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15007048845291138, + "step": 230, + "valid_targets_mean": 2354.4, + "valid_targets_min": 1504 + }, + { + "epoch": 0.45764362220058424, + "grad_norm": 1.0330846998899283, + "learning_rate": 2.6000000000000002e-05, + "loss": 0.3932, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18381167948246002, + "step": 235, + "valid_targets_mean": 1898.6, + "valid_targets_min": 781 + }, + { + "epoch": 0.4673807205452775, + "grad_norm": 0.882733313530884, + "learning_rate": 2.6555555555555558e-05, + "loss": 0.396, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17795206606388092, + "step": 240, + "valid_targets_mean": 2748.1, + "valid_targets_min": 1364 + }, + { + "epoch": 0.4771178188899708, + "grad_norm": 0.7790421961369038, + "learning_rate": 2.7111111111111114e-05, + "loss": 0.3861, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1974773406982422, + "step": 245, + "valid_targets_mean": 3469.5, + "valid_targets_min": 1085 + }, + { + "epoch": 0.48685491723466406, + "grad_norm": 0.9770266940397602, + "learning_rate": 2.7666666666666667e-05, + "loss": 0.3841, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2519018054008484, + "step": 250, + "valid_targets_mean": 2484.8, + "valid_targets_min": 1504 + }, + { + "epoch": 0.49659201557935734, + "grad_norm": 0.868298351235294, + "learning_rate": 2.8222222222222227e-05, + "loss": 0.3897, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2724197208881378, + "step": 255, + "valid_targets_mean": 3457.8, + "valid_targets_min": 1065 + }, + { + "epoch": 0.5063291139240507, + "grad_norm": 0.7596864734968185, + "learning_rate": 2.877777777777778e-05, + "loss": 0.3422, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23726016283035278, + "step": 260, + "valid_targets_mean": 4169.0, + "valid_targets_min": 2438 + }, + { + "epoch": 0.5160662122687439, + "grad_norm": 0.8718334359023358, + "learning_rate": 2.9333333333333333e-05, + "loss": 0.3726, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2140669971704483, + "step": 265, + "valid_targets_mean": 3537.9, + "valid_targets_min": 992 + }, + { + "epoch": 0.5258033106134372, + "grad_norm": 0.6989164218061054, + "learning_rate": 2.9888888888888892e-05, + "loss": 0.3541, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.20658528804779053, + "step": 270, + "valid_targets_mean": 4186.6, + "valid_targets_min": 1458 + }, + { + "epoch": 0.5355404089581305, + "grad_norm": 0.874757587964136, + "learning_rate": 3.0444444444444445e-05, + "loss": 0.3688, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19685882329940796, + "step": 275, + "valid_targets_mean": 2467.0, + "valid_targets_min": 1414 + }, + { + "epoch": 0.5452775073028238, + "grad_norm": 0.8053998992861866, + "learning_rate": 3.1e-05, + "loss": 0.3908, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15753547847270966, + "step": 280, + "valid_targets_mean": 2525.6, + "valid_targets_min": 1493 + }, + { + "epoch": 0.555014605647517, + "grad_norm": 0.805790874423672, + "learning_rate": 3.155555555555556e-05, + "loss": 0.3506, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16402557492256165, + "step": 285, + "valid_targets_mean": 2415.4, + "valid_targets_min": 1408 + }, + { + "epoch": 0.5647517039922103, + "grad_norm": 0.9207062014606001, + "learning_rate": 3.2111111111111114e-05, + "loss": 0.3528, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13046719133853912, + "step": 290, + "valid_targets_mean": 1499.0, + "valid_targets_min": 871 + }, + { + "epoch": 0.5744888023369036, + "grad_norm": 0.7580062459322378, + "learning_rate": 3.266666666666667e-05, + "loss": 0.3548, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1915191262960434, + "step": 295, + "valid_targets_mean": 2890.1, + "valid_targets_min": 562 + }, + { + "epoch": 0.5842259006815969, + "grad_norm": 0.8333886934844644, + "learning_rate": 3.3222222222222226e-05, + "loss": 0.3643, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1909688115119934, + "step": 300, + "valid_targets_mean": 2758.6, + "valid_targets_min": 1306 + }, + { + "epoch": 0.5939629990262901, + "grad_norm": 0.8714843579233928, + "learning_rate": 3.377777777777778e-05, + "loss": 0.3757, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1878107488155365, + "step": 305, + "valid_targets_mean": 3332.0, + "valid_targets_min": 1009 + }, + { + "epoch": 0.6037000973709834, + "grad_norm": 0.8947388864456415, + "learning_rate": 3.433333333333333e-05, + "loss": 0.3929, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.21496254205703735, + "step": 310, + "valid_targets_mean": 2662.9, + "valid_targets_min": 763 + }, + { + "epoch": 0.6134371957156767, + "grad_norm": 0.8537461910420446, + "learning_rate": 3.4888888888888895e-05, + "loss": 0.3531, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.20809201896190643, + "step": 315, + "valid_targets_mean": 2802.1, + "valid_targets_min": 936 + }, + { + "epoch": 0.6231742940603701, + "grad_norm": 0.6719251437783023, + "learning_rate": 3.5444444444444445e-05, + "loss": 0.3503, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16037863492965698, + "step": 320, + "valid_targets_mean": 3631.4, + "valid_targets_min": 1046 + }, + { + "epoch": 0.6329113924050633, + "grad_norm": 0.7848437937857925, + "learning_rate": 3.6e-05, + "loss": 0.3651, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18178021907806396, + "step": 325, + "valid_targets_mean": 3496.6, + "valid_targets_min": 589 + }, + { + "epoch": 0.6426484907497566, + "grad_norm": 0.9302516156723694, + "learning_rate": 3.655555555555556e-05, + "loss": 0.3494, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17631950974464417, + "step": 330, + "valid_targets_mean": 2446.1, + "valid_targets_min": 1254 + }, + { + "epoch": 0.6523855890944499, + "grad_norm": 0.9528588397406593, + "learning_rate": 3.7111111111111113e-05, + "loss": 0.3732, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16663740575313568, + "step": 335, + "valid_targets_mean": 2470.2, + "valid_targets_min": 761 + }, + { + "epoch": 0.6621226874391432, + "grad_norm": 0.660648347651872, + "learning_rate": 3.766666666666667e-05, + "loss": 0.3491, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23378151655197144, + "step": 340, + "valid_targets_mean": 5074.6, + "valid_targets_min": 1774 + }, + { + "epoch": 0.6718597857838364, + "grad_norm": 0.7317060175922973, + "learning_rate": 3.8222222222222226e-05, + "loss": 0.3508, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18020671606063843, + "step": 345, + "valid_targets_mean": 3391.1, + "valid_targets_min": 747 + }, + { + "epoch": 0.6815968841285297, + "grad_norm": 0.7458049884451803, + "learning_rate": 3.877777777777778e-05, + "loss": 0.3635, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.22611746191978455, + "step": 350, + "valid_targets_mean": 4497.8, + "valid_targets_min": 560 + }, + { + "epoch": 0.691333982473223, + "grad_norm": 0.7028549167777147, + "learning_rate": 3.933333333333333e-05, + "loss": 0.3387, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18253403902053833, + "step": 355, + "valid_targets_mean": 3703.0, + "valid_targets_min": 1396 + }, + { + "epoch": 0.7010710808179162, + "grad_norm": 0.8105132831889875, + "learning_rate": 3.9888888888888895e-05, + "loss": 0.3615, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23544040322303772, + "step": 360, + "valid_targets_mean": 4297.5, + "valid_targets_min": 1307 + }, + { + "epoch": 0.7108081791626095, + "grad_norm": 0.9310499702219079, + "learning_rate": 3.9999849385947476e-05, + "loss": 0.359, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2279876470565796, + "step": 365, + "valid_targets_mean": 2907.0, + "valid_targets_min": 754 + }, + { + "epoch": 0.7205452775073028, + "grad_norm": 0.677986808450289, + "learning_rate": 3.999923752024691e-05, + "loss": 0.3444, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1297803521156311, + "step": 370, + "valid_targets_mean": 2690.5, + "valid_targets_min": 1284 + }, + { + "epoch": 0.7302823758519961, + "grad_norm": 0.8247430862079875, + "learning_rate": 3.999815500390824e-05, + "loss": 0.3404, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13174360990524292, + "step": 375, + "valid_targets_mean": 1874.4, + "valid_targets_min": 1231 + }, + { + "epoch": 0.7400194741966893, + "grad_norm": 0.7296077001600262, + "learning_rate": 3.9996601862406804e-05, + "loss": 0.3359, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1771552413702011, + "step": 380, + "valid_targets_mean": 3262.9, + "valid_targets_min": 1149 + }, + { + "epoch": 0.7497565725413826, + "grad_norm": 0.8158122022730464, + "learning_rate": 3.999457813229333e-05, + "loss": 0.3512, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11544027179479599, + "step": 385, + "valid_targets_mean": 1960.2, + "valid_targets_min": 1035 + }, + { + "epoch": 0.759493670886076, + "grad_norm": 0.7322606730903284, + "learning_rate": 3.999208386119314e-05, + "loss": 0.336, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.143889918923378, + "step": 390, + "valid_targets_mean": 3042.1, + "valid_targets_min": 973 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.7701997355670589, + "learning_rate": 3.998911910780497e-05, + "loss": 0.3252, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10486048460006714, + "step": 395, + "valid_targets_mean": 2053.1, + "valid_targets_min": 797 + }, + { + "epoch": 0.7789678675754625, + "grad_norm": 0.7202330659166157, + "learning_rate": 3.9985683941899654e-05, + "loss": 0.322, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18157202005386353, + "step": 400, + "valid_targets_mean": 4142.8, + "valid_targets_min": 1497 + }, + { + "epoch": 0.7887049659201558, + "grad_norm": 0.8024411032059268, + "learning_rate": 3.9981778444318414e-05, + "loss": 0.3468, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1389027237892151, + "step": 405, + "valid_targets_mean": 3080.4, + "valid_targets_min": 895 + }, + { + "epoch": 0.7984420642648491, + "grad_norm": 0.7764211510246699, + "learning_rate": 3.9977402706971e-05, + "loss": 0.3257, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1333743929862976, + "step": 410, + "valid_targets_mean": 2548.8, + "valid_targets_min": 1103 + }, + { + "epoch": 0.8081791626095424, + "grad_norm": 0.7770749008950054, + "learning_rate": 3.997255683283352e-05, + "loss": 0.3419, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16055616736412048, + "step": 415, + "valid_targets_mean": 3037.5, + "valid_targets_min": 774 + }, + { + "epoch": 0.8179162609542356, + "grad_norm": 0.9875453975255327, + "learning_rate": 3.9967240935946004e-05, + "loss": 0.3449, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2241707742214203, + "step": 420, + "valid_targets_mean": 2675.0, + "valid_targets_min": 1080 + }, + { + "epoch": 0.8276533592989289, + "grad_norm": 0.8008854116862701, + "learning_rate": 3.996145514140974e-05, + "loss": 0.3324, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.21498018503189087, + "step": 425, + "valid_targets_mean": 3361.9, + "valid_targets_min": 557 + }, + { + "epoch": 0.8373904576436222, + "grad_norm": 0.8712307832624219, + "learning_rate": 3.995519958538431e-05, + "loss": 0.3372, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18727466464042664, + "step": 430, + "valid_targets_mean": 2693.6, + "valid_targets_min": 1161 + }, + { + "epoch": 0.8471275559883155, + "grad_norm": 0.7823837544455564, + "learning_rate": 3.99484744150844e-05, + "loss": 0.3503, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18069647252559662, + "step": 435, + "valid_targets_mean": 3102.6, + "valid_targets_min": 889 + }, + { + "epoch": 0.8568646543330087, + "grad_norm": 0.6463271995153775, + "learning_rate": 3.9941279788776324e-05, + "loss": 0.3271, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18949922919273376, + "step": 440, + "valid_targets_mean": 5080.8, + "valid_targets_min": 1045 + }, + { + "epoch": 0.866601752677702, + "grad_norm": 0.9414900911940838, + "learning_rate": 3.9933615875774314e-05, + "loss": 0.3223, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.199024498462677, + "step": 445, + "valid_targets_mean": 2462.9, + "valid_targets_min": 764 + }, + { + "epoch": 0.8763388510223953, + "grad_norm": 0.8197953554869378, + "learning_rate": 3.992548285643653e-05, + "loss": 0.318, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1533162146806717, + "step": 450, + "valid_targets_mean": 2308.9, + "valid_targets_min": 973 + }, + { + "epoch": 0.8860759493670886, + "grad_norm": 0.8282523432739521, + "learning_rate": 3.99168809221608e-05, + "loss": 0.3219, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18272504210472107, + "step": 455, + "valid_targets_mean": 2762.1, + "valid_targets_min": 767 + }, + { + "epoch": 0.8958130477117819, + "grad_norm": 0.652073664820321, + "learning_rate": 3.9907810275380133e-05, + "loss": 0.3377, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12357011437416077, + "step": 460, + "valid_targets_mean": 2613.4, + "valid_targets_min": 1247 + }, + { + "epoch": 0.9055501460564752, + "grad_norm": 0.7821036719182141, + "learning_rate": 3.989827112955794e-05, + "loss": 0.3134, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19240909814834595, + "step": 465, + "valid_targets_mean": 3378.9, + "valid_targets_min": 1163 + }, + { + "epoch": 0.9152872444011685, + "grad_norm": 0.7504672949569424, + "learning_rate": 3.988826370918306e-05, + "loss": 0.3374, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13087493181228638, + "step": 470, + "valid_targets_mean": 2295.5, + "valid_targets_min": 1017 + }, + { + "epoch": 0.9250243427458618, + "grad_norm": 0.7954921854314398, + "learning_rate": 3.987778824976439e-05, + "loss": 0.3327, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.22488024830818176, + "step": 475, + "valid_targets_mean": 3514.4, + "valid_targets_min": 1207 + }, + { + "epoch": 0.934761441090555, + "grad_norm": 0.7890591234039893, + "learning_rate": 3.9866844997825415e-05, + "loss": 0.3029, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2003348469734192, + "step": 480, + "valid_targets_mean": 3809.4, + "valid_targets_min": 1026 + }, + { + "epoch": 0.9444985394352483, + "grad_norm": 0.7469531975242553, + "learning_rate": 3.985543421089839e-05, + "loss": 0.3513, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16702266037464142, + "step": 485, + "valid_targets_mean": 3206.9, + "valid_targets_min": 879 + }, + { + "epoch": 0.9542356377799416, + "grad_norm": 0.8923246595958493, + "learning_rate": 3.984355615751824e-05, + "loss": 0.33, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1329648494720459, + "step": 490, + "valid_targets_mean": 2201.8, + "valid_targets_min": 1126 + }, + { + "epoch": 0.9639727361246349, + "grad_norm": 0.8616912426190275, + "learning_rate": 3.983121111721631e-05, + "loss": 0.3458, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.21117722988128662, + "step": 495, + "valid_targets_mean": 2512.8, + "valid_targets_min": 1080 + }, + { + "epoch": 0.9737098344693281, + "grad_norm": 1.4342599215630012, + "learning_rate": 3.981839938051372e-05, + "loss": 0.3138, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16730377078056335, + "step": 500, + "valid_targets_mean": 2778.1, + "valid_targets_min": 1517 + }, + { + "epoch": 0.9834469328140214, + "grad_norm": 0.6581607561348938, + "learning_rate": 3.980512124891458e-05, + "loss": 0.3228, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13223397731781006, + "step": 505, + "valid_targets_mean": 2865.0, + "valid_targets_min": 388 + }, + { + "epoch": 0.9931840311587147, + "grad_norm": 0.7372438054026565, + "learning_rate": 3.979137703489883e-05, + "loss": 0.3199, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17514921724796295, + "step": 510, + "valid_targets_mean": 3584.5, + "valid_targets_min": 822 + }, + { + "epoch": 1.0019474196689386, + "grad_norm": 0.7858899547464105, + "learning_rate": 3.9777167061915e-05, + "loss": 0.3175, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16728319227695465, + "step": 515, + "valid_targets_mean": 3270.8, + "valid_targets_min": 944 + }, + { + "epoch": 1.011684518013632, + "grad_norm": 0.9705259234628123, + "learning_rate": 3.976249166437243e-05, + "loss": 0.2926, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15755394101142883, + "step": 520, + "valid_targets_mean": 2423.5, + "valid_targets_min": 355 + }, + { + "epoch": 1.0214216163583252, + "grad_norm": 0.6806683609325453, + "learning_rate": 3.974735118763359e-05, + "loss": 0.301, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15417352318763733, + "step": 525, + "valid_targets_mean": 3597.0, + "valid_targets_min": 1566 + }, + { + "epoch": 1.0311587147030186, + "grad_norm": 0.8739140435404488, + "learning_rate": 3.97317459880058e-05, + "loss": 0.302, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11484783887863159, + "step": 530, + "valid_targets_mean": 1932.6, + "valid_targets_min": 793 + }, + { + "epoch": 1.0408958130477117, + "grad_norm": 0.8457500312718982, + "learning_rate": 3.971567643273292e-05, + "loss": 0.2973, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16622629761695862, + "step": 535, + "valid_targets_mean": 2864.4, + "valid_targets_min": 815 + }, + { + "epoch": 1.0506329113924051, + "grad_norm": 0.7340752528530782, + "learning_rate": 3.9699142899986685e-05, + "loss": 0.2905, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11975589394569397, + "step": 540, + "valid_targets_mean": 3145.0, + "valid_targets_min": 1661 + }, + { + "epoch": 1.0603700097370983, + "grad_norm": 0.6371936175766221, + "learning_rate": 3.968214577885782e-05, + "loss": 0.2761, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1377580165863037, + "step": 545, + "valid_targets_mean": 3580.8, + "valid_targets_min": 684 + }, + { + "epoch": 1.0701071080817917, + "grad_norm": 0.7077946139993203, + "learning_rate": 3.966468546934687e-05, + "loss": 0.3011, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12562993168830872, + "step": 550, + "valid_targets_mean": 3473.5, + "valid_targets_min": 1306 + }, + { + "epoch": 1.0798442064264848, + "grad_norm": 0.7019427006203095, + "learning_rate": 3.964676238235479e-05, + "loss": 0.2989, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14723308384418488, + "step": 555, + "valid_targets_mean": 3323.2, + "valid_targets_min": 291 + }, + { + "epoch": 1.0895813047711782, + "grad_norm": 0.6320463460477083, + "learning_rate": 3.9628376939673275e-05, + "loss": 0.2786, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06414983421564102, + "step": 560, + "valid_targets_mean": 1992.5, + "valid_targets_min": 625 + }, + { + "epoch": 1.0993184031158714, + "grad_norm": 0.7782569308843855, + "learning_rate": 3.960952957397484e-05, + "loss": 0.2856, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14603619277477264, + "step": 565, + "valid_targets_mean": 2900.5, + "valid_targets_min": 975 + }, + { + "epoch": 1.1090555014605648, + "grad_norm": 0.7063719109391404, + "learning_rate": 3.9590220728802625e-05, + "loss": 0.3081, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17670801281929016, + "step": 570, + "valid_targets_mean": 4031.0, + "valid_targets_min": 875 + }, + { + "epoch": 1.1187925998052581, + "grad_norm": 0.7654501871062778, + "learning_rate": 3.9570450858559975e-05, + "loss": 0.2826, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13301081955432892, + "step": 575, + "valid_targets_mean": 2697.2, + "valid_targets_min": 991 + }, + { + "epoch": 1.1285296981499513, + "grad_norm": 0.8053526181867805, + "learning_rate": 3.9550220428499725e-05, + "loss": 0.2952, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11767854541540146, + "step": 580, + "valid_targets_mean": 2463.9, + "valid_targets_min": 1309 + }, + { + "epoch": 1.1382667964946447, + "grad_norm": 0.7055726427752043, + "learning_rate": 3.9529529914713266e-05, + "loss": 0.3074, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17200462520122528, + "step": 585, + "valid_targets_mean": 3473.1, + "valid_targets_min": 1285 + }, + { + "epoch": 1.1480038948393378, + "grad_norm": 0.7796259491074381, + "learning_rate": 3.950837980411934e-05, + "loss": 0.2979, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1489405483007431, + "step": 590, + "valid_targets_mean": 2696.5, + "valid_targets_min": 1311 + }, + { + "epoch": 1.1577409931840312, + "grad_norm": 0.6673286671425366, + "learning_rate": 3.948677059445257e-05, + "loss": 0.2733, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12524479627609253, + "step": 595, + "valid_targets_mean": 3187.4, + "valid_targets_min": 1479 + }, + { + "epoch": 1.1674780915287244, + "grad_norm": 0.7241193327707911, + "learning_rate": 3.946470279425177e-05, + "loss": 0.2954, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13313235342502594, + "step": 600, + "valid_targets_mean": 3273.0, + "valid_targets_min": 1199 + }, + { + "epoch": 1.1772151898734178, + "grad_norm": 0.7914833715074661, + "learning_rate": 3.944217692284794e-05, + "loss": 0.2866, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13487054407596588, + "step": 605, + "valid_targets_mean": 2979.8, + "valid_targets_min": 1731 + }, + { + "epoch": 1.186952288218111, + "grad_norm": 0.7822511631725727, + "learning_rate": 3.941919351035208e-05, + "loss": 0.2902, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13327060639858246, + "step": 610, + "valid_targets_mean": 2522.2, + "valid_targets_min": 672 + }, + { + "epoch": 1.1966893865628043, + "grad_norm": 0.734117202012451, + "learning_rate": 3.9395753097642703e-05, + "loss": 0.307, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15425297617912292, + "step": 615, + "valid_targets_mean": 3389.9, + "valid_targets_min": 1414 + }, + { + "epoch": 1.2064264849074975, + "grad_norm": 0.8355556162478627, + "learning_rate": 3.937185623635309e-05, + "loss": 0.3072, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15557613968849182, + "step": 620, + "valid_targets_mean": 2553.2, + "valid_targets_min": 749 + }, + { + "epoch": 1.2161635832521909, + "grad_norm": 0.7047142278427146, + "learning_rate": 3.934750348885835e-05, + "loss": 0.2815, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19004464149475098, + "step": 625, + "valid_targets_mean": 4268.2, + "valid_targets_min": 1013 + }, + { + "epoch": 1.225900681596884, + "grad_norm": 0.7237181867450329, + "learning_rate": 3.932269542826211e-05, + "loss": 0.2897, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16163642704486847, + "step": 630, + "valid_targets_mean": 3763.0, + "valid_targets_min": 705 + }, + { + "epoch": 1.2356377799415774, + "grad_norm": 0.6541968706944473, + "learning_rate": 3.9297432638383133e-05, + "loss": 0.2998, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15847386419773102, + "step": 635, + "valid_targets_mean": 3743.4, + "valid_targets_min": 1411 + }, + { + "epoch": 1.2453748782862708, + "grad_norm": 0.7551346347644923, + "learning_rate": 3.9271715713741484e-05, + "loss": 0.2892, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17334289848804474, + "step": 640, + "valid_targets_mean": 3369.0, + "valid_targets_min": 1937 + }, + { + "epoch": 1.255111976630964, + "grad_norm": 0.7989231073679077, + "learning_rate": 3.9245545259544575e-05, + "loss": 0.2914, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13173508644104004, + "step": 645, + "valid_targets_mean": 2341.4, + "valid_targets_min": 654 + }, + { + "epoch": 1.2648490749756571, + "grad_norm": 0.6713664995677392, + "learning_rate": 3.9218921891672955e-05, + "loss": 0.2937, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1702912300825119, + "step": 650, + "valid_targets_mean": 4714.0, + "valid_targets_min": 1272 + }, + { + "epoch": 1.2745861733203505, + "grad_norm": 0.839712903367956, + "learning_rate": 3.919184623666575e-05, + "loss": 0.2806, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15325814485549927, + "step": 655, + "valid_targets_mean": 2320.9, + "valid_targets_min": 822 + }, + { + "epoch": 1.284323271665044, + "grad_norm": 0.6641240558454146, + "learning_rate": 3.916431893170596e-05, + "loss": 0.2996, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13758313655853271, + "step": 660, + "valid_targets_mean": 3853.6, + "valid_targets_min": 1057 + }, + { + "epoch": 1.294060370009737, + "grad_norm": 0.6883117123729979, + "learning_rate": 3.913634062460548e-05, + "loss": 0.2867, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13828516006469727, + "step": 665, + "valid_targets_mean": 3092.4, + "valid_targets_min": 1104 + }, + { + "epoch": 1.3037974683544304, + "grad_norm": 0.644971806948973, + "learning_rate": 3.910791197378982e-05, + "loss": 0.3031, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09337528049945831, + "step": 670, + "valid_targets_mean": 2732.9, + "valid_targets_min": 664 + }, + { + "epoch": 1.3135345666991236, + "grad_norm": 0.8550709015445958, + "learning_rate": 3.9079033648282595e-05, + "loss": 0.3224, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1829337328672409, + "step": 675, + "valid_targets_mean": 2290.4, + "valid_targets_min": 683 + }, + { + "epoch": 1.323271665043817, + "grad_norm": 0.8135393271299947, + "learning_rate": 3.9049706327689844e-05, + "loss": 0.2723, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11390693485736847, + "step": 680, + "valid_targets_mean": 2498.4, + "valid_targets_min": 1532 + }, + { + "epoch": 1.3330087633885102, + "grad_norm": 0.6943074012950569, + "learning_rate": 3.9019930702183995e-05, + "loss": 0.2744, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1324174702167511, + "step": 685, + "valid_targets_mean": 3205.0, + "valid_targets_min": 622 + }, + { + "epoch": 1.3427458617332035, + "grad_norm": 0.6843356225052816, + "learning_rate": 3.8989707472487605e-05, + "loss": 0.2835, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1364578902721405, + "step": 690, + "valid_targets_mean": 3158.2, + "valid_targets_min": 1224 + }, + { + "epoch": 1.352482960077897, + "grad_norm": 0.819539606411039, + "learning_rate": 3.8959037349856915e-05, + "loss": 0.2898, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10035207122564316, + "step": 695, + "valid_targets_mean": 2056.0, + "valid_targets_min": 580 + }, + { + "epoch": 1.36222005842259, + "grad_norm": 0.776933978950489, + "learning_rate": 3.8927921056065085e-05, + "loss": 0.2916, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10031980276107788, + "step": 700, + "valid_targets_mean": 1935.1, + "valid_targets_min": 1267 + }, + { + "epoch": 1.3719571567672832, + "grad_norm": 0.7965490972208926, + "learning_rate": 3.8896359323385196e-05, + "loss": 0.296, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13414305448532104, + "step": 705, + "valid_targets_mean": 2537.6, + "valid_targets_min": 1122 + }, + { + "epoch": 1.3816942551119766, + "grad_norm": 0.8531863308487779, + "learning_rate": 3.886435289457306e-05, + "loss": 0.2983, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1738170087337494, + "step": 710, + "valid_targets_mean": 2457.0, + "valid_targets_min": 779 + }, + { + "epoch": 1.39143135345667, + "grad_norm": 0.731101519911782, + "learning_rate": 3.883190252284969e-05, + "loss": 0.2662, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11143383383750916, + "step": 715, + "valid_targets_mean": 2316.5, + "valid_targets_min": 317 + }, + { + "epoch": 1.4011684518013632, + "grad_norm": 0.6662025242467065, + "learning_rate": 3.879900897188361e-05, + "loss": 0.2824, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1794198453426361, + "step": 720, + "valid_targets_mean": 4965.1, + "valid_targets_min": 1986 + }, + { + "epoch": 1.4109055501460563, + "grad_norm": 0.8360114160936927, + "learning_rate": 3.876567301577287e-05, + "loss": 0.267, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15071365237236023, + "step": 725, + "valid_targets_mean": 2652.9, + "valid_targets_min": 973 + }, + { + "epoch": 1.4206426484907497, + "grad_norm": 0.6719303787687317, + "learning_rate": 3.873189543902683e-05, + "loss": 0.2857, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17039050161838531, + "step": 730, + "valid_targets_mean": 3710.0, + "valid_targets_min": 1899 + }, + { + "epoch": 1.4303797468354431, + "grad_norm": 0.8200941586506575, + "learning_rate": 3.86976770365477e-05, + "loss": 0.2859, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17211097478866577, + "step": 735, + "valid_targets_mean": 3249.9, + "valid_targets_min": 1512 + }, + { + "epoch": 1.4401168451801363, + "grad_norm": 0.7647285702362657, + "learning_rate": 3.8663018613611824e-05, + "loss": 0.2832, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1194126158952713, + "step": 740, + "valid_targets_mean": 2530.5, + "valid_targets_min": 1285 + }, + { + "epoch": 1.4498539435248297, + "grad_norm": 0.7598532952082641, + "learning_rate": 3.862792098585074e-05, + "loss": 0.2751, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1360061913728714, + "step": 745, + "valid_targets_mean": 3239.8, + "valid_targets_min": 909 + }, + { + "epoch": 1.4595910418695228, + "grad_norm": 0.8903098563066778, + "learning_rate": 3.8592384979231984e-05, + "loss": 0.2623, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11483582854270935, + "step": 750, + "valid_targets_mean": 2928.1, + "valid_targets_min": 1210 + }, + { + "epoch": 1.4693281402142162, + "grad_norm": 0.8495650293717296, + "learning_rate": 3.8556411430039656e-05, + "loss": 0.2949, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19118264317512512, + "step": 755, + "valid_targets_mean": 2806.5, + "valid_targets_min": 936 + }, + { + "epoch": 1.4790652385589094, + "grad_norm": 0.9628487252912299, + "learning_rate": 3.852000118485471e-05, + "loss": 0.3034, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1747865378856659, + "step": 760, + "valid_targets_mean": 2188.1, + "valid_targets_min": 1392 + }, + { + "epoch": 1.4888023369036028, + "grad_norm": 0.7004325923802018, + "learning_rate": 3.8483155100535074e-05, + "loss": 0.2768, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12580302357673645, + "step": 765, + "valid_targets_mean": 3205.1, + "valid_targets_min": 1514 + }, + { + "epoch": 1.4985394352482961, + "grad_norm": 0.74062192186481, + "learning_rate": 3.8445874044195475e-05, + "loss": 0.2734, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11894209682941437, + "step": 770, + "valid_targets_mean": 2435.5, + "valid_targets_min": 1201 + }, + { + "epoch": 1.5082765335929893, + "grad_norm": 0.8816120657903092, + "learning_rate": 3.840815889318699e-05, + "loss": 0.2707, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16805878281593323, + "step": 775, + "valid_targets_mean": 2607.0, + "valid_targets_min": 1426 + }, + { + "epoch": 1.5180136319376825, + "grad_norm": 0.6447763694072602, + "learning_rate": 3.837001053507645e-05, + "loss": 0.2731, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11964346468448639, + "step": 780, + "valid_targets_mean": 3388.4, + "valid_targets_min": 568 + }, + { + "epoch": 1.5277507302823758, + "grad_norm": 0.7624960908344781, + "learning_rate": 3.833142986762556e-05, + "loss": 0.2886, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1300625503063202, + "step": 785, + "valid_targets_mean": 2600.6, + "valid_targets_min": 1059 + }, + { + "epoch": 1.5374878286270692, + "grad_norm": 0.8002057073411786, + "learning_rate": 3.82924177987697e-05, + "loss": 0.2832, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.121633380651474, + "step": 790, + "valid_targets_mean": 1773.9, + "valid_targets_min": 603 + }, + { + "epoch": 1.5472249269717624, + "grad_norm": 0.6878158152271249, + "learning_rate": 3.8252975246596634e-05, + "loss": 0.284, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11953762918710709, + "step": 795, + "valid_targets_mean": 2672.4, + "valid_targets_min": 304 + }, + { + "epoch": 1.5569620253164556, + "grad_norm": 0.6397890876132153, + "learning_rate": 3.821310313932486e-05, + "loss": 0.3103, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1216995120048523, + "step": 800, + "valid_targets_mean": 3221.5, + "valid_targets_min": 1079 + }, + { + "epoch": 1.566699123661149, + "grad_norm": 0.7994201671438256, + "learning_rate": 3.8172802415281796e-05, + "loss": 0.2919, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1517755687236786, + "step": 805, + "valid_targets_mean": 2454.0, + "valid_targets_min": 1178 + }, + { + "epoch": 1.5764362220058423, + "grad_norm": 0.8475574220657818, + "learning_rate": 3.813207402288167e-05, + "loss": 0.2841, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12581583857536316, + "step": 810, + "valid_targets_mean": 2915.9, + "valid_targets_min": 1215 + }, + { + "epoch": 1.5861733203505355, + "grad_norm": 0.78684352089434, + "learning_rate": 3.8090918920603214e-05, + "loss": 0.2652, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10042424499988556, + "step": 815, + "valid_targets_mean": 2571.6, + "valid_targets_min": 989 + }, + { + "epoch": 1.5959104186952289, + "grad_norm": 0.7306518474461766, + "learning_rate": 3.804933807696712e-05, + "loss": 0.2753, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11414317041635513, + "step": 820, + "valid_targets_mean": 2278.1, + "valid_targets_min": 961 + }, + { + "epoch": 1.6056475170399223, + "grad_norm": 0.6843854426395394, + "learning_rate": 3.80073324705132e-05, + "loss": 0.295, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12278768420219421, + "step": 825, + "valid_targets_mean": 3124.6, + "valid_targets_min": 1350 + }, + { + "epoch": 1.6153846153846154, + "grad_norm": 0.7572889779601448, + "learning_rate": 3.7964903089777443e-05, + "loss": 0.2655, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11568583548069, + "step": 830, + "valid_targets_mean": 2210.9, + "valid_targets_min": 560 + }, + { + "epoch": 1.6251217137293086, + "grad_norm": 0.6441562182965196, + "learning_rate": 3.792205093326865e-05, + "loss": 0.276, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11288474500179291, + "step": 835, + "valid_targets_mean": 2595.0, + "valid_targets_min": 392 + }, + { + "epoch": 1.634858812074002, + "grad_norm": 0.6518505921760807, + "learning_rate": 3.7878777009445025e-05, + "loss": 0.284, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11960506439208984, + "step": 840, + "valid_targets_mean": 3584.9, + "valid_targets_min": 1431 + }, + { + "epoch": 1.6445959104186954, + "grad_norm": 0.8324832568062053, + "learning_rate": 3.7835082336690376e-05, + "loss": 0.2742, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1086902916431427, + "step": 845, + "valid_targets_mean": 1682.0, + "valid_targets_min": 715 + }, + { + "epoch": 1.6543330087633885, + "grad_norm": 0.899466301772704, + "learning_rate": 3.7790967943290186e-05, + "loss": 0.3167, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19043229520320892, + "step": 850, + "valid_targets_mean": 2796.4, + "valid_targets_min": 300 + }, + { + "epoch": 1.6640701071080817, + "grad_norm": 0.6833506865776794, + "learning_rate": 3.7746434867407405e-05, + "loss": 0.2733, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1205308735370636, + "step": 855, + "valid_targets_mean": 3405.9, + "valid_targets_min": 1141 + }, + { + "epoch": 1.673807205452775, + "grad_norm": 0.6735391267174385, + "learning_rate": 3.7701484157058016e-05, + "loss": 0.2809, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14146654307842255, + "step": 860, + "valid_targets_mean": 3473.8, + "valid_targets_min": 1788 + }, + { + "epoch": 1.6835443037974684, + "grad_norm": 0.7478096289483627, + "learning_rate": 3.7656116870086376e-05, + "loss": 0.2884, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13985690474510193, + "step": 865, + "valid_targets_mean": 2415.1, + "valid_targets_min": 1055 + }, + { + "epoch": 1.6932814021421616, + "grad_norm": 0.6917897462416348, + "learning_rate": 3.761033407414032e-05, + "loss": 0.2819, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13799864053726196, + "step": 870, + "valid_targets_mean": 3341.5, + "valid_targets_min": 1898 + }, + { + "epoch": 1.7030185004868548, + "grad_norm": 0.7825581458353967, + "learning_rate": 3.756413684664602e-05, + "loss": 0.2881, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17821389436721802, + "step": 875, + "valid_targets_mean": 3530.2, + "valid_targets_min": 1414 + }, + { + "epoch": 1.7127555988315482, + "grad_norm": 0.6296436510408459, + "learning_rate": 3.751752627478266e-05, + "loss": 0.266, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09756700694561005, + "step": 880, + "valid_targets_mean": 2915.2, + "valid_targets_min": 1017 + }, + { + "epoch": 1.7224926971762415, + "grad_norm": 0.6844544440869792, + "learning_rate": 3.747050345545683e-05, + "loss": 0.273, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11896862089633942, + "step": 885, + "valid_targets_mean": 2729.0, + "valid_targets_min": 641 + }, + { + "epoch": 1.7322297955209347, + "grad_norm": 0.6474680504011459, + "learning_rate": 3.7423069495276705e-05, + "loss": 0.283, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15866486728191376, + "step": 890, + "valid_targets_mean": 4134.0, + "valid_targets_min": 1708 + }, + { + "epoch": 1.741966893865628, + "grad_norm": 0.7020327824367882, + "learning_rate": 3.737522551052603e-05, + "loss": 0.2665, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10592299699783325, + "step": 895, + "valid_targets_mean": 2089.9, + "valid_targets_min": 800 + }, + { + "epoch": 1.7517039922103215, + "grad_norm": 0.6709837312667878, + "learning_rate": 3.732697262713783e-05, + "loss": 0.2681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17397888004779816, + "step": 900, + "valid_targets_mean": 4001.9, + "valid_targets_min": 968 + }, + { + "epoch": 1.7614410905550146, + "grad_norm": 0.7029093037875962, + "learning_rate": 3.727831198066791e-05, + "loss": 0.2891, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18118925392627716, + "step": 905, + "valid_targets_mean": 3871.5, + "valid_targets_min": 993 + }, + { + "epoch": 1.7711781888997078, + "grad_norm": 0.6670349375759711, + "learning_rate": 3.722924471626815e-05, + "loss": 0.2733, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11348405480384827, + "step": 910, + "valid_targets_mean": 2493.2, + "valid_targets_min": 1225 + }, + { + "epoch": 1.7809152872444012, + "grad_norm": 0.857871388664443, + "learning_rate": 3.7179771988659526e-05, + "loss": 0.2766, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1357949823141098, + "step": 915, + "valid_targets_mean": 1648.1, + "valid_targets_min": 411 + }, + { + "epoch": 1.7906523855890946, + "grad_norm": 0.8701206332552374, + "learning_rate": 3.712989496210497e-05, + "loss": 0.2827, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11637701839208603, + "step": 920, + "valid_targets_mean": 1869.1, + "valid_targets_min": 1118 + }, + { + "epoch": 1.8003894839337877, + "grad_norm": 0.7733174755020938, + "learning_rate": 3.707961481038196e-05, + "loss": 0.2852, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16073478758335114, + "step": 925, + "valid_targets_mean": 2269.2, + "valid_targets_min": 614 + }, + { + "epoch": 1.810126582278481, + "grad_norm": 0.9844262069470177, + "learning_rate": 3.702893271675487e-05, + "loss": 0.2849, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13788703083992004, + "step": 930, + "valid_targets_mean": 1586.2, + "valid_targets_min": 750 + }, + { + "epoch": 1.8198636806231743, + "grad_norm": 0.8214724111954528, + "learning_rate": 3.697784987394716e-05, + "loss": 0.2681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18131202459335327, + "step": 935, + "valid_targets_mean": 2752.5, + "valid_targets_min": 1511 + }, + { + "epoch": 1.8296007789678677, + "grad_norm": 0.6915571590018944, + "learning_rate": 3.692636748411329e-05, + "loss": 0.2617, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1040545403957367, + "step": 940, + "valid_targets_mean": 2429.0, + "valid_targets_min": 949 + }, + { + "epoch": 1.8393378773125608, + "grad_norm": 0.6643771603137818, + "learning_rate": 3.687448675881043e-05, + "loss": 0.2695, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10466742515563965, + "step": 945, + "valid_targets_mean": 2981.1, + "valid_targets_min": 1571 + }, + { + "epoch": 1.849074975657254, + "grad_norm": 0.7594173087993755, + "learning_rate": 3.682220891896995e-05, + "loss": 0.276, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1362399160861969, + "step": 950, + "valid_targets_mean": 3018.8, + "valid_targets_min": 1849 + }, + { + "epoch": 1.8588120740019474, + "grad_norm": 0.8226705248019373, + "learning_rate": 3.676953519486868e-05, + "loss": 0.2941, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1400984525680542, + "step": 955, + "valid_targets_mean": 1782.9, + "valid_targets_min": 819 + }, + { + "epoch": 1.8685491723466408, + "grad_norm": 0.716706641470173, + "learning_rate": 3.671646682609999e-05, + "loss": 0.2916, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16295509040355682, + "step": 960, + "valid_targets_mean": 3783.8, + "valid_targets_min": 1576 + }, + { + "epoch": 1.878286270691334, + "grad_norm": 0.6680265782602849, + "learning_rate": 3.666300506154455e-05, + "loss": 0.2929, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16173237562179565, + "step": 965, + "valid_targets_mean": 4008.1, + "valid_targets_min": 1997 + }, + { + "epoch": 1.8880233690360273, + "grad_norm": 0.7900756008320604, + "learning_rate": 3.660915115934103e-05, + "loss": 0.2989, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14800488948822021, + "step": 970, + "valid_targets_mean": 2644.2, + "valid_targets_min": 980 + }, + { + "epoch": 1.8977604673807207, + "grad_norm": 0.6749709974662765, + "learning_rate": 3.6554906386856394e-05, + "loss": 0.2749, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12733563780784607, + "step": 975, + "valid_targets_mean": 3121.5, + "valid_targets_min": 1292 + }, + { + "epoch": 1.9074975657254138, + "grad_norm": 0.6981467453263979, + "learning_rate": 3.650027202065617e-05, + "loss": 0.2905, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1113966852426529, + "step": 980, + "valid_targets_mean": 2686.6, + "valid_targets_min": 1086 + }, + { + "epoch": 1.917234664070107, + "grad_norm": 0.5759410728202508, + "learning_rate": 3.644524934647432e-05, + "loss": 0.2736, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11122055351734161, + "step": 985, + "valid_targets_mean": 3869.5, + "valid_targets_min": 974 + }, + { + "epoch": 1.9269717624148004, + "grad_norm": 0.6088691920046002, + "learning_rate": 3.6389839659183064e-05, + "loss": 0.2846, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1156475841999054, + "step": 990, + "valid_targets_mean": 3328.2, + "valid_targets_min": 1268 + }, + { + "epoch": 1.9367088607594938, + "grad_norm": 0.6429286727874445, + "learning_rate": 3.633404426276234e-05, + "loss": 0.2777, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12231488525867462, + "step": 995, + "valid_targets_mean": 3249.6, + "valid_targets_min": 1012 + }, + { + "epoch": 1.946445959104187, + "grad_norm": 0.7228518779689982, + "learning_rate": 3.6277864470269164e-05, + "loss": 0.2712, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12440205365419388, + "step": 1000, + "valid_targets_mean": 2283.8, + "valid_targets_min": 1294 + }, + { + "epoch": 1.95618305744888, + "grad_norm": 0.6995191271444386, + "learning_rate": 3.622130160380668e-05, + "loss": 0.2662, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1588570922613144, + "step": 1005, + "valid_targets_mean": 3194.8, + "valid_targets_min": 1316 + }, + { + "epoch": 1.9659201557935735, + "grad_norm": 0.6375267751555264, + "learning_rate": 3.616435699449311e-05, + "loss": 0.2711, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08528415113687515, + "step": 1010, + "valid_targets_mean": 2210.2, + "valid_targets_min": 729 + }, + { + "epoch": 1.9756572541382669, + "grad_norm": 0.639572278608358, + "learning_rate": 3.6107031982430374e-05, + "loss": 0.2974, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.193544402718544, + "step": 1015, + "valid_targets_mean": 4963.1, + "valid_targets_min": 1505 + }, + { + "epoch": 1.98539435248296, + "grad_norm": 0.865702036273128, + "learning_rate": 3.604932791667258e-05, + "loss": 0.2775, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15228410065174103, + "step": 1020, + "valid_targets_mean": 3556.8, + "valid_targets_min": 858 + }, + { + "epoch": 1.9951314508276532, + "grad_norm": 0.6528761878848074, + "learning_rate": 3.599124615519427e-05, + "loss": 0.2423, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11325448751449585, + "step": 1025, + "valid_targets_mean": 2563.0, + "valid_targets_min": 1192 + }, + { + "epoch": 2.0038948393378773, + "grad_norm": 0.7377205804499293, + "learning_rate": 3.5932788064858454e-05, + "loss": 0.2514, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1272452026605606, + "step": 1030, + "valid_targets_mean": 2710.0, + "valid_targets_min": 931 + }, + { + "epoch": 2.0136319376825704, + "grad_norm": 0.8091893084525955, + "learning_rate": 3.5873955021384464e-05, + "loss": 0.2485, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11636292934417725, + "step": 1035, + "valid_targets_mean": 2944.6, + "valid_targets_min": 1276 + }, + { + "epoch": 2.023369036027264, + "grad_norm": 0.7003258404063661, + "learning_rate": 3.5814748409315545e-05, + "loss": 0.2476, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14736023545265198, + "step": 1040, + "valid_targets_mean": 3605.8, + "valid_targets_min": 810 + }, + { + "epoch": 2.033106134371957, + "grad_norm": 0.6474036924563122, + "learning_rate": 3.575516962198632e-05, + "loss": 0.242, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07607968896627426, + "step": 1045, + "valid_targets_mean": 2740.0, + "valid_targets_min": 1238 + }, + { + "epoch": 2.0428432327166504, + "grad_norm": 0.824017837195956, + "learning_rate": 3.5695220061489935e-05, + "loss": 0.2429, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12290918081998825, + "step": 1050, + "valid_targets_mean": 2670.5, + "valid_targets_min": 1133 + }, + { + "epoch": 2.0525803310613435, + "grad_norm": 0.7035984369000724, + "learning_rate": 3.563490113864514e-05, + "loss": 0.2464, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09453026950359344, + "step": 1055, + "valid_targets_mean": 2812.5, + "valid_targets_min": 1225 + }, + { + "epoch": 2.062317429406037, + "grad_norm": 0.7983219444234277, + "learning_rate": 3.557421427296302e-05, + "loss": 0.2339, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0978788286447525, + "step": 1060, + "valid_targets_mean": 2292.6, + "valid_targets_min": 1485 + }, + { + "epoch": 2.0720545277507303, + "grad_norm": 0.7973735559663347, + "learning_rate": 3.551316089261363e-05, + "loss": 0.2394, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13901948928833008, + "step": 1065, + "valid_targets_mean": 3392.8, + "valid_targets_min": 1411 + }, + { + "epoch": 2.0817916260954235, + "grad_norm": 0.9870859614285062, + "learning_rate": 3.545174243439236e-05, + "loss": 0.228, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13286170363426208, + "step": 1070, + "valid_targets_mean": 3008.1, + "valid_targets_min": 297 + }, + { + "epoch": 2.091528724440117, + "grad_norm": 0.7644411930451773, + "learning_rate": 3.538996034368615e-05, + "loss": 0.2415, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16460761427879333, + "step": 1075, + "valid_targets_mean": 4027.2, + "valid_targets_min": 1542 + }, + { + "epoch": 2.1012658227848102, + "grad_norm": 0.7956734974517503, + "learning_rate": 3.5327816074439445e-05, + "loss": 0.243, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11771304905414581, + "step": 1080, + "valid_targets_mean": 2529.5, + "valid_targets_min": 879 + }, + { + "epoch": 2.1110029211295034, + "grad_norm": 0.6436131645357464, + "learning_rate": 3.526531108912e-05, + "loss": 0.2331, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12277089059352875, + "step": 1085, + "valid_targets_mean": 3293.4, + "valid_targets_min": 591 + }, + { + "epoch": 2.1207400194741965, + "grad_norm": 0.9138861433788087, + "learning_rate": 3.520244685868446e-05, + "loss": 0.2261, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1006382405757904, + "step": 1090, + "valid_targets_mean": 1795.1, + "valid_targets_min": 295 + }, + { + "epoch": 2.13047711781889, + "grad_norm": 0.7538499512907332, + "learning_rate": 3.513922486254373e-05, + "loss": 0.237, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0947801023721695, + "step": 1095, + "valid_targets_mean": 2311.6, + "valid_targets_min": 1122 + }, + { + "epoch": 2.1402142161635833, + "grad_norm": 0.6105399013697741, + "learning_rate": 3.507564658852817e-05, + "loss": 0.2408, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1116853579878807, + "step": 1100, + "valid_targets_mean": 4266.0, + "valid_targets_min": 1560 + }, + { + "epoch": 2.1499513145082765, + "grad_norm": 0.8544955534319482, + "learning_rate": 3.501171353285258e-05, + "loss": 0.2488, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.116363525390625, + "step": 1105, + "valid_targets_mean": 2383.8, + "valid_targets_min": 1061 + }, + { + "epoch": 2.1596884128529696, + "grad_norm": 0.7574503962153785, + "learning_rate": 3.494742720008102e-05, + "loss": 0.24, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09310120344161987, + "step": 1110, + "valid_targets_mean": 2279.9, + "valid_targets_min": 323 + }, + { + "epoch": 2.1694255111976632, + "grad_norm": 0.7054565876739526, + "learning_rate": 3.488278910309131e-05, + "loss": 0.2348, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11796418577432632, + "step": 1115, + "valid_targets_mean": 2481.4, + "valid_targets_min": 1256 + }, + { + "epoch": 2.1791626095423564, + "grad_norm": 0.8624432463998876, + "learning_rate": 3.481780076303953e-05, + "loss": 0.2477, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13654282689094543, + "step": 1120, + "valid_targets_mean": 3006.1, + "valid_targets_min": 1040 + }, + { + "epoch": 2.1888997078870496, + "grad_norm": 0.7729213149253524, + "learning_rate": 3.475246370932419e-05, + "loss": 0.2484, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1039193719625473, + "step": 1125, + "valid_targets_mean": 2482.0, + "valid_targets_min": 880 + }, + { + "epoch": 2.1986368062317427, + "grad_norm": 0.8306922531898471, + "learning_rate": 3.468677947955019e-05, + "loss": 0.2347, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09656044840812683, + "step": 1130, + "valid_targets_mean": 1547.6, + "valid_targets_min": 734 + }, + { + "epoch": 2.2083739045764363, + "grad_norm": 0.6807291213869415, + "learning_rate": 3.4620749619492704e-05, + "loss": 0.2236, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11662845313549042, + "step": 1135, + "valid_targets_mean": 3677.1, + "valid_targets_min": 1293 + }, + { + "epoch": 2.2181110029211295, + "grad_norm": 0.773377055591104, + "learning_rate": 3.455437568306076e-05, + "loss": 0.2428, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13956855237483978, + "step": 1140, + "valid_targets_mean": 3065.8, + "valid_targets_min": 753 + }, + { + "epoch": 2.2278481012658227, + "grad_norm": 0.6679841589471326, + "learning_rate": 3.448765923226069e-05, + "loss": 0.2152, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10934606194496155, + "step": 1145, + "valid_targets_mean": 4065.2, + "valid_targets_min": 727 + }, + { + "epoch": 2.2375851996105163, + "grad_norm": 0.9727047265171258, + "learning_rate": 3.442060183715935e-05, + "loss": 0.2371, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13547798991203308, + "step": 1150, + "valid_targets_mean": 1850.2, + "valid_targets_min": 1051 + }, + { + "epoch": 2.2473222979552094, + "grad_norm": 1.4802333281654472, + "learning_rate": 3.4353205075847195e-05, + "loss": 0.2553, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1198016032576561, + "step": 1155, + "valid_targets_mean": 2603.4, + "valid_targets_min": 903 + }, + { + "epoch": 2.2570593962999026, + "grad_norm": 0.6955103270734062, + "learning_rate": 3.428547053440113e-05, + "loss": 0.2327, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1287013739347458, + "step": 1160, + "valid_targets_mean": 3495.5, + "valid_targets_min": 1050 + }, + { + "epoch": 2.2667964946445958, + "grad_norm": 0.7088985386321863, + "learning_rate": 3.421739980684718e-05, + "loss": 0.2368, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10113312304019928, + "step": 1165, + "valid_targets_mean": 2847.8, + "valid_targets_min": 1382 + }, + { + "epoch": 2.2765335929892894, + "grad_norm": 0.7514819406593808, + "learning_rate": 3.4148994495123e-05, + "loss": 0.2442, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1315937042236328, + "step": 1170, + "valid_targets_mean": 3536.6, + "valid_targets_min": 1575 + }, + { + "epoch": 2.2862706913339825, + "grad_norm": 0.7145930585200598, + "learning_rate": 3.408025620904012e-05, + "loss": 0.2294, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13585850596427917, + "step": 1175, + "valid_targets_mean": 4002.5, + "valid_targets_min": 1390 + }, + { + "epoch": 2.2960077896786757, + "grad_norm": 0.7845317543314521, + "learning_rate": 3.401118656624613e-05, + "loss": 0.2448, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12422779947519302, + "step": 1180, + "valid_targets_mean": 3427.6, + "valid_targets_min": 1350 + }, + { + "epoch": 2.305744888023369, + "grad_norm": 0.6726307033660588, + "learning_rate": 3.3941787192186584e-05, + "loss": 0.2507, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11506038904190063, + "step": 1185, + "valid_targets_mean": 3105.5, + "valid_targets_min": 762 + }, + { + "epoch": 2.3154819863680625, + "grad_norm": 0.5479126063821994, + "learning_rate": 3.387205972006671e-05, + "loss": 0.2282, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1155364066362381, + "step": 1190, + "valid_targets_mean": 5606.8, + "valid_targets_min": 1619 + }, + { + "epoch": 2.3252190847127556, + "grad_norm": 0.7289512246085008, + "learning_rate": 3.3802005790813046e-05, + "loss": 0.2157, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08744939416646957, + "step": 1195, + "valid_targets_mean": 2208.5, + "valid_targets_min": 317 + }, + { + "epoch": 2.334956183057449, + "grad_norm": 0.790510816502957, + "learning_rate": 3.373162705303476e-05, + "loss": 0.2426, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11945021152496338, + "step": 1200, + "valid_targets_mean": 2726.6, + "valid_targets_min": 1387 + }, + { + "epoch": 2.344693281402142, + "grad_norm": 0.7673786848762931, + "learning_rate": 3.366092516298489e-05, + "loss": 0.2414, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11489750444889069, + "step": 1205, + "valid_targets_mean": 2555.6, + "valid_targets_min": 1796 + }, + { + "epoch": 2.3544303797468356, + "grad_norm": 0.7113294796141776, + "learning_rate": 3.358990178452135e-05, + "loss": 0.2357, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11983954906463623, + "step": 1210, + "valid_targets_mean": 3281.4, + "valid_targets_min": 1324 + }, + { + "epoch": 2.3641674780915287, + "grad_norm": 0.777022529971605, + "learning_rate": 3.351855858906778e-05, + "loss": 0.2617, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09583482146263123, + "step": 1215, + "valid_targets_mean": 2159.6, + "valid_targets_min": 674 + }, + { + "epoch": 2.373904576436222, + "grad_norm": 0.6562812529073736, + "learning_rate": 3.344689725557422e-05, + "loss": 0.2463, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12501710653305054, + "step": 1220, + "valid_targets_mean": 4507.2, + "valid_targets_min": 1699 + }, + { + "epoch": 2.3836416747809155, + "grad_norm": 0.8459375828858623, + "learning_rate": 3.337491947047757e-05, + "loss": 0.2442, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14401216804981232, + "step": 1225, + "valid_targets_mean": 2357.1, + "valid_targets_min": 735 + }, + { + "epoch": 2.3933787731256086, + "grad_norm": 0.8326039552981002, + "learning_rate": 3.330262692766193e-05, + "loss": 0.2421, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09895455837249756, + "step": 1230, + "valid_targets_mean": 2117.0, + "valid_targets_min": 1140 + }, + { + "epoch": 2.403115871470302, + "grad_norm": 0.7636805929064753, + "learning_rate": 3.323002132841875e-05, + "loss": 0.2603, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15124696493148804, + "step": 1235, + "valid_targets_mean": 3200.6, + "valid_targets_min": 1942 + }, + { + "epoch": 2.412852969814995, + "grad_norm": 0.7622703732144163, + "learning_rate": 3.315710438140674e-05, + "loss": 0.2351, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0872589498758316, + "step": 1240, + "valid_targets_mean": 2806.6, + "valid_targets_min": 434 + }, + { + "epoch": 2.4225900681596886, + "grad_norm": 0.8028872445443147, + "learning_rate": 3.308387780261169e-05, + "loss": 0.2439, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1435239613056183, + "step": 1245, + "valid_targets_mean": 2922.1, + "valid_targets_min": 764 + }, + { + "epoch": 2.4323271665043817, + "grad_norm": 0.6764755919211342, + "learning_rate": 3.3010343315306106e-05, + "loss": 0.2336, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1212049052119255, + "step": 1250, + "valid_targets_mean": 4281.9, + "valid_targets_min": 1472 + }, + { + "epoch": 2.442064264849075, + "grad_norm": 0.8208423036570491, + "learning_rate": 3.293650265000864e-05, + "loss": 0.252, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12002964317798615, + "step": 1255, + "valid_targets_mean": 2410.8, + "valid_targets_min": 369 + }, + { + "epoch": 2.451801363193768, + "grad_norm": 0.8169817154882845, + "learning_rate": 3.286235754444335e-05, + "loss": 0.2395, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1182420626282692, + "step": 1260, + "valid_targets_mean": 2966.0, + "valid_targets_min": 1388 + }, + { + "epoch": 2.4615384615384617, + "grad_norm": 0.7317850822612334, + "learning_rate": 3.278790974349882e-05, + "loss": 0.2413, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08225102722644806, + "step": 1265, + "valid_targets_mean": 2597.4, + "valid_targets_min": 870 + }, + { + "epoch": 2.471275559883155, + "grad_norm": 0.688118158837201, + "learning_rate": 3.271316099918709e-05, + "loss": 0.2318, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1168079525232315, + "step": 1270, + "valid_targets_mean": 3529.0, + "valid_targets_min": 1027 + }, + { + "epoch": 2.481012658227848, + "grad_norm": 0.6621925099955391, + "learning_rate": 3.2638113070602436e-05, + "loss": 0.237, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10795827209949493, + "step": 1275, + "valid_targets_mean": 3523.0, + "valid_targets_min": 1461 + }, + { + "epoch": 2.4907497565725416, + "grad_norm": 0.7147418152885615, + "learning_rate": 3.256276772387997e-05, + "loss": 0.2367, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10587899386882782, + "step": 1280, + "valid_targets_mean": 3488.6, + "valid_targets_min": 406 + }, + { + "epoch": 2.5004868549172348, + "grad_norm": 0.5826637028854669, + "learning_rate": 3.248712673215405e-05, + "loss": 0.2395, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08619245886802673, + "step": 1285, + "valid_targets_mean": 3449.2, + "valid_targets_min": 347 + }, + { + "epoch": 2.510223953261928, + "grad_norm": 0.7131189592981495, + "learning_rate": 3.241119187551661e-05, + "loss": 0.2463, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11175684630870819, + "step": 1290, + "valid_targets_mean": 2802.9, + "valid_targets_min": 1079 + }, + { + "epoch": 2.519961051606621, + "grad_norm": 0.5797857681941478, + "learning_rate": 3.233496494097521e-05, + "loss": 0.2416, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1312026083469391, + "step": 1295, + "valid_targets_mean": 4597.2, + "valid_targets_min": 2042 + }, + { + "epoch": 2.5296981499513143, + "grad_norm": 0.7921243303267592, + "learning_rate": 3.2258447722411e-05, + "loss": 0.2465, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10868055373430252, + "step": 1300, + "valid_targets_mean": 2683.8, + "valid_targets_min": 716 + }, + { + "epoch": 2.539435248296008, + "grad_norm": 0.774825290380697, + "learning_rate": 3.2181642020536536e-05, + "loss": 0.2383, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12708991765975952, + "step": 1305, + "valid_targets_mean": 2726.1, + "valid_targets_min": 1300 + }, + { + "epoch": 2.549172346640701, + "grad_norm": 0.762900017416662, + "learning_rate": 3.210454964285333e-05, + "loss": 0.2257, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13067397475242615, + "step": 1310, + "valid_targets_mean": 3118.5, + "valid_targets_min": 1296 + }, + { + "epoch": 2.558909444985394, + "grad_norm": 0.7072153569488926, + "learning_rate": 3.202717240360941e-05, + "loss": 0.2169, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.135477215051651, + "step": 1315, + "valid_targets_mean": 4222.9, + "valid_targets_min": 683 + }, + { + "epoch": 2.568646543330088, + "grad_norm": 0.6623399885812072, + "learning_rate": 3.194951212375654e-05, + "loss": 0.2327, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10676338523626328, + "step": 1320, + "valid_targets_mean": 3114.4, + "valid_targets_min": 1354 + }, + { + "epoch": 2.578383641674781, + "grad_norm": 0.7492045989504769, + "learning_rate": 3.18715706309074e-05, + "loss": 0.2357, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13046498596668243, + "step": 1325, + "valid_targets_mean": 2841.6, + "valid_targets_min": 1208 + }, + { + "epoch": 2.588120740019474, + "grad_norm": 0.726912759253883, + "learning_rate": 3.17933497592926e-05, + "loss": 0.2502, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12865900993347168, + "step": 1330, + "valid_targets_mean": 3095.1, + "valid_targets_min": 879 + }, + { + "epoch": 2.5978578383641677, + "grad_norm": 1.1433374470740283, + "learning_rate": 3.171485134971748e-05, + "loss": 0.2534, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15513890981674194, + "step": 1335, + "valid_targets_mean": 2641.1, + "valid_targets_min": 1152 + }, + { + "epoch": 2.607594936708861, + "grad_norm": 0.7637229854220811, + "learning_rate": 3.163607724951877e-05, + "loss": 0.2361, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09301097691059113, + "step": 1340, + "valid_targets_mean": 1989.9, + "valid_targets_min": 1201 + }, + { + "epoch": 2.617332035053554, + "grad_norm": 0.8127208648335544, + "learning_rate": 3.155702931252119e-05, + "loss": 0.2381, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1383160650730133, + "step": 1345, + "valid_targets_mean": 3118.1, + "valid_targets_min": 832 + }, + { + "epoch": 2.627069133398247, + "grad_norm": 0.6903895599708771, + "learning_rate": 3.147770939899376e-05, + "loss": 0.2203, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07707555592060089, + "step": 1350, + "valid_targets_mean": 2399.9, + "valid_targets_min": 1426 + }, + { + "epoch": 2.6368062317429404, + "grad_norm": 0.7897696866609453, + "learning_rate": 3.1398119375606046e-05, + "loss": 0.2286, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11783762276172638, + "step": 1355, + "valid_targets_mean": 2069.9, + "valid_targets_min": 1264 + }, + { + "epoch": 2.646543330087634, + "grad_norm": 0.7747773221903235, + "learning_rate": 3.13182611153842e-05, + "loss": 0.2414, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13851211965084076, + "step": 1360, + "valid_targets_mean": 3161.2, + "valid_targets_min": 1458 + }, + { + "epoch": 2.656280428432327, + "grad_norm": 0.8478030732090569, + "learning_rate": 3.1238136497666944e-05, + "loss": 0.2344, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12053705751895905, + "step": 1365, + "valid_targets_mean": 3051.6, + "valid_targets_min": 947 + }, + { + "epoch": 2.6660175267770203, + "grad_norm": 0.6743401573821601, + "learning_rate": 3.115774740806128e-05, + "loss": 0.2388, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1084040105342865, + "step": 1370, + "valid_targets_mean": 3590.5, + "valid_targets_min": 1565 + }, + { + "epoch": 2.675754625121714, + "grad_norm": 0.8102210531374561, + "learning_rate": 3.1077095738398153e-05, + "loss": 0.2362, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10908074676990509, + "step": 1375, + "valid_targets_mean": 2245.8, + "valid_targets_min": 1055 + }, + { + "epoch": 2.685491723466407, + "grad_norm": 0.8612666953860721, + "learning_rate": 3.099618338668791e-05, + "loss": 0.2473, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08933654427528381, + "step": 1380, + "valid_targets_mean": 1772.0, + "valid_targets_min": 761 + }, + { + "epoch": 2.6952288218111002, + "grad_norm": 0.7294410115256623, + "learning_rate": 3.0915012257075635e-05, + "loss": 0.2214, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12363948673009872, + "step": 1385, + "valid_targets_mean": 3711.4, + "valid_targets_min": 1066 + }, + { + "epoch": 2.704965920155794, + "grad_norm": 0.7820610472085976, + "learning_rate": 3.083358425979637e-05, + "loss": 0.2472, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08485744148492813, + "step": 1390, + "valid_targets_mean": 1744.8, + "valid_targets_min": 942 + }, + { + "epoch": 2.714703018500487, + "grad_norm": 0.8743217316124318, + "learning_rate": 3.075190131113009e-05, + "loss": 0.2428, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12635160982608795, + "step": 1395, + "valid_targets_mean": 2463.0, + "valid_targets_min": 825 + }, + { + "epoch": 2.72444011684518, + "grad_norm": 0.7151099603336041, + "learning_rate": 3.0669965333356695e-05, + "loss": 0.2436, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0974293202161789, + "step": 1400, + "valid_targets_mean": 2545.4, + "valid_targets_min": 907 + }, + { + "epoch": 2.7341772151898733, + "grad_norm": 0.8709721011876739, + "learning_rate": 3.058777825471071e-05, + "loss": 0.2423, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0986693874001503, + "step": 1405, + "valid_targets_mean": 1826.6, + "valid_targets_min": 603 + }, + { + "epoch": 2.7439143135345665, + "grad_norm": 0.5683439455404877, + "learning_rate": 3.0505342009335898e-05, + "loss": 0.2221, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.04642200842499733, + "step": 1410, + "valid_targets_mean": 2629.6, + "valid_targets_min": 1201 + }, + { + "epoch": 2.75365141187926, + "grad_norm": 0.7939068734660043, + "learning_rate": 3.0422658537239823e-05, + "loss": 0.2509, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10494433343410492, + "step": 1415, + "valid_targets_mean": 2548.2, + "valid_targets_min": 1609 + }, + { + "epoch": 2.7633885102239533, + "grad_norm": 0.728119544785679, + "learning_rate": 3.0339729784248103e-05, + "loss": 0.2306, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10721099376678467, + "step": 1420, + "valid_targets_mean": 3043.4, + "valid_targets_min": 415 + }, + { + "epoch": 2.7731256085686464, + "grad_norm": 0.6664958468000649, + "learning_rate": 3.025655770195866e-05, + "loss": 0.246, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10482774674892426, + "step": 1425, + "valid_targets_mean": 3292.0, + "valid_targets_min": 828 + }, + { + "epoch": 2.78286270691334, + "grad_norm": 0.7866347340028299, + "learning_rate": 3.0173144247695796e-05, + "loss": 0.2365, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08502174913883209, + "step": 1430, + "valid_targets_mean": 2277.9, + "valid_targets_min": 1290 + }, + { + "epoch": 2.792599805258033, + "grad_norm": 0.7338395704422648, + "learning_rate": 3.008949138446413e-05, + "loss": 0.2161, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1111348569393158, + "step": 1435, + "valid_targets_mean": 2493.5, + "valid_targets_min": 1220 + }, + { + "epoch": 2.8023369036027264, + "grad_norm": 0.6653501894540796, + "learning_rate": 3.0005601080902376e-05, + "loss": 0.2319, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13429823517799377, + "step": 1440, + "valid_targets_mean": 4067.4, + "valid_targets_min": 726 + }, + { + "epoch": 2.8120740019474195, + "grad_norm": 0.6643398691845159, + "learning_rate": 2.9921475311237055e-05, + "loss": 0.2283, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1200399249792099, + "step": 1445, + "valid_targets_mean": 3554.1, + "valid_targets_min": 353 + }, + { + "epoch": 2.8218111002921127, + "grad_norm": 0.7001948444160422, + "learning_rate": 2.9837116055236e-05, + "loss": 0.2414, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07412829995155334, + "step": 1450, + "valid_targets_mean": 2387.1, + "valid_targets_min": 1514 + }, + { + "epoch": 2.8315481986368063, + "grad_norm": 0.8364387339858684, + "learning_rate": 2.975252529816178e-05, + "loss": 0.2334, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12055306136608124, + "step": 1455, + "valid_targets_mean": 2205.9, + "valid_targets_min": 490 + }, + { + "epoch": 2.8412852969814995, + "grad_norm": 0.8070899662966771, + "learning_rate": 2.9667705030724976e-05, + "loss": 0.2257, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11533784866333008, + "step": 1460, + "valid_targets_mean": 2544.1, + "valid_targets_min": 1081 + }, + { + "epoch": 2.8510223953261926, + "grad_norm": 0.7655668406337863, + "learning_rate": 2.9582657249037335e-05, + "loss": 0.225, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09084555506706238, + "step": 1465, + "valid_targets_mean": 2362.1, + "valid_targets_min": 1568 + }, + { + "epoch": 2.8607594936708862, + "grad_norm": 0.682547283492675, + "learning_rate": 2.9497383954564807e-05, + "loss": 0.2405, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12286686152219772, + "step": 1470, + "valid_targets_mean": 2804.1, + "valid_targets_min": 688 + }, + { + "epoch": 2.8704965920155794, + "grad_norm": 0.6403813646612935, + "learning_rate": 2.941188715408041e-05, + "loss": 0.2342, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10485698282718658, + "step": 1475, + "valid_targets_mean": 3492.6, + "valid_targets_min": 873 + }, + { + "epoch": 2.8802336903602725, + "grad_norm": 0.834361603882908, + "learning_rate": 2.9326168859617044e-05, + "loss": 0.2276, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09768801182508469, + "step": 1480, + "valid_targets_mean": 1849.0, + "valid_targets_min": 698 + }, + { + "epoch": 2.889970788704966, + "grad_norm": 0.8142513172358198, + "learning_rate": 2.9240231088420112e-05, + "loss": 0.2319, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16129133105278015, + "step": 1485, + "valid_targets_mean": 3350.1, + "valid_targets_min": 1870 + }, + { + "epoch": 2.8997078870496593, + "grad_norm": 0.6938599150985236, + "learning_rate": 2.9154075862900047e-05, + "loss": 0.2321, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1012616902589798, + "step": 1490, + "valid_targets_mean": 3434.6, + "valid_targets_min": 1459 + }, + { + "epoch": 2.9094449853943525, + "grad_norm": 0.8285559001767225, + "learning_rate": 2.9067705210584764e-05, + "loss": 0.2476, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14174655079841614, + "step": 1495, + "valid_targets_mean": 3295.6, + "valid_targets_min": 1011 + }, + { + "epoch": 2.9191820837390456, + "grad_norm": 0.8908683413393621, + "learning_rate": 2.898112116407186e-05, + "loss": 0.2307, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1453523486852646, + "step": 1500, + "valid_targets_mean": 2463.2, + "valid_targets_min": 676 + }, + { + "epoch": 2.928919182083739, + "grad_norm": 0.7010616312835074, + "learning_rate": 2.889432576098087e-05, + "loss": 0.2436, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1279774010181427, + "step": 1505, + "valid_targets_mean": 3304.6, + "valid_targets_min": 1697 + }, + { + "epoch": 2.9386562804284324, + "grad_norm": 0.6383882094213501, + "learning_rate": 2.8807321043905253e-05, + "loss": 0.2196, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0935613363981247, + "step": 1510, + "valid_targets_mean": 3910.6, + "valid_targets_min": 1412 + }, + { + "epoch": 2.9483933787731256, + "grad_norm": 0.6063463816177003, + "learning_rate": 2.872010906036435e-05, + "loss": 0.2466, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10311927646398544, + "step": 1515, + "valid_targets_mean": 3975.9, + "valid_targets_min": 1086 + }, + { + "epoch": 2.9581304771178187, + "grad_norm": 0.765783427016532, + "learning_rate": 2.863269186275519e-05, + "loss": 0.2383, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14790189266204834, + "step": 1520, + "valid_targets_mean": 2920.8, + "valid_targets_min": 1026 + }, + { + "epoch": 2.9678675754625123, + "grad_norm": 0.6912276152303378, + "learning_rate": 2.8545071508304193e-05, + "loss": 0.2358, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14214438199996948, + "step": 1525, + "valid_targets_mean": 3831.4, + "valid_targets_min": 1235 + }, + { + "epoch": 2.9776046738072055, + "grad_norm": 0.6832518900842646, + "learning_rate": 2.8457250059018765e-05, + "loss": 0.2333, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13680899143218994, + "step": 1530, + "valid_targets_mean": 3673.4, + "valid_targets_min": 1012 + }, + { + "epoch": 2.9873417721518987, + "grad_norm": 0.6389373131378439, + "learning_rate": 2.836922958163875e-05, + "loss": 0.2399, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09092959761619568, + "step": 1535, + "valid_targets_mean": 2189.8, + "valid_targets_min": 986 + }, + { + "epoch": 2.9970788704965923, + "grad_norm": 0.7702760592823965, + "learning_rate": 2.8281012147587816e-05, + "loss": 0.2403, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15108105540275574, + "step": 1540, + "valid_targets_mean": 3401.4, + "valid_targets_min": 1125 + }, + { + "epoch": 3.005842259006816, + "grad_norm": 0.7884417020648354, + "learning_rate": 2.8192599832924677e-05, + "loss": 0.2116, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11154694110155106, + "step": 1545, + "valid_targets_mean": 2741.4, + "valid_targets_min": 1141 + }, + { + "epoch": 3.015579357351509, + "grad_norm": 0.932462463859451, + "learning_rate": 2.810399471829429e-05, + "loss": 0.2033, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1208382397890091, + "step": 1550, + "valid_targets_mean": 2476.8, + "valid_targets_min": 790 + }, + { + "epoch": 3.0253164556962027, + "grad_norm": 0.8686517940140047, + "learning_rate": 2.8015198888878835e-05, + "loss": 0.2044, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08823367953300476, + "step": 1555, + "valid_targets_mean": 2029.2, + "valid_targets_min": 832 + }, + { + "epoch": 3.035053554040896, + "grad_norm": 0.7193788183596941, + "learning_rate": 2.7926214434348645e-05, + "loss": 0.2216, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11511079967021942, + "step": 1560, + "valid_targets_mean": 3413.9, + "valid_targets_min": 805 + }, + { + "epoch": 3.044790652385589, + "grad_norm": 0.8078546054093075, + "learning_rate": 2.78370434488131e-05, + "loss": 0.2148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07618220895528793, + "step": 1565, + "valid_targets_mean": 1835.6, + "valid_targets_min": 331 + }, + { + "epoch": 3.0545277507302826, + "grad_norm": 0.66507455067397, + "learning_rate": 2.7747688030771253e-05, + "loss": 0.2039, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0778750628232956, + "step": 1570, + "valid_targets_mean": 3440.4, + "valid_targets_min": 1193 + }, + { + "epoch": 3.0642648490749758, + "grad_norm": 0.9655649886690748, + "learning_rate": 2.765815028306249e-05, + "loss": 0.2026, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.130184605717659, + "step": 1575, + "valid_targets_mean": 2262.8, + "valid_targets_min": 889 + }, + { + "epoch": 3.074001947419669, + "grad_norm": 0.7563092839122303, + "learning_rate": 2.756843231281707e-05, + "loss": 0.2062, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11092890053987503, + "step": 1580, + "valid_targets_mean": 3115.4, + "valid_targets_min": 1238 + }, + { + "epoch": 3.083739045764362, + "grad_norm": 1.0075562492746872, + "learning_rate": 2.747853623140646e-05, + "loss": 0.2066, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11350446939468384, + "step": 1585, + "valid_targets_mean": 3914.6, + "valid_targets_min": 1350 + }, + { + "epoch": 3.0934761441090557, + "grad_norm": 0.8524230873224183, + "learning_rate": 2.7388464154393735e-05, + "loss": 0.2087, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11565576493740082, + "step": 1590, + "valid_targets_mean": 2432.1, + "valid_targets_min": 1141 + }, + { + "epoch": 3.103213242453749, + "grad_norm": 0.9501433370814879, + "learning_rate": 2.7298218201483725e-05, + "loss": 0.2075, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11040070652961731, + "step": 1595, + "valid_targets_mean": 2233.0, + "valid_targets_min": 1012 + }, + { + "epoch": 3.112950340798442, + "grad_norm": 0.7471389146890169, + "learning_rate": 2.7207800496473157e-05, + "loss": 0.2003, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08324768394231796, + "step": 1600, + "valid_targets_mean": 2868.4, + "valid_targets_min": 1412 + }, + { + "epoch": 3.122687439143135, + "grad_norm": 0.8740749170157414, + "learning_rate": 2.7117213167200696e-05, + "loss": 0.2042, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13211554288864136, + "step": 1605, + "valid_targets_mean": 2986.2, + "valid_targets_min": 1504 + }, + { + "epoch": 3.132424537487829, + "grad_norm": 0.7082769479304671, + "learning_rate": 2.7026458345496826e-05, + "loss": 0.2217, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08424953371286392, + "step": 1610, + "valid_targets_mean": 3818.8, + "valid_targets_min": 1278 + }, + { + "epoch": 3.142161635832522, + "grad_norm": 1.1956579863090437, + "learning_rate": 2.6935538167133702e-05, + "loss": 0.2121, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10671257227659225, + "step": 1615, + "valid_targets_mean": 2870.0, + "valid_targets_min": 874 + }, + { + "epoch": 3.151898734177215, + "grad_norm": 0.7418229133223673, + "learning_rate": 2.6844454771774898e-05, + "loss": 0.2083, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10029250383377075, + "step": 1620, + "valid_targets_mean": 3097.9, + "valid_targets_min": 879 + }, + { + "epoch": 3.1616358325219083, + "grad_norm": 0.834880643912381, + "learning_rate": 2.6753210302925044e-05, + "loss": 0.201, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11291643977165222, + "step": 1625, + "valid_targets_mean": 2799.5, + "valid_targets_min": 760 + }, + { + "epoch": 3.171372930866602, + "grad_norm": 0.8248332557311356, + "learning_rate": 2.6661806907879385e-05, + "loss": 0.2071, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06772299110889435, + "step": 1630, + "valid_targets_mean": 1523.8, + "valid_targets_min": 784 + }, + { + "epoch": 3.181110029211295, + "grad_norm": 0.850479033976409, + "learning_rate": 2.6570246737673216e-05, + "loss": 0.2029, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11443410068750381, + "step": 1635, + "valid_targets_mean": 2366.2, + "valid_targets_min": 996 + }, + { + "epoch": 3.190847127555988, + "grad_norm": 0.8214978220609186, + "learning_rate": 2.6478531947031325e-05, + "loss": 0.2102, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08211231231689453, + "step": 1640, + "valid_targets_mean": 2102.9, + "valid_targets_min": 848 + }, + { + "epoch": 3.2005842259006814, + "grad_norm": 0.7642459302531467, + "learning_rate": 2.638666469431723e-05, + "loss": 0.2055, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09310499578714371, + "step": 1645, + "valid_targets_mean": 2598.9, + "valid_targets_min": 996 + }, + { + "epoch": 3.210321324245375, + "grad_norm": 0.7279570609997414, + "learning_rate": 2.6294647141482413e-05, + "loss": 0.2032, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10722191631793976, + "step": 1650, + "valid_targets_mean": 3704.6, + "valid_targets_min": 1141 + }, + { + "epoch": 3.220058422590068, + "grad_norm": 0.6325703321144482, + "learning_rate": 2.6202481454015426e-05, + "loss": 0.1842, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08083665370941162, + "step": 1655, + "valid_targets_mean": 2974.0, + "valid_targets_min": 1032 + }, + { + "epoch": 3.2297955209347613, + "grad_norm": 0.8206360624533692, + "learning_rate": 2.6110169800890937e-05, + "loss": 0.2165, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10948295146226883, + "step": 1660, + "valid_targets_mean": 2686.6, + "valid_targets_min": 1341 + }, + { + "epoch": 3.239532619279455, + "grad_norm": 0.6636697940759875, + "learning_rate": 2.6017714354518696e-05, + "loss": 0.2061, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11225951462984085, + "step": 1665, + "valid_targets_mean": 4094.9, + "valid_targets_min": 1625 + }, + { + "epoch": 3.249269717624148, + "grad_norm": 0.6399187741843992, + "learning_rate": 2.592511729069239e-05, + "loss": 0.1873, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11166227608919144, + "step": 1670, + "valid_targets_mean": 4579.6, + "valid_targets_min": 888 + }, + { + "epoch": 3.2590068159688412, + "grad_norm": 1.0378728175418352, + "learning_rate": 2.5832380788538457e-05, + "loss": 0.2108, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1347440779209137, + "step": 1675, + "valid_targets_mean": 4024.6, + "valid_targets_min": 930 + }, + { + "epoch": 3.2687439143135344, + "grad_norm": 0.7778994020586357, + "learning_rate": 2.573950703046479e-05, + "loss": 0.2054, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09186933189630508, + "step": 1680, + "valid_targets_mean": 2595.4, + "valid_targets_min": 1197 + }, + { + "epoch": 3.278481012658228, + "grad_norm": 0.6882678466561416, + "learning_rate": 2.564649820210939e-05, + "loss": 0.2147, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09283961355686188, + "step": 1685, + "valid_targets_mean": 3646.6, + "valid_targets_min": 1005 + }, + { + "epoch": 3.288218111002921, + "grad_norm": 0.8323723770495097, + "learning_rate": 2.5553356492288933e-05, + "loss": 0.208, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14615213871002197, + "step": 1690, + "valid_targets_mean": 3665.2, + "valid_targets_min": 828 + }, + { + "epoch": 3.2979552093476143, + "grad_norm": 0.8864387309698119, + "learning_rate": 2.5460084092947233e-05, + "loss": 0.191, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10318636894226074, + "step": 1695, + "valid_targets_mean": 2623.8, + "valid_targets_min": 290 + }, + { + "epoch": 3.3076923076923075, + "grad_norm": 0.8413830842891964, + "learning_rate": 2.5366683199103692e-05, + "loss": 0.1957, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11743028461933136, + "step": 1700, + "valid_targets_mean": 2832.2, + "valid_targets_min": 1831 + }, + { + "epoch": 3.317429406037001, + "grad_norm": 0.8730154800545424, + "learning_rate": 2.5273156008801614e-05, + "loss": 0.2008, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09757040441036224, + "step": 1705, + "valid_targets_mean": 1879.2, + "valid_targets_min": 562 + }, + { + "epoch": 3.3271665043816943, + "grad_norm": 0.6255478267813253, + "learning_rate": 2.517950472305651e-05, + "loss": 0.1971, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08264683187007904, + "step": 1710, + "valid_targets_mean": 4320.8, + "valid_targets_min": 1713 + }, + { + "epoch": 3.3369036027263874, + "grad_norm": 0.7830805843533527, + "learning_rate": 2.508573154580427e-05, + "loss": 0.206, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08903951942920685, + "step": 1715, + "valid_targets_mean": 2675.0, + "valid_targets_min": 1307 + }, + { + "epoch": 3.346640701071081, + "grad_norm": 0.7058135583859145, + "learning_rate": 2.4991838683849306e-05, + "loss": 0.2081, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16063743829727173, + "step": 1720, + "valid_targets_mean": 4538.2, + "valid_targets_min": 1549 + }, + { + "epoch": 3.356377799415774, + "grad_norm": 0.8558203857749066, + "learning_rate": 2.4897828346812637e-05, + "loss": 0.2063, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08233465254306793, + "step": 1725, + "valid_targets_mean": 2467.9, + "valid_targets_min": 1055 + }, + { + "epoch": 3.3661148977604674, + "grad_norm": 0.8346203949558786, + "learning_rate": 2.4803702747079858e-05, + "loss": 0.2123, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09843271225690842, + "step": 1730, + "valid_targets_mean": 2434.5, + "valid_targets_min": 579 + }, + { + "epoch": 3.3758519961051605, + "grad_norm": 0.6408101681065131, + "learning_rate": 2.4709464099749093e-05, + "loss": 0.1934, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11571668833494186, + "step": 1735, + "valid_targets_mean": 4195.4, + "valid_targets_min": 627 + }, + { + "epoch": 3.385589094449854, + "grad_norm": 0.7510417907281103, + "learning_rate": 2.4615114622578875e-05, + "loss": 0.2036, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08446644991636276, + "step": 1740, + "valid_targets_mean": 3063.1, + "valid_targets_min": 896 + }, + { + "epoch": 3.3953261927945473, + "grad_norm": 0.6380803714432992, + "learning_rate": 2.4520656535935927e-05, + "loss": 0.2045, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06864303350448608, + "step": 1745, + "valid_targets_mean": 2743.0, + "valid_targets_min": 1665 + }, + { + "epoch": 3.4050632911392404, + "grad_norm": 0.7549085112117736, + "learning_rate": 2.4426092062742933e-05, + "loss": 0.1952, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10727421194314957, + "step": 1750, + "valid_targets_mean": 2954.8, + "valid_targets_min": 1493 + }, + { + "epoch": 3.4148003894839336, + "grad_norm": 0.9668943886987618, + "learning_rate": 2.4331423428426227e-05, + "loss": 0.1975, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12314222753047943, + "step": 1755, + "valid_targets_mean": 2293.2, + "valid_targets_min": 1388 + }, + { + "epoch": 3.424537487828627, + "grad_norm": 0.7866864607213474, + "learning_rate": 2.4236652860863396e-05, + "loss": 0.2202, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10131599009037018, + "step": 1760, + "valid_targets_mean": 2515.0, + "valid_targets_min": 815 + }, + { + "epoch": 3.4342745861733204, + "grad_norm": 0.6962218941955951, + "learning_rate": 2.414178259033087e-05, + "loss": 0.2102, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09250883758068085, + "step": 1765, + "valid_targets_mean": 2914.4, + "valid_targets_min": 1009 + }, + { + "epoch": 3.4440116845180135, + "grad_norm": 0.8444040237557682, + "learning_rate": 2.404681484945143e-05, + "loss": 0.2097, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09887482225894928, + "step": 1770, + "valid_targets_mean": 2721.4, + "valid_targets_min": 1022 + }, + { + "epoch": 3.453748782862707, + "grad_norm": 0.6780117367085235, + "learning_rate": 2.3951751873141683e-05, + "loss": 0.1976, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08050332963466644, + "step": 1775, + "valid_targets_mean": 3568.2, + "valid_targets_min": 1270 + }, + { + "epoch": 3.4634858812074003, + "grad_norm": 0.8098629508625784, + "learning_rate": 2.385659589855944e-05, + "loss": 0.2127, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1392337828874588, + "step": 1780, + "valid_targets_mean": 4161.6, + "valid_targets_min": 1532 + }, + { + "epoch": 3.4732229795520935, + "grad_norm": 0.6886434383200422, + "learning_rate": 2.3761349165051075e-05, + "loss": 0.1887, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10259492695331573, + "step": 1785, + "valid_targets_mean": 3790.5, + "valid_targets_min": 1148 + }, + { + "epoch": 3.4829600778967866, + "grad_norm": 0.7575123339265709, + "learning_rate": 2.366601391409884e-05, + "loss": 0.2166, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10067093372344971, + "step": 1790, + "valid_targets_mean": 2686.6, + "valid_targets_min": 1520 + }, + { + "epoch": 3.49269717624148, + "grad_norm": 0.806811733165281, + "learning_rate": 2.35705923892681e-05, + "loss": 0.2002, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09538982808589935, + "step": 1795, + "valid_targets_mean": 2691.8, + "valid_targets_min": 1233 + }, + { + "epoch": 3.5024342745861734, + "grad_norm": 0.7246311962378337, + "learning_rate": 2.3475086836154547e-05, + "loss": 0.2096, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0840829387307167, + "step": 1800, + "valid_targets_mean": 2930.0, + "valid_targets_min": 1572 + }, + { + "epoch": 3.5121713729308666, + "grad_norm": 0.6538910297866242, + "learning_rate": 2.3379499502331347e-05, + "loss": 0.1772, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10728005319833755, + "step": 1805, + "valid_targets_mean": 3528.4, + "valid_targets_min": 1381 + }, + { + "epoch": 3.5219084712755597, + "grad_norm": 0.7855769486272002, + "learning_rate": 2.328383263729624e-05, + "loss": 0.1856, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09734708815813065, + "step": 1810, + "valid_targets_mean": 2871.0, + "valid_targets_min": 1470 + }, + { + "epoch": 3.5316455696202533, + "grad_norm": 0.8783046030095799, + "learning_rate": 2.318808849241861e-05, + "loss": 0.2185, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11639495193958282, + "step": 1815, + "valid_targets_mean": 2588.6, + "valid_targets_min": 1005 + }, + { + "epoch": 3.5413826679649465, + "grad_norm": 0.7676279432819073, + "learning_rate": 2.3092269320886497e-05, + "loss": 0.2051, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08277629315853119, + "step": 1820, + "valid_targets_mean": 2333.9, + "valid_targets_min": 1259 + }, + { + "epoch": 3.5511197663096397, + "grad_norm": 0.7720477396893514, + "learning_rate": 2.299637737765358e-05, + "loss": 0.2096, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11269618570804596, + "step": 1825, + "valid_targets_mean": 3290.0, + "valid_targets_min": 1317 + }, + { + "epoch": 3.5608568646543333, + "grad_norm": 0.6419534081897562, + "learning_rate": 2.290041491938612e-05, + "loss": 0.1907, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06976529955863953, + "step": 1830, + "valid_targets_mean": 2259.8, + "valid_targets_min": 324 + }, + { + "epoch": 3.5705939629990264, + "grad_norm": 3.0439876407749433, + "learning_rate": 2.280438420440981e-05, + "loss": 0.2121, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1164056658744812, + "step": 1835, + "valid_targets_mean": 4026.6, + "valid_targets_min": 1391 + }, + { + "epoch": 3.5803310613437196, + "grad_norm": 0.8720642252712207, + "learning_rate": 2.2708287492656675e-05, + "loss": 0.2186, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12395752966403961, + "step": 1840, + "valid_targets_mean": 2965.1, + "valid_targets_min": 875 + }, + { + "epoch": 3.5900681596884128, + "grad_norm": 0.7036801924845678, + "learning_rate": 2.2612127045611867e-05, + "loss": 0.2088, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09772051870822906, + "step": 1845, + "valid_targets_mean": 2779.5, + "valid_targets_min": 953 + }, + { + "epoch": 3.599805258033106, + "grad_norm": 0.8487045108395053, + "learning_rate": 2.2515905126260455e-05, + "loss": 0.2022, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1014087051153183, + "step": 1850, + "valid_targets_mean": 2140.2, + "valid_targets_min": 546 + }, + { + "epoch": 3.6095423563777995, + "grad_norm": 0.8643947332086274, + "learning_rate": 2.2419623999034146e-05, + "loss": 0.2092, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10153505206108093, + "step": 1855, + "valid_targets_mean": 2834.2, + "valid_targets_min": 1513 + }, + { + "epoch": 3.6192794547224927, + "grad_norm": 0.833500456749334, + "learning_rate": 2.232328592975802e-05, + "loss": 0.1893, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08012451976537704, + "step": 1860, + "valid_targets_mean": 2420.8, + "valid_targets_min": 1175 + }, + { + "epoch": 3.629016553067186, + "grad_norm": 0.7209308141157239, + "learning_rate": 2.2226893185597204e-05, + "loss": 0.1946, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08700944483280182, + "step": 1865, + "valid_targets_mean": 2764.1, + "valid_targets_min": 1449 + }, + { + "epoch": 3.6387536514118795, + "grad_norm": 0.7333785716703014, + "learning_rate": 2.213044803500351e-05, + "loss": 0.196, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10017083585262299, + "step": 1870, + "valid_targets_mean": 3329.2, + "valid_targets_min": 1173 + }, + { + "epoch": 3.6484907497565726, + "grad_norm": 0.6004357753526836, + "learning_rate": 2.2033952747662052e-05, + "loss": 0.2029, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.04709317535161972, + "step": 1875, + "valid_targets_mean": 2364.6, + "valid_targets_min": 1576 + }, + { + "epoch": 3.6582278481012658, + "grad_norm": 0.6759321598294998, + "learning_rate": 2.193740959443782e-05, + "loss": 0.1914, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05771394073963165, + "step": 1880, + "valid_targets_mean": 2007.0, + "valid_targets_min": 942 + }, + { + "epoch": 3.667964946445959, + "grad_norm": 0.8065156038990903, + "learning_rate": 2.1840820847322274e-05, + "loss": 0.2055, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0800083726644516, + "step": 1885, + "valid_targets_mean": 2349.8, + "valid_targets_min": 1212 + }, + { + "epoch": 3.6777020447906525, + "grad_norm": 0.8633743003683529, + "learning_rate": 2.1744188779379844e-05, + "loss": 0.2075, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09043803811073303, + "step": 1890, + "valid_targets_mean": 1872.9, + "valid_targets_min": 741 + }, + { + "epoch": 3.6874391431353457, + "grad_norm": 0.9285052612789028, + "learning_rate": 2.1647515664694455e-05, + "loss": 0.2103, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09387189149856567, + "step": 1895, + "valid_targets_mean": 1954.0, + "valid_targets_min": 1135 + }, + { + "epoch": 3.697176241480039, + "grad_norm": 0.765084208446341, + "learning_rate": 2.155080377831599e-05, + "loss": 0.1933, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1057450920343399, + "step": 1900, + "valid_targets_mean": 2513.1, + "valid_targets_min": 947 + }, + { + "epoch": 3.706913339824732, + "grad_norm": 0.773037410109213, + "learning_rate": 2.145405539620677e-05, + "loss": 0.2161, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12816593050956726, + "step": 1905, + "valid_targets_mean": 4268.6, + "valid_targets_min": 1579 + }, + { + "epoch": 3.7166504381694256, + "grad_norm": 0.7251729159914246, + "learning_rate": 2.135727279518798e-05, + "loss": 0.2129, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08925525844097137, + "step": 1910, + "valid_targets_mean": 2894.1, + "valid_targets_min": 1667 + }, + { + "epoch": 3.726387536514119, + "grad_norm": 0.8781856514746874, + "learning_rate": 2.1260458252886117e-05, + "loss": 0.2094, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0762854665517807, + "step": 1915, + "valid_targets_mean": 1900.8, + "valid_targets_min": 514 + }, + { + "epoch": 3.736124634858812, + "grad_norm": 0.7861955365536748, + "learning_rate": 2.1163614047679322e-05, + "loss": 0.2045, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12471576035022736, + "step": 1920, + "valid_targets_mean": 3699.2, + "valid_targets_min": 1566 + }, + { + "epoch": 3.7458617332035056, + "grad_norm": 0.6865624618114795, + "learning_rate": 2.1066742458643853e-05, + "loss": 0.2036, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08267700672149658, + "step": 1925, + "valid_targets_mean": 3012.9, + "valid_targets_min": 936 + }, + { + "epoch": 3.7555988315481987, + "grad_norm": 1.0641850208530184, + "learning_rate": 2.096984576550037e-05, + "loss": 0.2101, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13514497876167297, + "step": 1930, + "valid_targets_mean": 2025.6, + "valid_targets_min": 969 + }, + { + "epoch": 3.765335929892892, + "grad_norm": 0.7327848270033929, + "learning_rate": 2.087292624856033e-05, + "loss": 0.1981, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07251231372356415, + "step": 1935, + "valid_targets_mean": 1982.9, + "valid_targets_min": 610 + }, + { + "epoch": 3.775073028237585, + "grad_norm": 0.8591447330742223, + "learning_rate": 2.0775986188672324e-05, + "loss": 0.2074, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10901618003845215, + "step": 1940, + "valid_targets_mean": 2344.4, + "valid_targets_min": 1603 + }, + { + "epoch": 3.7848101265822782, + "grad_norm": 0.6459578970389677, + "learning_rate": 2.067902786716836e-05, + "loss": 0.1962, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09305711090564728, + "step": 1945, + "valid_targets_mean": 3765.6, + "valid_targets_min": 1404 + }, + { + "epoch": 3.794547224926972, + "grad_norm": 0.7623980367866929, + "learning_rate": 2.0582053565810242e-05, + "loss": 0.2023, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1069234162569046, + "step": 1950, + "valid_targets_mean": 3184.4, + "valid_targets_min": 1089 + }, + { + "epoch": 3.804284323271665, + "grad_norm": 0.6865944868768693, + "learning_rate": 2.0485065566735795e-05, + "loss": 0.1887, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0835491418838501, + "step": 1955, + "valid_targets_mean": 2863.6, + "valid_targets_min": 1303 + }, + { + "epoch": 3.814021421616358, + "grad_norm": 0.8714331071602269, + "learning_rate": 2.038806615240523e-05, + "loss": 0.2155, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12258565425872803, + "step": 1960, + "valid_targets_mean": 2687.0, + "valid_targets_min": 820 + }, + { + "epoch": 3.8237585199610518, + "grad_norm": 0.6495564993161029, + "learning_rate": 2.0291057605547393e-05, + "loss": 0.1964, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07098618894815445, + "step": 1965, + "valid_targets_mean": 3116.6, + "valid_targets_min": 1017 + }, + { + "epoch": 3.833495618305745, + "grad_norm": 1.3323923060061573, + "learning_rate": 2.0194042209106023e-05, + "loss": 0.2138, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12420941144227982, + "step": 1970, + "valid_targets_mean": 2603.6, + "valid_targets_min": 1056 + }, + { + "epoch": 3.843232716650438, + "grad_norm": 0.7575990456041161, + "learning_rate": 2.0097022246186092e-05, + "loss": 0.1954, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09626314043998718, + "step": 1975, + "valid_targets_mean": 2141.5, + "valid_targets_min": 764 + }, + { + "epoch": 3.8529698149951317, + "grad_norm": 0.7947575483779744, + "learning_rate": 2e-05, + "loss": 0.2095, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09834505617618561, + "step": 1980, + "valid_targets_mean": 3005.2, + "valid_targets_min": 1381 + }, + { + "epoch": 3.862706913339825, + "grad_norm": 0.7163570110860416, + "learning_rate": 1.9902977753813918e-05, + "loss": 0.2051, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10482780635356903, + "step": 1985, + "valid_targets_mean": 3228.2, + "valid_targets_min": 903 + }, + { + "epoch": 3.872444011684518, + "grad_norm": 0.6547426313348985, + "learning_rate": 1.980595779089398e-05, + "loss": 0.2073, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08434221148490906, + "step": 1990, + "valid_targets_mean": 2802.6, + "valid_targets_min": 862 + }, + { + "epoch": 3.882181110029211, + "grad_norm": 0.8338109259355849, + "learning_rate": 1.9708942394452613e-05, + "loss": 0.2031, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09361699223518372, + "step": 1995, + "valid_targets_mean": 2393.4, + "valid_targets_min": 1246 + }, + { + "epoch": 3.8919182083739043, + "grad_norm": 0.7966795271064488, + "learning_rate": 1.9611933847594775e-05, + "loss": 0.1919, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11265958845615387, + "step": 2000, + "valid_targets_mean": 2988.9, + "valid_targets_min": 1391 + }, + { + "epoch": 3.901655306718598, + "grad_norm": 0.717835201092483, + "learning_rate": 1.951493443326421e-05, + "loss": 0.1994, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09257732331752777, + "step": 2005, + "valid_targets_mean": 3231.4, + "valid_targets_min": 1164 + }, + { + "epoch": 3.911392405063291, + "grad_norm": 0.7700485323990515, + "learning_rate": 1.9417946434189768e-05, + "loss": 0.2083, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10518110543489456, + "step": 2010, + "valid_targets_mean": 3154.5, + "valid_targets_min": 565 + }, + { + "epoch": 3.9211295034079843, + "grad_norm": 0.8320741111362749, + "learning_rate": 1.9320972132831643e-05, + "loss": 0.2127, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10657656937837601, + "step": 2015, + "valid_targets_mean": 2612.1, + "valid_targets_min": 1660 + }, + { + "epoch": 3.930866601752678, + "grad_norm": 0.7083370381604185, + "learning_rate": 1.9224013811327686e-05, + "loss": 0.1982, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08739273250102997, + "step": 2020, + "valid_targets_mean": 3379.9, + "valid_targets_min": 1267 + }, + { + "epoch": 3.940603700097371, + "grad_norm": 0.8258382734735102, + "learning_rate": 1.912707375143967e-05, + "loss": 0.1968, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10398921370506287, + "step": 2025, + "valid_targets_mean": 3194.2, + "valid_targets_min": 441 + }, + { + "epoch": 3.950340798442064, + "grad_norm": 1.0294293750196222, + "learning_rate": 1.9030154234499635e-05, + "loss": 0.1988, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07358142733573914, + "step": 2030, + "valid_targets_mean": 1779.0, + "valid_targets_min": 727 + }, + { + "epoch": 3.960077896786758, + "grad_norm": 0.6679677522098728, + "learning_rate": 1.8933257541356154e-05, + "loss": 0.2011, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1221453845500946, + "step": 2035, + "valid_targets_mean": 3645.1, + "valid_targets_min": 1627 + }, + { + "epoch": 3.969814995131451, + "grad_norm": 0.7182017298628506, + "learning_rate": 1.8836385952320685e-05, + "loss": 0.2216, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07548783719539642, + "step": 2040, + "valid_targets_mean": 2344.6, + "valid_targets_min": 1120 + }, + { + "epoch": 3.979552093476144, + "grad_norm": 0.7840004882557964, + "learning_rate": 1.873954174711389e-05, + "loss": 0.1897, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12413585186004639, + "step": 2045, + "valid_targets_mean": 3969.9, + "valid_targets_min": 2425 + }, + { + "epoch": 3.9892891918208373, + "grad_norm": 0.7584589071211272, + "learning_rate": 1.8642727204812025e-05, + "loss": 0.1966, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10364937037229538, + "step": 2050, + "valid_targets_mean": 3282.2, + "valid_targets_min": 1595 + }, + { + "epoch": 3.9990262901655305, + "grad_norm": 0.7077299563219889, + "learning_rate": 1.854594460379324e-05, + "loss": 0.1898, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09577010571956635, + "step": 2055, + "valid_targets_mean": 4010.4, + "valid_targets_min": 1128 + }, + { + "epoch": 4.0077896786757545, + "grad_norm": 0.6714162811777626, + "learning_rate": 1.8449196221684015e-05, + "loss": 0.1806, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07479497790336609, + "step": 2060, + "valid_targets_mean": 3739.5, + "valid_targets_min": 1494 + }, + { + "epoch": 4.017526777020448, + "grad_norm": 0.8151018805872865, + "learning_rate": 1.835248433530555e-05, + "loss": 0.1745, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07561913132667542, + "step": 2065, + "valid_targets_mean": 2534.4, + "valid_targets_min": 1118 + }, + { + "epoch": 4.027263875365141, + "grad_norm": 0.6788404341948994, + "learning_rate": 1.825581122062016e-05, + "loss": 0.1803, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08429465442895889, + "step": 2070, + "valid_targets_mean": 4065.4, + "valid_targets_min": 317 + }, + { + "epoch": 4.0370009737098345, + "grad_norm": 0.7313335639788372, + "learning_rate": 1.815917915267773e-05, + "loss": 0.1797, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07114744186401367, + "step": 2075, + "valid_targets_mean": 3094.9, + "valid_targets_min": 1534 + }, + { + "epoch": 4.046738072054528, + "grad_norm": 0.7934180657619316, + "learning_rate": 1.8062590405562185e-05, + "loss": 0.175, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09695077687501907, + "step": 2080, + "valid_targets_mean": 3541.2, + "valid_targets_min": 1576 + }, + { + "epoch": 4.056475170399221, + "grad_norm": 0.94573730100302, + "learning_rate": 1.7966047252337958e-05, + "loss": 0.1788, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10533531755208969, + "step": 2085, + "valid_targets_mean": 2514.2, + "valid_targets_min": 956 + }, + { + "epoch": 4.066212268743914, + "grad_norm": 0.7362176811630884, + "learning_rate": 1.78695519649965e-05, + "loss": 0.1825, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07893982529640198, + "step": 2090, + "valid_targets_mean": 3586.1, + "valid_targets_min": 989 + }, + { + "epoch": 4.075949367088608, + "grad_norm": 0.7981259595246765, + "learning_rate": 1.7773106814402796e-05, + "loss": 0.1801, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08646465092897415, + "step": 2095, + "valid_targets_mean": 3420.0, + "valid_targets_min": 1093 + }, + { + "epoch": 4.085686465433301, + "grad_norm": 0.725014254374623, + "learning_rate": 1.7676714070241987e-05, + "loss": 0.1579, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07800125330686569, + "step": 2100, + "valid_targets_mean": 3832.5, + "valid_targets_min": 963 + }, + { + "epoch": 4.095423563777994, + "grad_norm": 0.7915625107935261, + "learning_rate": 1.7580376000965868e-05, + "loss": 0.1745, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09259450435638428, + "step": 2105, + "valid_targets_mean": 3586.2, + "valid_targets_min": 1209 + }, + { + "epoch": 4.105160662122687, + "grad_norm": 0.6952194628686748, + "learning_rate": 1.748409487373955e-05, + "loss": 0.1724, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09008082747459412, + "step": 2110, + "valid_targets_mean": 3660.9, + "valid_targets_min": 1594 + }, + { + "epoch": 4.114897760467381, + "grad_norm": 0.7269176114716939, + "learning_rate": 1.7387872954388136e-05, + "loss": 0.1797, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10437388718128204, + "step": 2115, + "valid_targets_mean": 4212.4, + "valid_targets_min": 745 + }, + { + "epoch": 4.124634858812074, + "grad_norm": 0.8316587832578521, + "learning_rate": 1.729171250734333e-05, + "loss": 0.1781, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07686236500740051, + "step": 2120, + "valid_targets_mean": 2780.6, + "valid_targets_min": 2031 + }, + { + "epoch": 4.134371957156767, + "grad_norm": 0.854473682482018, + "learning_rate": 1.71956157955902e-05, + "loss": 0.1902, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07336435467004776, + "step": 2125, + "valid_targets_mean": 2311.0, + "valid_targets_min": 726 + }, + { + "epoch": 4.144109055501461, + "grad_norm": 0.929705165504535, + "learning_rate": 1.7099585080613884e-05, + "loss": 0.1846, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10245804488658905, + "step": 2130, + "valid_targets_mean": 2843.9, + "valid_targets_min": 803 + }, + { + "epoch": 4.153846153846154, + "grad_norm": 0.9340385662936084, + "learning_rate": 1.7003622622346424e-05, + "loss": 0.166, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10968220233917236, + "step": 2135, + "valid_targets_mean": 2766.9, + "valid_targets_min": 1151 + }, + { + "epoch": 4.163583252190847, + "grad_norm": 0.7442177721335224, + "learning_rate": 1.6907730679113513e-05, + "loss": 0.1665, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0798567458987236, + "step": 2140, + "valid_targets_mean": 2954.2, + "valid_targets_min": 1231 + }, + { + "epoch": 4.1733203505355405, + "grad_norm": 0.8795685109074634, + "learning_rate": 1.6811911507581394e-05, + "loss": 0.1691, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07578116655349731, + "step": 2145, + "valid_targets_mean": 2428.5, + "valid_targets_min": 734 + }, + { + "epoch": 4.183057448880234, + "grad_norm": 0.6600606559095514, + "learning_rate": 1.6716167362703764e-05, + "loss": 0.1743, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07112875580787659, + "step": 2150, + "valid_targets_mean": 2903.2, + "valid_targets_min": 994 + }, + { + "epoch": 4.192794547224927, + "grad_norm": 0.8807610133387657, + "learning_rate": 1.6620500497668656e-05, + "loss": 0.1895, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06985434889793396, + "step": 2155, + "valid_targets_mean": 2146.6, + "valid_targets_min": 1199 + }, + { + "epoch": 4.2025316455696204, + "grad_norm": 0.7491123700968002, + "learning_rate": 1.6524913163845456e-05, + "loss": 0.1656, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10740687698125839, + "step": 2160, + "valid_targets_mean": 3712.4, + "valid_targets_min": 1252 + }, + { + "epoch": 4.212268743914313, + "grad_norm": 0.8696828626973055, + "learning_rate": 1.6429407610731902e-05, + "loss": 0.1892, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11889979243278503, + "step": 2165, + "valid_targets_mean": 3022.0, + "valid_targets_min": 1606 + }, + { + "epoch": 4.222005842259007, + "grad_norm": 0.9981169218888921, + "learning_rate": 1.6333986085901167e-05, + "loss": 0.1775, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08759283274412155, + "step": 2170, + "valid_targets_mean": 1711.8, + "valid_targets_min": 727 + }, + { + "epoch": 4.2317429406037, + "grad_norm": 0.793574430063338, + "learning_rate": 1.6238650834948935e-05, + "loss": 0.1723, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08245506882667542, + "step": 2175, + "valid_targets_mean": 2374.1, + "valid_targets_min": 1013 + }, + { + "epoch": 4.241480038948393, + "grad_norm": 0.851176631900123, + "learning_rate": 1.614340410144056e-05, + "loss": 0.1879, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11328478157520294, + "step": 2180, + "valid_targets_mean": 3784.9, + "valid_targets_min": 1349 + }, + { + "epoch": 4.251217137293087, + "grad_norm": 0.8926459640954965, + "learning_rate": 1.604824812685832e-05, + "loss": 0.1729, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13632264733314514, + "step": 2185, + "valid_targets_mean": 3229.4, + "valid_targets_min": 1558 + }, + { + "epoch": 4.26095423563778, + "grad_norm": 0.8828024040800031, + "learning_rate": 1.5953185150548574e-05, + "loss": 0.1829, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11008234322071075, + "step": 2190, + "valid_targets_mean": 3296.6, + "valid_targets_min": 1173 + }, + { + "epoch": 4.270691333982473, + "grad_norm": 0.6219157530963444, + "learning_rate": 1.5858217409669138e-05, + "loss": 0.1753, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09567482769489288, + "step": 2195, + "valid_targets_mean": 4047.8, + "valid_targets_min": 1394 + }, + { + "epoch": 4.280428432327167, + "grad_norm": 0.5984832035464692, + "learning_rate": 1.576334713913661e-05, + "loss": 0.1718, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08540554344654083, + "step": 2200, + "valid_targets_mean": 4458.5, + "valid_targets_min": 1064 + }, + { + "epoch": 4.29016553067186, + "grad_norm": 1.0053890597595496, + "learning_rate": 1.566857657157378e-05, + "loss": 0.1777, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10111808776855469, + "step": 2205, + "valid_targets_mean": 2122.9, + "valid_targets_min": 905 + }, + { + "epoch": 4.299902629016553, + "grad_norm": 0.7592102668598462, + "learning_rate": 1.557390793725707e-05, + "loss": 0.1692, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09460680186748505, + "step": 2210, + "valid_targets_mean": 3360.1, + "valid_targets_min": 903 + }, + { + "epoch": 4.309639727361247, + "grad_norm": 0.8125751626562816, + "learning_rate": 1.5479343464064077e-05, + "loss": 0.1867, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08566434681415558, + "step": 2215, + "valid_targets_mean": 2385.2, + "valid_targets_min": 992 + }, + { + "epoch": 4.319376825705939, + "grad_norm": 0.9700126039051541, + "learning_rate": 1.5384885377421132e-05, + "loss": 0.1711, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0993308424949646, + "step": 2220, + "valid_targets_mean": 2004.8, + "valid_targets_min": 782 + }, + { + "epoch": 4.329113924050633, + "grad_norm": 0.8320811442092875, + "learning_rate": 1.5290535900250914e-05, + "loss": 0.176, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08704739809036255, + "step": 2225, + "valid_targets_mean": 2471.2, + "valid_targets_min": 706 + }, + { + "epoch": 4.3388510223953265, + "grad_norm": 0.8370133721221221, + "learning_rate": 1.5196297252920142e-05, + "loss": 0.1898, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07865718752145767, + "step": 2230, + "valid_targets_mean": 1931.5, + "valid_targets_min": 412 + }, + { + "epoch": 4.348588120740019, + "grad_norm": 0.7705056096763342, + "learning_rate": 1.5102171653187367e-05, + "loss": 0.1777, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08623531460762024, + "step": 2235, + "valid_targets_mean": 2429.9, + "valid_targets_min": 771 + }, + { + "epoch": 4.358325219084713, + "grad_norm": 0.9845962520737138, + "learning_rate": 1.5008161316150699e-05, + "loss": 0.1856, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10871316492557526, + "step": 2240, + "valid_targets_mean": 2547.9, + "valid_targets_min": 612 + }, + { + "epoch": 4.368062317429406, + "grad_norm": 0.8289568853379979, + "learning_rate": 1.4914268454195737e-05, + "loss": 0.1752, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06594522297382355, + "step": 2245, + "valid_targets_mean": 2254.0, + "valid_targets_min": 753 + }, + { + "epoch": 4.377799415774099, + "grad_norm": 0.8395243990768753, + "learning_rate": 1.4820495276943491e-05, + "loss": 0.1808, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09926756471395493, + "step": 2250, + "valid_targets_mean": 2749.0, + "valid_targets_min": 847 + }, + { + "epoch": 4.387536514118793, + "grad_norm": 0.7530052418500517, + "learning_rate": 1.4726843991198389e-05, + "loss": 0.1743, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10601849108934402, + "step": 2255, + "valid_targets_mean": 3593.2, + "valid_targets_min": 816 + }, + { + "epoch": 4.3972736124634855, + "grad_norm": 0.8654911495042591, + "learning_rate": 1.4633316800896316e-05, + "loss": 0.1727, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08109898865222931, + "step": 2260, + "valid_targets_mean": 2410.8, + "valid_targets_min": 1440 + }, + { + "epoch": 4.407010710808179, + "grad_norm": 0.738439387301167, + "learning_rate": 1.453991590705277e-05, + "loss": 0.1779, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06938911974430084, + "step": 2265, + "valid_targets_mean": 2660.4, + "valid_targets_min": 1141 + }, + { + "epoch": 4.416747809152873, + "grad_norm": 0.7377441598826128, + "learning_rate": 1.4446643507711068e-05, + "loss": 0.1667, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08617185056209564, + "step": 2270, + "valid_targets_mean": 3686.0, + "valid_targets_min": 1726 + }, + { + "epoch": 4.426484907497565, + "grad_norm": 0.8555737565964598, + "learning_rate": 1.4353501797890617e-05, + "loss": 0.1853, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08488436043262482, + "step": 2275, + "valid_targets_mean": 2369.0, + "valid_targets_min": 1055 + }, + { + "epoch": 4.436222005842259, + "grad_norm": 0.8121396018190382, + "learning_rate": 1.4260492969535218e-05, + "loss": 0.1761, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11650624871253967, + "step": 2280, + "valid_targets_mean": 2974.4, + "valid_targets_min": 1081 + }, + { + "epoch": 4.445959104186953, + "grad_norm": 0.739472764979638, + "learning_rate": 1.416761921146155e-05, + "loss": 0.1866, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08772695064544678, + "step": 2285, + "valid_targets_mean": 3347.6, + "valid_targets_min": 755 + }, + { + "epoch": 4.455696202531645, + "grad_norm": 0.94241742553367, + "learning_rate": 1.4074882709307617e-05, + "loss": 0.1825, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10868126899003983, + "step": 2290, + "valid_targets_mean": 2952.5, + "valid_targets_min": 1445 + }, + { + "epoch": 4.465433300876339, + "grad_norm": 0.8620357853018356, + "learning_rate": 1.398228564548131e-05, + "loss": 0.1776, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08930197358131409, + "step": 2295, + "valid_targets_mean": 2496.9, + "valid_targets_min": 1159 + }, + { + "epoch": 4.4751703992210325, + "grad_norm": 0.8238165602040978, + "learning_rate": 1.3889830199109064e-05, + "loss": 0.1734, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06826405972242355, + "step": 2300, + "valid_targets_mean": 2244.8, + "valid_targets_min": 1102 + }, + { + "epoch": 4.484907497565725, + "grad_norm": 0.7448771421831222, + "learning_rate": 1.3797518545984577e-05, + "loss": 0.1775, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08596190810203552, + "step": 2305, + "valid_targets_mean": 3395.1, + "valid_targets_min": 1302 + }, + { + "epoch": 4.494644595910419, + "grad_norm": 0.6459287688321765, + "learning_rate": 1.3705352858517596e-05, + "loss": 0.1676, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06672492623329163, + "step": 2310, + "valid_targets_mean": 3345.8, + "valid_targets_min": 1065 + }, + { + "epoch": 4.504381694255112, + "grad_norm": 0.937622099826767, + "learning_rate": 1.3613335305682769e-05, + "loss": 0.1875, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13400760293006897, + "step": 2315, + "valid_targets_mean": 2776.0, + "valid_targets_min": 1585 + }, + { + "epoch": 4.514118792599805, + "grad_norm": 0.7440001368095495, + "learning_rate": 1.3521468052968682e-05, + "loss": 0.1664, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08430090546607971, + "step": 2320, + "valid_targets_mean": 3821.0, + "valid_targets_min": 1560 + }, + { + "epoch": 4.523855890944499, + "grad_norm": 0.9340245915074052, + "learning_rate": 1.3429753262326793e-05, + "loss": 0.1729, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08000458776950836, + "step": 2325, + "valid_targets_mean": 2148.0, + "valid_targets_min": 676 + }, + { + "epoch": 4.5335929892891915, + "grad_norm": 0.8249789412174031, + "learning_rate": 1.3338193092120628e-05, + "loss": 0.1775, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11837929487228394, + "step": 2330, + "valid_targets_mean": 2965.6, + "valid_targets_min": 411 + }, + { + "epoch": 4.543330087633885, + "grad_norm": 0.6986712399643566, + "learning_rate": 1.3246789697074958e-05, + "loss": 0.1708, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07143793255090714, + "step": 2335, + "valid_targets_mean": 3674.0, + "valid_targets_min": 1680 + }, + { + "epoch": 4.553067185978579, + "grad_norm": 0.8287706359385277, + "learning_rate": 1.3155545228225104e-05, + "loss": 0.1806, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09657485783100128, + "step": 2340, + "valid_targets_mean": 2945.1, + "valid_targets_min": 973 + }, + { + "epoch": 4.5628042843232715, + "grad_norm": 0.6624365352547787, + "learning_rate": 1.3064461832866304e-05, + "loss": 0.1765, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08598685264587402, + "step": 2345, + "valid_targets_mean": 5114.0, + "valid_targets_min": 1194 + }, + { + "epoch": 4.572541382667965, + "grad_norm": 0.7677101034035914, + "learning_rate": 1.297354165450318e-05, + "loss": 0.1689, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08426067978143692, + "step": 2350, + "valid_targets_mean": 3179.9, + "valid_targets_min": 749 + }, + { + "epoch": 4.582278481012658, + "grad_norm": 1.1268953134679656, + "learning_rate": 1.2882786832799305e-05, + "loss": 0.1845, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08948741853237152, + "step": 2355, + "valid_targets_mean": 1726.6, + "valid_targets_min": 727 + }, + { + "epoch": 4.592015579357351, + "grad_norm": 0.7420321706926357, + "learning_rate": 1.2792199503526848e-05, + "loss": 0.1814, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06331359595060349, + "step": 2360, + "valid_targets_mean": 2387.6, + "valid_targets_min": 339 + }, + { + "epoch": 4.601752677702045, + "grad_norm": 0.8931184358236555, + "learning_rate": 1.2701781798516288e-05, + "loss": 0.1835, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10443165898323059, + "step": 2365, + "valid_targets_mean": 2841.5, + "valid_targets_min": 1609 + }, + { + "epoch": 4.611489776046738, + "grad_norm": 0.8297066622004157, + "learning_rate": 1.2611535845606273e-05, + "loss": 0.1777, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07365868240594864, + "step": 2370, + "valid_targets_mean": 1751.5, + "valid_targets_min": 944 + }, + { + "epoch": 4.621226874391431, + "grad_norm": 0.7460221119378259, + "learning_rate": 1.2521463768593545e-05, + "loss": 0.1719, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05629142001271248, + "step": 2375, + "valid_targets_mean": 2663.4, + "valid_targets_min": 1070 + }, + { + "epoch": 4.630963972736125, + "grad_norm": 0.7931377268019347, + "learning_rate": 1.2431567687182939e-05, + "loss": 0.1782, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10078290104866028, + "step": 2380, + "valid_targets_mean": 3410.9, + "valid_targets_min": 1454 + }, + { + "epoch": 4.640701071080818, + "grad_norm": 0.7596015233716497, + "learning_rate": 1.234184971693751e-05, + "loss": 0.1763, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0730675607919693, + "step": 2385, + "valid_targets_mean": 2975.4, + "valid_targets_min": 1671 + }, + { + "epoch": 4.650438169425511, + "grad_norm": 1.0400260562656607, + "learning_rate": 1.2252311969228754e-05, + "loss": 0.184, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07677740603685379, + "step": 2390, + "valid_targets_mean": 2467.1, + "valid_targets_min": 1192 + }, + { + "epoch": 4.660175267770205, + "grad_norm": 0.7959513719005028, + "learning_rate": 1.2162956551186904e-05, + "loss": 0.1777, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0791003629565239, + "step": 2395, + "valid_targets_mean": 2879.6, + "valid_targets_min": 922 + }, + { + "epoch": 4.669912366114898, + "grad_norm": 0.9105837151780105, + "learning_rate": 1.2073785565651362e-05, + "loss": 0.1732, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09069986641407013, + "step": 2400, + "valid_targets_mean": 2383.4, + "valid_targets_min": 669 + }, + { + "epoch": 4.679649464459591, + "grad_norm": 0.84640137944273, + "learning_rate": 1.1984801111121173e-05, + "loss": 0.1698, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07074621319770813, + "step": 2405, + "valid_targets_mean": 2490.0, + "valid_targets_min": 727 + }, + { + "epoch": 4.689386562804284, + "grad_norm": 0.8136174798674647, + "learning_rate": 1.1896005281705717e-05, + "loss": 0.1814, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09965146332979202, + "step": 2410, + "valid_targets_mean": 2786.1, + "valid_targets_min": 1413 + }, + { + "epoch": 4.6991236611489775, + "grad_norm": 0.8586390881863123, + "learning_rate": 1.180740016707533e-05, + "loss": 0.1842, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09565109759569168, + "step": 2415, + "valid_targets_mean": 3106.9, + "valid_targets_min": 1276 + }, + { + "epoch": 4.708860759493671, + "grad_norm": 0.9158651878377554, + "learning_rate": 1.1718987852412193e-05, + "loss": 0.1739, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09477080404758453, + "step": 2420, + "valid_targets_mean": 3704.9, + "valid_targets_min": 1847 + }, + { + "epoch": 4.718597857838364, + "grad_norm": 1.0263917100943944, + "learning_rate": 1.1630770418361255e-05, + "loss": 0.1767, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10718078911304474, + "step": 2425, + "valid_targets_mean": 2840.9, + "valid_targets_min": 361 + }, + { + "epoch": 4.728334956183057, + "grad_norm": 0.9774967088484693, + "learning_rate": 1.154274994098124e-05, + "loss": 0.1828, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11131705343723297, + "step": 2430, + "valid_targets_mean": 2695.8, + "valid_targets_min": 839 + }, + { + "epoch": 4.738072054527751, + "grad_norm": 1.0630175252221974, + "learning_rate": 1.1454928491695803e-05, + "loss": 0.1603, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08804786950349808, + "step": 2435, + "valid_targets_mean": 1911.8, + "valid_targets_min": 533 + }, + { + "epoch": 4.747809152872444, + "grad_norm": 0.7434551987653799, + "learning_rate": 1.1367308137244817e-05, + "loss": 0.1797, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06759445369243622, + "step": 2440, + "valid_targets_mean": 1998.9, + "valid_targets_min": 641 + }, + { + "epoch": 4.757546251217137, + "grad_norm": 0.7688618376697021, + "learning_rate": 1.1279890939635659e-05, + "loss": 0.1758, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09798973798751831, + "step": 2445, + "valid_targets_mean": 3507.9, + "valid_targets_min": 546 + }, + { + "epoch": 4.767283349561831, + "grad_norm": 0.9291345312132047, + "learning_rate": 1.1192678956094753e-05, + "loss": 0.1702, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0608474463224411, + "step": 2450, + "valid_targets_mean": 1835.0, + "valid_targets_min": 749 + }, + { + "epoch": 4.777020447906524, + "grad_norm": 0.924147763117448, + "learning_rate": 1.1105674239019132e-05, + "loss": 0.193, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0942969024181366, + "step": 2455, + "valid_targets_mean": 2208.0, + "valid_targets_min": 1245 + }, + { + "epoch": 4.786757546251217, + "grad_norm": 0.7370127844190928, + "learning_rate": 1.1018878835928142e-05, + "loss": 0.1864, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09319327771663666, + "step": 2460, + "valid_targets_mean": 3291.6, + "valid_targets_min": 1340 + }, + { + "epoch": 4.79649464459591, + "grad_norm": 0.7645770396885143, + "learning_rate": 1.093229478941525e-05, + "loss": 0.169, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07110965996980667, + "step": 2465, + "valid_targets_mean": 2410.2, + "valid_targets_min": 930 + }, + { + "epoch": 4.806231742940604, + "grad_norm": 0.6581970408354032, + "learning_rate": 1.0845924137099953e-05, + "loss": 0.1568, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08824768662452698, + "step": 2470, + "valid_targets_mean": 4368.1, + "valid_targets_min": 1502 + }, + { + "epoch": 4.815968841285297, + "grad_norm": 0.7513095850383397, + "learning_rate": 1.0759768911579896e-05, + "loss": 0.1896, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09617326408624649, + "step": 2475, + "valid_targets_mean": 3501.6, + "valid_targets_min": 892 + }, + { + "epoch": 4.82570593962999, + "grad_norm": 0.9663199577680193, + "learning_rate": 1.0673831140382962e-05, + "loss": 0.1786, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09308472275733948, + "step": 2480, + "valid_targets_mean": 2080.4, + "valid_targets_min": 793 + }, + { + "epoch": 4.8354430379746836, + "grad_norm": 0.7754613234125508, + "learning_rate": 1.0588112845919594e-05, + "loss": 0.1823, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0970308929681778, + "step": 2485, + "valid_targets_mean": 2594.6, + "valid_targets_min": 361 + }, + { + "epoch": 4.845180136319377, + "grad_norm": 0.925597519137906, + "learning_rate": 1.0502616045435202e-05, + "loss": 0.1756, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09768900275230408, + "step": 2490, + "valid_targets_mean": 2550.8, + "valid_targets_min": 590 + }, + { + "epoch": 4.85491723466407, + "grad_norm": 0.8492111273552944, + "learning_rate": 1.0417342750962671e-05, + "loss": 0.1774, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06129820644855499, + "step": 2495, + "valid_targets_mean": 1947.1, + "valid_targets_min": 1349 + }, + { + "epoch": 4.8646543330087635, + "grad_norm": 0.7507411206504069, + "learning_rate": 1.0332294969275032e-05, + "loss": 0.1812, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06962542235851288, + "step": 2500, + "valid_targets_mean": 2455.6, + "valid_targets_min": 1232 + }, + { + "epoch": 4.874391431353457, + "grad_norm": 0.6262081583851395, + "learning_rate": 1.0247474701838222e-05, + "loss": 0.1753, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08543452620506287, + "step": 2505, + "valid_targets_mean": 5083.9, + "valid_targets_min": 1348 + }, + { + "epoch": 4.88412852969815, + "grad_norm": 0.934766709104454, + "learning_rate": 1.0162883944763999e-05, + "loss": 0.1799, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0959821417927742, + "step": 2510, + "valid_targets_mean": 2501.4, + "valid_targets_min": 1468 + }, + { + "epoch": 4.893865628042843, + "grad_norm": 0.8238816881001866, + "learning_rate": 1.0078524688762955e-05, + "loss": 0.1725, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09571737051010132, + "step": 2515, + "valid_targets_mean": 3524.8, + "valid_targets_min": 1513 + }, + { + "epoch": 4.903602726387536, + "grad_norm": 0.7956379233948672, + "learning_rate": 9.994398919097634e-06, + "loss": 0.1811, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10006579011678696, + "step": 2520, + "valid_targets_mean": 3144.8, + "valid_targets_min": 1350 + }, + { + "epoch": 4.91333982473223, + "grad_norm": 1.0031505931340245, + "learning_rate": 9.910508615535876e-06, + "loss": 0.1808, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1176973432302475, + "step": 2525, + "valid_targets_mean": 3230.6, + "valid_targets_min": 1483 + }, + { + "epoch": 4.923076923076923, + "grad_norm": 0.7961411053530816, + "learning_rate": 9.826855752304209e-06, + "loss": 0.1763, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06230532005429268, + "step": 2530, + "valid_targets_mean": 2015.2, + "valid_targets_min": 761 + }, + { + "epoch": 4.932814021421616, + "grad_norm": 0.8159616941356099, + "learning_rate": 9.743442298041348e-06, + "loss": 0.1804, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08193934708833694, + "step": 2535, + "valid_targets_mean": 2771.4, + "valid_targets_min": 968 + }, + { + "epoch": 4.94255111976631, + "grad_norm": 0.7849148331720305, + "learning_rate": 9.660270215751905e-06, + "loss": 0.1838, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09219025075435638, + "step": 2540, + "valid_targets_mean": 3240.4, + "valid_targets_min": 1143 + }, + { + "epoch": 4.952288218111003, + "grad_norm": 0.8024739199361195, + "learning_rate": 9.577341462760182e-06, + "loss": 0.1819, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1043066531419754, + "step": 2545, + "valid_targets_mean": 3759.1, + "valid_targets_min": 427 + }, + { + "epoch": 4.962025316455696, + "grad_norm": 0.7191457501116008, + "learning_rate": 9.494657990664105e-06, + "loss": 0.1795, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.061282359063625336, + "step": 2550, + "valid_targets_mean": 2609.5, + "valid_targets_min": 1201 + }, + { + "epoch": 4.97176241480039, + "grad_norm": 0.7332575836465272, + "learning_rate": 9.412221745289296e-06, + "loss": 0.1837, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08472214639186859, + "step": 2555, + "valid_targets_mean": 2834.2, + "valid_targets_min": 784 + }, + { + "epoch": 4.981499513145083, + "grad_norm": 0.9439541516349363, + "learning_rate": 9.330034666643304e-06, + "loss": 0.1748, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09852138161659241, + "step": 2560, + "valid_targets_mean": 2607.1, + "valid_targets_min": 672 + }, + { + "epoch": 4.991236611489776, + "grad_norm": 0.8536336779010893, + "learning_rate": 9.248098688869917e-06, + "loss": 0.1809, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09487920999526978, + "step": 2565, + "valid_targets_mean": 2370.9, + "valid_targets_min": 1042 + }, + { + "epoch": 5.0, + "grad_norm": 1.0434807392539245, + "learning_rate": 9.166415740203644e-06, + "loss": 0.1823, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18129360675811768, + "step": 2570, + "valid_targets_mean": 3821.0, + "valid_targets_min": 992 + }, + { + "epoch": 5.009737098344694, + "grad_norm": 0.8788739617158123, + "learning_rate": 9.084987742924365e-06, + "loss": 0.1469, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07336068898439407, + "step": 2575, + "valid_targets_mean": 2166.5, + "valid_targets_min": 667 + }, + { + "epoch": 5.019474196689386, + "grad_norm": 0.7132646052444165, + "learning_rate": 9.003816613312097e-06, + "loss": 0.1381, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07973846793174744, + "step": 2580, + "valid_targets_mean": 4526.4, + "valid_targets_min": 2500 + }, + { + "epoch": 5.02921129503408, + "grad_norm": 0.8584741060701055, + "learning_rate": 8.922904261601853e-06, + "loss": 0.1565, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06728353351354599, + "step": 2585, + "valid_targets_mean": 2732.1, + "valid_targets_min": 1147 + }, + { + "epoch": 5.0389483933787735, + "grad_norm": 0.7731473340495661, + "learning_rate": 8.842252591938725e-06, + "loss": 0.1567, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08202247321605682, + "step": 2590, + "valid_targets_mean": 3961.6, + "valid_targets_min": 760 + }, + { + "epoch": 5.048685491723466, + "grad_norm": 0.8694597268692199, + "learning_rate": 8.76186350233306e-06, + "loss": 0.1518, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10540521144866943, + "step": 2595, + "valid_targets_mean": 2925.5, + "valid_targets_min": 1548 + }, + { + "epoch": 5.05842259006816, + "grad_norm": 1.040810076249955, + "learning_rate": 8.681738884615805e-06, + "loss": 0.1638, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07151223719120026, + "step": 2600, + "valid_targets_mean": 1545.4, + "valid_targets_min": 816 + }, + { + "epoch": 5.068159688412853, + "grad_norm": 0.7793137982906794, + "learning_rate": 8.601880624393964e-06, + "loss": 0.1596, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07597843557596207, + "step": 2605, + "valid_targets_mean": 3288.5, + "valid_targets_min": 890 + }, + { + "epoch": 5.077896786757546, + "grad_norm": 0.8369156935610657, + "learning_rate": 8.522290601006245e-06, + "loss": 0.164, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06252319365739822, + "step": 2610, + "valid_targets_mean": 2056.4, + "valid_targets_min": 1449 + }, + { + "epoch": 5.08763388510224, + "grad_norm": 0.9475812466325354, + "learning_rate": 8.442970687478815e-06, + "loss": 0.1575, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0576450489461422, + "step": 2615, + "valid_targets_mean": 2248.6, + "valid_targets_min": 1311 + }, + { + "epoch": 5.0973709834469325, + "grad_norm": 0.7916662534612631, + "learning_rate": 8.363922750481239e-06, + "loss": 0.1572, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06718336790800095, + "step": 2620, + "valid_targets_mean": 2878.2, + "valid_targets_min": 1456 + }, + { + "epoch": 5.107108081791626, + "grad_norm": 0.9983523544701802, + "learning_rate": 8.28514865028253e-06, + "loss": 0.1532, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06368249654769897, + "step": 2625, + "valid_targets_mean": 2964.4, + "valid_targets_min": 1247 + }, + { + "epoch": 5.11684518013632, + "grad_norm": 0.8540179789550341, + "learning_rate": 8.2066502407074e-06, + "loss": 0.1631, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08504103124141693, + "step": 2630, + "valid_targets_mean": 2839.8, + "valid_targets_min": 1418 + }, + { + "epoch": 5.1265822784810124, + "grad_norm": 0.7428245796076182, + "learning_rate": 8.128429369092599e-06, + "loss": 0.1534, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07919472455978394, + "step": 2635, + "valid_targets_mean": 3592.6, + "valid_targets_min": 1514 + }, + { + "epoch": 5.136319376825706, + "grad_norm": 0.9250428955359546, + "learning_rate": 8.050487876243473e-06, + "loss": 0.1597, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07924531400203705, + "step": 2640, + "valid_targets_mean": 2436.1, + "valid_targets_min": 592 + }, + { + "epoch": 5.146056475170399, + "grad_norm": 0.9951745935177713, + "learning_rate": 7.972827596390594e-06, + "loss": 0.161, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.061942871659994125, + "step": 2645, + "valid_targets_mean": 1668.6, + "valid_targets_min": 674 + }, + { + "epoch": 5.155793573515092, + "grad_norm": 0.9734815345149869, + "learning_rate": 7.89545035714667e-06, + "loss": 0.163, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09595105051994324, + "step": 2650, + "valid_targets_mean": 2707.8, + "valid_targets_min": 945 + }, + { + "epoch": 5.165530671859786, + "grad_norm": 0.788230065562802, + "learning_rate": 7.81835797946347e-06, + "loss": 0.1562, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06886913627386093, + "step": 2655, + "valid_targets_mean": 2837.9, + "valid_targets_min": 801 + }, + { + "epoch": 5.175267770204479, + "grad_norm": 0.8515098494145404, + "learning_rate": 7.741552277589e-06, + "loss": 0.1532, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06747442483901978, + "step": 2660, + "valid_targets_mean": 2596.0, + "valid_targets_min": 444 + }, + { + "epoch": 5.185004868549172, + "grad_norm": 0.7160182068063499, + "learning_rate": 7.665035059024794e-06, + "loss": 0.1555, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06444238871335983, + "step": 2665, + "valid_targets_mean": 3043.0, + "valid_targets_min": 781 + }, + { + "epoch": 5.194741966893866, + "grad_norm": 0.7213961492220646, + "learning_rate": 7.5888081244833936e-06, + "loss": 0.1467, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0880516916513443, + "step": 2670, + "valid_targets_mean": 4118.0, + "valid_targets_min": 976 + }, + { + "epoch": 5.204479065238559, + "grad_norm": 0.8900388790037114, + "learning_rate": 7.512873267845948e-06, + "loss": 0.1535, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06810447573661804, + "step": 2675, + "valid_targets_mean": 2510.4, + "valid_targets_min": 1402 + }, + { + "epoch": 5.214216163583252, + "grad_norm": 0.9916836488767273, + "learning_rate": 7.437232276120037e-06, + "loss": 0.1598, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09773782640695572, + "step": 2680, + "valid_targets_mean": 2611.5, + "valid_targets_min": 996 + }, + { + "epoch": 5.223953261927946, + "grad_norm": 0.8593083409555594, + "learning_rate": 7.361886929397568e-06, + "loss": 0.1569, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08319692313671112, + "step": 2685, + "valid_targets_mean": 2737.4, + "valid_targets_min": 986 + }, + { + "epoch": 5.233690360272639, + "grad_norm": 0.8931821415391766, + "learning_rate": 7.286839000812922e-06, + "loss": 0.1572, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0685073584318161, + "step": 2690, + "valid_targets_mean": 2461.8, + "valid_targets_min": 1149 + }, + { + "epoch": 5.243427458617332, + "grad_norm": 0.9300425765001682, + "learning_rate": 7.212090256501185e-06, + "loss": 0.1582, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08733557164669037, + "step": 2695, + "valid_targets_mean": 3048.6, + "valid_targets_min": 646 + }, + { + "epoch": 5.253164556962025, + "grad_norm": 0.8576363814070205, + "learning_rate": 7.137642455556654e-06, + "loss": 0.1507, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0786384642124176, + "step": 2700, + "valid_targets_mean": 2628.5, + "valid_targets_min": 1201 + }, + { + "epoch": 5.2629016553067185, + "grad_norm": 0.8180043273409469, + "learning_rate": 7.063497349991364e-06, + "loss": 0.1593, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.043701425194740295, + "step": 2705, + "valid_targets_mean": 1487.5, + "valid_targets_min": 328 + }, + { + "epoch": 5.272638753651412, + "grad_norm": 0.904986148543234, + "learning_rate": 6.9896566846939e-06, + "loss": 0.1556, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08402790129184723, + "step": 2710, + "valid_targets_mean": 2780.1, + "valid_targets_min": 1045 + }, + { + "epoch": 5.282375851996105, + "grad_norm": 0.8469792592809527, + "learning_rate": 6.916122197388322e-06, + "loss": 0.1641, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08842235803604126, + "step": 2715, + "valid_targets_mean": 2929.1, + "valid_targets_min": 837 + }, + { + "epoch": 5.292112950340798, + "grad_norm": 0.8764054913729791, + "learning_rate": 6.842895618593275e-06, + "loss": 0.1651, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09535115212202072, + "step": 2720, + "valid_targets_mean": 3143.9, + "valid_targets_min": 1751 + }, + { + "epoch": 5.301850048685492, + "grad_norm": 0.8068241577486652, + "learning_rate": 6.769978671581257e-06, + "loss": 0.1555, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08451789617538452, + "step": 2725, + "valid_targets_mean": 3457.8, + "valid_targets_min": 1521 + }, + { + "epoch": 5.311587147030185, + "grad_norm": 1.285539080281442, + "learning_rate": 6.697373072338065e-06, + "loss": 0.16, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08277589082717896, + "step": 2730, + "valid_targets_mean": 2291.1, + "valid_targets_min": 609 + }, + { + "epoch": 5.321324245374878, + "grad_norm": 0.7401126717811027, + "learning_rate": 6.625080529522432e-06, + "loss": 0.1569, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08714379370212555, + "step": 2735, + "valid_targets_mean": 4592.0, + "valid_targets_min": 977 + }, + { + "epoch": 5.331061343719572, + "grad_norm": 0.8569685478901633, + "learning_rate": 6.55310274442579e-06, + "loss": 0.1636, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1178111732006073, + "step": 2740, + "valid_targets_mean": 3467.9, + "valid_targets_min": 1854 + }, + { + "epoch": 5.340798442064265, + "grad_norm": 0.8759406556500658, + "learning_rate": 6.481441410932221e-06, + "loss": 0.1714, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12362238764762878, + "step": 2745, + "valid_targets_mean": 4463.8, + "valid_targets_min": 946 + }, + { + "epoch": 5.350535540408958, + "grad_norm": 0.6505264845417847, + "learning_rate": 6.410098215478655e-06, + "loss": 0.1583, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.060885533690452576, + "step": 2750, + "valid_targets_mean": 3237.9, + "valid_targets_min": 749 + }, + { + "epoch": 5.360272638753651, + "grad_norm": 0.7802849044383483, + "learning_rate": 6.339074837015115e-06, + "loss": 0.1537, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06824596226215363, + "step": 2755, + "valid_targets_mean": 3272.1, + "valid_targets_min": 1764 + }, + { + "epoch": 5.370009737098345, + "grad_norm": 0.701379782837056, + "learning_rate": 6.268372946965245e-06, + "loss": 0.1645, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07687725126743317, + "step": 2760, + "valid_targets_mean": 3820.2, + "valid_targets_min": 589 + }, + { + "epoch": 5.379746835443038, + "grad_norm": 0.7326761676590664, + "learning_rate": 6.197994209186959e-06, + "loss": 0.157, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05246296897530556, + "step": 2765, + "valid_targets_mean": 2954.2, + "valid_targets_min": 1334 + }, + { + "epoch": 5.389483933787731, + "grad_norm": 0.7773206771174859, + "learning_rate": 6.1279402799332934e-06, + "loss": 0.1522, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06810751557350159, + "step": 2770, + "valid_targets_mean": 2217.6, + "valid_targets_min": 562 + }, + { + "epoch": 5.3992210321324245, + "grad_norm": 0.8495851051586549, + "learning_rate": 6.058212807813426e-06, + "loss": 0.1668, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07084648311138153, + "step": 2775, + "valid_targets_mean": 2546.6, + "valid_targets_min": 1245 + }, + { + "epoch": 5.408958130477118, + "grad_norm": 0.9497312924955319, + "learning_rate": 5.988813433753869e-06, + "loss": 0.1567, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07291065156459808, + "step": 2780, + "valid_targets_mean": 1611.0, + "valid_targets_min": 658 + }, + { + "epoch": 5.418695228821811, + "grad_norm": 1.0621931684135297, + "learning_rate": 5.919743790959888e-06, + "loss": 0.1661, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06611327081918716, + "step": 2785, + "valid_targets_mean": 1717.2, + "valid_targets_min": 545 + }, + { + "epoch": 5.4284323271665045, + "grad_norm": 0.8152250708964739, + "learning_rate": 5.851005504877012e-06, + "loss": 0.1558, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09362748265266418, + "step": 2790, + "valid_targets_mean": 3392.8, + "valid_targets_min": 901 + }, + { + "epoch": 5.438169425511198, + "grad_norm": 0.9746220014286787, + "learning_rate": 5.782600193152819e-06, + "loss": 0.1597, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09429764747619629, + "step": 2795, + "valid_targets_mean": 2643.0, + "valid_targets_min": 981 + }, + { + "epoch": 5.447906523855891, + "grad_norm": 0.8351011417902146, + "learning_rate": 5.714529465598872e-06, + "loss": 0.1663, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09567970037460327, + "step": 2800, + "valid_targets_mean": 3979.8, + "valid_targets_min": 1707 + }, + { + "epoch": 5.457643622200584, + "grad_norm": 0.8049061085235286, + "learning_rate": 5.646794924152808e-06, + "loss": 0.1666, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09440994262695312, + "step": 2805, + "valid_targets_mean": 3631.0, + "valid_targets_min": 627 + }, + { + "epoch": 5.467380720545277, + "grad_norm": 0.7490940611455418, + "learning_rate": 5.579398162840655e-06, + "loss": 0.1544, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07636860013008118, + "step": 2810, + "valid_targets_mean": 3577.4, + "valid_targets_min": 353 + }, + { + "epoch": 5.477117818889971, + "grad_norm": 0.9000108148792098, + "learning_rate": 5.512340767739315e-06, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08217882364988327, + "step": 2815, + "valid_targets_mean": 2592.8, + "valid_targets_min": 1436 + }, + { + "epoch": 5.486854917234664, + "grad_norm": 0.8305266244336463, + "learning_rate": 5.445624316939244e-06, + "loss": 0.1641, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08523901551961899, + "step": 2820, + "valid_targets_mean": 3182.4, + "valid_targets_min": 1404 + }, + { + "epoch": 5.496592015579357, + "grad_norm": 0.8454176944891794, + "learning_rate": 5.379250380507302e-06, + "loss": 0.1538, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.061266086995601654, + "step": 2825, + "valid_targets_mean": 2720.6, + "valid_targets_min": 1358 + }, + { + "epoch": 5.506329113924051, + "grad_norm": 0.859366502571961, + "learning_rate": 5.313220520449811e-06, + "loss": 0.1647, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08275482058525085, + "step": 2830, + "valid_targets_mean": 2100.1, + "valid_targets_min": 1103 + }, + { + "epoch": 5.516066212268744, + "grad_norm": 0.805421207634601, + "learning_rate": 5.247536290675821e-06, + "loss": 0.1672, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10261974483728409, + "step": 2835, + "valid_targets_mean": 4486.8, + "valid_targets_min": 1335 + }, + { + "epoch": 5.525803310613437, + "grad_norm": 0.844029206448696, + "learning_rate": 5.182199236960476e-06, + "loss": 0.1661, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10321641713380814, + "step": 2840, + "valid_targets_mean": 4219.5, + "valid_targets_min": 1153 + }, + { + "epoch": 5.535540408958131, + "grad_norm": 0.905655733799374, + "learning_rate": 5.117210896908702e-06, + "loss": 0.1655, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07245047390460968, + "step": 2845, + "valid_targets_mean": 2070.0, + "valid_targets_min": 641 + }, + { + "epoch": 5.545277507302824, + "grad_norm": 0.8213330423903124, + "learning_rate": 5.0525727999189865e-06, + "loss": 0.158, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08456528931856155, + "step": 2850, + "valid_targets_mean": 3331.4, + "valid_targets_min": 415 + }, + { + "epoch": 5.555014605647517, + "grad_norm": 0.9709341086367071, + "learning_rate": 4.988286467147416e-06, + "loss": 0.1617, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08191390335559845, + "step": 2855, + "valid_targets_mean": 1917.4, + "valid_targets_min": 956 + }, + { + "epoch": 5.5647517039922105, + "grad_norm": 1.1701719048090364, + "learning_rate": 4.924353411471834e-06, + "loss": 0.1618, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09931658208370209, + "step": 2860, + "valid_targets_mean": 2297.6, + "valid_targets_min": 1473 + }, + { + "epoch": 5.574488802336903, + "grad_norm": 0.9281303675067485, + "learning_rate": 4.860775137456275e-06, + "loss": 0.1668, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07922400534152985, + "step": 2865, + "valid_targets_mean": 2203.0, + "valid_targets_min": 361 + }, + { + "epoch": 5.584225900681597, + "grad_norm": 0.7949441894303411, + "learning_rate": 4.797553141315543e-06, + "loss": 0.1642, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08693256229162216, + "step": 2870, + "valid_targets_mean": 2908.5, + "valid_targets_min": 1384 + }, + { + "epoch": 5.5939629990262905, + "grad_norm": 0.8561484147763357, + "learning_rate": 4.734688910880001e-06, + "loss": 0.164, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06763464212417603, + "step": 2875, + "valid_targets_mean": 2343.9, + "valid_targets_min": 1005 + }, + { + "epoch": 5.603700097370983, + "grad_norm": 0.8806999977127576, + "learning_rate": 4.672183925560559e-06, + "loss": 0.1572, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07673318684101105, + "step": 2880, + "valid_targets_mean": 2955.1, + "valid_targets_min": 1311 + }, + { + "epoch": 5.613437195715677, + "grad_norm": 0.6319972257202355, + "learning_rate": 4.610039656313854e-06, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06642003357410431, + "step": 2885, + "valid_targets_mean": 4784.6, + "valid_targets_min": 3072 + }, + { + "epoch": 5.62317429406037, + "grad_norm": 0.916217047093341, + "learning_rate": 4.548257565607643e-06, + "loss": 0.1541, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08463971316814423, + "step": 2890, + "valid_targets_mean": 2950.9, + "valid_targets_min": 1101 + }, + { + "epoch": 5.632911392405063, + "grad_norm": 0.8021495560370561, + "learning_rate": 4.486839107386378e-06, + "loss": 0.1533, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09788113832473755, + "step": 2895, + "valid_targets_mean": 3202.8, + "valid_targets_min": 992 + }, + { + "epoch": 5.642648490749757, + "grad_norm": 0.8588283420614421, + "learning_rate": 4.425785727036979e-06, + "loss": 0.1683, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07578335702419281, + "step": 2900, + "valid_targets_mean": 2622.9, + "valid_targets_min": 822 + }, + { + "epoch": 5.65238558909445, + "grad_norm": 0.9176614969113452, + "learning_rate": 4.365098861354862e-06, + "loss": 0.1696, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07465335726737976, + "step": 2905, + "valid_targets_mean": 2558.9, + "valid_targets_min": 1013 + }, + { + "epoch": 5.662122687439143, + "grad_norm": 0.7619193048169172, + "learning_rate": 4.304779938510073e-06, + "loss": 0.1768, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0974918007850647, + "step": 2910, + "valid_targets_mean": 4078.2, + "valid_targets_min": 1683 + }, + { + "epoch": 5.671859785783837, + "grad_norm": 0.7718442613293865, + "learning_rate": 4.244830378013689e-06, + "loss": 0.1629, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.084866963326931, + "step": 2915, + "valid_targets_mean": 2901.4, + "valid_targets_min": 1304 + }, + { + "epoch": 5.681596884128529, + "grad_norm": 0.8043628746607439, + "learning_rate": 4.185251590684458e-06, + "loss": 0.1516, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07334727048873901, + "step": 2920, + "valid_targets_mean": 2889.6, + "valid_targets_min": 592 + }, + { + "epoch": 5.691333982473223, + "grad_norm": 0.9400775178899726, + "learning_rate": 4.126044978615544e-06, + "loss": 0.1523, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06582826375961304, + "step": 2925, + "valid_targets_mean": 3248.0, + "valid_targets_min": 484 + }, + { + "epoch": 5.701071080817917, + "grad_norm": 0.73244731105116, + "learning_rate": 4.067211935141551e-06, + "loss": 0.1552, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07959744334220886, + "step": 2930, + "valid_targets_mean": 3028.8, + "valid_targets_min": 859 + }, + { + "epoch": 5.710808179162609, + "grad_norm": 0.8805977975136875, + "learning_rate": 4.008753844805737e-06, + "loss": 0.1727, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07804688066244125, + "step": 2935, + "valid_targets_mean": 2540.8, + "valid_targets_min": 877 + }, + { + "epoch": 5.720545277507303, + "grad_norm": 0.7987023815464585, + "learning_rate": 3.950672083327425e-06, + "loss": 0.1658, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0801348090171814, + "step": 2940, + "valid_targets_mean": 2479.9, + "valid_targets_min": 1414 + }, + { + "epoch": 5.730282375851996, + "grad_norm": 0.8299655079384172, + "learning_rate": 3.8929680175696315e-06, + "loss": 0.1479, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07135441899299622, + "step": 2945, + "valid_targets_mean": 2472.5, + "valid_targets_min": 684 + }, + { + "epoch": 5.740019474196689, + "grad_norm": 0.7969720132913237, + "learning_rate": 3.835643005506893e-06, + "loss": 0.1554, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08523745834827423, + "step": 2950, + "valid_targets_mean": 3318.8, + "valid_targets_min": 1012 + }, + { + "epoch": 5.749756572541383, + "grad_norm": 0.7356512604683877, + "learning_rate": 3.7786983961933234e-06, + "loss": 0.1507, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06636009365320206, + "step": 2955, + "valid_targets_mean": 3269.4, + "valid_targets_min": 930 + }, + { + "epoch": 5.759493670886076, + "grad_norm": 0.9844827711902144, + "learning_rate": 3.7221355297308483e-06, + "loss": 0.1645, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10001415759325027, + "step": 2960, + "valid_targets_mean": 2881.2, + "valid_targets_min": 1582 + }, + { + "epoch": 5.769230769230769, + "grad_norm": 1.011006569694185, + "learning_rate": 3.665955737237665e-06, + "loss": 0.176, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07234857231378555, + "step": 2965, + "valid_targets_mean": 2019.0, + "valid_targets_min": 844 + }, + { + "epoch": 5.778967867575463, + "grad_norm": 0.7013473242141751, + "learning_rate": 3.6101603408169373e-06, + "loss": 0.1526, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07861198484897614, + "step": 2970, + "valid_targets_mean": 3716.4, + "valid_targets_min": 805 + }, + { + "epoch": 5.7887049659201555, + "grad_norm": 0.7663032724768525, + "learning_rate": 3.5547506535256825e-06, + "loss": 0.1481, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07591184973716736, + "step": 2975, + "valid_targets_mean": 3134.9, + "valid_targets_min": 770 + }, + { + "epoch": 5.798442064264849, + "grad_norm": 0.9982055407518172, + "learning_rate": 3.4997279793438387e-06, + "loss": 0.16, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08362538367509842, + "step": 2980, + "valid_targets_mean": 2185.9, + "valid_targets_min": 1351 + }, + { + "epoch": 5.808179162609543, + "grad_norm": 0.8369557811189366, + "learning_rate": 3.445093613143611e-06, + "loss": 0.1663, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06555432081222534, + "step": 2985, + "valid_targets_mean": 2496.9, + "valid_targets_min": 829 + }, + { + "epoch": 5.817916260954235, + "grad_norm": 0.7826913703379452, + "learning_rate": 3.390848840658978e-06, + "loss": 0.1517, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05012332648038864, + "step": 2990, + "valid_targets_mean": 2353.4, + "valid_targets_min": 1369 + }, + { + "epoch": 5.827653359298929, + "grad_norm": 0.9167281481100252, + "learning_rate": 3.3369949384554514e-06, + "loss": 0.1547, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0563075914978981, + "step": 2995, + "valid_targets_mean": 2255.0, + "valid_targets_min": 766 + }, + { + "epoch": 5.837390457643622, + "grad_norm": 0.807847869234867, + "learning_rate": 3.2835331739000178e-06, + "loss": 0.1445, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08572809398174286, + "step": 3000, + "valid_targets_mean": 3547.1, + "valid_targets_min": 931 + }, + { + "epoch": 5.847127555988315, + "grad_norm": 0.9042350803251398, + "learning_rate": 3.2304648051313146e-06, + "loss": 0.1614, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.060835100710392, + "step": 3005, + "valid_targets_mean": 1769.5, + "valid_targets_min": 1011 + }, + { + "epoch": 5.856864654333009, + "grad_norm": 0.6067934336851712, + "learning_rate": 3.177791081030057e-06, + "loss": 0.1567, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.048806332051754, + "step": 3010, + "valid_targets_mean": 2325.2, + "valid_targets_min": 766 + }, + { + "epoch": 5.866601752677702, + "grad_norm": 0.8063814297615702, + "learning_rate": 3.125513241189575e-06, + "loss": 0.1591, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07505717128515244, + "step": 3015, + "valid_targets_mean": 3204.6, + "valid_targets_min": 879 + }, + { + "epoch": 5.876338851022395, + "grad_norm": 0.8193782903352972, + "learning_rate": 3.073632515886711e-06, + "loss": 0.1526, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07415521889925003, + "step": 3020, + "valid_targets_mean": 3054.9, + "valid_targets_min": 629 + }, + { + "epoch": 5.886075949367089, + "grad_norm": 0.9230925758861487, + "learning_rate": 3.0221501260528406e-06, + "loss": 0.1519, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07016010582447052, + "step": 3025, + "valid_targets_mean": 2294.0, + "valid_targets_min": 1459 + }, + { + "epoch": 5.895813047711782, + "grad_norm": 1.6643358275637952, + "learning_rate": 2.9710672832451303e-06, + "loss": 0.1532, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0754982978105545, + "step": 3030, + "valid_targets_mean": 1577.0, + "valid_targets_min": 585 + }, + { + "epoch": 5.905550146056475, + "grad_norm": 0.7882969996439173, + "learning_rate": 2.9203851896180422e-06, + "loss": 0.1693, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.072642020881176, + "step": 3035, + "valid_targets_mean": 2888.5, + "valid_targets_min": 1482 + }, + { + "epoch": 5.915287244401169, + "grad_norm": 0.8164072771583177, + "learning_rate": 2.8701050378950303e-06, + "loss": 0.1614, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07803675532341003, + "step": 3040, + "valid_targets_mean": 2868.0, + "valid_targets_min": 726 + }, + { + "epoch": 5.9250243427458615, + "grad_norm": 0.9813071989138026, + "learning_rate": 2.820228011340478e-06, + "loss": 0.1605, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07952482998371124, + "step": 3045, + "valid_targets_mean": 1763.4, + "valid_targets_min": 722 + }, + { + "epoch": 5.934761441090555, + "grad_norm": 0.8810999081460665, + "learning_rate": 2.7707552837318584e-06, + "loss": 0.1589, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10277026146650314, + "step": 3050, + "valid_targets_mean": 3749.4, + "valid_targets_min": 1459 + }, + { + "epoch": 5.944498539435248, + "grad_norm": 0.8466147093454865, + "learning_rate": 2.7216880193320915e-06, + "loss": 0.1628, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11054025590419769, + "step": 3055, + "valid_targets_mean": 3374.1, + "valid_targets_min": 1362 + }, + { + "epoch": 5.9542356377799415, + "grad_norm": 0.8089472605814677, + "learning_rate": 2.673027372862178e-06, + "loss": 0.1666, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07416535913944244, + "step": 3060, + "valid_targets_mean": 3063.5, + "valid_targets_min": 1618 + }, + { + "epoch": 5.963972736124635, + "grad_norm": 0.8188091275314583, + "learning_rate": 2.624774489473978e-06, + "loss": 0.1535, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07103756070137024, + "step": 3065, + "valid_targets_mean": 2943.8, + "valid_targets_min": 1430 + }, + { + "epoch": 5.973709834469328, + "grad_norm": 0.8544686193773303, + "learning_rate": 2.5769305047233004e-06, + "loss": 0.1594, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08920057862997055, + "step": 3070, + "valid_targets_mean": 3097.4, + "valid_targets_min": 1756 + }, + { + "epoch": 5.983446932814021, + "grad_norm": 0.9325324164166047, + "learning_rate": 2.5294965445431775e-06, + "loss": 0.1599, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07478688657283783, + "step": 3075, + "valid_targets_mean": 2376.0, + "valid_targets_min": 830 + }, + { + "epoch": 5.993184031158715, + "grad_norm": 0.7790233347072747, + "learning_rate": 2.4824737252173447e-06, + "loss": 0.1573, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.058365389704704285, + "step": 3080, + "valid_targets_mean": 2365.9, + "valid_targets_min": 836 + }, + { + "epoch": 6.001947419668939, + "grad_norm": 0.7062322366525114, + "learning_rate": 2.4358631533539902e-06, + "loss": 0.1501, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06631424278020859, + "step": 3085, + "valid_targets_mean": 3813.0, + "valid_targets_min": 837 + }, + { + "epoch": 6.011684518013632, + "grad_norm": 0.6129874851931072, + "learning_rate": 2.3896659258596877e-06, + "loss": 0.1446, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06790415942668915, + "step": 3090, + "valid_targets_mean": 3828.0, + "valid_targets_min": 701 + }, + { + "epoch": 6.021421616358325, + "grad_norm": 0.8268875045690981, + "learning_rate": 2.3438831299136265e-06, + "loss": 0.1506, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08013726770877838, + "step": 3095, + "valid_targets_mean": 3162.8, + "valid_targets_min": 1046 + }, + { + "epoch": 6.031158714703018, + "grad_norm": 0.9247544535256333, + "learning_rate": 2.2985158429419885e-06, + "loss": 0.15, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06007855385541916, + "step": 3100, + "valid_targets_mean": 2378.1, + "valid_targets_min": 1293 + }, + { + "epoch": 6.040895813047712, + "grad_norm": 0.7287184304341361, + "learning_rate": 2.2535651325926012e-06, + "loss": 0.1479, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09027600288391113, + "step": 3105, + "valid_targets_mean": 5085.2, + "valid_targets_min": 1460 + }, + { + "epoch": 6.050632911392405, + "grad_norm": 0.7509599811577314, + "learning_rate": 2.209032056709821e-06, + "loss": 0.1459, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07516921311616898, + "step": 3110, + "valid_targets_mean": 3181.6, + "valid_targets_min": 1125 + }, + { + "epoch": 6.060370009737098, + "grad_norm": 0.915474952725836, + "learning_rate": 2.1649176633096312e-06, + "loss": 0.1688, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10079792141914368, + "step": 3115, + "valid_targets_mean": 3642.8, + "valid_targets_min": 1414 + }, + { + "epoch": 6.070107108081792, + "grad_norm": 0.7814533531585784, + "learning_rate": 2.121222990554981e-06, + "loss": 0.1532, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.062230873852968216, + "step": 3120, + "valid_targets_mean": 2256.4, + "valid_targets_min": 722 + }, + { + "epoch": 6.079844206426485, + "grad_norm": 0.7473012000861389, + "learning_rate": 2.0779490667313485e-06, + "loss": 0.1558, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05086394399404526, + "step": 3125, + "valid_targets_mean": 2438.0, + "valid_targets_min": 1017 + }, + { + "epoch": 6.089581304771178, + "grad_norm": 0.8582424039064526, + "learning_rate": 2.0350969102225603e-06, + "loss": 0.1494, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05906301736831665, + "step": 3130, + "valid_targets_mean": 2134.6, + "valid_targets_min": 735 + }, + { + "epoch": 6.099318403115872, + "grad_norm": 0.8437783516310411, + "learning_rate": 1.9926675294868024e-06, + "loss": 0.1524, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06787307560443878, + "step": 3135, + "valid_targets_mean": 2764.5, + "valid_targets_min": 877 + }, + { + "epoch": 6.109055501460565, + "grad_norm": 0.8328816937841937, + "learning_rate": 1.9506619230328884e-06, + "loss": 0.1457, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0856187641620636, + "step": 3140, + "valid_targets_mean": 3496.8, + "valid_targets_min": 936 + }, + { + "epoch": 6.118792599805258, + "grad_norm": 0.9309557209194895, + "learning_rate": 1.9090810793967885e-06, + "loss": 0.1451, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06270939111709595, + "step": 3145, + "valid_targets_mean": 1664.8, + "valid_targets_min": 289 + }, + { + "epoch": 6.1285296981499515, + "grad_norm": 0.8748339479233537, + "learning_rate": 1.8679259771183322e-06, + "loss": 0.1555, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.054008278995752335, + "step": 3150, + "valid_targets_mean": 1955.2, + "valid_targets_min": 585 + }, + { + "epoch": 6.138266796494644, + "grad_norm": 0.8704660174157336, + "learning_rate": 1.8271975847182056e-06, + "loss": 0.1507, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0936213806271553, + "step": 3155, + "valid_targets_mean": 3036.5, + "valid_targets_min": 1241 + }, + { + "epoch": 6.148003894839338, + "grad_norm": 0.9192586335083177, + "learning_rate": 1.7868968606751425e-06, + "loss": 0.1436, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0694252997636795, + "step": 3160, + "valid_targets_mean": 2225.6, + "valid_targets_min": 762 + }, + { + "epoch": 6.1577409931840315, + "grad_norm": 0.7622929705588467, + "learning_rate": 1.7470247534033703e-06, + "loss": 0.1445, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08810773491859436, + "step": 3165, + "valid_targets_mean": 3285.2, + "valid_targets_min": 1078 + }, + { + "epoch": 6.167478091528724, + "grad_norm": 0.9722240434806685, + "learning_rate": 1.7075822012303022e-06, + "loss": 0.1661, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10509815067052841, + "step": 3170, + "valid_targets_mean": 2533.6, + "valid_targets_min": 1122 + }, + { + "epoch": 6.177215189873418, + "grad_norm": 0.8410518297648337, + "learning_rate": 1.6685701323744386e-06, + "loss": 0.1512, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05361660569906235, + "step": 3175, + "valid_targets_mean": 1968.9, + "valid_targets_min": 256 + }, + { + "epoch": 6.186952288218111, + "grad_norm": 0.7240547569836189, + "learning_rate": 1.6299894649235448e-06, + "loss": 0.1548, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07740481942892075, + "step": 3180, + "valid_targets_mean": 4169.2, + "valid_targets_min": 1535 + }, + { + "epoch": 6.196689386562804, + "grad_norm": 0.8960600592655978, + "learning_rate": 1.591841106813017e-06, + "loss": 0.1457, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05455697700381279, + "step": 3185, + "valid_targets_mean": 1794.1, + "valid_targets_min": 339 + }, + { + "epoch": 6.206426484907498, + "grad_norm": 0.9394917842854902, + "learning_rate": 1.5541259558045308e-06, + "loss": 0.1446, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08670862019062042, + "step": 3190, + "valid_targets_mean": 3084.0, + "valid_targets_min": 1261 + }, + { + "epoch": 6.21616358325219, + "grad_norm": 0.9303783534345266, + "learning_rate": 1.5168448994649266e-06, + "loss": 0.1526, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10066616535186768, + "step": 3195, + "valid_targets_mean": 3411.4, + "valid_targets_min": 590 + }, + { + "epoch": 6.225900681596884, + "grad_norm": 0.8661424560699829, + "learning_rate": 1.479998815145296e-06, + "loss": 0.1609, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06600621342658997, + "step": 3200, + "valid_targets_mean": 2468.0, + "valid_targets_min": 1121 + }, + { + "epoch": 6.235637779941578, + "grad_norm": 0.7212205370395781, + "learning_rate": 1.4435885699603524e-06, + "loss": 0.1498, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06191493943333626, + "step": 3205, + "valid_targets_mean": 3401.9, + "valid_targets_min": 1330 + }, + { + "epoch": 6.24537487828627, + "grad_norm": 0.7318902055656283, + "learning_rate": 1.4076150207680185e-06, + "loss": 0.1466, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07200630009174347, + "step": 3210, + "valid_targets_mean": 2824.1, + "valid_targets_min": 1297 + }, + { + "epoch": 6.255111976630964, + "grad_norm": 0.8089814338499913, + "learning_rate": 1.3720790141492657e-06, + "loss": 0.1455, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06058313325047493, + "step": 3215, + "valid_targets_mean": 2880.1, + "valid_targets_min": 1603 + }, + { + "epoch": 6.264849074975658, + "grad_norm": 0.8956533258704256, + "learning_rate": 1.336981386388183e-06, + "loss": 0.141, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0910784900188446, + "step": 3220, + "valid_targets_mean": 2851.8, + "valid_targets_min": 973 + }, + { + "epoch": 6.27458617332035, + "grad_norm": 0.9983186521247152, + "learning_rate": 1.3023229634523027e-06, + "loss": 0.1516, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0565696582198143, + "step": 3225, + "valid_targets_mean": 2139.9, + "valid_targets_min": 749 + }, + { + "epoch": 6.284323271665044, + "grad_norm": 0.7931983927885173, + "learning_rate": 1.2681045609731756e-06, + "loss": 0.1476, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.057150743901729584, + "step": 3230, + "valid_targets_mean": 2575.1, + "valid_targets_min": 1146 + }, + { + "epoch": 6.2940603700097375, + "grad_norm": 0.8917068940461859, + "learning_rate": 1.2343269842271366e-06, + "loss": 0.143, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06814391911029816, + "step": 3235, + "valid_targets_mean": 2241.4, + "valid_targets_min": 1401 + }, + { + "epoch": 6.30379746835443, + "grad_norm": 0.9725488559567614, + "learning_rate": 1.2009910281163962e-06, + "loss": 0.1602, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09598910808563232, + "step": 3240, + "valid_targets_mean": 2813.2, + "valid_targets_min": 726 + }, + { + "epoch": 6.313534566699124, + "grad_norm": 0.8168973650270686, + "learning_rate": 1.168097477150314e-06, + "loss": 0.1447, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09560485184192657, + "step": 3245, + "valid_targets_mean": 3842.6, + "valid_targets_min": 2181 + }, + { + "epoch": 6.3232716650438165, + "grad_norm": 0.8274857149268511, + "learning_rate": 1.1356471054269447e-06, + "loss": 0.147, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08190575242042542, + "step": 3250, + "valid_targets_mean": 3535.4, + "valid_targets_min": 331 + }, + { + "epoch": 6.33300876338851, + "grad_norm": 0.7704740440870161, + "learning_rate": 1.1036406766148033e-06, + "loss": 0.1452, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07610350102186203, + "step": 3255, + "valid_targets_mean": 3209.8, + "valid_targets_min": 658 + }, + { + "epoch": 6.342745861733204, + "grad_norm": 0.8315281394348871, + "learning_rate": 1.0720789439349177e-06, + "loss": 0.1418, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06037437915802002, + "step": 3260, + "valid_targets_mean": 3152.1, + "valid_targets_min": 843 + }, + { + "epoch": 6.3524829600778965, + "grad_norm": 0.8531153468072658, + "learning_rate": 1.0409626501430847e-06, + "loss": 0.1502, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0811975970864296, + "step": 3265, + "valid_targets_mean": 2895.6, + "valid_targets_min": 772 + }, + { + "epoch": 6.36222005842259, + "grad_norm": 0.9454462934134843, + "learning_rate": 1.0102925275123954e-06, + "loss": 0.1411, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08641975373029709, + "step": 3270, + "valid_targets_mean": 2877.4, + "valid_targets_min": 1097 + }, + { + "epoch": 6.371957156767284, + "grad_norm": 0.9225800431729173, + "learning_rate": 9.800692978160064e-07, + "loss": 0.1528, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06825999915599823, + "step": 3275, + "valid_targets_mean": 1949.8, + "valid_targets_min": 550 + }, + { + "epoch": 6.381694255111976, + "grad_norm": 0.7936902856352361, + "learning_rate": 9.502936723101563e-07, + "loss": 0.1591, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08225861936807632, + "step": 3280, + "valid_targets_mean": 3964.6, + "valid_targets_min": 1418 + }, + { + "epoch": 6.39143135345667, + "grad_norm": 0.7581193556943887, + "learning_rate": 9.209663517174094e-07, + "loss": 0.1399, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0483221635222435, + "step": 3285, + "valid_targets_mean": 2562.1, + "valid_targets_min": 1065 + }, + { + "epoch": 6.401168451801363, + "grad_norm": 0.868024536322805, + "learning_rate": 8.920880262101894e-07, + "loss": 0.154, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11282053589820862, + "step": 3290, + "valid_targets_mean": 3496.5, + "valid_targets_min": 1513 + }, + { + "epoch": 6.410905550146056, + "grad_norm": 0.8888575978246757, + "learning_rate": 8.636593753945188e-07, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06756648421287537, + "step": 3295, + "valid_targets_mean": 2240.8, + "valid_targets_min": 682 + }, + { + "epoch": 6.42064264849075, + "grad_norm": 0.8529665918478495, + "learning_rate": 8.356810682940386e-07, + "loss": 0.1575, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11254268884658813, + "step": 3300, + "valid_targets_mean": 4025.2, + "valid_targets_min": 730 + }, + { + "epoch": 6.430379746835443, + "grad_norm": 0.9647205012714647, + "learning_rate": 8.081537633342562e-07, + "loss": 0.1518, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08208870887756348, + "step": 3305, + "valid_targets_mean": 3272.1, + "valid_targets_min": 1331 + }, + { + "epoch": 6.440116845180136, + "grad_norm": 0.964544059777689, + "learning_rate": 7.810781083270491e-07, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.060784801840782166, + "step": 3310, + "valid_targets_mean": 1944.2, + "valid_targets_min": 944 + }, + { + "epoch": 6.44985394352483, + "grad_norm": 0.9764587921394783, + "learning_rate": 7.544547404554236e-07, + "loss": 0.1542, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1391175389289856, + "step": 3315, + "valid_targets_mean": 2892.4, + "valid_targets_min": 1259 + }, + { + "epoch": 6.459591041869523, + "grad_norm": 0.9060986854710543, + "learning_rate": 7.282842862585204e-07, + "loss": 0.1519, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06699791550636292, + "step": 3320, + "valid_targets_mean": 2136.5, + "valid_targets_min": 1072 + }, + { + "epoch": 6.469328140214216, + "grad_norm": 0.7302814899655221, + "learning_rate": 7.025673616168704e-07, + "loss": 0.1493, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08695138990879059, + "step": 3325, + "valid_targets_mean": 4740.8, + "valid_targets_min": 557 + }, + { + "epoch": 6.47906523855891, + "grad_norm": 0.996714422784996, + "learning_rate": 6.773045717378912e-07, + "loss": 0.1613, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10565310716629028, + "step": 3330, + "valid_targets_mean": 2909.1, + "valid_targets_min": 1536 + }, + { + "epoch": 6.4888023369036025, + "grad_norm": 0.7573297467786815, + "learning_rate": 6.524965111416604e-07, + "loss": 0.1466, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09427705407142639, + "step": 3335, + "valid_targets_mean": 5019.6, + "valid_targets_min": 1891 + }, + { + "epoch": 6.498539435248296, + "grad_norm": 0.863330378987873, + "learning_rate": 6.281437636469135e-07, + "loss": 0.1598, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09574584662914276, + "step": 3340, + "valid_targets_mean": 3279.8, + "valid_targets_min": 550 + }, + { + "epoch": 6.508276533592989, + "grad_norm": 0.7794865658148458, + "learning_rate": 6.042469023573016e-07, + "loss": 0.1437, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06418845057487488, + "step": 3345, + "valid_targets_mean": 3014.4, + "valid_targets_min": 683 + }, + { + "epoch": 6.5180136319376825, + "grad_norm": 0.9881048043200676, + "learning_rate": 5.808064896479227e-07, + "loss": 0.1531, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08860848844051361, + "step": 3350, + "valid_targets_mean": 2982.4, + "valid_targets_min": 743 + }, + { + "epoch": 6.527750730282376, + "grad_norm": 0.9447565572161739, + "learning_rate": 5.578230771520665e-07, + "loss": 0.1486, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08905159682035446, + "step": 3355, + "valid_targets_mean": 3384.9, + "valid_targets_min": 862 + }, + { + "epoch": 6.537487828627069, + "grad_norm": 0.9284282146812189, + "learning_rate": 5.352972057482375e-07, + "loss": 0.1419, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.081339031457901, + "step": 3360, + "valid_targets_mean": 2548.0, + "valid_targets_min": 889 + }, + { + "epoch": 6.547224926971762, + "grad_norm": 0.6784066397902585, + "learning_rate": 5.132294055474307e-07, + "loss": 0.1358, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.04548973590135574, + "step": 3365, + "valid_targets_mean": 3270.2, + "valid_targets_min": 761 + }, + { + "epoch": 6.556962025316456, + "grad_norm": 0.7775427498027039, + "learning_rate": 4.916201958806621e-07, + "loss": 0.1354, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07671080529689789, + "step": 3370, + "valid_targets_mean": 3316.1, + "valid_targets_min": 1243 + }, + { + "epoch": 6.566699123661149, + "grad_norm": 0.6637750998088501, + "learning_rate": 4.704700852867361e-07, + "loss": 0.1494, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06787963211536407, + "step": 3375, + "valid_targets_mean": 4622.5, + "valid_targets_min": 1390 + }, + { + "epoch": 6.576436222005842, + "grad_norm": 1.0260630365672823, + "learning_rate": 4.497795715002795e-07, + "loss": 0.1508, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06825481355190277, + "step": 3380, + "valid_targets_mean": 1388.2, + "valid_targets_min": 612 + }, + { + "epoch": 6.586173320350536, + "grad_norm": 0.9363676312851329, + "learning_rate": 4.295491414400288e-07, + "loss": 0.1468, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.055219776928424835, + "step": 3385, + "valid_targets_mean": 1722.1, + "valid_targets_min": 1112 + }, + { + "epoch": 6.595910418695229, + "grad_norm": 0.8129631138954565, + "learning_rate": 4.0977927119737695e-07, + "loss": 0.1451, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.047071993350982666, + "step": 3390, + "valid_targets_mean": 2574.9, + "valid_targets_min": 1204 + }, + { + "epoch": 6.605647517039922, + "grad_norm": 0.8516137553282794, + "learning_rate": 3.9047042602516237e-07, + "loss": 0.153, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06578527390956879, + "step": 3395, + "valid_targets_mean": 3201.1, + "valid_targets_min": 1211 + }, + { + "epoch": 6.615384615384615, + "grad_norm": 0.772193092512818, + "learning_rate": 3.716230603267268e-07, + "loss": 0.1397, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06762434542179108, + "step": 3400, + "valid_targets_mean": 3307.2, + "valid_targets_min": 1614 + }, + { + "epoch": 6.625121713729309, + "grad_norm": 0.7368043674681918, + "learning_rate": 3.532376176452146e-07, + "loss": 0.1611, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07038801908493042, + "step": 3405, + "valid_targets_mean": 3472.5, + "valid_targets_min": 1613 + }, + { + "epoch": 6.634858812074002, + "grad_norm": 0.6416048814973079, + "learning_rate": 3.353145306531347e-07, + "loss": 0.1452, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08920369297266006, + "step": 3410, + "valid_targets_mean": 5517.2, + "valid_targets_min": 1445 + }, + { + "epoch": 6.644595910418695, + "grad_norm": 0.8370135405169448, + "learning_rate": 3.17854221142182e-07, + "loss": 0.1577, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07018916308879852, + "step": 3415, + "valid_targets_mean": 2871.5, + "valid_targets_min": 507 + }, + { + "epoch": 6.6543330087633885, + "grad_norm": 0.7498795681365913, + "learning_rate": 3.008571000133187e-07, + "loss": 0.1526, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.057709187269210815, + "step": 3420, + "valid_targets_mean": 3584.9, + "valid_targets_min": 1275 + }, + { + "epoch": 6.664070107108082, + "grad_norm": 0.9347846602773136, + "learning_rate": 2.843235672670841e-07, + "loss": 0.1619, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.058967068791389465, + "step": 3425, + "valid_targets_mean": 1801.9, + "valid_targets_min": 978 + }, + { + "epoch": 6.673807205452775, + "grad_norm": 0.8516991908572321, + "learning_rate": 2.682540119942023e-07, + "loss": 0.148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06559383124113083, + "step": 3430, + "valid_targets_mean": 2611.9, + "valid_targets_min": 896 + }, + { + "epoch": 6.6835443037974684, + "grad_norm": 0.9086726382103076, + "learning_rate": 2.5264881236640947e-07, + "loss": 0.1451, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07148931175470352, + "step": 3435, + "valid_targets_mean": 2447.8, + "valid_targets_min": 806 + }, + { + "epoch": 6.693281402142162, + "grad_norm": 0.7984514670137559, + "learning_rate": 2.375083356275676e-07, + "loss": 0.1542, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07836025953292847, + "step": 3440, + "valid_targets_mean": 3478.2, + "valid_targets_min": 822 + }, + { + "epoch": 6.703018500486855, + "grad_norm": 0.8100452297505701, + "learning_rate": 2.2283293808501138e-07, + "loss": 0.1493, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07283401489257812, + "step": 3445, + "valid_targets_mean": 3027.8, + "valid_targets_min": 1204 + }, + { + "epoch": 6.712755598831548, + "grad_norm": 0.7780103039934312, + "learning_rate": 2.086229651011684e-07, + "loss": 0.1605, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0804707333445549, + "step": 3450, + "valid_targets_mean": 3938.6, + "valid_targets_min": 1086 + }, + { + "epoch": 6.722492697176241, + "grad_norm": 0.6753064529876447, + "learning_rate": 1.9487875108542997e-07, + "loss": 0.145, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0617658756673336, + "step": 3455, + "valid_targets_mean": 3657.9, + "valid_targets_min": 729 + }, + { + "epoch": 6.732229795520935, + "grad_norm": 0.743711913805455, + "learning_rate": 1.8160061948628406e-07, + "loss": 0.1497, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08515971899032593, + "step": 3460, + "valid_targets_mean": 3728.2, + "valid_targets_min": 1911 + }, + { + "epoch": 6.741966893865628, + "grad_norm": 0.9202450054044959, + "learning_rate": 1.6878888278369255e-07, + "loss": 0.1492, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08222606778144836, + "step": 3465, + "valid_targets_mean": 2622.2, + "valid_targets_min": 1032 + }, + { + "epoch": 6.751703992210321, + "grad_norm": 0.8220976561569973, + "learning_rate": 1.5644384248176159e-07, + "loss": 0.1472, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.057911958545446396, + "step": 3470, + "valid_targets_mean": 2889.1, + "valid_targets_min": 1585 + }, + { + "epoch": 6.761441090555015, + "grad_norm": 0.8209075786638838, + "learning_rate": 1.445657891016161e-07, + "loss": 0.1511, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0630555972456932, + "step": 3475, + "valid_targets_mean": 3064.6, + "valid_targets_min": 1219 + }, + { + "epoch": 6.771178188899708, + "grad_norm": 0.7691294952327506, + "learning_rate": 1.3315500217458533e-07, + "loss": 0.1486, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0932154655456543, + "step": 3480, + "valid_targets_mean": 4289.0, + "valid_targets_min": 1582 + }, + { + "epoch": 6.780915287244401, + "grad_norm": 0.815710348733706, + "learning_rate": 1.2221175023561238e-07, + "loss": 0.1483, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07741531729698181, + "step": 3485, + "valid_targets_mean": 3159.1, + "valid_targets_min": 1049 + }, + { + "epoch": 6.790652385589095, + "grad_norm": 1.0470107205372206, + "learning_rate": 1.1173629081694392e-07, + "loss": 0.1412, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06302547454833984, + "step": 3490, + "valid_targets_mean": 2552.5, + "valid_targets_min": 631 + }, + { + "epoch": 6.800389483933788, + "grad_norm": 0.9122606227891553, + "learning_rate": 1.0172887044205936e-07, + "loss": 0.1635, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1148647665977478, + "step": 3495, + "valid_targets_mean": 3472.0, + "valid_targets_min": 1362 + }, + { + "epoch": 6.810126582278481, + "grad_norm": 0.9626562702605224, + "learning_rate": 9.21897246198733e-08, + "loss": 0.1495, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09417961537837982, + "step": 3500, + "valid_targets_mean": 1924.6, + "valid_targets_min": 622 + }, + { + "epoch": 6.8198636806231745, + "grad_norm": 0.6352208483338943, + "learning_rate": 8.311907783920437e-08, + "loss": 0.1501, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.047910600900650024, + "step": 3505, + "valid_targets_mean": 3332.6, + "valid_targets_min": 1171 + }, + { + "epoch": 6.829600778967867, + "grad_norm": 0.9136507988334205, + "learning_rate": 7.451714356347062e-08, + "loss": 0.1541, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08161959797143936, + "step": 3510, + "valid_targets_mean": 3096.0, + "valid_targets_min": 930 + }, + { + "epoch": 6.839337877312561, + "grad_norm": 1.2551160126692982, + "learning_rate": 6.638412422568463e-08, + "loss": 0.1445, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08053211867809296, + "step": 3515, + "valid_targets_mean": 2590.6, + "valid_targets_min": 1004 + }, + { + "epoch": 6.849074975657254, + "grad_norm": 1.0257326889744056, + "learning_rate": 5.8720211223677324e-08, + "loss": 0.1609, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10879699140787125, + "step": 3520, + "valid_targets_mean": 2752.0, + "valid_targets_min": 891 + }, + { + "epoch": 6.858812074001947, + "grad_norm": 0.9127077478996932, + "learning_rate": 5.1525584915605996e-08, + "loss": 0.1438, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0697256401181221, + "step": 3525, + "valid_targets_mean": 2522.8, + "valid_targets_min": 969 + }, + { + "epoch": 6.868549172346641, + "grad_norm": 0.9104831182013515, + "learning_rate": 4.480041461569773e-08, + "loss": 0.1444, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0774691104888916, + "step": 3530, + "valid_targets_mean": 2399.1, + "valid_targets_min": 1011 + }, + { + "epoch": 6.878286270691334, + "grad_norm": 0.8826495691129895, + "learning_rate": 3.854485859026591e-08, + "loss": 0.1439, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0804499015212059, + "step": 3535, + "valid_targets_mean": 2748.5, + "valid_targets_min": 909 + }, + { + "epoch": 6.888023369036027, + "grad_norm": 0.913417678940946, + "learning_rate": 3.2759064054002086e-08, + "loss": 0.1471, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05592145025730133, + "step": 3540, + "valid_targets_mean": 2153.2, + "valid_targets_min": 1253 + }, + { + "epoch": 6.897760467380721, + "grad_norm": 0.8502418717979767, + "learning_rate": 2.744316716648543e-08, + "loss": 0.1493, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07393450289964676, + "step": 3545, + "valid_targets_mean": 3059.4, + "valid_targets_min": 1200 + }, + { + "epoch": 6.907497565725414, + "grad_norm": 0.7870789251167677, + "learning_rate": 2.259729302900304e-08, + "loss": 0.1481, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06616164743900299, + "step": 3550, + "valid_targets_mean": 3009.4, + "valid_targets_min": 760 + }, + { + "epoch": 6.917234664070107, + "grad_norm": 1.1878012555440394, + "learning_rate": 1.822155568159012e-08, + "loss": 0.1542, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07589468359947205, + "step": 3555, + "valid_targets_mean": 1543.5, + "valid_targets_min": 986 + }, + { + "epoch": 6.926971762414801, + "grad_norm": 1.0618609799102148, + "learning_rate": 1.431605810034764e-08, + "loss": 0.1528, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08335718512535095, + "step": 3560, + "valid_targets_mean": 1765.9, + "valid_targets_min": 698 + }, + { + "epoch": 6.936708860759493, + "grad_norm": 0.6868086489492625, + "learning_rate": 1.0880892195028747e-08, + "loss": 0.135, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06128156557679176, + "step": 3565, + "valid_targets_mean": 4373.9, + "valid_targets_min": 2048 + }, + { + "epoch": 6.946445959104187, + "grad_norm": 0.9793651224632165, + "learning_rate": 7.916138806864927e-09, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06675729155540466, + "step": 3570, + "valid_targets_mean": 3163.9, + "valid_targets_min": 1613 + }, + { + "epoch": 6.9561830574488805, + "grad_norm": 0.9938042141448199, + "learning_rate": 5.421867706671968e-09, + "loss": 0.1454, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07806721329689026, + "step": 3575, + "valid_targets_mean": 2147.5, + "valid_targets_min": 879 + }, + { + "epoch": 6.965920155793573, + "grad_norm": 0.9088693887620701, + "learning_rate": 3.3981375932001792e-09, + "loss": 0.157, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08775289356708527, + "step": 3580, + "valid_targets_mean": 3062.0, + "valid_targets_min": 1572 + }, + { + "epoch": 6.975657254138267, + "grad_norm": 0.7280671741052036, + "learning_rate": 1.8449960917599207e-09, + "loss": 0.1479, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06314358115196228, + "step": 3585, + "valid_targets_mean": 3273.1, + "valid_targets_min": 816 + }, + { + "epoch": 6.98539435248296, + "grad_norm": 0.9776504526888148, + "learning_rate": 7.624797530958461e-10, + "loss": 0.1611, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06692498922348022, + "step": 3590, + "valid_targets_mean": 2156.6, + "valid_targets_min": 844 + }, + { + "epoch": 6.995131450827653, + "grad_norm": 0.8885675545896405, + "learning_rate": 1.5061405252980721e-10, + "loss": 0.1441, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08266542851924896, + "step": 3595, + "valid_targets_mean": 2982.6, + "valid_targets_min": 1288 + }, + { + "epoch": 7.0, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11300152540206909, + "step": 3598, + "total_flos": 8.035723720943206e+17, + "train_loss": 0.23149373163243941, + "train_runtime": 29019.8306, + "train_samples_per_second": 1.981, + "train_steps_per_second": 0.124, + "valid_targets_mean": 2509.4, + "valid_targets_min": 1446 + } + ], + "logging_steps": 5, + "max_steps": 3598, + "num_input_tokens_seen": 0, + "num_train_epochs": 7, + "save_steps": 600, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 8.035723720943206e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000..40a1833 --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:acf4d997c7c9a62d10bb24576cbdd3d90d756e9c9a0cb95c461e37504f75b5a2 +size 8593 diff --git a/training_loss.png b/training_loss.png new file mode 100644 index 0000000..71265df Binary files /dev/null and b/training_loss.png differ diff --git a/vocab.json b/vocab.json new file mode 100644 index 0000000..6c49fc6 --- /dev/null +++ b/vocab.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910 +size 2776833